“mud; . . . . . .. has, {vii—1.4 :7. .

:12... .. a

.L. ,
a n ii
0.1:!) halt:

i. 7.. 2:.

t . 23.3: v!

.. a
A .. uh. ‘
. m..a...§5.m...é5«$
adnuflrh... 1’91. Q4.».Q.5§

.I “but: .1“ it

«:6
sir.“
J22;

\Izatiit.

1 unfit:
2...?

”wwwhiuu

.
I i

 

 

Ir éflis‘nnl infi. \r
.il. r ‘I. ii. “In.

(#5546
Z- l

100}

This is to certify that the

dissertation entitled

ELECTROCATALYTIC HYDROGENATION OF a-
FUNCTIONALIZED CARBOXYLIC ACIDS

 

presented by

Tulika Sanjeev Dalavoy

LIBRARY
Michigan State
University

has been accepted towards fulfillment
of the requirements for the

 

 

 

PhD. degree in Chemistry

QM... {Quorum

Major P/ofessor’s Signature

2.5" pr Mcjycst. 2-006
Date

MSU is an Affirmative Action/Equal Opportunity Institution

PLACE IN RETURN BOX to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

36581935303:

 

 

 

 

 

 

 

 

 

 

 

 

 

6/07 p:/CIRC/Date0ue.indd-p.1

 

 

ELECTROCATALYTIC HYDROGENATION OF a-FUNCTIONALIZED
CARBOXYLIC ACIDS

By

Tulika Sanjeev Dalavoy

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Chemistry

2006

ABSTRACT

ELECTROCATALYTIC HYDROGENATION OF a-FUNCTIONALIZED
CARBOXYLIC ACIDS

By

Tulika Sanjeev Dalavoy

The goal of our research is to develop an alternative, greener, safer and milder process for
the metal catalyzed hydrogenation of biomass-derived carboxylic acids to value-added
products. Direct catalytic reduction of biomass-derived carboxylic acids avoids the
additional esterification step, apart from other advantages such as easier product recovery
and 100% atom economy. Ru/C catalyzed hydrogenation of lactic acid to propylene
glycol requires high temperature and pressure (150°C, 1200 psi H2).

Electrocatalytic hydrogenation (ECH) offers an alternative approach to traditional
chemical catalysis. In this case the hydrogen is produced by electrolytic decomposition of
water and the reaction occurs between hydrogen and the target molecule, both adsorbed
on the electrode surface. This eliminates the kinetic barrier to the dissolution of the
poorly soluble hydrogen gas in aqueous medium, mass transport of hydrogen and the
splitting of hydrogen molecules into hydrogen atoms, and enables us to perform
heterogeneous catalytic hydro genations at ambient temperature and pressure. Advantages
of this process include conservation of energy, possibility of stereoretention at low
temperature and increased safety due to elimination of high pressure and hydrogen
storage requirements. It also enables us to study catalytic surface mechanisms with in-situ
spectroscopic techniques, such as ATR-IR and Raman. Electrocatalytic hydrogenation

has previously been carried out on various substrates such as phenol, esters and

diketones, using electrodes, such as Raney Ni, Pt and Raney Co. However, to the best of
our knowledge, this is the first study of the ECH of carboxylic acids. In this study, the
ECH of various a-functionalized carboxylic acids has been attempted and Optimized.
Initial studies using the model compound, benzoyl formic acid showed that the ECH
reaction is dependent more on the availability of surface sites, rather than the rate of mass
transport, which is consistent with the results obtained with the catalytic hydrogenation of
lactic acid. ECH of lactic acid using electrodeposited Ru on carbon felt, though giving a
poor yield of product, enabled the determination of kinetic parameters, such as order of
reaction, fractional surface coverages of lactic acid and hydrogen, rate constants,
adsorption constants and activation energy using the Langmuir-Hinshelwood model.
ECH of lactic acid using 5%Ru/C agglomerated in Reticulated Vitreous Carbon (RVC)
resulted in much higher conversion. The effect of various parameters such as
temperature, ECH current and nature of electrolyte on the yields and distribution of
products was studied. The unexpected finding from this work is that lactaldehyde (2-
hydroxypropanal), a proposed intermediate in the path of LA hydrogenation, is the major
product instead of the expected propylene glycol, with small quantities of propylene
glycol obtained only with 5%Ru/C/RVC. This thermodynamically unusual result reflects
the relative surface adsorption and aqueous hydration properties of the starting materials
and the two products. For the purpose of comparison, similar studies were carried using
Ru deposited on non-carbon supports. In-situ ATR-IR spectroscopy has been used to
probe the mode of binding of the substrate on the electrode surface, the effect of
electrolyte anions and the mechanism by which the ct-position functional group facilitates

the hydrogenation of carboxylic acids.

ACKNOWLEDGEMENTS
I am gratefiil to all the people who have helped me during my graduate career. First of all
I would like to thank my advisor, Dr. Jackson for his constant support and encouragement
during the course of the project and for inspiring me to take up new challenges and try to
solve them. Through frequent interactions with him, I have learnt to think critically and
look at problems from various angles. I would also like to thank our collaborator, Prof.
Dennis Miller from the Department of Chemical Engineering and Materials Science for
his inputs. Our collaboration with the chemical engineering group has given me the
unique opportunity to gain insights, develop critical thinking, work among diverse
professionals and communicate effectively across disciplines. I am grateful to our
collaborator, Dr. Swain for providing me the labspace and instrumentation to carry out
my experiments and for sharing his expertise in electrochemistry. The occasional
discussions with him on various aspects of the project has played an important role in
expanding my knowledge. A special thanks to my committee members at Michigan State
University, Dr. Bruening, Dr. Baker and Dr. Pinnavaia for reading my thesis and
providing helpful suggestions.
I would like to acknowledge Prof. J acek Lipkowski and his group from the University of
Guelph for their help with the ATR-FTIR experiments. I am also grateful to Prof. Carol
Kowzeniewski (Texas Tech Univ.) and Prof. Hugues Menard (Univ. of Sherbrooke) for
their generous help and advice.
I would also like to thank the Jackson and Miller group members for their help and
support and the Swain group members for helping me with the electrochemical

techniques and instrumentation. I gratefully acknowledge the National Science

iv

 

Foundation for financial support and the Department of Chemistry at Michigan State
University for the teaching and research assistantships. Last, but not the least, I would

like to thank my husband, parents, in-laws and friends for their support.

TABLE OF CONTENTS

LIST OF TABLES .................................................................................. ix
LIST OF FIGURES ................................................................................ xi
CHAPTER 1 INTRODUCTION ................................................................. 1
1.1 Importance of lactic acid and biomass based raw materials .............................. 1
1.2 Advantages of heterogeneous catalytic hydrogenation .................................... 2
1.3 Previous work on catalytic hydrogenation of esters and carboxylic acids .............. 2
1.4 Ru/C catalyzed hydrogenation of lactic acid ................................................ 6
1.4.1 Mechanistic studies ....................................................................... 7
1.4.2 Kinetic studies ........................................................................... 13
1.4.3 Studies on other a substituted carboxylic acids ..................................... 16

1.5 Introduction to electrocatalytic hydrogenation ............................................ 17
1.6 Motivation for research ....................................................................... 25
1.7 Outline of dissertation ......................................................................... 27

CHAPTER 2 ELECTROCATALYTIC HYDROGENATION OF A MODEL «-

FUNCTIONALIZED CARBOXYLIC ACID, BENZOYL FORMIC ACID ......... .30
2.1 Introduction ..................................................................................... 30
2.2 Experimental ................................................................................... 33

2.2.1 Electrochemical cell ..................................................................... 33
2.2.2 Electrode preparation .................................................................... 35
2.2.3 ECH method ............................................................................... 36
2.2.4 Analysis of products ..................................................................... 36
2.3 Results and Discussion ....................................................................... 45
2.3.1 Cyclic voltammetry of benzoyl formic acid .......................................... 45
2.3.2 Determination of active metal surface area .......................................... 48
2.3.3 Current-time transients .................................................................. 49
2.3.4 ECH of BFA on Pt polycrystalline electrode ........................................ 50
2.3.5 Determination of optimum analyte concentration ................................... 52
2.3.6 Effect of step potential on ECH ....................................................... 54
2.3.7 ECH of BFA using boron-doped diamond (BDD) composite electrodes ........ 55
2.3.8 ECH of MA .............................................................................. 59
2.3.9 Effect of mass transport ................................................................ 61
2.4 Conclusions .................................................................................... 62

CHAPTER 3 ELECTROCATALYTIC HYDROGENATION OF LACTIC

ACID USING RUTHENIUM ON CARBON FELT ELECTRODE .................... 64
3.1 Introduction ..................................................................................... 64
3.2 Experimental ................................................................................... 67

3.2.1 Set up of electrochemical cell and ECH method .................................... 67
3.2.2 Electrode preparation .................................................................... 68
3.2.3 Characterization of the electrode surface ............................................. 69
3.2.4 Analysis of the products of ECH ...................................................... 69

vi

3.3 Results and Discussion ....................................................................... 77

3.3.1 Electrodeposition of Ru on carbon felt ................................................ 77
3.3.2 Factors governing lactic acid conversion .............................................. 87
3.3.3 Determination of the order of reaction ................................................. 95
3.3.4 Hydrogen evolution on Ru .............................................................. 98
3.3.5 Determination of the rate constant of lactic acid electrohydrogenation .......... 99
3.4 Conclusions ................................................................................... 105

CHAPTER 4 BULK ELECTROCATALYTIC HYDROGENATION OF
LACTIC ACID USING RETICULATED VITREOUS CARBON

AGGLOMERATED WITH RUTHENIUM 0N CARBON ............................. 107
4.1 Introduction .................................................................................. 107
4.2 Experimental ................................................................................. 110

4.2.1 Electrochemical cell setup ............................................................ 110
4.2.2 Electrocatalytic hydrogenation ....................................................... 111
4.2.3 Characterization of the electrode surface ........................................... 112
4.2.4 Analysis of products ................................................................... 114
4.3 Results and Discussion ..................................................................... 114
4.3.1 Factors governing lactic acid conversion ............................................ 115
4.3.2 Anion adsorption on electrode surfaces ............................................. 125
4.3.3 Study of anion and hydrogen adsorption on Ru using cyclic voltammetry.....128
4.3.4 Hydrogen adsorption on metal surfaces ............................................. 132
4.3.5 ECH of other a-functionalized carboxylic acids .................................... 135
4.3.6 Reason for lactaldehyde formation ................................................... 136
4.4 Conclusrons/138

CHAPTER 5 BULK ELECTROCATALYTIC HYDROGENATION OF
LACTIC ACID USING RETICULATED VITREOUS CARBON
AGGLOMERATED WITH RUTHENIUM ON NON-CARBON SUPPORTS.....141

5.1 Introduction ................................................................................... 141
5.2 Experimental ................................................................................. 143
5.2.1 Catalyst preparation .................................................................... 143
5.2.2 Electrochemical cell setup ............................................................ 144
5.2.3 Electrocatalytic hydrogenation ....................................................... 148
5.2.4 Characterization of the electrode surface ........................................... 149
5.2.5 Analysis of products ................................................................... 149
5.3 Results and Discussion ..................................................................... 149
5.3.1 Catalyst structure ....................................................................... 149
5.3.2 Factors governing lactic acid conversion (5% Ru/SiOz) .......................... 151
5.3.3 Factors governing lactic acid conversion (5% Ru/BaSO4) ........................ 156
5.4 Conclusions ................................................................................... 161

CHAPTER 6 IN-SITU VIBRATIONAL SPECTROSCOPIC STUDIES OF

ELECTROCATALYTIC HYDROGEN ATION OF LACTIC ACID ................. 162
6.1 Introduction .................................................................................... 162
6.2 ATR-IR and the SEIRAS effect ............................................................ 165

vii

6.3 Experimental ................................................................................. 173

6.3.1 ATR cell configuration ................................................................ 173
6.3.2 Preparation of Au thin film ............................................................ 174
6.3.3 Deposition of Ru ........................................................................ 176
6.3.4 Experimental setup ...................................................................... 176

6.4 Results and Discussion ..................................................................... 177
6.4.1 Optimum concentration of lactic acid ................................................ 177
6.4.2 IR spectra of electrolytes .............................................................. 178
6.4.3 In-situ FTIR study of the ECH of lactic acid ....................................... 184

6.5 Conclusions ................................................................................... 195
CHAPTER 7 CONCLUSIONS AND FUTURE WORK ................................. 197
7.1 Conclusions ................................................................................... 197
7.2 Suggestions for future work ................................................................ 200
APPENDIX ........................................................................................ 205
REFERENCES .................................................................................... 235

viii

Table 2.1:

Table 3.1:

Table 3.2:

Table 3.3:

Table 3.4:

Table 3.5:

Table 3.6:

Table 3.7:

Table 4.1:

Table 4.2:

Table 5.1:

Table 5.2:

LIST OF TABLES

Active Pt surface areas for the Pt electrode and composite electrodes
calculated from the area under the hydrogen desorption peaks in

figure 2.9 ................................................................................ 49
Response factors of analytes used for quantifying unknown samples from
ECH ....................................................................................... 76
Summary of Ru electrodeposition conditions ...................................... 87
ECH of lactic acid using Ru/carbon felt prepared using two Ru

electrodeposition times (11.1 mM lactic acid in 0.01 M H2804 for 21 h at
70°C, i= 100 mA). The error in the HPLC is estimated to be :1:3%. The
standard deviation of the yield of lactaldehyde at 11.1 mM initial
concentration of lactic acid is i5.2% ............................................................... 88

ECH of lactic acid using Ru/carbon felt at various temperatures (11.1 mM
lactic acid in 0.01 M H2804 for 21 h, i= 40 mA). The error in the HPLC is
estimated to be :l:3%. The standard deviation of the yield of lactaldehyde at
11.1 mM initial concentration of lactic acid is 35.2% ...................................... 90

ECH of lactic acid using Ru/carbon felt at various currents (11.1 mM lactic
acid in 0.01 M H2804 for 21 h at 70°C). The error in the HPLC is estimated
to be i3%. The standard deviation of the yield of lactaldehyde at 11.1 mM
initial concentration of lactic acid is 3:5.20/0 .................................................... 91

Final yields of lactaldehyde after ECH of lactic acid at various initial
concentrations (Electrolyte: 0.01 M HCl, T=70°C, time=21h, i=100mA).....95

Kinetic parameters for the EC H of lactic acid at two different
temperatures ........................................................................... 104

Yields of products of ECH of lactic acid at various temperatures (11.1 mM
lactic acid in 0.01 M H2804 after 9 h; i= 40 mA). The error in the HPLC is
estimated to be i3% ....................................................................................... 116

Final yields (%) of products after lactic acid electrohydrogenation in
various electrolytes (T= 70°C, i= 100 mA) ........................................ 123

Comparison of the characteristics of the silica support in 5% Ru/SiOz
catalyst ................................................................................. 150

ECH of lactic acid using 5% Ru/SiOg/RVC at various temperatures

(11.1 mM lactic acid in 0.01 M HCl for 21h. i= 40 mA). The error in the
HPLC-determined concentrations is estimated to be i3% ............................ 152

ix

Table 5.3: ECH of lactic acid using 5% Ru/SiOZ/RVC at various currents
(11.1 mM lactic acid in 0.01 M HCl for 21h at 70°C). The error in the
HPLC is estimated to be 21:3% ........................................................................ 153

Table 5.4: ECH of lactic acid using 5% Ru/BaSO4/RVC at various currents
(11.1 mM lactic acid in 0.01 M H2804 for 21h at 70°C). The error in the
HPLC is estimated to be i3% ........................................................................ 157

Table 5.5: ECH of lactic acid using 5% Ru/BaSO4/RVC at various currents
(11.1 mM lactic acid in 0.01 M H2804 for 21 h at 70°C). The error in the
HPLC is estimated to be i3% ........................................................................ 158

LIST OF FIGURES

Figure 1.1: Conversion of biomass to propylene glycol ........................................ 2
Figure 1.2: Catalytic hydrogenation of u-substituted esters to alcohols in water ............ 4
Figure 1.3: Catalytic hydrogenation of carboxylic acids to alcohols in water ............... 5
Figure 1.4: Aqueous phase hydrogenation of lactic acid to propylene glycol ............... 6
Figure 1.5: Comparison of the catalytic hydrogenation of substituted and

unsubstituted carboxylic acids ........................................................ 8
Figure 1.6: Mechanism of ester hydrogenation proposed by Scrum and Onsager‘7 ........ 8
Figure 1.7: Mechanism of acetate hydrogenolysis described by Kohler et a1.‘9 ............. 9

Figure 1.8: Mechanism of Ru catalyzed ethyl lactate hydrogenation involving
promotion by Sn (Luo and coworkersz‘) .......................................... 11

Figure 1.9: Proposed mechanism of lactic acid hydrogenation from deuterium labeling
studies (Jackson and coworkers)24 .................................................. 13

Figure 1.10: Langmuir-Hinshelwood mechanism for lactic acid hydrogenation...........14

Figure 1.11: General scheme of electrocatalytic hydrogenation and hydrogen
evolution .............................................................................. 18

Figure 2.1: Schematic diagram of electrochemical cell (adapted from Granger, M.C.;
Witek, M.; Xu, J .; Wang, J .; Hupert, M.; Hanks, A.; Koppang, M.D.;
Butler, J .E.; Lucazeau, G.; Mermoux, M.; Strojek, J .W.; Swain, G.M.

Anal. Chem, 2000, 72, 3793-3804) ................................................ 34
Figure 2.2: Schematic diagram of the electrochemical cell used with the rotating disc

electrode (RDE) (R-reference, W-working. C-counter) .......................... 35
Figure 2.3: HPLC-UV chromatogram_standard showing peaks for BFA, MA and PED

(concentration=l.33 X 10" M) ...................................................... 37
Figure 2.4a: Calibration curve for BF A (R2=0.9951) .......................................... 38
Figure 2.4b: Calibration curve for MA (R2=0.9986) ........................................... 39
Scheme 2.1: Batch synthesis of CEO ............................................................ 40

Figure 2.5a: GC-FID chromatogram standard showing peaks for the TMS derivatives

xi

of CEG and HMA (concentration: 2.0 mg/L) .................................... 42

Figure 2.5b: GC-FID chromatogram standard showing peaks from the blank reagent
(BSTFA + TMCS) run under similar conditions ................................ 43

Figure 2.6: Calibration curves for HMA (top) and CEG (bottom) ........................... 44
Figure 2.7: Cyclic voltammogram of 6.67 X 10'5 M (10 ppm) BFA in 0.01 M H2804

on Pt polycrystalline working electrode at a scan rate of 10 mV/s.

Arrows indicate direction of scan ................................................... 46
Figure 2.8: Cyclic voltammograms in 0.1 M H2804 in the presence (top) and absence

(bottom) of BF A at a scan rate of 50 mV/s (data obtained by Prof. Carol

Korzeniewski at Department of Chemistry, Texas Tech University).........47

Figure 2.9: Cyclic voltammograms of Pt polycrystalline, Pt/BDD, and Pt/Ru/BDD in

0.1 M HClO4 at 100 mV/s ............................................................. 48

Figure 2.10: Current-time response of an EC H experiment with Pt/Ru/BDD in 0.01 M
H2804 (E.=400 mV, E2=-400 mV) ............................................... 50
Scheme 2.2: ECH of benzoyl formic acid (possible products) ................................ 51

Figure 2.11: Conversion of BFA as a function of initial analyte concentration during
ECH on polycrystalline Pt electrode (electrolyte: 0.01M H2804,
temperature= 25°C, El: 400 mV, E2: -425 mV, time = 2 h). Error bars
indicate uncertainty in the analysis of products (i10%) using GC-FID.....53

Figure 2.12: Conversion of BFA as a function of final step potential (E2) during ECH
on polycrystalline Pt electrode (electrolyte: 0.01 M H2804, initial [BFA]
= 6.67 X 10'5 M, temperature: 25°C , time = 2 h) ................................ 54

Figure 2.13: Conversion of BF A as a function of final step potential (E2) during ECH
on Pt and Pt/Ru composite electrodes (electrolyte: 0.01 M H2804, initial
[BFA]= 6.67 X 10'5 M, temperature: 25°C, time = 2 h) ........................ 56

Figure 2.14: Comparison between adsorption patterns on Pt/BDD and Pt/Ru/BDD
electrodes ............................................................................. 57

Figure 2.15: Conversion of BFA as a function of pH during ECH on Pt and Pt/Ru
composite electrodes (electrolyte: 0.01 M H2804, initial [BFA]=
6.67 X 10'5 M, E2: -400 mV, temperature: 25°C, time = 2 h) ................ 58

Figure 2.16: Conversion of BF A as a function of temperature during ECH on Pt and

Pt/Ru composite electrodes (electrolyte: 0.01 M H2804, initial [BFA]=
6.67 X 10’ M, E2: -400 mV, time = 2 h) ........................................ 59

xii

Figure 2.17: Conversion of MA as a function of step potential (E2) during ECH on a
polycrystalline Pt and composite electrodes (electrolyte: 0.01 M H2804,
initial [MA]= 6.58 X 10" M, temperature =25°C, time = 2 h). Graph
shows yields of CEG ............................................................... 60

Figure 2.18: Conversion of BFA as a function of rotation rate during ECH on a
polycrystalline Pt rotatingdisc electrode (electrolyte: 0.01 M H2804,
initial [BFA]= 6.67 X 10'3 M, E2: -400 mV, temperature =25°C, time
= 2 h). Graph shows yields of C EG .............................................. 62

Figure 3.1: Schematic diagram of the electrochemical cell used for ECH
(W-Working, A-Auxiliary, R-Reference) ......................................... 68

Figure 3.2: HPLC chromatogram (standard 0.1 wt %) showing peaks for lactic acid
(LA) (retention time = 13.4 min.) and propylene glycol (PG)

(retention time= 17.4 min.) ........................................................... 70
Figure 3.3: Calibration curve for lactic acid (R2: 0.9966) .................................... 71
Figure 3.4: Calibration curve for propylene glycol (R2: 0.9937) ........................... 72
Scheme 3.1: Synthesis of lactaldehyde .......................................................... 74

Figure 3.5: HPLC chromatogram of the D20 extract from the synthesis of
lactaldehyde using Method 1. The mixture contains the product
lactaldehyde (LAL) (retention time= 14.7 min.) and reactant, propylene
glycol (PG) ............................................................................ 76

Figure 3.6: (A) SEM of bare carbon felt; (B) SEM of carbon felt after 100 s of Ru
electrodeposition from 5 mM RuCl3.xH2O in 0.01 M H2804 (i= 10 mA)....78

Figure 3.7: Low magnification image of Ru/carbon felt deposited from 5 mM
RuCl3.xH20 in 0.01 M H2804 (i= 10 mA) with a deposition time of 90 min
(top). This is the magnification used for the quantitative analysis with
EDS.
Energy dispersive spectrum (EDS) of Ru/carbon felt showing the X-ray
signal for the Ru L line (bottom) ................................................... 79

Figure 3.8: SEM images of Ru/carbon felt after electrodeposition of Ru from 5 mM
RuCl3.xH2O in 0.01 M H2804 (i= 10 mA) for (A) 1 min; (B) 5 min;
(C) 30 min; (D) 90 min ............................................................... 80

Figure 3.9: Cyclic voltammetry (I-E curve) of carbon felt in 5 mM RuC13.xH20 in
0.01 M H2804 at three different scan rates. Arrows indicate the direction
of scan ................................................................................. 83

xiii

Figure 3.10: Amperometric i-t curve of carbon felt in 5 mM RuCl3.xH20 in
0.01 M H2804 at three different applied potentials .............................. 85

Figure 3.11: Amperometric i-t curve of carbon felt in 5 mM RuCl3.xH20 in
0.01 M H2804 at three different applied potentials; expansions of the
shorter time regions showing rapid current increase due to nucleation. .....86

Figure 3.12: Yield of lactaldehyde as a function of time in various electrolytes
(11.1 mM lactic acid in 0.01 M electrolyte concentration at 70°C for 21 h,
i=100 mA) ............................................................................ 92

Figure 3.13: Moles of lactaldehyde as a function of time in various electrolytes
(11.1 mM lactic acid in 0.01 M electrolyte at 70°C for 21 h,
i=100 mA) ............................................................................. 93

Figure 3.14: Moles of lactaldehyde formed as a function of time at various initial
lactic acid concentrations (lactic acid in 0.01 M HCl at 70°C for 21 h,
i=100 mA) ............................................................................ 94

Figure 3.15: Moles of lactaldehyde formed as a function of time at various ECH
currents (11.1 mM lactic acid in 0.01 M HCl at 70°C for 21 h) ............. 96

Figure 3.16: TOF for lactaldehyde formation as a function of initial moles of
lactic acid in 0.01 M HCl at 70°C after 21 h, i=100 mA ...................... 97

Figure 3.17: TOF for lactaldehyde formation as a function of moles of H atoms
formed per hour (assumed to be directly proportional to current)
(11.1 mM lactic acid in 0.01 M llCl at 70°C after 21 h) ...................... 97

Scheme 3.2: Proposed mechanism for hydrogen evolution on Ru ........................... 99

Figure 3.18: Moles of lactaldehyde formed as a function of time at 50°C (323 K)
(11.1 mM lactic acid in 0.01 M HCl. i=100 mA) .............................. 103

Figure 3.19: Arrhenius plot for ECH of lactic acid ........................................... 104

Figure 4.1: SEM images of an RVC working electrode (100 ppi): bare RVC
electrode (top); two regions of the RVC electrode after entrapment of
Ru/C via agglomeration for 2 h at 20 mA under cathodic polarization
(bottom) ............................................................................... 113

Figure 4.2: Yield of lactic acid reduction products after electrohydrogenation
(11.1 mM lactic acid in 0.01 M H2804 for 9 h at 90 °C). Only
lactaldehyde and propylene glycol were detected. The error in the
HPLC is estimated to be 13% ...................................................................... 117

xiv

Scheme 4.1: Sequence of reactions in lactic acid hydrogenation ........................... 118

Figure 4.3: Comparison of rates of ECH of lactic acid (11.1 mM) with 5% Ru/C/RVC
in various electrolytes at a concentration of 0.1 M (top) and 0.01 M
(bottom) (T= 70°C. i= 100 mA). The error in HPLC is estimated to
be :l:3% .......................................................................................................... 122

Scheme 4.2: Electrocatalytic reduction of perchlorate to chloride on Ru proposed
by Colom80 .......................................................................... 127

Figure 4.4a: Cyclic voltammogram of a Ru polycrystalline electrode in various
electrolytes (0.01 M) (A= 0.2 cm°. I): 10 mV/s) ............................... 130

Figure 4.4b: Cyclic voltammogram of a Ru polycrystalline electrode in various
electrolytes (0.01 M) in the presence of lactic acid (0.01 M)
(A= 0.2 cm2, 1): 10 mV/s) .......................................................... 131

Figure 4.5: Volcano curve for electrocatalysis in the hydrogen evolution region
(HER) in terms of log i0 plotted as a function of heat of adsorption of H
(from ref. 86) ......................................................................... 134

Scheme 4.3: ECH of lactic acid analogs using 5% Ru/C agglomerated in RVC
(T= 70°C, i= IOOmA, electrolyte: 0.01 M HCl). Complete conversion
of the reactants was obtained, but actual yields of the aldehydes were
not determined ...................................................................... 135

Scheme 4.4: ECH of mandelic acid ............................................................ 137

Figure 5.1: SEM images of 5% Ru/SiO2 (commercial catalyst) (A) Secondary
Electron image (B) Backscattered electron image .............................. 145

Figure 5.2: SEM images of 5% Ru/8i02 prepared by wet impregnation method.
Arrows indicate Ru particles ...................................................... 145

Figure 5.3: Energy Dispersive Spectra (EDS) of 5% Ru/Si02: commercial catalyst
(top); in-house prepared catalyst (bottom) ....................................... 146

Figure 5.4: SEM images of 5% Ru/Ba804 prepared by wet impregnation method,
showing both low magnification (top) and high magnification (bottom)
images ................................................................................ 147
Figure 5.5: Energy Dispersive Spectra (EDS) of 5% Ru/Ba804 ........................... 148
Figure 5.6: Comparison of rates of ECH of lactic acid (11.1 mM) with

5% Ru/Si02/RVC in various electrolytes at a concentration of 0.01 M
(top) and 0.1 M (bottom) (T= 70°C. i2 100 mA). The error in HPLC is

XV

estimated to be i3% ...................................................................................... 154

Figure 5.7: ECH of lactic acid (1 1.1 mM, T= 70°C. i= 100 mA) in HCl, with 5%
Ru/SiO2/RVC prepared in-house. The error in HPLC is estimated to be
:l:3% .............................................................................................................. 156

Figure 5.8: Comparison ofrates of ECH of lactic acid (1 1.1 mM) with
5% Ru/BaSO4/RVC in various electrolytes at a concentration of 0.01 M
(top) and 0.1 M (bottom) (T= 70°C. i= 100 mA). The error in HPLC is
estimated to be i3% ..................................................................................... 159

Figure 6.1: Schematic diagram of Attenuated Total Reflection (ATR) in the
horizontal configuration showing the important parameters for spectral
acquisition and calculation of penetration depth ................................ 166

Figure 6.2: Electromagnetic enhancement effect of Surface enhanced infrared
absorption spectroscopy (SEIRAS) showing the metal particles of the
island film modeled as a set of rotating ellipsoids with the axis of
rotation normal to the substrate surface (from ref. 140) (E: is the electric
field of the incident radiation) ..................................................... 168

Figure 6.3: Schematic diagram of the cell used in the in-situ ATR-FTIR studies
of the ECH of lactic acid ............................................................ 175

Figure 6.4: ATR-IR spectra of the electrolyte solutions, obtained using the spectrum
of water as the reference (T = 30°C, E: -400 mV) (solid lines) and
ATR-IR spectra taken during the ECH of lactic acid with the water
spectrum as the reference (T= 30°C , E: —400 mV) (dotted lines) ............. 182

Figure 6.5: Schematic diagram showing different modes of adsorption of water
molecules on the electrode surface at various potentials ....................... 184

Figure 6.6: Real time ATR-IR spectra of the EC 1-1 of lactic acid in 0.01 M H2804
(T=30°C, PF -400 mV) using the spectrum of 0.01M H2804 taken under
identical conditions as the reference .............................................. 186

Figure 6.7: Schematic diagram showing the mode of adsorption of hydronium ion
on the electrode surface ............................................................ 187

Figure 6.8: Possible modes of adsorption of lactic acid on a metal surface ............... 189

Figure 6.9: Symmetric and asymmetric stretching modes of the carboxylate group
and their resultant transition dipoles ............................................. 189

Figure 6.10: Real time ATR-IR spectra of the ECH of lactic acid in 0.01 M HCl
(T=30°C, E= -400 mV) using the spectrum of 0.01M HCl taken under

identical conditions as the reference ............................................ 191
Figure 6.11: Real time ATR-IR spectra of the ECH of lactic acid in 0.01 M HC104

(T=30°C, E= -400 mV) using the spectrum of 0.01 M HC104 taken

under identical conditions as the reference .................................... 195

Figure 7.1: Use of a cationic surfactant to modify the electrode surface .................. 203

Figure Al: NMR spectrum of the D20 extract obtained from PCC oxidation of

1,2-propanediol ('H, D20) ......................................................... 205
Figure A2: NMR spectrum of the D20 extract obtained from PCC oxidation of

1,2- propanediol (l3 C, D20) ........................................................ 206
Figure A3: NMR spectrum of lactaldehyde dimethyl acetal ('H, D20) ................... 207
Figure A4: NMR spectrum of lactaldehyde obtained by the hydrolysis of

lactaldehyde dimethyl acetal in 0.05 M also. for 2 h ('H, D20) ............. 208
Figure A5: NMR spectrum of lactaldehyde obtained by hydrolysis of lactaldehyde

dimethyl acetal in 0.05 M H2804 for 2 n (‘-’c. 020) ............................ 209
Figure A6: FTIR spectrum of lactaldehyde obtained by hydrolysis of lactaldehyde

dimethyl acetal in 0.05 M H2804 for 2 h (NaCl plate) ......................... 210
Figure A7: NMR spectrum of product mixture obtained after electrocatalytic

hydrogenation of lactic acid in 0.01 M HC104 for 21 h (1H, 0.0) ........... 211
Figure A8.l: Expansion ofthe peaks in figure A7 (5 0.5-3.0 ppm) ........................ 212
Figure A8.2: Expansion of the peaks in figure A7 (5 2.6-4.4 ppm) ........................ 212
Figure A8.3: Expansion of the peaks in figure A7 (5 8.24-8.40 ppm) ..................... 213
Figure A9: NMR spectrum of product mixture obtained after electrohydrogenation

of lactic acid in 0.01 M HC104 for 21 it ("C, 1320) ........................... 214
Figure A10: ESI-MS of chemically synthesized lactaldehyde .............................. 215
Figure All: ESI-MS of lactaldehyde obtained after electrohydrogenation of lactic

acid in 0.01 M HC104 ............................................................. 216
Figure A12: NMR spectrum of product mixture obtained after electrocatalytic
hydrogenation of glycolic acid in 0.01 M HCl for 21 h ('H, 020) ......... 217

Figure A13.l: Expansion of the peaks in figure A12 (0 3.16328 ppm) .................. 218

xvii

Figure A13.2: Expansion of the peaks in figure A12 (5 8.28-8.39 ppm) .................. 219

Figure A14: ESI-MS of sample obtained after electrohydrogenation of glycolic acid

in 0.01 M HCl ...................................................................... 220
Figure A15: NMR spectrum of alanine ('H, D20) ............................................ 221
Figure A16: NMR spectrum of product mixture obtained after electrocatalytic

hydrogenation of alanine in 0.01 M HCl for 21 h (1H, D20) ................ 222
Figure A17.1: Expansion of the peaks in ft gure A 16 (6 1.2-3.8 ppm) ..................... 223
Figure Al7.2: Expansion of the peaks in figure A16 (5 8.30-8.37 ppm) .................. 224
Figure A18: NMR spectrum of the product of batch reduction of l-phenyl

1,2—ethanediol (PED) ('11. com.) .............................................. 225
Figure Al9.1: Expansion of the peaks in figure A18 (5 0.6-2.6 ppm) ..................... 226
Figure A19.2: Expansion of the peaks in figure A18 (8 3.1-4.0 ppm) ..................... 227
Figure A20: NMR spectrum of the product of batch reduction of l-phenyl

1,2-ethanediol (PED) (”c, CDCI3) .............................................. 228
Figure A21.l: Expansion of the peaks in figure A20 (0 20-34 ppm) ...................... 229
Figure A21.2: Expansion of the peaks in figure A20 (0 73—81 ppm) ...................... 230

Figure A22: FTIR spectrum of cyclohexyl ethylene glycol (CEG) synthesized
by batch reduction of phenyl ethanediol (PED) (KBr pellet) ................ 231

Figure A23: Electron ionization mass spectrum (El-MS) of trimethyl silyl (TMS)
derivative of cyclohexyl ethylene glycol (CEG) synthesized by batch
reduction of phenyl ethanediol (PED) derivatized using BSTF A and
separated by GC (DBl column) ................................................. 232

Figure A24: Electron ionization mass spectrum (El-MS) of CEG obtained by
ECH of mandelic acid, after precipitation of sulfate, evaporation
of water, derivatization using BSTFA and separation by GC
(DB1 column) ...................................................................... 233

Figure A25: Electron ionization mass spectrum (El-MS) of 2-cyclohexyl
2-hydroxy acetaldehyde (CHA) obtained by ECH of mandelic acid,
after precipitation of sulfate, evaporation of water, derivatization
using BSTFA and separation by (‘16 (D81 column) .......................... 234

xviii

CHAPTER 1

INTRODUCTION
1.1 Importance of lactic acid and biomass based raw materials
Due to the increasing scarcity of petroleum and petroleum products, there is a great need
to find alternative sources of raw materials for commodity chemicals. One such
alternative source is biomass. Biomass-based organic acids such as lactic acid can be
easily obtained by the fermentation of com-derived glucose. Lactic acid can then be
converted to propylene glycol (1,2-propanediol) by heterogeneous catalytic
hydrogenation using metals or metal oxides (figure 1.1). Propylene glycol, an important
commodity chemical, used at a rate of a billion kilograms per year in the US. is
important in chemical and food industries“. It is also used as a coolant in automobiles and
as a medium in formulations of cosmetics“. Currently, propylene glycol is commercially
produced by the hydration of propylene oxide (2-methyl oxirane), which is obtained by
the oxidation of propene, a petroleum based raw material.
Heterogeneous catalytic hydrogenation has the advantage of being an atom economical
process, comprising of the addition of a molecule of hydrogen across a double or triple
bond. Although a large number of processes are known for the heterogeneous catalytic
hydrogenation of olefins, aldehydes and ketones, there are relative fewer examples of

heterogeneous catalytic hydrogenation of carboxylate compounds.

 

‘ Information from http://en.wikipedia.org/wiki/Propylene_glycol

CHO

 

 

 

——OH
Hg—L—H fermentation_ OH H2, catalyst _ OH
HTFOH ACOOH NO”
HflCEOgH lactic acid propylene glycol
2
D- glucose

Figure 1.1: Conversion of biomass to propylene glycol

1.2 Advantages of heterogeneous catalytic hydrogenation

Carboxylic acid hydrogenations have been carried out using metal hydride reagents like
LiAlH4 and NaBH4 and hydrogen in the presence of organometallic catalysts. But these
methods suffer from several disadvantages. Metal hydrides are highly toxic and
expensive and some of them are pyrophoric. They may bring about racemization of the
stereogenic sites in the alcohol products. They are also difficult to use industrially or on a
large scale because of the problems involved in separating the catalyst and the reductant’s
inorganic byproducts from the target compounds and the problems of waste disposal.
Heterogeneous catalytic hydrogenations, on the other hand, do not involve any problems
in the separation of the product from the catalyst. Such catalysts are also easier to handle
compared to homogeneous catalysts and can be regenerated and reused. However, they
may suffer from low selectivities due to the high temperature and pressure conditions
involved, though this definitely can be overcome in some cases by the optimization of

conditions and use of better catalysts.

1.3 Previous work on catalytic hydrogenation of esters and carboxylic acids
In the 1930’s, Adkins and coworkers reported the heterogeneous catalytic hydrogenation

of esters such as ethyl lactate, ethyl [i-phenyl propanoate, diethyl sebacate, etc. over

Cu/Ba/Cr02 at 250°C at 2800-4300 psi with yields in the range of 90.98%,la
hydrogenation of ethyl mandelate and ethyl phenyl acetate over Raney Ni and Cu
chromite at 200°C and 1400-2400 psi with yields over 90%lb and the stereoretentive
hydrogenation of optically active esters such as menthyl phenyl butyrates over Cu/Cr02
at zoo-250°C with yields ranging from 17 to 73%.2

Preparation of ethylene glycol from glycolic acid esters was reported by Um and
coworkers3 using oxide catalysts such as CuO, ZnO, C00, etc. at 300-40,000 psi in
alcohol solvent. Roesky and coworkers4 prepared phenyl ethanols with 95% selectivity
(ratio of the % yield of the desired product to the % conversion) and 65% conversion by
the catalytic hydrogenation of phenylacetate esters at 200°C and 1305 psi over a
Cu0/Zr02 catalyst. Similarly cyclohexylmethanol was prepared by the gas phase
hydrogenation of cyclohexyl carboxylic acid or the ethyl ester in the presence of oxides
of Group IV A or IV B elements (eg. Zr02) at 250-400°C and 360-870 psi. Recently, Luo
and coworkers5 used Ru-B/y-A1203, containing small amounts of Sn for the
hydrogenation of ethyl lactate to 1,2-propanediol in n—heptane at 150°C and 800 psi with
91.5% selectivity at 90.7% conversion after 10 h. In this case, the hydrogenation to 1,2-
propanediol was found to proceed directly from the ester and not through the lactic acid
intermediate. Enantioselective hydrogenation of ethylpyruvate was reported by Schurch
and coworkers6 using 1-(9-anthracenyl)-2-(l-pyrrolidinyl)-ethanol modified Pt. The
enantioselective hydrogenation of a-keto esters over cinchona modified Pt was
discovered by Orito et al.7 About 90-92% enantioselectivity was achieved using this

catalyst.

The earliest examples of aqueous phase heterogeneous catalytic hydrogenation of
carboxylate compounds were in 1948 by Adkins and coworkers8 who carried out the
hydrogenation of (it-substituted esters like ethyl 2-amino propanoate, ethyl a-(N-
piperidyl) acetate and ethyl lactate to the corresponding primary alcohols, 2 amino
propanol, B (N -piperidyl) ethanol and propylene glycol respectively. The catalyst used
was Raney Ni at 100°C and a pressure of 5000 psi. The products were obtained in a yield
of 80% in all cases (figure 1.2). By using high catalyst loading, esters of alpha
substituted acids were hydrogenated to the corresponding alcohols at a low temperature

of 50°C with retention of configuration.

 

 

NH? RalNi H NH2
0000 H 2: JVOH Yield=80%
2 5 100°C 5000 i
ethyl 2 amino ’ ps 2 amino propanol
propionate
OH Ra/Ni’ H2 ‘ OH Y' ld -80‘V
”coca 100°C, sooo'psi A/OH '6 ‘ °
ethyl lactate propylene glycol

Figure 1.2: Catalytic hydrogenation of tit-substituted esters to alcohols in water

Even though hydrogenation of esters is relatively easier than that of carboxylic acids, it is
important to develop processes for the hydrogenation of carboxylic acids to the
corresponding alcohols because carboxylic acids are more directly available from
biomass processing and the direct hydrogenation of carboxylic acids eliminates the
intermediate esterification step for the production of alcohols and therefore, is of high

value in renewables-based chemicals production. The first aqueous phase hydrogenation

of free carboxylic acids was carried out in 1955 by Carnahan and coworkers.9 Several
carboxylic acids like acetic acid and a hydroxy acetic acid were hydrogenated to ethanol
and ethylene glycol respectively in the presence of the ruthenium based catalysts and
reaction conditions shown in figure 1.3.

RuO , H - ._
CH30H2OOH 2 2 , CH3CH20H Yield —88%

147-17o°c, 10h
10000-14000 psi

 

 

10% Ru/C
HOCH2COOH 2 H0. ,0“ Yie.d=30%
145-150°c, 1h
950040.500 psi

Figure 1.3: Catalytic hydrogenation of carboxylic acids to alcohols in water

Broadbent et a1. '0 carried out the first catalytic hydrogenation studies of lactic acid using
an unsupported Re black catalyst at 150°C and 3791 psi for 8 h, giving propylene glycol
with an yield of 84%.

Mendes and coworkersll studied the hydrogenation of oleic acid to stearyl alcohol in
tetradecane over Ti02 and A1203 supported Ru catalysts. Hydrogenation over supported
bimetallic Ru-Sn catalyst was found to selectively reduce the carboxylic acid group,
leading to the formation of unsaturated alcohols. Toba and coworkers12 reported the
hydrogenation of dicarboxylic acids and saturated fatty acids to the corresponding diols
and alcohols respectively using Ru-Sn-Al203 catalysts. In case of fatty acids the
conversion and yield increased with increase in the carbon chain length. Rachmady and
coworkers13 studied the vapor phase hydrogenation of acetic acid over Pt supported on
various oxides at ISO-300°C, 100-700 Torr hydrogen and 50-70 Torr acetic acid. The

product selectivity was found to be strongly dependent on the nature of the oxide support,

with the various being ethanol, ethyl acetate, acetaldehyde, methane, C0 and C02. The

Ti02 support showed the highest activity.

1.4 Ru/C catalyzed hydrogenation of lactic acid

Recently, an efficient aqueous phase hydrogenation of lactic acid to propylene glycol,
catalyzed by 5% Ru/C has been developed. This reaction shows a higher reaction rate and
better selectivity to propylene glycol under milder conditions compared to the ones used
earlier, offering real potential for industrial application.'4 The reaction is shown in figure
1.4. The aqueous environment in this case is favourable for the reaction and has the
advantage of being non-hazardous and a good solvent for biomass-derived carboxylic
acids such as lactic acid. Ru based catalysts were found to show good intrinsic
hydrogenation activity for carbonyl compounds in earlier studies“ "'2 and Ru has been
found to be a highly efficient catalyst for hydrogenation especially in aqueous solution.

Carbon is a good support for the catalyst due to its inertness in aqueous medium.

H .‘\OH H2, 5%) RU/C H \OH
)\coou t 2H2 ’ ‘

lactic acid

OH '1’ H20

 

150°C, 1200 psi
propylene glycol

Figure 1.4: Aqueous phase hydrogenation of lactic acid to propylene glycol

Control experiments were carried out to prove that no propylene glycol was formed in the
absence of either catalyst or hydrogen. Several catalysts were tested for efficiency of
hydrogenation of lactic acid. It was found that the ruthenium on carbon catalyst is the

most active catalyst. Almost 95-98% conversion and more than 90% ee was achieved

after 5 hours at 150°C and 1200 psi pressure. These conditions are mild compared to
those used earlier for the hydrogenation of lactate esters.

The reaction both with respect to the rate of conversion and selectivity to propylene
glycol are favoured by high pressure, but at temperatures above 170°C, there are more
secondary reactions leading to the formation of gaseous and liquid by-products such as

ethanol, propanol and hydrocarbons like ethane, propane, etc.

1.4.1 Mechanistic studies

The conversion of lactic acid to propylene glycol was found to occur 3-5 times faster than
the conversion of propanoic acid to 1- propanol.” However, the mechanism by which the
—OH group at the a position activates the hydrogenation is unclear. It may be due to the
electron withdrawing effect of the —OH group which enhances the neighbouring carbonyl
group’s preference for sp3 hybridization. This explanation is supported by the observation
that protonated amino acids16 and 2- chloro propanoic acid are also more reactive towards
hydrogenation, and that the charged ammonium group in the case of protonated amino
acids is more activating than the hydroxy] group in lactic acid. Stabilization of
intermediates on the catalyst surface by intramolecular hydrogen bonding must play an
important role.

As shown in figure 1.5, the measured heats of reaction for the hydrogenation of
propanoic acid and lactic acid are almost the same.14 Therefore, it was concluded that

kinetic factors may play an important role in the reaction.

 

Acoon . 2H2 ——> N0“ + H20
propanoic acid 1-propanol

AH: -10.49 kcal/mol

H .OH to H .OH
ACOOH 4‘ 2H2 ‘_——’ )K/OH + H20
lactic acid propylene glycol

AH= -11.00 kcal/mol

Figure 1.5: Comparison of the catalytic hydrogenation of substituted and unsubstituted

carboxylic acids

In earlier studies, mechanisms for the hydrogenolysis of esters to alcohols were proposed.
In the gas phase hydrogenolysis of methyl fonnate on Cu/SiO2, it was suggested by
Scrum and Onsagerl7 that there is a formation of a hemiacetal intermediate followed by
dissociation to formaldehyde and methanol. The aldehyde is then rapidly hydrogenated to
methanol as shown in figure 1.6. This mechanism was confirmed by Monti et al.18 by in

situ infrared spectroscopy and labeling studies.

 

o
+ H r H c A H
Haco’lJ‘H 2 3 O O
H3COAOH CH30H + HCHO

 

HCHO + H2 ——> CHaOH

Figure 1.6: Mechanism of gas phase ester hydrogenation proposed by Sorum and

Onsager17

In the case of Cu/Si02 catalyzed acetate hydrogenolysis, it was proved by Kohler et al.19
using FTIR that the acetate is adsorbed dissociatively on the catalyst surface as shown in
figure 1.7 where 8 indicates a catalytic site on the surface.

RCOOR' + S RCO* + R'O“

 

Figure 1.7: Mechanism of acetate hydrogenolysis described by Kohler et al.19

0n carrying out isotopic labeling studies it was found that the alkoxy fragment, R’0*
reacts quickly to form R’OH, while the acyl group, RC0“ is adsorbed longer. The
hydrogenation of the acyl fragment is believed to be the rate determining step, which can
be hydrogenated to either the desired alcohol or to the aldehyde which is then
hydrogenated to the alcohol.

In the hydrogenation of acetic acid on Cu/Si02, studied by Dumesic and coworkers,20
adsorption of aliphatic carboxylic acids and esters was proposed to occur by the cleavage
of the C-0 bond of the carboxylate group, giving rise to acyl species, RCO". The
existence of this species on Cu/8i02 surface was supported by FTIR studies and was
further confirmed by the high initial heats of adsorption observed in rnicrocalorimetric
studies of methyl acetate, ethyl acetate and acetaldehyde adsorption on Cu/Si02 at 300K
as well as the observed production of gaseous species upon ester adsorption at higher
temperatures. The surface bound acetyl is therefore a relatively stable species and does
not immediately hydrogenate to the aldehyde and alcohol. DFT calculations20 suggest
that the activation energy required for dissociative adsorption increase in the order ethyl

acetate < methyl acetate < acetic acid. This is due to differences in the C-0 bond adjacent

to the carbonyl group, whose strength increases from ethyl acetate to acetic acid, due to
the ability of the R group to donate electron density to the alkoxy oxygen atom. 0n the
basis of DFT calculations, Neurock and coworkers21 proposed a model like those above
for acetic acid hydrogenation on Pd(1 1 1), which involves dissociation of acetic acid via
C-OH bond breaking to form surface acetyl and hydroxyl species, and is followed by
either addition of H to the a-carbon atom forming acetaldehyde and subsequent
hydrogenation to ethanol or the decomposition of the acetyl species to form C0 and
methyl species via C-C bond scission. According to the calculations, the addition of H to
the oxygen end of the acetyl species to form hydroxyl ethylidene’ is an endothermic
reaction and is less likely to occur. It was also found that the initial adsorption of acetic
acid prefers the 111 mode through the carbonyl oxygen, which is more favorable than the
di-o mode involving two oxygen atoms.

In the Ru-B/y-Al203 catalyzed hydrogenation of ethyl lactate to 1,2-propanediol, it was
observed by Luo and coworkers22 that the addition of small quantities of Sn significantly
increased the selectivity and conversion to the diol. The most active catalyst was the one
containing a Sn/Ru ratio of 7%. It was proposed that the effect of Sn is based on the
formation of an oxidized Sn surface (Sn2+) in interaction with metallic Ru, in which 8n
plays the role of activating and polarizing the carbonyl group, whereas the electron rich
Ru site activates hydrogen. The activated hydrogen on each Ru site then attacks the
positively charged C atom of the carbonyl group, forming an acetal of Sn, which is then
converted to aldehyde, followed by rapid hydrogenation to the alcohol as shown in figure

1.8. A similar mechanism was proposed for the hydrogenation of fatty acid esters by

10

23 24

Deshpande et a1 and Pouilloux et a1 and for the hydrogenation of oleic acid over Ru-

Sn catalysts by Mendes et al.1]

H2 C2H50H

CszO R 0sz0 R U RCHO

 

o a? H
l
Ru——Sn0x
H R= CH3CH(OH)-
Rt't—Snox
Ru—SnOx
l'

Figure 1.8: Mechanism of Ru catalyzed ethyl lactate hydrogenation involving promotion

by 8n (Luo and coworkers”)

Mechanistic studies were carried out on the hydrogenation of lactic acid using isotopic
labeling, NMR analysis and chiral GC-MS by Jackson and coworkers.25 The
heterogeneous catalytic hydrogenation of lactic acid leads primarily to the formation of
1,2 propylene glycol with retention of stereochemistry at C2. There are several possible
ways in which the hydrogenation might occur as shown in figure 1.9. One involves only
the carboxylic carbon, C1 which leads to the formation of 1,2 propylene glycol via 1,1,2
propanetriol (lactaldehyde hydrate) and lactaldehyde with retention of configuration at

C2. If this path is followed, then the reaction carried out using D2 in D20 should lead to

11

 

deuterium incorporation only at C l , leading to dideuterated propylene glycol. 0n the
other hand, if fast tautomerization of the lactic acid to the enol form takes place followed
by C=C hydrogenation and then the dehydration of the geminal diol, intermediate
formation of pyruvaldehyde, and hydrogenation of the C=0 bond leading to propylene
glycol, C2 will also be involved in the mechanism of hydrogenation. In such a case
hydrogenation in the presence of D2/D20 would lead to deuterium incorporation at both
C1 and C2 of propylene glycol and a loss of stereochemistry as shown in scheme 2.
However, the tautomerization of lactic acid shown in Route B is less likely to occur as it
has a very high activation energy. Instead, based on standard pKa data, the
tautomerization of lactaldehyde shown in Route C appears more likely and chemical
scrambling would lead to the same result with respect to D incorporation and
stereoselectivity of the reaction.

In a key control experiment, propylene glycol was exposed to hydrogenation conditions
in the presence of D2/D2O and the deuterium incorporation in the product was followed
by '3 C NMR. The '3 C NMR spectra collected at different times after the beginning of the
reaction showed D incorporation at both C1 and C2 of the product. The H/D exchange
was much faster at C1 than at C2, giving almost complete incorporation at C1 well before
significant exchange at C2. No H/D exchange was detected at C3”.

The retention of stereochemistry and the slow exchange of D onto C2 even in propylene
glycol suggest that the first proposed mechanism (path A) shown in figure 1.9 dominates
over the alternative competing paths for lactic acid hydrogenation. Therefore, direct
hydrogenation occurs at the carboxylic carbon, C1 only. The others (path B and C)

involve tautomerization to enol forms followed by hydrogenation of the C=C double

12

a.

bonds or equilibrated C=0 bonds. All these alternative paths however reflect

stereochemical scrambling at C2.

Route A (Complete D incorporation at C1)

 

+D IRu 00 o +Dleu oo oo
oo o 2 oo on +120 M >——éo
H DIR ; EODD D 0
OD ' 2 u lactaldehyde
i + DO 00
Dleu DO on 'H20 Do 0 2A. “(Ho
D D

DO __ ; < = ..
Hon 0 0% 0‘ 0

Route B (Partial D incorporation at C2)

\ +02/RU
O

o oo 0
D _. )1 /<
4 :oo 0
pyruvaldehyde
Routec
DO 0 DO _ OD DZIRU DO ODD
4 :o / D ”o 0
lactaldehyde

Figure 1.9: Proposed mechanism of lactic acid hydrogenation from deuterium labeling

studies (Jackson and coworkers)24

1.4.2 Kinetic studies
In kinetic studies of the Ru/C catalyzed hydrogenation of lactic acid to propylene

glycol,ls

the system was treated as a three-phase reaction involving H2 gas, an aqueous
solution of lactic acid and product propylene glycol in the liquid phase and the Ru/C
catalyst in the solid phase. The reaction involves four types of mass transfer processes,

namely hydrogen mass transfer from the gas to the liquid phase, lactic acid and hydrogen

mass transfer from the liquid to the solid catalyst phase, diffusion of hydrogen and lactic

13

 

acid within the porous catalyst and the conversion of lactic acid to propylene glycol via a
series of surface chemical reactions. In hydrogen solubility studies, the solubility of
hydrogen in water was found to increase with increase in pressure, but the solubility of
hydrogen in lactic acid solution was found to be 10% lower than that in water. ' Mass
transfer analysis revealed very minor effects of liquid-solid and gas-liquid mass transfer
rates and of the hydrogen partial pressure on the rate of hydrogenation. From initial rate
analysis, the apparent activation energy of the process was calculated to be 94 kJ/mol.

A Langrnuir-Hinshelwood model was proposed for lactic acid hydrogenation in which
molecular hydrogen and lactic acid adsorb on the catalyst surface, followed by reaction to
form propylene glycol and desorption of propylene glycol from the surface, as shown in

figure 1.10. The rate limiting step is irreversible.

 

 

 

 

 

 

 

 

LA + S LA-S S- surface site

H2 + S H2-S

”(.3 + H2-S [31.3 4. S Rate determining step
P1-S + H2-S PG-S

PG-S PG + S

Figure 1.10: Langmuir-Hinshelwood mechanism for lactic acid hydrogenation

In accordance with the assumptions of the Langmuir-Hinshelwood model, the adsorption
of water is neglected, the total catalyst site density is considered constant and all sites are
assumed to be equivalent. The final rate expression in terms of kmol/kg of catalyst/s is

given as follows.

-R (kmol/kg of cat/s) = KCLAPH2

 

(1 + KHZPHZ + KLACLA)2

14

where K is the rate constant, CLA is the lactic acid concentration, Pm is the hydrogen
partial pressure and KH2 and KM are the adsorption constants of lactic acid and hydrogen
on Ru. The term for propylene glycol was not included in the denominator because the
presence of propylene glycol was found to have negligible effect on the rate of lactic acid
hydrogenation. The heats of adsorption of lactic acid and hydrogen were determined to be
47 and 79 kJ/mol respectively.

A Langmuir-Hinshelwood model was also proposed by Luo and coworkers5 for Ru/Sn
catalyzed ethyl lactate hydrogenation and by Rachmady and coworkers13 for the
hydrogenation of acetic acid over Pt/Ti02. The apparent reaction order was found to be
04-06 with respect to H2 and 0.2-0.4 with respect to acetic acid. Unlike the lactic acid
hydrogenation, this model incorporates dissociated hydrogen and acetic acid adsorption
on one type of site and only molecular acetic acid adsorption at another type of site on the
oxide surface and only the latter species was considered catalytically significant for
ethanol, acetaldehyde and ethane formation. Only Pt supported on Ti02 and n-Al203
yielded ethanol and acetaldehyde as products whereas Pt/ Si02 caused decarbonylation,
leading to CH4 and C0 and Pt/Fe203 yielded acetone as the major product, indicating that
catalyst activity and product selectivity are controlled by the adsorption behavior of
acetic acid on the various oxide surfaces and the availability of activated atomic

hydrogen on Pt.
1.4.3 Studies on other a substituted carboxylic acids

In order to study the trends in the reactivity of various acids to hydrogenation and to gain

a better understanding of the mechanism, a number of a substituted carboxylic acids were

15

subjected to hydrogenation26 and the rates and percentage conversion in each case was
measured. The compounds can be listed in the order of increasing reactivity as follows.
Propanoic acid < 2- methoxy propanoic acid < methoxy acetic acid < glycolic acid <
lactic acid < ethyl lactate.

2-chloro propanoic acid and 2-acetoxy propanoic acid were converted to propanoic acid
and lactic acid respectively under the hydrogenation conditions (150°C, 1200 psi H2).
Based on these studies, it was concluded that hydrogen bonding may have a significant
effect on the rate of hydrogenation and selectivity of the product. This is proved by the
higher reactivity of the tit-hydroxy substituted lactic and glycolic acid relative to their
CH3-capped analogs. The lower reactivity of 2-methoxy propanoic acid, even in the
presence of a higher catalyst to substrate ratio may be because hydrogen bonding with the
carboxyl group is no longer possible when the -0H of lactic acid is replaced by —0CH3.
The much higher reactivity of methoxy acetic acid compared to that of methoxy
propanoic acid indicates that steric factors may also affect the reactivity. However, the
fact that isobutyric acid reacted much faster than propanoic acid indicates that the
mechanism is more dependent on the exact mode of binding of the substrate to the
catalyst surface which is believed to vary considerably with the structure of the substrate.
However many of these results were somewhat qualitative in nature and no definite
conclusions could be drawn from them.

The aqueous phase hydrogenation of alanine to alaninol over 5% Ru/C catalyst was
studied by Jere and coworkers. 16 A selectivity of 95% to L-alaninol and cc of 99% was
achieved at 150°C and 1000 psi in 4 h in the presence of 0.29 M H3P04 (required to

convert the zwitterionic form of the amino acid to the completely protonated form). In

16

more recent studies,27 the replacement of the amine hydrogens with —CH3 groups was
found to significantly decrease the rate of hydrogenation to the corresponding alcohol,

just as had been found in the 0H/0CH3 systems.

1.5 Introduction to electrocatalytic hydrogenation

The use of electrochemical methods to produce organic compounds could provide an
alternative approach to the traditional chemical synthesis methods. Substantial work has
been done in the past few years on electrocatalytic hydrogenation (ECH) or
electrohydrogenation of various substrates like phenol, esters, 01- and B-diketones,
conjugated enones, etc. on electrodes such as Raney Ni, Pt, Raney Co, etc.

The electrocatalytic hydrogenation is thought to occur in two consecutive steps. The first
step involves the electrolytic decomposition of the solvent, usually water, with the
formation of hydrogen atoms adsorbed on the electrode. The second step involves
hydrogen atom migration on the electrode surface to react with the organic substrate also
adsorbed on the electrode surface to give the final reduction product. Therefore, ECH has
a lot in common with heterogeneous catalytic hydrogenation, but it offers an advantage in
bypassing the kinetic barrier to the dissolution of the poorly soluble hydrogen gas
(especially in aqueous medium), the mass transport of hydrogen and the splitting of
hydrogen molecules to hydrogen atoms. This obviates use of high pressure conditions to
get the hydrogen into the aqueous phase and provides a method of carrying out
heterogeneous catalytic hydrogenation under milder conditions. In most of the cases, the

mechanism on the catalyst surface and the mode of binding of the substrate to the catalyst

17

surface in electrohydrogenation was found to be similar to that of heterogeneous catalytic

hydrogenation carried out under high temperature and pressure condition828'30.

The various steps involved in ECH can be represented as in figure 1.11.

 

 

 

 

 

 

 

 

 

 

2H30” 21420
2e"
\ / tr
=Z
\ /YH-ZH
H H Y—Z H H H HY—ZH H H H
1 l
' '\'\1\'\'\ ——~ kr'rkxk ——-> \\\\\\
H2
Tl? Tafelreactiion
\\\\\\ :::<<\\
Competinghydrogen
evolutionreactions
H301 _ H2+H20 /
\ H H Heyrovskireaction‘ 1L1
\\KK\\ ' \\\\\\

Figure 1.11: General scheme of electrocatalytic hydrogenation and hydrogen evolution

From the above steps it can be seen that for a given electrode at a given potential, both
electrohydrogenation and hydrogen evolution can occur simultaneously and there is
always competition between the two processes. Once the electrode surface is saturated

with hydrogen atoms, some of them desorb and evolve as hydrogen gas. However the

18

relative amounts of hydrogenation and hydrogen evolution depend on several factors like
the electrode material, morphology, applied potential, substrate concentration, reaction
medium, pH and temperature. Since ECH occurs as a result of the reaction between the
adsorbed substrate and adsorbed hydrogen and not with hydrogen gas, the process will be
more efficient if the various parameters are optimized to a point where the
electroreduction of protons to hydrogen atoms can occur, but where hydrogen evolution
is minimized.

Though ECH has been carried out successfully on carbonyl compounds like diketones
and enones, to the best of our knowledge, no such studies have been reported on
carboxylic acids. This technique might be a greener, milder and more efficient alternative
to the batch process developed earlier both with respect to industrial production and
mechanistic studies.

Beyond the milder conditions, ECH might offer additional advantages; in some cases, the
adsorption of poisons and the consequent fouling of the catalyst surface is prevented by
the cathodic polarization of the electrode. In comparison to hydrogenation by
electroreduction followed by protonation, ECH requires less energy because the
formation of chemisorbed hydrogen by the electrolysis of water on a low hydrogen
overvoltage cathode requires a less negative potential compared to the electroreduction of
most organic functional groups. In some cases, ECH has also been found to be more
selective compared to chemical catalytic hydrogenation due to the milder conditions
involved. For example, Lessard and coworkers28 achieved the ECH of 2-cyclohexen-1-
one to cyclohexanone with high selectivity using electrodeposited Ni and Cu electrodes

(about 100% selectivity at 94% conversion). In some cases the stereoselectivity of the

19

reaction can be controlled by controlling the potential. For example, in the ECH of m-
xylene on platinized platinum electrodes in 0.5M H2804, Quiroz and coworkers29 found
that the stereoisomeric ratio of the products, cis and trans 1,3-dimethyl cyclohexane was
strongly dependent on the potential. The cis/trans ratio was 79:92.] at 0.24V (vs RHE)
and decreased at more negative potentials.This was explained by the change in
orientation of the adsorbed m-xylene from edgewise to parallel orientation as the
concentration of H atoms and water molecules on the electrode surface increased at more
negative potentials.

Many ECH reactions have been carried out successfully at room temperature and have
produced reasonably good yields of desired products and high enantioselectivity. The
conditions need to be optimized to obtain the maximum rate of hydrogenation and to
minimize the rate of the competing hydrogen evolution reaction (HER) by the Heyrovsky
and Tafel reactions. The current efficiency is a measure of the competition between
hydrogenation and HER which depends on a number of factors such as substrate,
electrode, electrolyte, pH and temperature. Most ECH reactions reported in the literature
were found to be most efficient at low currents and moderate temperature and high
currents and temperatures were found to increase the rate of HER.

Mahdavi and coworkers30 studied the ECH of conjugated enones using nickel boride,
fractal nickel and Raney nickel electrodes in aqueous methanol. In this case ECH was
found to be more selective than catalytic hydrogenation. They showed that in the ECH of
cyclohex-Z-ene-l-one, the selectivity was quite low with the more active Raney Ni
electrodes, but at nickel boride electrodes, the selectivity of the C=C hydrogenation and

the hydrogenation efficiency can be increased to 100% at 95-100% conversion by

20

decreasing the current density and increasing the substrate concentration. Fractal nickel
electrodes were found to give higher selectivity than nickel boride electrodes under
certain conditions and were found to be useful for the selective ECH of C=C of a number
of conjugated enones such as alkyl substituted cyclohexenones.

Grimshaw and coworkers3 ' reported ECH of 4-phenylbuten-2-ones to 4-phenyl-butan-2-
ones at Ni black cathode in aqueous ethanol containing H2804. The maximum current
efficiency of 63-68% was obtained using high substrate concentrations.

Senda and coworkers32 studied the ECH of 4-t-butyl cyclohexanone to the corresponding
alcohol using Raney Ni, Raney Co and Pt/C powder electrodes. A high conversion of
ketone to alcohol was observed on the Raney catalysts when the pKa of the proton source
employed was around 5. The conversion over Pt/C was low (15%), but increased with
increasing current density.

Navaro and coworkers33 studied the ECH of diethyl fumarate to diethyl succinate in an
undivided cell using electrodes such as Fe, Pt, Ni, Pb, Al, Cu, Zn and carbon felt. The
conditions were optimized with respect to current density, supporting electrolyte, pH and
co-solvent. The best results for hydrogenation were obtained using Fe electrode using a
current density of 175 mA/dm3 in H20/CH3CN (4:1) at pH 5.5 with dimerization of the
ester being the major competing reaction

Cheong and coworkers34 studied the ECH of ketones and a and B diketones using Raney
nickel, Raney cobalt or Devarda copper electrodes embedded in a nickel matrix in an
aqueous medium. They achieved chemoselective reductions of only one carbonyl group
of a-diketones and showed that the ECH of benzil gave either either benzoin or

hydrobenzoin depending on the electrode used. Also the diastereoselectivity observed for

21

the diol formation in the ECH of a and B diketones was comparable to that obtained in
catalytic hydrogenation. Using modified Raney nickel, the ECH of benzil produced
hydrobenzoin with a threozmeso ratio of 7:93.

Romulus and coworkers35 have shown that ECH of the heterocyclic aromatic ring can
occur in competition with the electroreduction of carboxylic acids and esters of pyridine
to the corresponding alcohols in a sulfuric acid medium using a lead cathode. ECH of the
aromatic nucleus occurred mainly on non-reducible derivatives such as hemiacetal or
hydrate. Hu and coworkers36 carried out the ECH of 4-amino 5-nitrosodimethyl uracil to
4,5 diamino dimethyl uracil using a foamed nickel cathode. This reaction is an important
step in the synthesis of caffeine and has the potential for replacing the currently used
industrial method of Fe powder catalyzed hydrogenation which leads to poor
hydrogenation efficiency and a high level of pollution due to the resulting F e304 slurry.
The foamed nickel was found to give better current efficiency compared to Raney nickel
and other electrodes studied, due to its higher surface area and porosity and its reticulated
three dimensional structure. Conversions of over 98% and current efficiencies of 95-
100% were obtained at the optimum conditions i.e. pH 3-4, a current density of 84-
168Am'2 and a temperature of 20-30°C.

Polcaro and coworkers37 studied the ECH of benzaldehyde and acetophenone in acidic
and alkaline ethanol/water mixture using Pd supported on carbon felt due to its high
surface area and good fluid permeability. In this case the total current efficiency and the
yields of the products were found to be greatly influenced by the method of preparation

and the morphology of the electrode.

22

Thomalla and coworkers have carried out the ECH of phenol,38 alkyl substituted
phenols,39 carvone and limonene4O using Raney nickel electrodes in the presence of non-
micelle fonning cationic surfactants. The presence of a surfactant was found to increase
the efficiency of ECH by increasing the substrate solubility and by the formation of an
adsorbed hydrophobic surfactant layer on the electrode surface, thus increasing the
efficiency of adsorption of the hydrophobic substrates. These hydrogenation products are
widely used in the perfume and fragrance industry.

Roessier and coworkers41 studied the ECH of an aqueous suspension of indigo to water
soluble leuco indigo in a divided flow cell using Raney nickel particles fixed in a porous
bed of plated nickel. The reaction could be scaled upto IOg/L. The reaction rate was
enhanced by high pH, the presence of an organic cosolvent such as methanol and
ultrasonic waves, due to the increase in the active reaction surface as as result of decrease
in particle diameter. Under optimum conditions of current density and temperature a
conversion of 95% and a current efficiency of 12.7% was achieved. However, the current
efficiency could be increased to 19.2% by doping the Ranay nickel electrodes with Pt
particles. In this case, the low current efficiencies are mainly due to the poor contact
between the suspended indigo particles and the electrode surface.

Raney nickel has been widely used over the years for ECH reactions due to its easy
availability, low cost and high stability in a wide range of solvents and pH, and its
efficiency for the hydrogenation of carbonyl groups. Recently, however, a number of
noble metal catalysts such as Pt, Ru, Rh, etc. are being used for the ECH of specific
ftmctional groups. Since these catalysts are expensive, they can be more efficiently

utilized if they are dispersed in a support such as alumina or carbon. Amouzegar and

23

coworkers42 reported the ECH of phenol to cyclohexanol at 60°C using Pt/C electrodes
where a current efficieny of 85% was achieved with a Pt loading of 2%, higher than that
on platinized Pt, Pt on carbon rod or smooth Pt. The rate determining step was found to
be the reaction between adsorbed species on the electrode surface. In many studies,
polymers were used to support noble metal catalysts used for ECH, providing a high
surface area electrode with good chemical and mechanical resistence. Recently Lofrano
and coworkers43 reported ECH of isosafrole (olefin) and benzaldehyde in good yields
using Pt dispersed on poly (allyl ether p-benzeneammonium derivatives). The metal was
deposited using ion exchange involving the metal salt, followed by electrochemical
reduction to the metal. Pd deposited on poly-[N-(S-hydroxypentyl) pyrrole]44 and poly-
[N-(5-carboxyhexyl) pyrrole]45 film coated on carbon material was used for the ECH of
acetylene and other organic compounds such as cyclohexenone and methyl-a-
oxobenzeneacetate by Takano and coworkers. Other examples of ECH using noble metal
catalysts dispersed on high surface area supports will be discussed in subsequent
chapters.

Finally, several examples demonstrate the preparative and commercial applications of
ECH. Navarro and coworkers46 recently reported the selective ECH of several organic
compounds such as cyclohexene, cyclohexenone, benzaldehyde, acetophenone, styrene,
1,3-cyclohexadiene, citral, linalool and geraniol using H2O/Me0H, NH40Ac, or NH4C1
(0.2M) as the supporting electrolyte in an undivided cell using Ni deposited on Fe. They
were the first to report the use of a sacrificial Ni anode which enabled deposition of fresh
Ni on the cathode and avoided catalyst poisoning. A current density gradient was used to

reduce electrolysis time and obtain good current efficiency. ECH of vat dyes such as

24

indigo was reported by Roessler and coworkers47 as an alternative method to reduce vat
dyes using electrodes comprised of a thin grid coated with a layer of Ni in which fine
particles of Raney Ni are dispersed. ECH is convenient and environmentally beneficial in
this case because it reduces the use of chemicals by avoiding the use of a chemical
reducing agent such as sodium dithionite, which releases vast quantities of excess
dithionite and its oxidation products, sulfite, sulfate, thiosulfate and toxic sulfide into the
environment which are difficult to regenerate. Raney Ni was more active compared to Pt
and Pd electrode, but showed a lower current efficiency. Raju and coworkers48 reported

the synthesis of non-steroidal anti-inflammatory drugs such as ibuprofen using ECH.

1.6 Motivation for research

The goal of this project is to develop an alternative method for the electrocatalytic
hydrogenation of lactic acid and other a-functionalized carboxylic acids. The
electrocatalytic method is milder, more convenient and environmentally friendly
compared to the chemical catalytic or batch hydrogenation process currently in use. The
generation of hydrogen on the catalyst surface eliminates the need for an external source
of hydrogen gas, making the process much safer. The high pressure requirement which
arises from the poor solubility of hydrogen gas in water is also eliminated because the
hydrogen is generated where it is needed i.e. at the catalyst surface. Thus, the kinetic
barrier for the dissolution of hydrogen gas and splitting into H atoms is eliminated and
this enables the use of lower hydrogenation temperatures as well, compared to the batch
process. Milder and environmentally friendlier conditions are certainly desirable if a

process is to be used on a preparative scale in the laboratory or industry, especially for

25

the production of value added chemicals from easily available starting materials. Though
this thesis describes the development of an electrocatalytic hydrogenation process,
specifically for lactic acid and certain other a-functionalized carboxylic acids, our long
term goal is to extend the process to other carboxylic acids as well.

The strategies and techniques used for electrocatalytic hydrogenation are ones that are
reported in literature and have been used for the ECH of other compounds such as
phenols, olefins, aromatic compounds, enones, ketones, etc. To the best of our knowledge
however, no studies of ECH of carboxylic acids have been reported in literature.
Carboxylic acids being some of the raw materials most readily available from biomass, it
is important and interesting to explore their behavior under electrocatalytic conditions. A
successful and practical reduction would enable the effective and controllable
conversions of carboxylic acids into value added products. The work may also help to
uncover the reaction mechanisms in situ during catalytic hydrogenation. From the
previous work described, it is clear that though a few mechanistic paths have been probed
for the Ru/C catalyzed hydrogenation of lactic acid using indirect deuterium labeling
methods, there have been no studies to date of the nature of surface-bound intermediates
involved and the exact catalyst surface mechanism. Recent studies27 have shown that
water is required for the Ru/C catalyzed hydrogenation and the use of organic solvents
such as ethanol or THF either in the pure form or in aqueous mixtures significantly
reduces the conversion efficiency of lactic acid. This suggests a unique mechanism
involving the reaction of adsorbed water molecules on the catalyst surface, which
significantly differs from hydrogenation mechanisms in the vapor phase or in other

solvents. Electrocatalytic hydrogenation would make the process more amenable to in

26

situ spectroscopic studies. The reaction on the electrode surface can be better controlled
by applying a suitable potential in the hydrogen adsorption/evolution and the
electrochemical process can be easily coupled with surface spectroscopic techniques such
as Surface Enhanced Raman Spectroscopy (SERS) and Attenuated Total Reflection
Infrared (ATR-IR) spectroscopy to study the surface reaction. Part of this work involves
exploring the feasibility of such spectroelectrochemical studies of lactic acid

hydrogenation.

1.7 Outline of dissertation

This dissertation describes studies on the electrocatalytic hydrogenation of a-
functionalized carboxylic acids. The study is based on the process optimized for the
catalytic hydrogenation of aqueous lactic acid to propylene glycol using 5% Ru/C
catalyst, 150°C and 1200 psi H2 gas at a substrate concentration of 10-15g/100mL. It was
observed from earlier studies that the hydrogenation of carboxylic acids having certain
filnctional groups at the a position is much faster than the corresponding unsubstituted
carboxylic acids. Chapter 2 describes some initial studies of electrocatalytic
hydrogenation with a model a-functionalized carboxylic acid, benzoyl formic acid at low
concentration (10 ppm) using boron doped diamond (BDD) electrodes modified with Pt
and Pt/Ru. This compound was chosen in order to facilitate the detection and
quantification of the reactant using HPLC-UV at low concentrations. A potential in the
hydrogen evolution region for Pt was applied using chronoamperometry and a plot of

current vs. time was obtained. The effect of varying the conditions such as analyte

27

concentration, step potential, pH and temperature on the yields of the various
electrohydrogenation products was studied.

For the ECH to be useful at a preparative and commercial scale, the process needs to be
carried out at much higher concentrations of carboxylic acid. Chapter 3 describes the
ECH of lactic acid in mM concentrations in a bulk electrolysis cell using constant current
electrolysis and Ru deposited on carbon felt as a cathode. The effects of variation in
temperature, current, nature of electrolyte and lactic acid concentration are discussed
followed by the calculation of kinetic parameters. In this case, the only product of the
ECH of lactic acid is lactaldehyde and not propylene glycol. Since the conversion was
very low, other strategies were developed to improve the conversion of lactic acid.
Chapter 4 describes the ECH of lactic acid using 5% Ru/C powder catalyst entrapped in
the pores of Reticulated Vitreous Carbon (RVC), a macroporous material with a high
hydrogen overpotential. This method was developed by Menard and coworkers, and was
used to study the hydrogenation of phenol to cyclohexanol. This method allows the use of
the same powder catalysts that effect chemical hydrogenation and therefore, retains the
essential features of the catalytic process. The effect of variations in temperature, current
and nature of electrolyte on lactic acid conversion are reported. In this case, the major
ECH product is lactaldehyde, but small quantities of propylene glycol (<10%) are also
obtained. Also included is a discussion about studies on the adsorption of electrolyte
anions on electrode surfaces reported in the literature and the possible reasons for the
effects of electrolyte anions on the efficiency of lactic acid conversion. Chapter 5
describes the ECH of lactic acid using a similar strategy to that used in chapter 4, but

using non-carbon supports for Ru. In this case, lactaldehyde is the major ECH product

28

along with small quantities of acrylic acid formed in some cases, but no propylene glycol
formation was observed, indicating the importance of the precise structure of the catalyst
in determining the nature of products.

Chapter 6 describes in situ Attenuated Total Reflection-Infrared (ATR-IR) spectroscopic
studies of the ECH of lactic acid. The advantage of the ATR mode over other modes of
FTIR spectroscopy is that it is highly selective to species very close to the electrode
surface and minimizes interference from bulk species. The ECH of lactic acid in various
electrolytes was followed in real-time and the IR spectra were recorded at regular
intervals. We will see that the infrared spectra provide valuable information on the mode
of adsorption of lactic acid and effect of electrolyte anions, thus proving the feasibility of
using vibrational spectroscopic techniques to study the mechanistic details of the
hydrogenation reaction. Finally chapter 7 presents the conclusions of the study and some

directions for future research.

29

CHAPTER 2
ELECTROCATALYTIC HYDROGENATION OF A MODEL (1-
FUNCTIONALIZED CARBOXYLIC ACID, BENZOYL FORMIC ACID
2.] Introduction
This chapter deals with studies on the ECH of a model a-functionalized carboxylic acid,
benzoyl formic acid (BFA) using boron doped diamond (BDD) electrodes modified with
Pt and Pt/Ru. The goal of this study was to determine the feasibility of carrying out ECH
of carboxylic acids. It is important to carry out experiments at low concentration of
analyte and to look at the extent of reaction and distribution of products before moving
on to bulk electrolysis. Diamond as a material is known for its unique mechanical,
thermal, optical and electrical properties and is currently being investigated as an
electrode material by many groups.
Swain and coworkers”50 have developed boron doped diamond electrodes containing Pt
particles for use in various types of electrocatalysis. These newly developed conductive
sp3 carbon electrodes are superb materials for various electrochemical applications. They
are made by microwave assisted chemical vapor deposition (CVD) in a CH4/H2 gas
mixture in a reactor, and have unique properties like high conductivity, chemical
inertness, thermal conductivity, corrosion resistence and electrochemical stability. The
high atomic density and the strong sp3 bonding within the lattice inhibit morphological
and microstructural damage due to oxidation by anodic polarization and chemical
reagents as the oxidation is confined only to the surface. These properties enable
diamond to withstand current densities on the order of 1A cm'2 for days, in both acidic

and alkaline media without significant structural degradation that would result in

30

dislodging or instability of the deposited metal particles. Incorporation of metal adlayers
like Pt, Pb, Hg, etc. into the surface of the boron doped diamond support by
electrochemical deposition allows higher current densities and temperatures to be used
for electrocatalysis without any effect on the structural characteristics of the electrodes.
Pt can also be codeposited with Ru nanoparticles51 to form a bimetallic catalytic surface
for electrocatalysis.

These specially designed electrodes should be ideal for carrying out electrocatalytic
hydrogenation of carboxylic acids and their derivatives, especially a-functionalized
carboxylic acids. This would enable us to develop a method for the electrosynthesis of
the reduction products using milder conditions compared to the batch process. It also
opens up the possibility of using spectroscopic techniques coupled with electrochernistry
to investigate the mechanism of the reaction on the catalytic surface and the nature of the
adsorbed layer. This approach is much more feasible than a direct in situ spectroscopic
study of the heterogeneous catalytic surface under the high temperature and pressure
conditions of chemical hydrogenation, which could suffer from hurdles such as the
instability of the Internal Reflection Element (IRE) in the harsh conditions.

The Pt/BDD composite electrodes used here were prepared by galvanostatic
electrodeposition50 from a solution containing lmM K2PtC16 in 0.1M HC104 using a
current density of 0.1 mA/cm2 for 400 s. The Pt//Ru/BDD composite electrodes were
prepared by galvanostatic electrodeposition using a similar method, from a solution
containing 1 mM K2PtC16 and 1 mM RUCI3. in 0.1 M HC104. The metal coated film was
then placed back in the reactor and boron-doped diamond was deposited for an additional

3 h. During this step, the diamond film grows around the base of the particles, anchoring

31

them and stabilizing them within the diamond microstructure. The secondary diamond
film thickness is less than lum. After preparation, the electrodes were cleaned by cycling
the potential between -400 and 1500 mV (vs. Ag/AgCl) 100 times at 50 mV/s in order to
oxidize non-diamond carbon impurities such as adsorbed carbon and hydrocarbons
(formed during CVD) from the surface. The nominal Pt particle size is 30-500 nm and the
distribution density is 109 cm'z. The secondary diamond growth reduces the total active Pt
exposed by about 30-40%.

The main goal of this project is to explore the possibility of applying electrocatalytic
hydrogenation effectively to aqueous free carboxylic acids as an alternative to high
pressure hydrogenation and to determine the nature of adsorbed species on the catalyst
surface in the electrocatalytic hydrogenation of (it-functionalized carboxylic acids like
benzoyl formic acid and lactic acid. These two compounds are good model compounds
for aromatic and aliphatic alpha functionalized carboxylic acids respectively and may
show differences in behaviour due to different modes of adsorption on the catalyst
surface. Since the alpha carbonyl group is easier to reduce, it is first converted to the
alcohol which then facilitates the hydrogenation of the carboxylic acid group. In case of
BFA, the expected intermediate is mandelic acid, the direct analogue of lactic acid, so the
subsequent reduction is presumably similar except for extra features introduced by the

presence of the phenyl group.

An analog of pyruvic acid, BF A, was chosen as the analyte in order to facilitate the
analysis of the reactant to determine the conversion under various conditions. The
analysis of BF A was performed using HPLC-UV at 210 nm. as BFA has a much higher

extinction co-efficient for u.v. absorption than do pyruvic or lactic acids, which could not

32

' be detected by HPLC with u.v. or refractive index detection in the range of
concentrations used. Also there are four possible products that can be obtained by the
hydrogenation of BFA and it is interesting to study their distribution under various ECH
conditions. BF A has three reducible groups with the ease of reduction decreasing in the
order, -COOH > -C6H5 > -C=O. For the purpose of comparison, mandelic acid, the
aromatic analog of lactic acid, was also subjected to ECH under similar conditions and

the results were compared to those of the ECH of BFA.

2.2 Experimental

2.2.1 Electrochemical cell

The ECH of BFA was carried out in a three necked glass cell of volume about 20 mL as
shown in figure 2.1. The auxiliary electrode was graphite and the reference electrode was
Ag/AgCl (E0 vs. SCE= -47mV). The working electrode was connected to the external
circuit through a Cu current collecting plate. The exposed surface area of the working
electrode was determined by the o-ring between the cell and the electrode and was about
0.2 cm2. The applied potential was controlled using a potentiostat, model CY2000
(Cypress Systems). All potentials reported are in reference to Ag/AgCl reference
electrode.

The experiments with the Pt rotating disc electrodes were carried out in a three-necked
cell of total volume about 40 mL (figure 2.2) containing an anodic and a cathodic
compartment separated by a glass frit. The auxiliary and reference electrodes were

similar to the ones used in the other experiments. The working electrode was a Pt rotating

33

disc electrode (A=0.07cm2). The rotation speed was controlled by a rotation controller

(Metrohm 628-10).

 

nitrogen inlet
Ag/AgCl reference

electrode
carbon rod
counter electrode

 

 

 

O-ring

 

 

 

Figure 2.1: Schematic diagram of electrochemical cell (adapted from Granger, M.C.;
Witek, M.; Xu, J .; Wang, J.; Hupert, M.; Hanks, A.; Koppang, M.D.; Butler, J .E.;

Lucazeau, G.; Merrnoux, M.; Strojek, J .W.; Swain, G.M. Anal. Chem, 2000, 72, 3793-
3804)

34

Ag/AgCl
reference R
electrode

C

W rotating shaft

  
  
  

carbon auxiliary
electrode

 

”23.1

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

L_u

teflon holder
anodic chamber cathodic chamber
Pt polycrystalline
electrode

Figure 2.2: Schematic diagram of the electrochemical cell used with the rotating disc

electrode (RDE) (R-reference, W-working, C-counter)

2.2.2 Electrode preparation

The Pt polycrystalline electrode was polished successively with a slurry of 1.0 pm, 0.8
pm and 0.05 pm alumina in water before use. After each step of the polishing process, the
electrode was sonicated in ultrapure water for 10 min. to remove residual alumina from
the surface followed by rinsing with water. The electrode was also polished with a slurry
of the 0.05 pm alumina in water before every run.

The Pt/BDD and Pt/Ru/BDD electrodes were cleaned by immersing in distilled isopropyl

alcohol in the cell for 20 min. The electrode was then rinsed with water and dried in a

35

 

stream of nitrogen, followed by cycling the potential between -200 mV and 1200 mV in
0.1M HC104 until the features of a clean Pt surface were clearly visible. The electrode

was again rinsed with water and dried in nitrogen.

2.2.3 ECH method

The cell was filled with the solution of BFA of required concentration in 0.01 M H2804.
The solution was degassed with nitrogen gas for 10 min. to remove dissolved oxygen.
The potential was stepped from 400 mV to the required potential in the cathodic region
and the current-time curve was recorded for 2 h at constant potential. After every
experiment, the solution was removed from the cell and analysed by HPLC-UV and GC-
FID.

For the experiments with the Pt rotating disc electrode, the procedure was the same,
except that the anodic chamber was filled with 0.01M H2804, and the cathodic chamber

was filled with 6.67 mM BF A in 0.01 M H2804.

2.2.4 Analysis of products
The standard compounds, benzoyl formic acid (99%), mandelic acid (99%) and
hexahydromandelic acid (98%) were purchased from Aldrich. Cyclohexyl ethylene

glycol was synthesized using the procedure described below.

36

2.2.4.1 Benzoyl formic acid (BFA), mandelic acid (MA) and l-phenyl 1,2-ethanediol
(PED)

Reaction mixtures were analyzed by HPLC with uv detection at 1:210 nm, using a
Biorad HPX87H ion exchange column at 65°C, a Spectra-Physics pump model SP8810
and Hitachi uv detector model L-4000H. The mobile phase used was 5 mM H2804, run

isocratically at a flow rate of 0.6 mL/min.

Solutions of BFA, MA and PED in 5 mM H2804 in the concentration range 2-10 mg/L
were used to obtain calibration curves (figure 2.4). A standard HPLC chromatogram
containing benzoyl formic acid, mandelic acid and phenyl ethylene glycol, each at 10

mg/L is shown in figure 2.3.

2-5- * BFA MA

0.0 - ll PED
-2.5 .

Signal
(m V)

 

 

 

 

 

 

 

 

o 10 20 30 to
Time (min.)

Figure 2.3: HPLC-UV chromatogram standard showing peaks for BF A, MA and PED

(concentration=l .33 X 10'5 M)

37

The response factors were calculated using the following expression.

Molar concentration of analyte
Response factor =

 

Peak area counts

1

 

slope of the calibration curve

The concentrations of the unknown samples were calculated by multiplying the peak area

counts of the respective sample with the response factors.

 

30000

25000
Peak 20000
size

15000

10000

5000

 

 

 

 

 

1 2 3 4
Amount (x 10'5 M)

Figure 2.4a: Calibration curve for BFA (R2=O.9951)

38

 

30000 -

20000-

Peak
aze

10000 -

 

 

 

 

 

:1-
J

l l
'0 ‘fi

1 2 3 4 5
Amount (x 105 M)

Figure 2.4b: Calibration curve for MA (R2=0.9986)

2.2.4.2 Synthesis of cyclohexyl ethylene glycol (CEG)

CEG used as a standard for analysis was synthesized by the 5% Ru/C catalyzed
hydrogenation of phenyl ethylene glycol in a Parr reactor (scheme 2.1). The 5% Ru/C
catalyst was reduced overnight in the reactor in the presence of H2 (300 psi) at 150°C.
100 mL of an aqueous solution containing 1.0 wt% of phenyl ethylene glycol was
charged into the reactor and reduced at 150°C and 1200 psi H2 for 7 h with constant
stirring at 1100 rpm. After the reaction, the reactor was cooled to room temperature and
the excess H2 was vented out of the reactor. The reaction mixture was transferred to a

250mL beaker and allowed to stand overnight. The product precipitated due to poor

39

solubility in water and was collected by filtration and dried in air (for characterization,

see appendix).

OH

OH
OH
: A .OH H2. RulC O/LV

150°C, 1200 psi
phenyl ethylene glycol 7h cyclohexyl ethylene glycol

 

Scheme 2.1: Batch synthesis of CEG

2.2.4.3 Hexahydromandelic acid (HMA) and cyclohexyl ethylene glycol (CEG)

HMA and CEG were detected and quantified using derivatization, followed by GC-FID
since these compounds do not absorb in the u.v. region.

For the derivatization, 2 mL of the sample was treated with 0.2 mL of 0.1 M Ba(NO3)2 to
precipitate the sulfate ions from the electrolyte. The solution was then evaporated to
dryness in a vacuum centrifuge. 0.5 mL of his trimethylsilyl trifluoro acetamide (BSTFA)
containing 1% trimethyl chloro silane (TMCS) (Supelco) was added to the sample, mixed
thoroughly, the mixture is transferred to a vial and allowed to stand for 1h at room
temperature.

The analysis was performed using a Perkin Elmer model 8500 gas chromatograph
equipped with a flame ionization detector, using a SPB-35 column (30m, 25m ID).

2 11L of the mixture was injected into the GC with the injector temperature at 300°C. The
column temperature was held at 50°C for 5 min., ramped to 250°C at 15°C/min. and held
at 250°C for 15 min. The detector was maintained at a temperature of 330°C and the total

runtime of each sample was 33.3 min.

40

Solutions of CEG and HMA in 5 mM H2804 in the concentration range 2-10 mg/L were
used to obtain calibration curves. A standard GC chromatogram containing the trimethyl
silyl (TMS) derivatives of CEG and HMA (10 mg/L) is shown in figure 2.5a. The blank
chromatogram is shown in figure 2.5b. The peaks at 13.5 and 14.0 min. are present in the
blank as well and are therefore assigned to species present in the derivatizing reagent.
The presence of two peaks for each derivative is probably due to the incomplete
derivatization, leading to the formation of both the monosilylated and disilylated
derivatives. For the quantification, only the disilylated derivatives were considered. The
TMCS peak (ret. time 8.59 min., not shown) was used as the reference peak. The
calibration curves for CEG and HMA are shown in figures 2.6. The concentration of
HMA and CEG in the unknown samples were determined by dividing the peak area ratio

by the response factors.

41

 

 

 

 

 

 

 

 

I
12.5 13.0 13.5 14.0 14.5

Time (minutes)

Figure 2.5a: GC-FID chromatogram standard showing peaks for the TMS derivatives of

CEG and HMA (concentration= 2.0 mg/L)

42

 

 

 

 

 

 

 

4 J
3 _
Signal
(mV)
2 a
1 _
0 l l I 1
12.5 13.0 13.5 14.0 14.5

Time (minutes)

Figure 2.5b: GC-F ID chromatogram standard showing peaks from the blank reagent

(BSTFA + TMCS) run under similar conditions

43

Peak area ratio

Peak area ratio

" " y = 0.1914x

 

l
l

R: = 0.8503

 

 

 

'9

0.03 -

1.5 2 2.5 3 3.5 4
Cone (X10'° M)

“y": 0.0072x'

 

0.02 .. —

0.01

 

R2=1

 

 

Cone. (x 10" 1111)

Figure 2.6: Calibration curves for HMA (top) and CEG (bottom)

44

2.3 Results and Discussion
2.3.1 Cyclic voltammetry of benzoyl formic acid
Shown in figure 2.7 is the cyclic voltammogram of 6.67 X 10'5 M (10 ppm) BFA in 0.01
M H2804 on Pt polycrystalline working electrode at a scan rate of 10 mV/s. The figure
shows scans 1, 2, 5 and 10 out of 10 scans in the range -600 to 1200 mV. After the first
scan, there is a significant increase in the anodic oxidation current due to oxidation of
other carbonaceous impurities, forming a cleaner surface and providing more sites for the
adsorption of BFA. This indicates that BF A is not oxidized easily even at higher
potential. The CV indicates an irreversible and strong adsorption of BFA on the Pt
surface, involving partial charge transfer from the n-orbital in the aromatic ring to the Pt
d-orbital. This explains the appearance of a broad oxidation peak and the absence of
sharp and distinct oxidation and reduction peaks.
Cyclic voltammograms were also obtained in 0.01 M H2804 on Pt polycrystalline
electrode by scanning the potential in the range -200 to 400 mV at 50 mV/s in the
presence and absence of BFA (70 11M). The results are shown in figure 2.8. There are no
features due to BFA in the potential range used, indicating that BFA does not undergo
oxidation or reduction by direct electron transfer in this potential range. However, the
peak areas of the hydrogen adsorption and desorption peaks at 200 mV are lower in the
presence of BF A, indicating that part of the surface is covered by BFA which decreases

the surface area available for hydrogen adsorption.

45

 

i (uA)

 

 

 

 

Figure 2.7: Cyclic voltammogram of 6.67 X 10'5 M (10 ppm) BFA in 0.01M H2804 on a

Pt polycrystalline working electrode at a scan rate of 10 mV/s. Arrows indicate direction

of scan.

46

 

 

 

M..—
am
off

.—

 

 

 

 

-0.20 0.00 0.20 0.40
EN (AglAgCl)
Figure 2.8: Cyclic voltammograms of Pt in 0.1 M H2804 in the presence (top) and
absence (bottom) of BFA (70 11M) at a scan rate of 50 mV/s (data obtained by Prof. Carol

Korzeniewski at Department of Chemistry, Texas Tech University)

2.3.2 Determination of active metal surface area

The active metal surface areas of the two working electrodes, namely Pt/BDD and
Pt/Ru/BDD were determined from the hydrogen desorption charge, obtained from the
area under the hydrogen desorption peaks at 200 mV. The hydrogen desorption charges
were calculated by integration to determine the area under the peaks followed by
subtraction of the area under the baseline due to double layer charging. The calculated
areas are reported in table 2.1. Figure 2.9 shows the cyclic voltammograms of the Pt
polycrystalline, Pt/BDD and Pt/Ru/BDD in 0.1 M HC104 in the potential range, -200 to
1200 mV, scanned at 100 mV/s, in comparison with Pt polycrystalline electrode. It can be
seen that both the composite electrodes show features similar to Pt, but have active

surface areas much lower than Pt polycrystalline electrode.

0.00012
0.0001 ~

 

l (A)

15 2, __W_-.__ .
,—Ptpolycrystallinet -

; - - - -Pt/diamond : l

'°'°°°°°i ------ Pt/Ru/diamond ‘ *
00001 - , ___-m__,__ -. *

E (V)

 

 

 

 

 

Figure 2.9: Cyclic voltammograms of Pt polycrystalline, Pt/BDD, and Pt/Ru/BDD in 0.1

M HC104 at 100 mV/s.

48

 

Electrode Actual Pt surface areas (cmz)

 

 

 

Pt polycrystalline 0.77
Pt/BDD 0.13
Pt/Ru/BDD 0.21

 

 

 

 

Table 2.1: Active Pt surface areas for the Pt electrode and composite electrodes

calculated from the area under the hydrogen desorption peaks in figure 2.9

2.3.3 Current-time transients

The i-t curve for one of the experiments on Pt/Ru/BDD is shown in figure 2.10 (Initial
potential (E1)= 0.4 V, Final potential (E2): 04 V). The i-t curve for the run without the
substrate and the theoretical curve as predicted by the Cottrell equation are also plotted.
The current response in chronoamperometry is proportional to the flux or the rate of
diffirsion of the electroactive material (H30+ in this case) to the electrode surface. As
time progresses, there is an increase in the length of the difiusion layer leading to a
decrease in the concentration gradient of the electroactive species between the electrode
surface and the bulk solution. This leads to a decrease in the flux of the electroactive
species to the electrode surface, and a decrease in current. In figure 2.10, it can be seen
that the theoretical curve shows a much sharper decay in current compared to the
experimental curve, thus indicating that the reaction is not completely diffusion
controlled, but may be controlled by convection or the availability of surface sites. The
total charge passed (obtained by integrating the area under the curve) is slightly lower in

the presence of the substrate, indicating that some of the surface sites may be occupied by

49

the substrate molecules even at the more negative hydrogen evolution region, thus

reducing the total surface area for hydrogen adsorption.

 

 

 

 

 

O O O O O O O l
O O O O O O O '
| o o o o o o o ;

O ‘- N ('0 V In (D N

o -2

I -50 i .
I 2 "0°“ 2 2 A . _______,f
j a -150 ~ I'M/4:" ---0.01M H2804 ,1
E '200 “ W. -~.-;:::::::::-::"”’ 1 .l
i g ’250 l ' , - -- - benzoylformic :1
' o -300 . aCIdIn 0.01M :‘
H2804 ll
. '350 A Cottrell curve 1.
I' 400 l l,‘

' Time (s)

Figure 2.10: Current-time response of an ECH experiment with Pt/Ru/BDD in 0.01 M

H2804 (Er—400 mV, E2=-400 mV)

2.3.4 ECH of BFA on Pt polycrystalline electrode

The reaction for the ECH of BFA and the various possible products are shown in scheme
2.2. BFA had three different functional groups with different ease of reduction. MA is
formed by the reduction of the a-keto group. HMA is formed by the reduction of the a-
keto group and the aromatic ring. PED is formed by the reduction of the a-keto group and
carboxylic acid group, and CEG is formed by complete reduction of all the three
functional groups. Therefore the optimum conditions for ECH were considered as the
ones giving maximum yield of CEG. In all the experiments, there was no PED detected

in any of the mixtures at the end of the ECH experiment and HMA was formed in

50

negligible amounts. However, though the total conversion of BFA was low (<25%), the
yields of MA and CEG vary with conditions.

OH

OH
OH OH
0 O o O 0”
OH Eledmhydmemam; Mandelic Acid (M A) 1-Phenyl—1,2-Ethanediol (PED)

O o 0” OH

OH OH
Benzoyl Fonnic Acid (BFA)

O OH

Hexahydromandelic Acid Cyclohexylethyleneglycol
(HMA) (CEG)

 

Scheme 2.2: ECH of benzoyl formic acid (possible products)

The next few sections describe the effect of ECH conditions and nature of electrode
material on the yields of MA and CEG. Several parameters such as analyte concentration,
step potential, pH and temperature were varied and the final concentration of the
products, determined by HPLC-UV and GOP ID, were plotted as a function of the
parameter.

The conditions used for the ECH experiments are summarized below.

- Method: Chronoamperometry

- Analyte: Benzoyl formic acid (6.67 X 10'5 M)

0 Cell volume: 20 ml.

0 Working electrode: Polycrystalline Pt (0.2 cmz), Pt/BDD ot Pt/Ru/BDD
° Reference electrode: Ag/AgCl (E0: -47 mV vs. SCE)

' Auxilliary electrode: Graphite

51

0 Electrolyte: 0.01 M H2804

0 Time: 2 h

2.3.5 Determination of optimum analyte concentration

Chronoamperometry was carried out using different concentrations of BFA in the range 1
to 100 ppm in 0.01 M H2804 at E2= -425 mV. The final concentrations of the products
and the percentage yields after 2 h are shown in figure 2.11.

From the results, we see that the yield of MA is maximum at an initial BFA concentration
of 6.67 M or 1 ppm and the yield of the desired product, CEG is maximum at 10 ppm or
6.67 X 10'5 M (10 ppm) analyte concentration. No significant quantities of HMA and
PED were detected in the product mixtures. It should be noted that while the final
concentration of CEG was maximum at a starting BFA concentration of 25 ppm, the
percentage yield at this concentration was lower than at 10 ppm. Beyond this limit, there
is a decrease in both turnover and yield. The decrease in the yields of products at higher
substrate concentration is probably due to the saturation of the electrode surface with
substrate molecules which compete with H atoms for surface sites. All further
experiments were carried out using 6.67 X 10'5 M BF A. In the case of Pt/BDD and
Pt/Ru/BDD, the active metal surface areas are much lower compared to polycrystalline Pt
and so we would expect the surface to be saturated at much lower concentrations than

that in polycrystalline Pt.

52

25

 

~—O:cchhexyl ethylene—l '
glycol ‘ j

l

‘ + mandelic acid ~ 1

 

 

 

 

0 50 100 150
( Analyte concentration (ppm)

Lvi, ,,A_,, 2 2 ,2,., 2,2,2 72,3

 

 

 

 

 

( i
- '9
' 3
51.
C
8 2 2- ,___.__-___. ,
g + cyclohexyl ethylene “
§ . glycol "
g + mandelic acid J
. 00.5 ‘
. 7.- _ L --__ _ _ .
. 5 l
i ll. 1
I o 50 100 150 .
Analyte concentration (ppm)

Figure 2.11: Conversion of BFA as a function of initial analyte concentration during
ECH on polycrystalline Pt electrode (electrolyte: 0.01 M H2804, temperature= 25°C, Et=
400 mV, E2= -425 mV, time=2 h). Error bars indicate uncertainty in the analysis of

Products (i10%) using GC-FID

53

2.3.6 Effect of step potential on ECH

Electrohydrogenation of benzoyl formic acid was carried out at various step potentials
ranging from —200 mV to —600 mV in 0.01 M H2804 for 2 h. The results are shown in
figure 2.12. Moving towards more negative potentials, the yield of products increases and
then decreases due to competition from the hydrogen evolution reaction at more negative
potentials, which leads to lesser availability of adsorbed H atoms for hydrogenation of
BFA. Maximum yield of the desired product (CEG) was obtained at —425 mV, at which
point there is an optimum balance between surface hydrogen concentration and rate of
hydrogen evolution. Conversions are indicated in terms of final concentrations of

products.

 

 

 

 

l
l 5* ~- -2 ,5-
In
1 a w .5
F
1 >5 -. ,4
l 1:
3 3
| H—
_ 5
c .22- -2 2
l 3 2 2 +0133}
= ;
I 8 A 1 +m l
'1 g +BFA
. E _ '0 _7_>____
-0.8 1t?

 

Step potential (E,)(V)

Figure 2.12: Conversion of BFA as a function of final step potential (E2) during ECH on
polycrystalline Pt electrode (electrolyte: 0.01M st0., initial [BFA]= 6.67 x 10'5 M,

temperature= 25°C, time = 2 h)

54

2.3.7 ECH of BFA using boron-doped diamond (BDD) composite electrodes

2.3.7.1 Effect of step potential

Study of electrohydrogenation of BFA at various step potentials ranging from —200 mV
to —600 mV indicates that the trends in the ECH turnovers and yields as a function of
potential are similar to that using a polycrystalline Pt electrode. The results are shown in
figure 2.13 for Pt/BDD and Pt/Ru/BDD. In the case of Pt/BDD, the yields are decreased
by a factor of 3 relative to polycrystalline Pt, which is roughly proportional to the
decrease in active Pt surface area. This suggests that Pt/BDD has insufficient surface sites
available for the adsorption of analyte, since Pt is known to favor hydrogen adsorption
and evolution. However, in the case of Pt/Ru/BDD, there is a significant increase in the
yield of the hydrogenated product at the optimum potential. This indicates that the Ru
electrocatalyst plays a role in enhancing the rate of hydrogenation of BF A in comparison
to hydrogen evolution.

The maximum yield of CEG is obtained at -425 mV and -3 75 mV for Pt/BDD and
Pt/Ru/BDD respectively. Thus, in all cases, the optimum cathode potential for ECH
appears to be between -3 75 and -425 mV. The maximum yield at the optimum potentials
for Pt/BDD and Pt/Ru/BDD are 6.0% and 25.3% respectively. The yield of CEG on
Pt/Ru/BDD is slightly higher than that obtained on the Pt polycrystalline electrode, even
though the active metal surface area on the Pt/Ru/BDD electrode was much lower than
that on the Pt polycrystalline electrode. Yield of mandelic acid is much lower at this point

compared to that on polycrystalline Pt.

55

 

-2 - - _. m: - - 4,3
— 1.6
1.4
5182* t" lama/Taco)“
1~ l +MA(PUBIID) ii

”0.3 l +cactn/Rwaoo) it
as ‘. """" A TEA L' U.'u__3qo)__-l
0.4

0.2

Final concentration (X 10E-5 M)

 

 

 

 

cp

 

l -0.8 -0.6 .04 -o.2 o 1
l ' Step potential (E,) (V) l
1

Figure 2.13: Conversion of BFA as a function of final step potential (E2) during ECH on

Pt and Pt/Ru composite electrodes (electrolyte: 0.01 M H2804, initial [BF A]= 6.67 X 10'5

M, temperature= 25°C, time = 2 h)

The possible reason for the enhanced hydrogenation activity of the Ru electrocatalyst
may be explained by the fact that Ru is less active than Pt with respect to hydrogen
adsorption and evolution52 due to differences in the electronic properties of the two
metals. Therefore, Ru provides sites for adsorption of the analyte adjacent to the adsorbed
H atoms on Pt, thus increasing the rate of hydrogenation, since the rate of hydrogenation
is dependent on the surface concentration of both the organic acid and hydrogen. This is

depicted in figure 2.14.

56

Pt/diamond
H H H H S H H H H H H

.4.-. .. .MH..-..._.....- .,.._...._...... .
.
.

. . . ‘.._.. . ‘1'...“ _.. _._-. .... ”a. .. —.-'..-. ...~.. . - -‘fi

Pt/Ru/diamond
H H s H H H S H H s H

Pt- - Ru- - H- hydrogen 8- substrate

Figure 2.14: Comparison between adsorption patterns on Pt/BDD and Pt/Ru/BDD

electrodes

2.3.7.2 Effect of pH

ECH of BFA was studied at pH values in the range of 2-12 at —400 mV. The pH of the
solution was varied using calculated amounts NaOH and KH2PO4. The results are shown
in figure 2.15. A study of product concentrations as a function of pH indicates that in case
of both Pt/BDD and Pt/Ru/BDD, the yield of CED decreases steadily as the pH increases,
indicating difficulty in hydrogenating the carboxylate group formed in basic medium.
The pKa of BFA is ~3, therefore at pH>3, the -COOH group is partially dissociated. The
optimum pH on both the electrodes was 2. The yield of CEG decreases sharply at higher
pH due to the formation of the carboxylate anion, which has a higher affinity for the polar
solvent rather than for the electrode surface. Also carboxylates resist hydrogenation
because the pKa values of the product alcohols are about 10 units higher than those of the

carboxylic acids, which corresponds to a difference in AG of about 15 kcal/mol. The

57

reason for the huge increase in the formation of mandelic acid at pH 7 is not yet

 

 

 

 

 

understood.
8 l
7...- 2 l
l g t
l 6..- _
i g
l as 5*" *
: s
l E 3* _._'oasi“puaoo;
1 § 2“ +MA(Pt/Boo) =
l 2 1-222 2 +CEG(PVRUIBCD):‘
‘ E 0» “315059091;
_1tp 5 1J5

 

pH

L,_L,-__,,_,,- -, 2.

Figure 2.15: Conversion of BFA as a function of pH during ECH on Pt and Pt/Ru
composite electrodes (electrolyte: 0.01 M H2804, E2= -400 mV, initial [BFA]= 6.67 X

10'5 M, temperature= 25°C, time = 2 h)

2.3.7.3 Effect of temperature

ECH of BFA was carried out at various temperatures from 25 to 90°C at -400 mV. The
results show completely different behavior between the two electrodes, as shown in
figure 2.16. In case of Pt/BDD, the yield of CEG decreases slightly at 50 and 70°C
probably due to the higher rate of desorption of reactants from the electrode surface
decreasing the efficiency of ECH. There is a slight increase in the yields at 90°C, which

could be due to the increase in the rate of ECH overcoming the effect of faster desorption

58

of reactants. In case of Pt/Ru/BDD, even though the product yields are much higher than
on Pt/BDD at room temperature, the yields sharply decrease at higher temperature. This
may be attributed to either faster desorption of reactants or the leaching of Ru from the

electrode surface at higher temperature.

0.9
0.8 a. - ,
0.7 <4 2

0.6 .2 ,

0.5.
0.4. - a
0.3 -- 2 ,

 

+cactl=uiaoof '
0.2 ,_ +MA(F1IBCD)

0.1 7 2 g 7 _f 7 +C$(PUR1IBUJ)

0 g _ _ _ XQ,__* -——--x-—MA(P1IRu/BCD) ‘
-0.1 0————- .———40—————60———.—~80———~—100

Tern pe rature (C)

 

 

Final concentration (X 10E-5 M)

 

Figure 2.16: Conversion of BF A as a function of temperature during ECH on Pt and
Pt/Ru composite electrodes (electrolyte: 0.01 M H2804, initial [BF A]= 6.67 X 10'5 M,

E2= 400 mV, time = 2 h)

2.3.8 ECH of MA

For the purpose of comparison, the ECH was carried out using MA (6.58 X 10'5 M) as the
analyte. Chronoamperometry was carried out at various step potentials. The results are
shown in figure 2.17 for polycrystalline Pt, Pt/BDD and Pt/Ru/BDD. In this case, unlike

BFA, there is an increase in the yield of CEG at more negative potentials with the highest

59

yield at -600 mV. Pt/Ru/BDD was much more active compared to the others due to the

enhancement of hydrogenation rate by Ru.

MA is expected to be the intermediate in the hydrogenation of BFA to HMA or CEG,

since the hydrogenation of the keto group is much more facile than that of the other two

fimctional groups. If this is the case, then one would expect the hydrogenation of MA to

follow the same trend with applied potential as BF A i.e. exhibit maximtun conversion at

an intermediate potential, followed by a decrease in conversion at more negative

potentials.

 

 

 

 

 

.. 42
s .
3
e "1”
X
c 043
3
t 0:6
‘ o t~ 0.4
‘ C
. 8
C
E - v 0
-0.8 -0.6 -0.4 -0.2

Step potential (E2) (V)

 

«JP—t polycrystalline ‘
—-— Pt/BDD

Figure 2.17: Conversion of MA as a function of step potential (E2) during ECH on

polycrystalline Pt and composite electrodes (electrolyte: 0.01 M H2804, initial [MA]=

6.58 X 10'5 M, temperature = 25°C, time = 2 h). Graph shows yields of CEG.

6O

The difference in the behavior of BFA and MA with increasing step potential indicates
that there is a difference in the efficiency of adsorption of the two acids on the electrode
surface. BFA is much more strongly adsorbed at all potentials due to its planar structure
and extended conjugation and the hydrogenation to CEG occurs very fast before the
intermediate MA desorbs from the surface. This is supported by the low yields of MA as
compared to CEG in all of the experiments with BFA. Therefore the hydrogenation of
BF A is only limited by the concentration of adsorbed H adjacent to adsorbed BFA. On
the other hand, hydrogenation of MA is probably limited by the concentrations of both

adsorbed MA and H.

2.3.9 Effect of mass transport

In order to study the effect of mass transport on the ECH of BF A (6.67 X 10'5 M),
experiments were carried using a Pt rotating disc electrode at various rotation speeds. In
each experiment, the potential was stepped to -400 mV and the ECH was carried out for 2
h in 0.01 M H2804. The results are shown in figure 2.18.

At all rotation rates, the yields of the products are similar and within the range of
experimental error. No MA was detected in any of the mixtures. Increasing the stirring
rate did not enhance the rate of ECH. From these results, it is clear that mass transport

does not significantly influence the ECH reaction.

61

 

 

0.6 4*“ , '2 ~ ~ 2 . 7 , - 1

 

 

 

0 a 2 t
0 500 1 000 1 500 2000

Stirring rate (rpm)

Final concentration (X1OE-5M)
O .
.5
l
l l
l m

Figure 2.18: Conversion of BFA as a function of rotation rate during ECH on a
polycrystalline Pt rotating disc electrode (electrolyte: 0.01 M H2804, initial [BFA]= 6.67

X 10'5 M, E2=-400 mV, temperature = 25°C, time = 2 h). Graph shows yields of CEG.

2.4 Conclusions

In this study, a model tit-functionalized carboxylic acid, namely benzoyl formic acid, an
analog of pyruvic acid, was subjected to ECH under various conditions. This study was
carried out to achieve proof of principal of electrocatalytic hydrogenation of carboxylic
acids and also to determine the effect of various conditions on the ECH of a compound
having multiple reducible groups. The technique used was chronoamperometry, in which
the potential was stepped from a value at which there is no hydrogen generation on Pt to
various potentials in the hydrogen evolution region, and the distribution of ECH products
was determined under various conditions of step potential, pH and temperature.

Maximum yield of the desired product, CEG was obtained at a potential close to -400 mV

62

on both Pt/BDD and Pt/Ru/BDD electrodes. However, the Ru containing electrode
showed a significant enhancement in the final yield of CEG, probably due to the Ru
atoms providing surface sites for the adsorption of BFA. The ECH was favored at more
acidic pH, at which the -COOH group is undissociated, and also by higher temperature.
However, an anomalous temperature effect on ECH was obtained on the Pt/Ru/BDD
probably due to the leaching of Ru from the diamond surface at higher temperature. ECH
of BFA using a Pt rotating disc electrode at various speeds of rotation showed that the
yields of the products did not vary significantly with the rotation rate. These results
indicate that the rate of the ECH reaction is influenced more by the availability of surface
sites for the adsorption and reaction, and the surface concentration of reactants rather than
the rates of mass transport, consistent with the findings in the batch hydrogenation of

lactic acid.

63

CHAPTER 3
ELECTROCATALYTIC HYDROGENATION OF LACTIC ACID USING
RUTHENIUM ON CARBON FELT ELECTRODE
3.1 Introduction
For the electrocatalytic hydrogenation to be useful on a preparative and commercial
scale, the process needs to be optimized at much higher substrate concentrations,
compared to those used for ECH of BF A, and different cell and electrode designs are
required for this purpose. Several approaches have been used in the past for bulk scale
electrocatalytic hydrogenation of organic substrates, some of which are described in
chapter 1. Raney Ni in various forms has been used for the hydrogenation of carbonyl
compounds, but when expensive noble metals such as Ru and Pt are required, it would
not be cost effective to use pure metal electrodes. The best approach in this case would be
to deposit the metal in a highly dispersed form on a high surface area support using
electrodeposition from a salt solution of the metal. The deposition can also be done by
physical vapor deposition or sputtering, but electrodeposition is known to result in
particles of more uniform size and distribution. In many cases, noble metals deposited on
high surface area materials have resulted in better electrochemical efficiencies for ECH,
when compared to smooth metal electrodes. The reasons for this may be the higher
surface area as well as the unique properties of highly dispersed metal micro or
nanoparticles.

Polcaro and coworkers”53

studied the ECH of benzaldehyde and acetophenone in acidic
and alkaline ethanol/water mixtures using Pd supported on carbon felt, a support chosen

for its high specific surface area and good fluid permeability. In the case of

64

benzaldehyde, two parallel reactions were found to occur, leading to two different
products, benzyl alcohol and toluene, These products were believed to form from the
adsorption of benzaldehyde at two different types of sites and therefore, the product
distribution was greatly influenced by the method of electrode preparation and the
resulting morphology and distribution of active sites. The acidity of the supporting
electrolyte also influenced the product distribution with acetophenone and benzaldehyde
forming phenylethanol and benzyl alcohol respectively as the only products in alkaline
medium, whereas ethyl benzene and toluene were also obtained in acidic medium.
Comparison between two different electrocatalysts, Pt and Pd, deposited on carbon felt
under identical conditions revealed that the same hydrocarbon/alcohol ratio (15%) in HCl
medium was obtained in both the cases, but the product ratio was different when the
electrodeposition time was increased. The current efficiency, influenced by the relative
rates of hydrogenation and hydrogen evolution, also varied with the nature of the metal
and the supporting electrolyte.

Iwakura and coworkerss‘l’55 studied the ECH of 4-methyl styrene to 4-ethyl toluene using
a Pd sheet electrode modified by Pd black, Pt and Au. The deposition was performed
chemically by using active hydrogen passing through the Pd sheet electrode as the
reducing agent. A current efficiency of 100% was obtained using Pd deposited on Pd
sheet, which was 40 times higher than that obtained using bare Pd sheet electrode,
indicating that hydrogenation occurs on the three dimensionally expanded surface of the
Pd black deposits. Similar enhancements in efficiency were obtained using Pt and Au

deposits as well, with the current efficiencies being 95% and 70% respectively, though

65

these deposits showed lower rates of permeation of active hydrogen with increases in
metal loading.

Savadogo and coworkers56 studied the ECH of phenol to cyclohexanol using Pt deposited
on Vulcan carbon. The current efficiency and product selectivity were found to depend
on the electrode material, but not on the electrochemically active surface area or metal
particle size. Pt/C showed the highest electrocatalytic activity, whereas Ru/C showed the
highest selectivity. A Pt-Co/C based alloy electrode showed the highest activity and
selectivity for cyclohexanol production among the alloy electrodes studied. Further, the
current efficiency did not change significantly for Pt loadings of 2-30%, but there was a
decrease in activity when 60% Pt was used.

In an earlier study, Savadago and coworkers57 reported ECH of phenol with current
efficiency as high as 85% using only 2% Pt deposited on graphite particles. Again, this
high efficiency was assumed to be due to the high surface area of dispersed metal
particles. The ECH was inhibited in the presence of quaternary ammonium salts due to
the blocking of active sites by the quaternary ammonium ions. The rate of ECH increased
with increase in temperature and was influenced by the Teflon (used as a binder) content
of the electrode, which was believed to affect the phenol concentration on the electrode
surface. Baturova and coworkers58 studied the ECH of acetophenone over Pd deposited
on carbon fiber in 1 N NaOH at 20°C. Phenyl carbinol was the only product obtained
with a yield of 50% and a current efficiency of 50-80%. In this case, only Pd deposited at
a potential close to the equilibrium potential for Pd reduction was electrocatalytically
active because this deposition occurs through the specific adsorption of Pd. Pd deposited

at more negative potential showed no activity.

66

In the present study, carbon felt was chosen as the support for deposition of Ru, due to its
high specific surface area, good fluid permeability, more rapid dynamic adsorption
characteristics (mass transfer rates) and lower resistance to the flow of liquids and gases

compared to other carbon supports.

3.2 Experimental

3.2.1 Set up of electrochemical cell and ECH method

Electrohydrogenation studies were carried out on 75 mL volumes of several electrolytes
containing 11.1 mM lactic acid (Aldrich). The glass electrochemical cell employed
(figure 3.1) is a two compartment cell with the anode and the cathode compartments
separated by a glass frit in order to prevent the diffusion of gaseous products. The
reference electrode was Ag/AgCl (E0: 035 V vs. SCE). The counter electrode was Pt
wire. The Carbon felt (CF) working electrode (dimensions 4.0 X 3.0 X 0.25 cm) was
connected to the external circuit via a copper wire, positioned to minimize surface area
exposed to solution; electrochernistry on the copper surface could thus be assumed to be
negligible. A voltmeter was connected between the reference and working electrode to
measure the potential of the working electrode. Controlled current was applied using a
model 273 potentiostat/galvanostat from Princeton Applied Research. The temperature of
the system was controlled using a heating tape wrapped around the cell and connected to
a temperature controller from Omega. The solution was stirred using a magnetic stirrer.
Samples were withdrawn for analysis immediately after the cell was switched on and at

intervals during the electrohydrogenation.

67

The electrolytes were prepared by the dilution of the concentrated acids, namely 98%
sulfuric acid (CCI), 36.5% hydrochloric acid (CCI), and 70% perchloric acid (Spectrum)
with HPLC grade water. S-Lactic acid and propylene glycol were purchased from

Aldrich and were used without further purification.

Ag/AgCl reference

copper wire
electrode

  
 

cathodic compartment

 

anodic compartment

\ Pt wire auxiliary
electrode

\ cell body

.: .- .1 .34;
-: ,
122' I 2. i111- :
. _ Lift-fl; ; 4'
32-; .
0
glass frlt

Figure 3.1: Schematic diagram of the electrochemical cell used for ECH (W—Working,

 

working electrode

 

 

 

 

 

 

stir bar

A-Auxiliary, R-Reference)

3.2.2 Electrode preparation

The Ru/carbon felt electrode (4.0 X 3.0 X 0.318 cm) was prepared by electrodeposition
of Ru from 5 mM RuCl3.xH20. A constant current of 10 mA for 90 min was used for the
electrodeposition, and the potential was maintained between 0 and -0.2 V to ensure

complete reduction of Ru3+ to the metallic state. After deposition the electrode was

68

thoroughly rinsed with ultrapure water and then with the electrolyte to ensure complete

removal of traces of RuCl3.

3.2.3 Characterization of the electrode surface

The electrode surface was characterized using a J SM-6400 V scanning electron
microscope (J EOL, Ltd., Tokyo, Japan). Micrographs were recorded using secondary
electron mode generated with an accelerating voltage of 20 kV. The presence of Ru was
verified by energy dispersive X-ray (EDX) analysis (Noran Instruments, Inc.,
Middletown, WI) connected to the J SM-64OO V. SEM images were obtained using the

AnalySIS software (Software Imaging System Corp., Lakewood, CO).

3.2.4 Analysis of the products of ECH

(a) Lactic acid and propylene glycol

Reaction mixtures were analyzed by HPLC with refractive index detection, using a
Biorad HPX87H ion exchange column at 65°C. The mobile phase used was 5 mM
H2804, run isocratically at a flow rate of 0.6 mL/min.

Solutions of lactic acid and propylene glycol in 5 mM H2804 in the concentration range
0.2-1.0 g/100 mL were used to obtain a calibration curve. A standard HPLC
chromatogram containing lactic acid and propylene glycol (0.1 g/mL) is shown in Figure
3.2. The calibration curves for lactic acid and propylene glycol are shown in Figure 3.3
and 3.4. The response factor was calculated using the following expression.

Concentration of analyte (wt%)
Response factor =

 

Peak area counts

69

1

 

slope of the calibration curve

The concentrations of the unknown samples were calculated by multiplying the peak area

counts of the respective sample with the response factors.

112.

1 1 1 .
Signal
(mV)

 

110‘

 

 

109 .

 

I —I j

10 11 12 13 14 15 16 17 18 19

Time (minutes)

Figure 3.2: HPLC chromatogram (standard 0.1 wt %) showing peaks for lactic acid (LA)

(retention time = 13.4 min.) and propylene glycol (PG) (retention time= 17.4 min.)

70

 

60000

50000

40000
Peak

size
30000 1

20000

1 0000

 

 

0 i 1 i

0.025 0.050 0.075
Amount (wt %)

Figure 3.3: Calibration curve for lactic acid (R2= 0.9966)

71

 

 

50000
50000
Peak 40000

$126

30000

20000

1 0000

 

0.

 

 

 

 

0.025 0.050 0.075
Amount (wt %)

Figure 3.4: Calibration curve for propylene glycol (R2= 0.993 7)

(b) Synthesis of lactaldehyde

Method 1

Lactaldehyde to be used as a standard for analysis was synthesized by the oxidation of
propylene glycol with pyridinium chlorochromate (PCC), using the procedure described
by George.59 A 50 mL round bottomed flask was charged with 3.23 g of PCC and 20 mL
of CH2C12 and fitted with a reflux condenser. Propylene glycol in 10 mL CH2C12 was
added slowly through the condenser with swirling to the suspended PCC. The reaction
mixture turned brown black and CH2C12 began to boil gently. After 1.5 h, 10 mL of

anhydrous diethyl ether was added, and the supernatant liquid was decanted from the

72

gummy black residue. The residue was washed twice with ether and the combined
CH2Cl2-ether solution was filtered through a Buchner funnel. The solvent was removed
by rotary evaporation. The product was extracted from the residue with 10 mL of D20.

The reaction scheme is shown in Scheme 3.1.

Method 2

The sample of lactaldehyde prepared using method 1 was used as a standard for HPLC
analysis of the ECH reaction mixtures. However, the NMR of the sample (see Appendix)
indicates the presence of pyruvic aldehyde, propylene glycol and pyridine in the sample
in addition to the desired lactaldehyde. An alternative method60 was used to obtain
lactaldehyde in a pure form. Pyruvic aldehyde dimethyl acetal (2g) in 10 mL CH30H was
placed in a 50 mL round bottomed flask and stirred in an ice bath for 30 min. to allow the
temperature of the mixture to reach 0°C. NaBH4 (0.64g) was added to the reaction
mixture, and the mixture was stirred for 20 h at room temperature after which the mixture
was neutralized with 10 mL of 0.1 M HCl, separated by vacuum filtration to remove the
solids and concentrated in a rotary evaporator. The residue was purified by vacuum
distillation. The fraction distilling at 68-70°C was collected. The NMR (see appendix) of
this compound clearly indicates the presence of pure lactaldehyde dimethyl acetal. The
purified product was then hydrolyzed in 0.05 M H2804 for 2 h to form lactaldehyde. The
final product was characterized by 1H and '3 C NMR and ESI-MS. The scheme for the
reaction is shown in Scheme 3.1.

This sample of lactaldehyde was used as a qualitative standard for NMR and ESI-MS, but

could not be used as a HPLC standard due to the presence of an intense peak overlapping

73

with the lactaldehyde peak which made the peak integration and quantification difficult.
Since the NMR and ESI-MS indicated the presence of lactaldehyde as the only organic
component of this sample, the impurity peak was assumed to be due to the presence of an
inorganic anion such as borate, which did not appear in the ESI-MS due to its low

volatility. No further attempt was made to remove this impurity or to characterize it.

 

 

 

 

 

Method 1
OH OH o
PCC, CH2C|2 +
OH > O O
/l\/ 1.5“ A? /u\¢
1,2-propanediol lactaldehyde pyruvic aldehyde
H20
OH
*8”
OH
Method 2
O 0 OH OH
NaBH .25 C .
oc1-13 ‘ g OCH3 005M H2304. OCH3
o
OCH3 2°“ OCH3 5° C' 2" OH
pyruvic aldehyde lactaldehyde lactaldehyde
dimethyl acetal dimethyl acetal hemiacetal
H4-
OH
Ar°“
OH
lactaldehyde hydrate
Scheme 3.1: Synthesis of lactaldehyde
74

(c) Quantification of lactaldehyde

Due to the difficulty involved in obtaining a 100% pure sample of lactaldehyde and the
fact that chemically synthesized lactaldehyde showed a relatively broad peak in the
HPLC probably due to the presence of lactaldehyde in more than one form in solution
(such as acetol, hydrate, dimers, etc.) the quantification of lactaldehyde posed a major
problem. Therefore, instead of using a calibration curve, all the quantification was carried
out using a single standard. Even though this method is less accurate, it gives us a good
estimate of the relative yields of lactaldehyde in the unknown samples. From the 1H
NMR of the D20 extract of lactaldehyde (see appendix) prepared using method 1, the
approximate concentration of lactaldehyde was estimated to be 0.14 g/ 100mL by
comparison of the relative intensities of the C2-H peaks Of lactaldehyde and propylene
glycol (whose concentration was determined from HPLC). The HPLC chromatogram of
the sample using conditions similar to those used for the analysis Of lactic acid and
propylene glycol (described in section 3.2.4 a) is shown in Figure 3.5. Since the
lactaldehyde could not be isolated from the D20 extract, the broadening of the peak may
be a result Of the reversible interaction of lactaldehyde with the other components in
solution such as reagents, byproducts, etc. which causes it to exist in more than one form.
The dip in the baseline is caused by the presence of D20. The peak area of the
lactaldehyde peak is 50161 units. Using this data, the response factor for lactaldehyde
was calculated. The concentration of lactaldehyde in the unknown samples were
determined by multiplying the peak area counts with the response factor.

The response factors for lactic acid, lactaldehyde and propylene glycol are shown in table

3.1.

75

 

102.5 _
reagent peak

100.0 -

 

97.5 ‘
Signal PG

(mV)
95.

1

92.5 _
D20

 

 

 

 

91.0

 

l l l l r

10 15 20 25 30
Time (minutes)

Figure 3.5: HPLC chromatogram Of the D20 extract from the synthesis of lactaldehyde
using Method 1. The mixture contains the product lactaldehyde (LAL) (retention time=

14.7 min.) and reactant, propylene glycol (PG)

 

 

 

 

Analyte Response factor (X 1045 wt%)
Lactic acid 1.74
Lactaldehyde 2.59
Propylene glycol 1.64

 

 

 

 

Table 3.1: Response factors of analytes used for quantifying unknown samples from

ECH

76

3.3 Results and Discussion

3.3.1 Electrodeposition of Ru on carbon felt

The electrodeposition Of Ru on carbon felt was carried out from 5 mM RuC13.xH2O in
0.01M H2804 using a constant current of 10 mA for 90 min. Figure 3.6B shows the
carbon felt surface after 10 s of Ru electrodeposition from 5 mM RuCl3 in 0.01M H2804
at 10 mA current. The SEM clearly shows the formation of hemispherical metal particles
serving as nuclei for further growth.

The Ru content Of the surface was determined using EDS, and the average value of three
different regions was found to be 22.4 wt. % and 3.8 atom % after 90 min of deposition.
This is higher than the calculated value (from the Faradaic current) Of 11.4% because
only the parts of the carbon felt that were rich in particles were used for the EDS analysis,
as shown in figure 3.7.

From the lower magnification SEM micrograph in figure 3.7, it is clear that most of the
deposition takes place on the fibers at the surface, whereas fibers that are deeper inside
have fewer or no metal particles on them. This may be due to the hydrophobic nature Of
carbon felt, which may prevent the penetration of electrolyte deeper into the felt material,
introducing a mass transport limitation on the rate Of deposition.

Figure 3.8 shows the scanning electron micrographs of the surface of a carbon felt fiber at
1, 5, 30 and 90 minutes of electrodeposition time. From the SEM’s, it is observed that Ru
particles grow on the carbon felt surface in the form of clusters with particle sizes in the
range Of 1-5 pm. The deposited metal does not appear as distinct particles, but overlap
with each other. After 1 min of deposition, significant particle growth is observed. The

SEM’s display the surface of a single fiber of carbon felt where a substantial amount Of

77

metal deposition has occurred, but a majority of the carbon felt fibers had no metal
particles deposited on them. With increasing time, the number of felt fibers deposited
with metal particles increases and some of them contain larger and more irregular
clusters. This indicates that nucleation and growth of particles occurs simultaneously in

the case of Ru deposition on carbon felt.

   

Figure 3.6: (A) SEM of bare carbon felt; (B) SEM of carbon felt after 10 s of Ru

electrodeposition from 5 mM RuC13.xH2O in 0.01 M H2804 (i= 10 mA)

78

 

 

16000

COUNTS

 

 

 

 

 

 

keV . 1 0.240

 

 

Figure 3.7 : Low magnification image of Ru/carbon felt deposited from 5 mM
RuC13.xH2O in 0.01 M H2804 (i= 10 mA) with a deposition time of 90 min (top). This is
the magnification used for the quantitative analysis with EDS

Energy dispersive spectrum (EDS) of Ru/carbon felt showing the X-ray signal for the Ru

L line (bottom)

79

 

Figure 3.8: SEM images of Ru/carbon felt after electrodeposition of Ru from 5 mM

RuC13.xH2O in 0.01 M H2804 (i= 10 mA) for (A) 1 min; (B) 5 min; (C) 30 min; (D) 90

min.

80

In the electrodeposition of Pd on carbon felt carried out by Polcaro and coworkers37 from
a solution of 1 mM PdCl2/O.1 M H2804, it was observed that a constant current
deposition under kinetic control produces smaller and more uniformly distributed nuclei
resulting in more uniform sized hemispherical particles with no cluster formation.
However, when the deposition was carried out using constant potential conditions under
diffusion control with the same amount of Pd, the crystal grth occurred prevalently on
the nuclei that were more exposed to the solution, resulting in center overlap and cluster
formation. By comparison with this model, the Ru deposition on carbon felt can be
assmned to be occurring under diffusion limited conditions under the conditions
employed. This might be expected, since the working electrode potential during the
electrodeposition was maintained in the negative potential region which is well below the
potential for the onset of hydrogen adsorption.

In an earlier study on the electrodeposition of Pd on carbon felt from PdCl2 in 0.1 M HCl
carried out by Polcaro and coworkers,53 the number density Of the Pd nuclei was found to
increase at more negative potentials and as the deposition time increased, the centers of
the nuclei on the outer fibers grew and partially overlapped, but there was no increase in
the size Of the nuclei on the inner fibers. From cyclic voltammetric studies of Pd
electrodeposition, it was also found that at more negative potentials, the current
efficiency for Pd deposition decreased due to the competition with hydrogen adsorption
and hydrogen gas evolution catalyzed by the deposited metal.

The cyclic voltammogram (CV) of the carbon felt electrode (3.0 X 2.0 X 0.318 cm)
obtained in 0.01 M H2804 containing 5 mM RuCl3.xH2O is shown in figure 3.9 at three

different scan rates in the potential range, 0.6 to -0.4 V. The current during the cathodic

81

scan consists of components due to both Ru 3+ electroreduction and hydrogen adsorption
and evolution on the deposited Ru. The voltammograms were recorded in a stirred
solution and resemble more closely those of a system which is not limited by semi-
infinite linear diffusion. Therefore the redox processes are not diffusion limited and
should be governed by the steady state kinetics of electron transfer. A current crossover is
Observed indicating a modification of the electrode surface due to Ru deposition. The
current in the negative potential region Observed during the cathodic scan is due to the
combined processes Of Ru3+ reduction and hydrogen adsorption and evolution on the
deposited Ru. The current in the positive potential region during the anodic scan is due to
the desorption and oxidation of adsorbed hydrogen on Ru. The current maximum due to
Ru3+ reduction/H adsorption shifts to more negative potential and currents on the reverse
(anodic) scan larger than on the forward one are observed with increase in scan rate. This
behavior indicates that the electrodeposition of Ru is controlled by the formation of
growth centers or nuclei which are thermodynamically stable. The simultaneous
reduction of H30" is catalyzed by the metal deposits formed on the carbon felt. In the
absence of Ru“, very low currents are Observed in the same potential range. However,
unlike Pd electrodeposition, there are no features due to co-deposition of hydrogen with
the metal, resulting in hydrogen diffusing into the metal lattice as Observed in the Polcaro
case, indicating that hydrogen is not incorporated in the metal lattice Of Ru during

electrodeposition.

82

 

1 (A)

i—---50mV/s I

.-_--:~:.. 10vas .I

L

 

-0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0 8
ease-ea
E (V)

 

 

 

Figure 3.9: Cyclic voltammetry (I-E curve) of carbon felt in 5 mM RuC13.xH2O in 0.01

M H2804 at three different scan rates. Arrows indicate the direction of scan

The current transients following the application Of various values of constant potentials
are shown in figure 3.10. The measurements were performed in 5 mM RuCl3 in 0.01 M
H2804 by applying the required potential, E, and recording the amperometric i-t curve.
The applied potentials were sufficiently cathodic to induce nucleation and growth of Ru
on carbon felt. The current increases immediately after the application Of the potential
and decreases at very short times as the electrode double layer fiilly charges (figure 3.11).
The current then increases as a result of the formation of stable nuclei and grth of the
metal deposits. As the individual diffusion zones of the growing crystallites coalesce, the
current passes through a maximum and then decays as a function Of t'”2 according to the

Cottrell equation.

83

Similar behaviour was Observed for the electrodeposition of Pd on carbon felt by Polcaro
and coworkers53 and for the electrodeposition Of Cu on boron-doped diamond by Wang
and coworkers.61 With an increase in the applied overpotential, there is a decrease in the
time at which the current maximum occurs, indicating that at more negative potentials,
the nucleation and grth is much faster and gets to the diffusion limited regime much
faster. Therefore at higher overpotentials, the time period required for the coalescence Of
the diffusion zones of the individual nuclei decreases, implying an increase in the number
density of growth centers. At E = -3 00 mV and lower, the nucleation step occurs within
the first 200 ms as shown in figure 3.11. The maximum current, on the other hand,
remains almost constant at different applied potentials. These results support the idea that
under the given Operating conditions, the nucleation and growth of Ru is occurring under

diffusion controlled conditions.

84

 

0.01192
0.01191

0.0119 . 1
0.01189 \\ ‘
0.01188 ”\42'; “jig-“2773:7593 ______ ‘
0.01187 #2 _,._- _ ‘i
. 0.01186 . w_ 5‘20”" ‘l

. - - - - E=-300mV (i

0.01185 ~ .._. . .. 3.4004“; 1'

0.01184 , , t
0 500 1000 1500 1

11"3‘31-

i (A)

 

 

 

 

Figure 3.10: Amperometric i-t curve Of carbon felt in 5 mM RuCl3.xH2O in 0.01 M

H2804 at three different applied potentials.

85

i (A)

i (A)

 

0.01192
0.01191 J
0.0119

0.01189 ' -

0.01188
0.01187

g I

 

0.01186 .
0.01185 5

 

 

 

—" _ E=-200mVTi
. — - - - E=-300mV
"":'§_=-40°.[“_Y l

 

 

0.01184 '
0

50 100 150
Time (8)

200

250

 

0.01192

0.0119
0.01188 5
0.01186 1
0.01184
0.01182 '

0.0118

 

 

» —+ E=-200mv7
- - - - E=-300mV 1
E=-400mV1

._ —_,..

 

 

 

 

0.01178
0

0.2 0.4 0.6

“m? “l e

0.8

Figure 3.11: Amperometric i-t curve Of carbon felt in 5 mM RuCl3.xH20 in 0.01 M

H2804 at three different applied potentials; expansions of the shorter time regions

showing rapid current increase due to nucleation

86

3.3.2 Factors governing lactic acid conversion

The optimized conditions for the electrodeposition of Ru is shown in table 3.2. These
were the conditions used for electrode preparation in all the experiments. A series of
experiments was carried out using the Ru/carbon felt electrodes in order to determine the
effect of various experimental conditions on lactic acid conversion. The only product of
the ECH of lactic acid using Ru/carbon felt is lactaldehyde. There was no conversion to
propylene glycol under any condition. A control experiment was carried out using lactic
acid (11.1 mM in 0.01 M H2804) on bare carbon felt for 21 h at 100 mA current. There
was negligible conversion of lactic acid, indicated that lactic acid is not hydrogenated on

bare carbon felt and the Ru electrocatalyst is required for ECH.

 

 

 

 

Electrolyte 0.01 M H2804
Ru“ concentration 5 mM

Current 10 mA

Time 30-90 min.

 

 

 

 

Table 3.2: Summary of Ru electrodeposition conditions

3.3.2.1 Ru deposition time

ECH of lactic acid was carried out using two different electrodes on which the Ru
electrodeposition was carried out for 30 and 90 min. The ECH was carried out using 11.1
mM lactic acid in 0.01 M H2804 for 21 h at 70°C and 100 mA current. The results are

shown in Table 3.3. The yield of lactaldehyde is higher on the electrode obtained from 90

87

min electrodeposition. From the characterization of Ru electrodeposition on carbon felt
in the previous section, it can be inferred that the amount of Ru metal on the electrode
surface is directly related to the electrodeposition time. Therefore, the higher yield of
lactaldehyde in the electrode prepared using a 90 min electrodeposition time can be
attributed to the greater amount of Ru electrocatalyst on the electrode surface. The results
also indicate that the ECH of lactic acid requires the presence of Ru and that carbon felt
alone cannot catalyze the ECH reaction. Therefore, a 90 min. electrodeposition time was
used for the preparation Of subsequent Ru/carbon felt electrodes.

An increase in the yield Of product with increase in the metal electrodeposition time was

also Observed by Polcaro and coworkers37 in the ECH of acetophenone using Pd/carbon

 

 

 

 

felt electrodes.
Ru LA (mM) LA (% LAL (mM) LAL (%
deposition remaining yield)
time Quin)
Initial 14.4 - - -
30 13.3 91.7 0.54 4.7
90 10.0 77.3 2.57 18.8

 

 

 

 

 

 

Table 3.3: ECH of lactic acid using Ru/carbon felt prepared using two Ru

 

electrodeposition times (11.1 mM lactic acid in 0.01 M H2804 for 21 h at 70°C, i= 100
mA). The error in the HPLC is estimated to be i3%. The standard deviation Of the yield

of lactaldehyde at 11.1 mM initial concentration Of lactic acid is 35.2%

88

3.3.2.2 Temperature

Lactic acid (11.1 mM in 0.01 M H2804) was subjected to ECH at 25, 50 and 70°C for 21
h. A constant current of 40 mA was used in order to prevent the heating of the system due
to resistence at high currents. The results are shown in Table 3.4. There is an increase in
the yield Of lactaldehyde with increase in temperature. At higher temperatures, there is an
increase in the the rates Of both ECH and hydrogen evolution by the Tafel reaction. In
this case, the increase in the rate Of hydrogenation appears to exceed the increase in the
rate of hydrogen evolution. There is negligible conversion of lactic acid at room
temperature.

The relative rates of hydrogenation and hydrogen evolution depend on the strength of the
bond to be hydrogenated and on various factors that influence the adsorption phenomena
and the activity Of chemisorbed hydrogen.38 A similar increase in the conversion and
current efficiency with increase in temperature was Observed by Hu and coworkers in the
ECH Of 4-amino 5-nitroso dimethyl uracil on a foamed nickel cathode in (NH4)2804 due
to greater increase in the rate of hydrogenation compared to hydrogen evolution”. An
increase in the rate Of ECH of lignin models such as guaicol62 and 4-phenoxy phenol63 to
phenol and cyclohexanol with increase in temperature at given current was Observed by
Menard and coworkers using Raney Ni powder and 5% Pd/C in Reticulated Vitreous
Carbon (RVC) and 1 M NaOH as the electrolyte. Roessler and coworkers“ Observed an
increase in conversion with increase in temperature up to 60-80°C in the ECH of vat dyes
such as indigo in 1M NaOH, but there was a significant drop in current efficiency and
conversion beyond this temperature. These findings were interpreted in terms of the

competition between hydrogen incorporation into the substrate and decreasing surface

89

concentration due to H2 desorption. Bryan and coworkers3 1 observed an increase in the
ECH rate and current efficiency with increase in temperature in the ECH of 4-
phenylbutene-Z-one to 4-phenylbutan-2-one with no change in product selectivity on Ni
cathode in ethanol containing H2804. Amouzegar and coworkers56 observed an increase
in the ECH Of phenol to cyclohexanol with increase in temperature on dispersed Pt
electrode in 0.05 M H2804. In all these cases, the reported trends were believed to be due

to greater increase in the rate of hydrogenation compared to hydrogen evolution.

 

 

 

 

 

Temperature LA (mM) LA (% LAL (mM) LAL (%
(°C) remaining) yield)
Initial 10.0 100 - -

25 9.3 88.6 0.17 1.7

50 9.1 87.1 1.01 7.7

70 7.8 77.3 1.89 14.5

 

 

 

 

 

 

 

Table 3.4: ECH of lactic acid using Ru/carbon felt at various temperatures (11.1 mM
lactic acid in 0.01 M H2804 for 21 h, 1: 40 mA). The error in the HPLC is estimated to
be :1:3%. The standard deviation of the yield of lactaldehyde at 11.1 mM initial

concentration Of lactic acid is 152%.

3.3.2.3 ECH current
Lactic acid (11.1 mM in 0.01 M H2804) was subjected to ECH at various currents in the

range Of 10-100 mA at 70°C for 21 h. . Though the rate of ECH was maximal at 90°C, the

70°C temperature was chosen in order to prevent excessive evaporation Of water from the

90

system during the long run times. The results are shown in Table 3.5. The yield of
lactaldehyde was found to increase with increase in the ECH current. These results
suggest that the rate of hydrogenation is dependent on the concentration of hydrogen
atoms on the catalyst surface at a given time, which increases with increasing current.
Thus, in spite Of the increased rate of hydrogen evolution, the rate of product formation is
also higher at higher currents. The surface concentration of hydrogen may be directly
related to the current because it is the surface H atoms that form gaseous H2 either by the

Tafel or Heyrovski mechanism.

 

 

 

 

 

 

 

 

Current (mA) LA (mM) LA (% LAL (mM) LAL (%
remaining) yield)

Initial 13.9 100 - -

10 13.3 95.4 0 0

20 13.0 93.6 0.17 1.2

40 12.9 92.6 0.22 1.6

60 12.5 90.6 0.84 6.0

80 10.4 75.1 1.11 8.0

100 10.8 77.3 1.89 14.5

 

 

 

 

 

 

 

Table 3.5: ECH of lactic acid using Ru/carbon felt at various currents (11.1 mM lactic
acid in 0.01 M H2804 for 21 h at 70°C). The error in the HPLC is estimated to be :1:3%.
The standard deviation of the yield of lactaldehyde at 11.1 mM initial concentration of

lactic acid is i5.2%.

91

The conversion Of lactic acid to lactaldehyde after 21 hours at low current (10 mA) is
close to zero, suggesting a requirement for adsorbed molecular hydrogen (or at least
closely spaced atoms) on the surface rather than isolated hydrogen atoms. This
mechanism was proposed earlier from kinetic studies of lactic acid hydrogenation”.
Thus, the low surface concentration Of hydrogen and the low surface area Of active
catalyst, along with mild temperatures, may explain why ECH of LA is slow compared to

classical high pressure hydrogenation.

3.3.2.4 Nature of electrolyte

ECH of lactic acid (11.1 mM) was carried out in three different electrolytes at 0.01 M
electrolyte concentration at 70°C for 21 h. The electrolytes studied were H2804, HC104
and HCl. The results are shown in Figure 3.12 and 3.13. Similar experiments were carried
out using 0.1 M concentrations of the same electrolytes, but the conversions were found
to be insignificant, probably due to competitive adsorption of electrolyte anions on the
electrode surface preventing adsorption and reaction of lactic acid, which is present in
much lower concentration. From the results, it can be seen that there is considerable
variation in the rates of formation of lactaldehyde in various electrolytes due to
differences in the nature of the electrolyte anions and surface modification of the catalyst
by these anions. In general, the relative rate of formation Of LAL was found to be higher
in HC104 and HCl and lowest in the presence Of H2804. A more detailed discussion of

anion effects on electrodes is included in Chapter 4.

92

(2’ __

Yield (%)

 

 

 

, +0.01M H0104 :

 

 

5 10 15 20
_ lime m

+0.01M H2SO4

. +0.01M HCI

__l

Figure 3.12: Yield of lactaldehyde as a function of time in various electrolytes (11.1 mM

lactic acid in 0.01 M electrolyte concentration at 70°C for 21 h, i=100 mA)

Moles of LAL

 

 

 

 

0.00025
0.0002 4 - .
0.00015 - 2 -
0.0001 - - 2 L
0
0.00005 -. ,2 a ’
I ‘ o
0 e—l 9 ° , p
0 5 10 15 20

T1me(h)

co.01M H2804
‘.0.01M H01 1*
E. 0.01M HC104 ? *

l

Figure 3.13: Moles Of lactaldehyde as a function of time in various electrolytes (11.1

mM lactic acid in 0.01 M electrolyte at 70°C for 21 h, i=100 mA)

93

3.3.2.5 Initial lactic acid concentration

ECH of lactic acid was carried out at three different lactic acid concentrations, 11.1 mM,
55.5 mM and 0.11 M in 0.01 M HCl at 70°C for 21 h. A constant current of 100 mA was
used. The results are shown in Figure 3.14. The final yields of lactaldehyde are shown in
Table 3.6. There is an increase in the moles of lactaldehyde formed with an increase in
the initial lactic acid concentration, though the % yield decreases with increase in the

initial substrate concentration.

 

  

 

 

 

 

0.0012

' R2 = 0.9987 _ __- .__ -- _ g
j 3' 0.0008 ~ - ~ =3E-05x ‘ 0 01% LA l
‘. - R2=0.9903 ‘ ' ”MA
1 § 0' ' *5 . 1.0% LA 3
l a _ —Linear (01% LA). I.
l 5 00004 A 7 y: 9596" ; —Linear(0.5% LA)! ,
f 00002 '37 09781 ’ —Linear(1.0% LA): 1
l 0 ,1 f .
‘ 0 5 10 15 20 25

Time (h)

Figure 3.14: Moles Of lactaldehyde formed as a function Of time at various initial lactic

acid concentrations (lactic acid in 0.01M HCl at 70°C for 21 h, i=100 mA)

94

 

 

 

 

LA initial LA initial molar LAL (wt.%) LAL (% yield)
concentration (wt. concentration (M)

0/0)

0.10 0.011 0.02 25.0

0.49 0.054 0.06 15.0

1.00 0.111 0.09 11.5

 

 

 

 

 

 

Table 3.6: Final yields of lactaldehyde after ECH of lactic acid at various initial

concentrations (Electrolyte: 0.01 M HCl, T=70°C, time = 21 h, i=100 mA)

3.3.3 Determination of the order of reaction

For the purpose Of calculating the order of the reaction, the data for the ECH of lactic
acid in 0.01 M HCl at 70°C were used. The order of the reaction was determined with
respect to both lactic acid and hydrogen. The order of reaction with respect to lactic acid
was determined from the log-log plot Of the turnover frequencies (TOF) at various initial
lactic acid concentrations in the range of 0.1 to 1 wt.% (11.1 mM to 0.11 M). The order
of reaction with respect to hydrogen was determined from the log-log plot of the TOFs at
various currents. Variation of the current, in this case is considered to be equivalent to
variation of H2 partial pressure in the batch hydrogenation because the surface hydrogen
concentration is directly proportional to the current and the H2 partial pressure,

respectively, in the two cases.

95

‘: 0.00025 .

 

 

 

 

 

 

' 0.0002 , *,
E .l l
i 5 0.00015 ~ 1
l '5 eeeeeeee ‘
j % o_ooo1 , o l=50mA l
l 5 . . l=100mA . 1
3 0.00005 - - , Linear (l=50mA) 3 ;
: —Linear(l=100mA)i .

0 ~ ” T T T “'”_" .
i 0 5 10 15 20 25 '
' Time (h)

Figure 3.15: Moles of lactaldehyde formed as a function of time at various ECH currents

(11.1 mM lactic acid in 0.01 M HCl at 70°C for 21 h)

A plot of the number of moles of lactaldehyde formed versus time at various currents is
shown in Figure 3.15. The turnover frequencies (TOF) were obtained from the slopes of
the plot. The plots of TOF versus initial lactic acid concentration and TOP versus moles
of H atoms produced per hour are shown in figures 3.16 and 3.17 respectively. From the
slopes Of the log-log plots of TOF versus concentration, the order Of reaction with respect

to lactic acid and hydrogen were found to be 0.86 and 0.84 respectively.

96

6.00E-05

 

5.00E-05 ,

4. 00E-05

3.005-05 -

TOP (11")

2.00E-05 ,.

1.00E-05

 

 

 

0.00E+00

 

0 0.002 0.004 0.006 0.008 0.01
Initial moles of LA 1

Figure 3.16: TOF for lactaldehyde formation as a function Of initial moles Of lactic acid

in 0.01 M HCl at 70°C after 21 h, i=100 mA

I? __ ._ ”2.2 2, 7 . . i2. 2 . 2. ,2 L L 7.

0.00001
0.000009
0.000008
0.000007 -. _--
.7- 0.000006 -. ,-,_
5 0.000005 ,
‘ § 0.000004: - .2- .- --
0.000003 «
0.000002 « - - -----
0.000001 4. __
0

 

 

 

 

0 0.001 0.002 0.003 0.004
Moles of H atoms formed per hour

Figure 3.17: TOF for lactaldehyde formation as a function of moles of H atoms formed
per hour (assumed to be directly proportional to current) (11.1 mM lactic acid in 0.01 M

HCl at 70°C after 21 h)

97

In case Of both lactic acid and hydrogen, a positive reaction order indicates that increase
in the surface coverage of either reactant is definitely beneficial in terms of TOF for
lactaldehyde formation by ECH. However, a fractional order of reaction indicates that the
surface coverage reaches a saturation value on the electrode surface, corresponding to
almost complete coverage at a certain concentration. This indicates strong adsorption of

lactic acid and hydrogen on the electrode surface.

3.3.4 Hydrogen evolution on Ru
Giles and coworkers64 studied the hydrogen evolution reaction (HER) using Ru on carbon
electrodes. Using plots Of the potential dependence of HER of various electrodes, it was
determined that HER on Ru is shifted in the cathodic direction compared to HER on Pt/C
prepared under similar conditions. The Tafel slope at a constant [Ru3+] and [W] was
given by the following expression.

i242 1

610g =
6E [RT/(1+a)F]

 

where a is the symmetry factor of the plot of AG versus the reaction co-ordinate.
Considering a = 0.5, the Tafel slope was calculated to be 40 mV. Based on these results,
the following mechanism (scheme 3.2) was proposed for HER on Ru deposited on

carbon.

 

H++e H.

 

H++H.+e———> H2

Scheme 3.2: Proposed mechanism for hydrogen evolution on Ru

98

The formation of H. is considered as the transition state Of the reaction and the stability of
this transition state determines the rate of HER and hence, the current.

Tafel curves of Ru deposited on vitreous carbon obtained by Breiter and coworkers“ at
various stages in the deposition revealed that the rate of HER at constant potential
depends on the size of the electrocatalyst centers. Since HER was delayed until relatively
large centers were formed, the rate detemining step was assumed to be the surface
diffusion of hydrogen from one type of site to the other and the hydrogen coverage of
both types Of sites was assumed to be negligible. Such a rate determining process was
also suggested by Grenness and coworkers66 in an earlier study. However due to
simultaneous occurrence of other phenomena such as anion adsorption on the electrode
surface, it is not possible to determine the surface hydrogen coverage by voltammetric
current-potential curves. Examination Of Tafel curves and exchange current densities
using smooth Ru in H2804 revealed that the rate determining step for HER on Ru at low
polarization is the Tafel reaction and convective diffusion Of H2 whereas at high

polarization, the rate determining step becomes the Volmer-Heyrovski reaction.

3.3.5 Determination of the rate constant of lactic acid electrohydrogenation
The rate expression for the ECH reaction can be described as follows.38’39
R= k9LA39Hb (1)
where k is the apparent rate constant, OLA and 0” are the fraction of surface sites occupied
by lactic acid and hydrogen respectively, and a and b are the order of the reaction with

respect to lactic acid and hydrogen, respectively. k is a function of the concentration of

the solution and the electrode surface area.

99

Substituting the values Of a and b determined in section 3.3.3, the rate expression can be
written as follows.
R= k0LA0'860H0‘84 (2)
The fractional coverage by the organic substrate and hydrogen were assumed to be
independent of each other in previous studies of ECH kinetics37 due to the difference in
the nature of the species involved. Also due to the low overpotential, the electrochemical
formation of adsorbed H (Volmer reaction) was considered at quasi-equilibrium state.
Mass transfer effects are assumed to be insignificant since in the ECH Of the model
compound, benzoyl formic acid, the rate of ECH was found to be dependent on the
availability Of surface sites and not on the rate of mass transfer. Mass transfer effects
were also found to be negligible in the batch hydrogenation of lactic acid since the rate of
hydrogenation was independent of the stirring speed”. A Langmuir-Hinshelwood rate
equation, similar to that described by Zhang‘5 was used tO describe the kinetics of the
ECH of lactic acid.
The rate of desorption Of adsorbed H or the rate Of formation Of molecular H2 by the
Tafel reaction, which is a parallel reaction to ECH can be expressed as follows”.
RH: 14.0.3 (3)
where kn is the rate constant for hydrogen evolution.
According the Langmuir adsorption isotherm, the fractional catalyst coverage by lactic
acid can be expressed as follows.

KLACLA

BLA = (4)
1 + KLACLA

100

where KLA is the adsorption constant Of lactic acid on Ru and CLA is the concentration of
lactic acid in solution.
The fractional catalyst coverage by hydrogen can be expressed as follows.
Kanilexm-nF/RT)

9n= (5)

1 + KH[H+]exp(-nF/RT)
where Kn represents the adsorption constant for hydrogen on Ru, 1] is the overpotential
during the ECH, F is Faraday’s constant, R is the universal gas constant and T is the
absolute temperature. 11 is given by the equation,
n= E-E°
where E is the potential Of the working electrode and E0 is the equilibrium potential Of the

proton reduction reaction written as
211* + 2e' =—_— H2 13°: -0197 v vs. Ag/AgCl
E0 is dependent on the pH Of the solution. The actual equilibrium potential, EO’ at the
solution pH can be calculated using Nemst equation.

0.059 aH+2
E0’ = E0 + __ log _ (6)

n 8112

The activity Of gaseous H2 can be assumed to be unity. At aH+=l, EO=EO’. In 0.01 M HCl,
E0’ is calculated as -0.433 V. The applied overpotential, n was calculated to be -0.767 V.
For calculating the temperature dependent adsorption constants, KLA and K”, the
following expression was used.15

AH:

Ki: K01 exp __ (7)
RT

101

Where, K4 is the adsorption equilibrium constant of a particular species at the given
temperature, K01 is the adsorption constant at O K, T is the absolute temperature and AH
is the heat of adsorption. The adsorption constants, KOLA and K0242 were determined to be
4.45 X 10'7 and 2.13 X 10''3 respectively, considering AHLA= 47 kJ mol'1 and AHH2= 79
kJ mol", KLA (403 K)= 0.55 and KHz (403 K): 3.7 x103 '5 and assuming hydrogenation
via adsorption of molecular hydrogen on the surface, as inferred from the increase in the
hydrogenation yield with increase in current. KLA and KH at 323 K and 343 K were then
calculated using equation 7.

The values of the exponential terms in equation 5 at 323 and 343K are 1.19 X 1010 and
4.28 X 108 respectively. Therefore, KH[H+]exp(-nF/RT) >>1, which gives a value for the
fractional coverage of hydrogen close to unity. The rate Of hydrogenation in this case is
determined only by the temperature and lactic acid coverage. However, at a lower current
of 50mA, the fractional hydrogen coverage decreases significantly (table 3.7) to 0.39 at
343 K which leads to a decrease in the rate of hydrogenation.

In order to calculate the apparent activation energy of the reaction, the rate constant of the
reaction was calculated at two different temperatures, 323 K and 353 K. The plot of the
number of moles of lactaldehyde formed as a function Of time at 323 K is shown in figure

3.18. The TOF is determined from the slope of the plot to be 4 X 10'6 h".

102

0.00009
0.00008 .. .2 - 2 ~ - .- .2 . . - -2 L 2.
0.00007 -.

0.00006 -. -- - - - - -. - - - _ - _ 2 , ,-,
0.00005 -. -7 .. - - - - -. .¥.=2.€E-.0_6X_ w - i

2 = I
000004 a B -3 0.9828 I

0.00003 -
0.000024 --
0.00001 1

 

 
 
 
   

WadLAL

 

 

 

0 5 10 15 20 25
Time (h) '

Figure 3.18: Moles of lactaldehyde formed as a function of time at 50°C (323 K) (11.1

mM lactic acid in 0.01 M HCl, i=100 mA)

The values of the TOF ’3, adsorption constants, fractional coverages and rate constants at
the two different temperatures, 323 K and 343 K are shown in Table 3.7. According to
Arrhenius equation,

Ea 1
ln k = In A - (8)

R T
where k is the rate constant, T is the absolute temperature, R is the universal gas constant,

Ea is the activation energy, A is the frequency factor.

Therefore, a plot of In k versus I/T is obtained with slope = -Ea/R and intercept = In A.

103

Temperature

 

Table 3.7: Kinetic parameters for the ECH of lactic acid at two different temperatures

 

 
    
 

y = -9143.7x + 17.454
R2 = 1

n

I
_;.
O

 

 

 

0.0029
003292
0.00294
am A
0.00304
00333
0.00308

0.CXI31
0 (13312

0.00298
0.003
2 0.(X)302

Figure 3.19: Arrhenius plot for the ECH of lactic acid

104

From the plot in figure 3.19, the activation energy, Ba and the frequency factor, A were
determined to be 76.0 kJ mol'1 and 3.8 X 107 h'l respectively. It should be noted that the
values of rate constant and activation energy are apparent values, as a major part of the
hydrogen adsorbed on the electrode surface is used up in HER. Since the rate of hydrogen
evolution was not considered in the calculations of the kinetic parameters, values of rate
constants higher than the actual values and that of activation energy lower than the actual
value are obtained. The apparent activation energy determined in this case is
approximately half the value of 138 kJmol", determined from the analysis of the batch

hydrogenation of lactic acid.'5

3.4 Conclusions

ECH of lactic acid was carried out using carbon felt modified with Ru. Carbon felt was
chosen as the support due to its high fluid permeability, specific surface area, more rapid
dynamic adsorption characteristics (mass transfer rates) and lower resistance to the flow
of liquids and gases. The electrodeposition of Ru was carried out at constant current
under diffusion limited conditions which resulted in formation of micrometer sized Ru
particles as well as larger clusters mainly on the carbon felt fibers closer to the surface.
The only product of ECH of lactic acid under these conditions is lactaldehyde, and the
yields are low (<20%). The yield of the product was found to increase with increase in
the Ru electrodeposition time, temperature and current. Studies on the ECH of lactic acid
in various electrolytes showed higher rates in HCl and HC104, compared to that in H2804
at 0.01 M concentration. However, when the electrolyte concentration was increased to

0.1 M, there was a significant inhibition of the ECH of lactic acid. The turnover

105

frequency with respect to lactaldehyde formation increased with increasing initial lactic
acid concentration.

TOF values at various initial lactic acid concentrations and various currents were used to
determine the order of the reaction. The order of the surface reaction was found to be
0.86 with respect to lactic acid and 0.84 with respect to hydrogen. The kinetics of the
surface reaction was described using the Langmuir Hinshelwood kinetic model, the
adsorption constants and fractional surface coverages of lactic acid and hydrogen and the
apparent rate constants of the reaction were calculated at 323 and 343 K. Using these data
the activation energy and frequency factor of the reaction were determined to be 76.0 kJ
mol'1 and 3.8 X 107 h'1 respectively. Comparison of the values for 6“ at i= 50 mA and
100 mA at constant temperature shows that the fractional surface coverage was almost
unity at 100 mA and decreased to 0.39 at 50 mA, which may be one of the reasons for the

lower rate of hydrogenation at lower current.

106

CHAPTER 4
BULK ELECTROCATALYTIC HYDROGENATION OF LACTIC ACID
USING RETICULATED VITREOUS CARBON AGGLOMERATED WITH
RUTHENIUM ON CARBON

4.1 Introduction
In chapter 3, ECH of lactic acid was achieved using Ru dispersed on a high surface area
carbon felt support. However, the yield was low and the major product was lactaldehyde.
In order to improve the yields, an alternative strategy was used for ECH in which 5%
Ru/C was agglomerated in a cathode made of a macroporous material, namely a
Reticulated Vitreous Carbon (RVC), by simply stirring the catalyst under cathodic
polarization. This chapter describes a series of ECH studies carried out with lactic acid
using RVC working electrode suffused with the same 5% Ru/C powder catalyst as was
used in the chemical reduction. Previous work using this type of electrode was reported
by Menard and coworkers.68'7' Trapped in the pores of the RVC, the catalyst particles are
in electrical contact with the electrode. The same powder catalyst was used previously in
our studies of the chemical catalytic hydrogenation. The RVC has a high overpotential
for hydrogen evolution and is a relatively inert material that does not participate in
electrohydrogenation and hydrogen evolution. It is also electrochemically stable and its
three dimensional network of large pores provides electrically conductive host sites to
support composite powder catalyst particles. The electrolyte penetrates the RVC by
forced convection and the suspended catalyst particles are trapped inside the pores. These
catalyst particles are in electrical contact with the RVC electrode and show

electrocatalytic activity.

107

The method provides an attractive way to electrochemically explore the reducing ability
of new supported metal powder catalysts, such as those used in chemical catalytic
hydrogenation, without changing their chemical or physical characteristics. Thus, the
physical form of the catalyst, and presumably the essential components of the catalytic
process can be retained, allowing maximum comparability between the electrochemical
and the chemical processes.

The rate of ECH depends on several parameters such as the catalyst particle size and
current density. The electrochemical cell design is also important in this case because the
catalyst agglomeration process is governed by the fluid dynamics. In these reactions, the
organic compound is usually thought to be adsorbed on the support, the H atoms are
adsorbed on the metals, and the hydrogenation takes place at the adlineation point
between the metal and the support particles.71 The efficiency of hydrogenation therefore
depends on the nature and adsorbing power of the support. The rate of the ECH process
also depends on the nature of the metal and the heat of adsorption of H on the metal.
Since the rate of ECH is determined by the surface concentration of both reactants, i.e.
organic substrate and hydrogen, supports such as carbon and alumina which are strong
adsorbers of organic molecules show high efficiency. The rate determining step is the
hydrogen transfer from the metal surface to the unsaturated molecule.

The electrocatalytic hydrogenation of lignin models68 to useful organic molecules and the

ECH of 4-phen0xy phenol to phenol69

showed the successful use of Pd/alumina, Pd/C
and Rh/alumina catalysts. The cleavage of the strong diphenyl ether bond is much more
difficult compared to the glycosidic bonds of lignin and thermal cracking and catalytic

hydrogenation give very low conversion rates. ECH provided higher efficiencies of 50-

108

60%. The cleavage of the diphenyl ether bond is very promising for the conversion of
biomass to useful organic molecules, as these types of bonds are found in Kraft lignin
produced by the paper and pulp industry. ECH of the pyrrole ring in pyrrole and its
carboxylic acid and ester derivatives was also studied using Pd/alumina, Pt/alumina and
Rh/alumina catalysts by Menard and coworkers.70 It was shown that the hydrogenation is
specific to the pyrrole ring of the alkyl pyrrole 2-carboxylate and does not modify the
ester function of the molecule. The ECH of the alkyl pyrrole carboxylates was found to
be more efficient with a Rh/alumina catalyst in 0.1 M NaCl at pH 1.2, whereas pyrrole
and pyrrole-Z-carboxylate gave better results in phosphate solution at pH 7. The ECH
process could be used to hydrogenate the pyrrole ring in a complex natural product,
ryanodine, but there was no diastereoselection in the formation of the reduced compound
due to the easy access of hydrogen to both sides of the pyrrole ring. The ECH was faster
with the Rh/C catalyst compared to Rh/alumina, but there was stronger adsorption of the
hydrogenated product on Rh/C, leading to lower recovery.

The ECH of phenol to cyclohexanol using RVC electrodes was studied in the presence of
various catalysts, namely Pd/A1203, Pd/BaCO3 and Pd/BaSO...67 Pd/A1203 was found to
be the most efficient catalyst due to the stronger adsorbing power of alumina. The ECH
of cyclohexanone to cyclohexanol was carried out using various metals supported on
carbon and alumina.71 In this case also, the ECH was found to be largely dependent on
the nature of both the metal and the support. The ECH efficiency is higher on carbon
support compared to alumina due to the greater adsorbing power of carbon and for both
types of matrices, the best performance was observed with Rh. In this study, the cell was

designed so as to allow the catalyst particles to flow through the RVC electrode. The

109

electrode seemed to work like a fluidized bed and could be compared to a “current pulse
electrochemical reaction”. This allowed the catalyst to be in open circuit condition when
it is circulating in the cell and in an applied current state when it is flowing through the
electrode. H atoms are produced only when the catalyst is in the applied current state.
When the catalyst is in the open circuit mode, the adsorbed atomic hydrogen favors the
hydrogenation of organic molecules rather than hydrogen evolution, leading to a higher
ECH efficiency compared to a constant current method because the Tafel reaction cannot

proceed.

4.2 Experimental
S-Lactic acid and propylene glycol were purchased from Aldrich and were used without

further purification. 5% Ru/C was purchased from PMC Catalysts.

4.2.1 Electrochemical cell setup

Electrohydrogenation studies of lactic acid in several aqueous electrolyte solutions were
carried out in the glass electrochemical cell described in chapter 3. It is a two
compartment cell with the anode and the cathode compartments separated by a glass flit
in order to prevent the diffusion of gaseous products. The reference electrode is Ag/AgCl
(E°= -0.35 V vs. SCE). The counter electrode is Pt wire. The Reticulated Vitreous Carbon
(RVC) working electrode (dimensions 4.0 X 3.0 X 0.5 cm); pore size 100 pores per inch
(ppi) is connected to the external circuit via a copper wire, positioned to minimize the
surface area of Cu exposed to solution; electrochernistry on the copper surface can thus

be assumed to be negligible. A voltmeter is connected between the reference and working

110

electrode to measure the potential of the working electrode. Controlled current Was
applied using a model 273 potentiostat/galvanostat from Princeton Applied Research.
The temperature of the system was controlled using a heating tape wrapped around the
cell and connected to a temperature controller from Omega. Samples were withdrawn for
analysis immediately after the cell was switched on and at intervals during the
electrohydrogenation.

The electrolytes were prepared by the dilution of the concentrated acids, namely 98%
sulfuric acid (CCI), 36.5% hydrochloric acid (CCI), and 70% perchloric acid (Spectrum)
with HPLC grade water. S-Lactic acid and propylene glycol were purchased from

Aldrich and were used without further purification.

4.2.2 Electrocatalytic hydrogenation

Before each experiment, 1.0 g (dry weight) of 5% Ru/C was added to the electrolyte
solution and the solution was stirred for 2 hours under cathodic polarization (I= 20-40
mA) in order to ensure proper agglomeration of the catalyst in the pores of the RVC. The
potential of the working electrode during this process was in the range of -100 to 400
mV (vs. Ag/AgCl) depending on the electrolyte and the temperature of the cell. The
stirring rate was kept low in order to prevent the agglomerated catalyst from being
washed off the pores. The RVC surfaces before and after agglomeration were examined
by Scanning Electron Microscopy (SEM) as shown in figure 4.1. Elemental composition
as measured via Energy Dispersive Spectral (EDS) analysis showed the average surface
Ru content to be about 5.5i0.l %. After the agglomeration, 1 mL of lactic acid solution

in the respective electrolyte was added so that the final concentration of lactic acid in

111

solution was 0.1 wt% or 11.1 mM, and the total volume of the solution was 75 mL.
Samples were withdrawn for analysis immediately after addition of lactic acid and at

intervals during the electrohydrogenation.

4.2.3 Characterization of the electrode surface

The electrode surface was characterized using a J SM-6400 V scanning electron
microscope (JEOL, Ltd., Tokyo, Japan). Micrographs were recorded using secondary
electron mode generated with an accelerating voltage of 20 kV. The presence of Ru was
verified by energy dispersive X-ray (EDX) analysis (N oran Instruments, Inc.,
Middletown, WI) connected to the J SM-6400 V. SEM images were obtained using the

AnalySIS sofiware (Software Imaging System Corp., Lakewood, CO).

112

 

Figure 4.1: SEM images of an RVC working electrode (100 ppi): bare RVC electrode

(top); two regions of the RVC electrode after entrapment of Ru/C via agglomeration for

2 h at 20 mA under cathodic polarization (bottom).

113

4.2.4 Analysis of products

Reaction mixtures were filtered using Millipore syringe filters (0.25 pm) in order to
remove the catalyst before analysis of the reaction products by HPLC with refractive
index detection, using a Biorad HPX87H ion exchange column at 65°C. The mobile

phase used was 5 mM H2804, run isocratically at a flow rate of 0.6 mL/min.

The method used for the analysis and quantification of reactant, lactic acid and products,

lactaldehyde and propylene glycol is similar to the one described in chapter 3.

4.3 Results and Discussion

ECH of lactic acid yields lactaldehyde and/or propylene glycol under all conditions
studied. The efficiency of electrocatalytic hydrogenation of lactic acid was found to be
highly dependent on the nature of the cell and the stability and dynamics of the system.
The reductive catalyst agglomeration process before substrate addition was used to
ensure that all Ru in contact with the RVC was in the metallic state and saturated with

hydrogen. This step is equivalent to catalyst reduction in traditional batch hydrogenation.

A control experiment carried out at 90°C using RVC with no catalyst in 10 mM H2804
showed no LA conversion, indicating that the RVC alone is not capable of
hydrogenation. Another control with Ru/C catalyst and gaseous hydrogen (1 atm) run at
90°C in the absence of the RVC similarly showed no reaction, indicating that Ru needs to
be in electrical contact for the hydrogenation of lactic acid to occur at the mild conditions

of this study. To maximize its current efficiency, the ECH should be carried out under

114

conditions that minimize the competing hydrogen evolution by the Tafel or Heyrovski
processes.

As an additional control experiment, lactaldehyde, synthesized chemically using the
procedure described by Nambiar and coworkers60 was subjected to electrohydrogenation
in 10 mM HC104 at 70°C and 100 mA current. There was negligible conversion of
lactaldehyde to propylene glycol. Lactaldehyde produced in situ from pyruvaldehyde

dimethyl acetal and from pyruvaldehyde was also not hydrogenated to propylene glycol.

4.3.1 Factors governing lactic acid conversion
4.3.1.1 Temperature

Electrohydrogenation experiments with 11.1 mM lactic acid in 0.01 M H2804 were
carried out at 25, 50, 70, and 90°C for 9 h using a constant current of 40 mA. The yields
of the products, namely lactaldehyde and propylene glycol, were found to increase at
higher temperatures as shown in table 4.1, indicating an increase in current efficiency
with temperature. At higher temperatures, there is an increase in the the rates of both
ECH and hydrogen evolution by the Tafel reaction. In this case, the increase in the rate of
hydrogenation appears to exceed the increase in the rate of hydrogen evolution. There is
no conversion of lactic acid at room temperature. The relative rates of hydrogenation and
hydrogen evolution are thought to depend on the strength of the bond to be hydrogenated
and on various factors that influence the adsorption phenomena and the activity of
chemisorbed hydrogen”. These trends are similar to those observed with Ru/carbon felt

in chapter 3.

115

 

 

 

 

 

Temperature LA remaining LAL (%) PG (%) Material

(°C) (%) balance
(%)

25 71.8 1.2 1.6 74.6

50 63.0 3.3 3.5 69.8

70 66.3 4.2 5.1 75.6

90 67.0 5.7 9.8 82.5

 

 

 

 

 

 

 

Table 4.1: Yields of products of ECH of lactic acid at various temperatures (11.1 mM
lactic acid in 0.01 M H2SO4 after 9 h; i= 40 mA). The error in the HPLC-determined

concentrations is estimated to be i3%.

4.3.1.2 ECH Current

Electrohydrogenation of 1 1.1 mM lactic acid in 0.01 M H2SO4 for 9 hours at 90°C at
different currents indicated that with increasing current, the yield of propylene glycol
increases and that of lactaldehyde decreases slightly, as shown in figure 4.2. These results
suggest that the rate of hydrogenation is dependent on the concentration of hydrogen
atoms on the catalyst surface at a given time, which increases with increasing current.
Thus, in spite of the increased rate of hydrogen evolution, the rate of propylene glycol
formation is also higher at higher currents. In the ECH of benzaldehyde and
acetophenone studied by Polcaro and coworkers37 using Pd/carbon felt electrodes, the
surface hydrogen atom concentration was assumed to increase with an increase in the
current. The conversion of lactic acid to propylene glycol at low current (10 mA) is close
to zero and rises with increasing current in spite of the attendant increased rate of
hydrogen evolution. Assuming the sequential mechanism of scheme 4.1, the aldehyde
reduction step appears to require adsorbed molecular hydrogen (or at least closely spaced

atoms) on the surface rather than isolated hydrogen atoms.

116

 

1 + lactaldehyde
. + propylene glycoljt i

 

% yield

 

 

 

 

 

0 20 40 60 80 100 120
Current (mA)

Figure 4.2: Yield of lactic acid reduction products after electrohydrogenation (11.1 mM
lactic acid in 0.01 M H2SO4 for 9 h at 90 °C). Only lactaldehyde and propylene glycol

were detected. The error in the HPLC is estimated to be i3%.

In the ECH of cyclohexenone using nickel boride electrode, Mahdavi and coworkers30
observed an increase in the conversion rate with a decrease in current density due to a
greater increase in the rate of hydrogen evolution compared to hydrogenation. However,
there was a decrease in the further hydrogenation of cyclohexanone to cyclohexanol. This
was explained by the fact that the more negative potential at higher current densities
would cause the chemisorbed hydrogen to be more active leading to faster hydrogenation
of cyclohexanone molecules generated at the electrode surface by cyclohexenone
hydrogenation. At lower current densities, the hydrogenation of cyclohexanone would
become slow enough to allow its diffusion to the bulk solution before being

hydrogenated. Cyclohexanone is expected to be more weakly adsorbed on the electrode

117

surface compared to the conjugated enone, cyclohexenone. In the ECH of phenanthrenes
on Raney Ni electrode studied by Robin and coworkers,72 an increase in the extent of
ECH (yield of octahydrophenanthrene) was observed with an increase in current up to a
certain value beyond which there was a decrease in yield. On the other hand, Senda and
coworkers32 reported an increase in the ECH of 4-t-butyl cyclohexanone to alcohol with

decrease in current.

OH OH OH OH
_. H -H O OH
Nor-1 2” NH ..___. /|\y( 2 . A/
HO OH 0 2H
0 lactaldehyde
lactic acid 132-9109311910“ 1.2-propanediol
(lactaldehyde hydrate)

Scheme 4.1: Sequence of reactions in lactic acid hydrogenation

The case of lactic acid hydrogenation is similar to the Mahdavi results above, but more
complex due to the intermediate dehydration step involved before lactaldehyde can be
further hydrogenated to propylene glycol. With increasing currents, yields of
lactaldehyde and propylene glycol decrease and increase, respectively. The simplest
interpretation for this observation is that lactic acid conversion to propylene glycol occurs
via two sequential hydrogen additions with lactaldehyde hydrate and lactaldehyde as
discrete intermediates. At low current, displacement of the relatively weakly bound
lactaldehyde hydrate (1,1,2-propanetriol) by lactic acid is faster than its dehydration and
capture by surface hydrogens. At higher currents lactaldehyde (also relatively weakly

bound) would be more quickly converted to propylene glycol due to the availability of

118

more surface hydrogen atoms. But the nearly complete absence of propylene glycol at
low currents suggests another factor: base-promoted acceleration of lactaldehyde
formation at the surface via dehydration of lactaldehyde hydrate in the diffusion layer.
Like other carbonyl hydrates, lactaldehyde hydrate’s equilibration with the free carbonyl
form is accelerated at pH values both above and below neutral. At higher currents,
solution protons are more quickly adsorbed and reduced, so the pH at the cathode surface
is also higher. Not only are more hydrogen atoms available for reaction but also the
concentration of free LAL in proximity to the surface is enhanced, increasing PG

formation.

An alternative to the sequential interpretation involves independent paths: parallel
reactions might form LAL and PG simultaneously from LA adsorbed on two different
types of catalytic sites. In the ECH of benzaldehyde and acetophenone by Polcaro and
coworkers,37 the aldehydes were converted to two different products, the corresponding
alcohol and hydrocarbon, and the formation of the two types of products were assumed to
involve parallel reactions of the aldehyde adsorbed on different catalytic sites.
Adsorption of aldehyde on low coordination sites such as steps and kinks was assumed to
lead to the formation of hydrocarbon whereas the alcohol was thought to form by
adsorption on high co-ordination sites. Moreover, the ratio of the two types of products
was found to depend on the particle size, morphology and the method of preparation of
the catalyst due to differences in the relative number of structural features such as edges
and terraces, resulting in different ratios of the two catalytic sites. Lacking information
on the reactivity of multiple catalyst morphologies, the two-site interpretation of the

lactaldehyde and propylene glycol production cannot be ruled out. However, the

119

sequential scheme is simple, consistent with all observations, and represents a
satisfactory working model for this reaction.

The increase in the rate of formation of propylene glycol at higher current may also be
attributed to higher electrode surface energy at more negative potential, resulting in
higher mobility of H atoms, similar to the phenomenon observed by Mahdavi and
coworkers30 in the ECH of cyclohexenone. The trend observed here also points to the
involvement of overpotential deposited hydrogen (OPD-H), rather than underpotential
deposited hydrogen (UPD-H) in the ECH of lactic acid. UPD-H atoms are strongly bound
to the electrode surface and are believed to be formed on Ru at around 0.0 V, whereas the
weakly bound OPD-H atoms are formed at more negative potentials and may co-exist
with UPD-H. These weakly bound H atoms were found to be involved in the hydrogen
evolution reaction and in the ECH of the olefinic double bond in substrates such as

maleic acid.73

4.3.1.3 Nature of electrolyte

Electrohydrogenation of 11.1 mM lactic acid was carried out in different electrolytes at
70°C and 100 mA current. Electrolytes studied were H2SO4, HC104 and HCl, at
concentrations of 0.01 M and 0.1 M. Though the rate of ECH was maximal at 90°C, the
70°C temperature was chosen in order to prevent excessive evaporation of water from the
system during the long run times. The experiments with 0.01 M electrolyte could be
carried out only for 21 hours because of the increase in resistance of the system due to

depletion of protons. The results are shown in figure 4.3.

120

From the results, it can be seen that there is considerable variation in the rates of
formation of lactaldehyde and propylene glycol in different electrolytes due to the
difference in the nature of the electrolyte anions and surface modification of the catalyst
by these anions. The final yields of the products are listed in table 4.2. The low material
balances may be due to the adsorption of the products in the carbon support of the
catalyst and the formation of volatile or polymerized products; volatiles such as methane
have been observed under severe conditions in the chemical hydrogenation of carboxylic

acids with 5% Ru/C.

121

 

l lactaldehyde
(0.1M H0104) l 1

+ propylene glycol 1

(0.1M H0104) 1
1 +lactaldehyde l l
(0.1M H2804) ' '

i ~—x—~ pmpylene glycol
1 (0.1M H2804) ’ l
l +lactaldehyde 1 ‘
(0.1M HCI) l
1 +propylene glycoli I
610 ~ (0.1M HCI)

% yield

 

 

 

 

 

 

 

 

‘ 90 pie—Tafctaldehyde (0.01M_ .
‘ HCIO4) 1;

8° * '* * * l-o—propylene glycol (0.01M H
70 1 A 44444 ~ ,- , , ‘ H0104)
, _ ,/ ' ee ; +lactaldehyde(0.01M 1.
xx---» , H2804) _‘

‘ Wx propylene glycol (0.01 M 11
w 7 L w . , , H2804) 11'

, + lactaldehyde (0.01 M H01) 1
l;
, HCI)

.-... ' ‘ ' :JI ‘ --- lactaldehyde (no

i l
10 15 20 25 LL . 9'99W2ww_ L _ _*
Time (h)

 

%yhm

 

 

 

 

 

 

 

Figure 4.3: Comparison of rates of ECH of lactic acid (11.1 mM) with 5% Ru/C/RVC in
various electrolytes at a concentration of 0.1 M (top) and 0.01 M (bottom) (T= 70°C, i=

100 mA). The error in HPLC is estimated to be i3%.

122

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Electrolyte Lactic acid Lactaldehyde 1,2- Material Current
(% propanediol balance efficiency
remaining)

Time = 48h

0.1M 91.8 2.6 2.7 97.1 7.48 X 10'7

H2804

0.1M HCl 8.9 64.0 6.6 79.5 7.17 X 10'15

0.1M 54.5 25.4 0.7 80.6 2.48 X 106

HC104
Time = 21h

0.01M 37.4 29.1 4.5 71.0 8.09 X 10"

H2S04

0.01M HCl 9.4 80.3 2.9 92.6 1.83 X 10'5

0.01M 0.2 58.6 1.1 59.9 1.29 X 10'5

HC104

 

 

 

 

 

 

 

 

Table 4.2: Final yields (%) of products after lactic acid (11.1 mM) electrohydrogenation

in various electrolytes (T= 70°C, i= 100 mA)

The overall current efficiency, 11p, for ECH may be calculated using the following

equation.74

2

(111' 1110)

 

Zi

 

1 t
—— 21.11
F to

where n1 and 11,0 represent the moles of product formed at the cathode at times t and to

respectively, zi is the number of electrons involved in the electrode reaction, and F is the

123

Faraday constant. The calculated current efficiencies in each electrolyte, calculated was
over the period of time between t0=0 h and t = 21 or 48 h are listed in table 4.2. Only
conversion to lactaldehyde and propylene glycol were considered in the calculations. It
can be seen that the current efficiencies are very low due to the long electrolysis time and
high current involved. The highest current efficiency of 1.83 X 10’5 was obtained in 0.01

M HCl.

In general, in case of both 0.1 M and 0.01 M electrolyte concentrations, the relative rate
of formation of lactaldehyde was found to decrease in the order, HCl>HClO4>H2SO4.
Formation of propylene glycol was generally slow, maximal in 0.1 M HCl and almost
negligible in 0.1 M HC104 due to an inhibitory effect of the C104' anion. In the case of
ECH of lactic acid in 0.01 M and 0.1M HCl, it was found that the rate of conversion of
lactic acid to lactaldehyde and propylene glycol was low initially and increased alter a
period of time. Apparently, as the electrohydrogenation proceeds, C12 evolution at the
anode decreases the Cl' concentration in solution, leading to the desorption of Cl from the
cathode surface and leaving more sites for the adsorption of lactic acid, hydrogen or other
active species. It is also possible that partial surface coverage by adsorbed Cl is beneficial
for the ECH reaction due to the electronegative Cl atoms modifying the electronic state of
the Ru atoms and enhancing adsorption of reactants. In order to test this hypothesis, an
experiment was conducted in which ECH of lactic acid was carried out in the absence of
an electrolyte. Since lactic acid is ionizable to some degree, it was expected to function as
the electrolyte. However, the resistence of the solution was very high and the ECH
current was reduced to half the value, i.e.50 mA. Theoretically, even at this current the

quantity of H atoms produced should be large enough to completely hydrogenate all the

124

 

lactic acid present in solution. As expected there was no propylene glycol formed in this
case due to the low concentration of protons in solution, leading to a low surface
coverage of H (figure 4.3). This is consistent with the requirement for a high surface
coverage of H for propylene glycol formation. Also there was no induction period like
that observed in HCl, supporting the idea that the initial low rate of ECH is due to Cl
adsorption. The rate of ECH during the later stages of the reaction was not as high as that
observed with HCl. From this it was concluded that the anomalous behavior observed

with HCl may be a result of the combination of two effects, namely, the partial

a.

desorption of Cl during the later stages of ECH and the electronic enhancement effect of

Cl partially covering the Ru electrode surface.

Earlier attempts at ECH of lactic acid using Ru/carbon felt and 10% Ru/Pt mesh
electrodes in 10 mM H2SO4 under similar conditions produced only lactaldehyde;
formation of propylene glycol was seen only when the electrohydrogenation was carried
out using 5% Ru/C catalyst agglomerated in the RVC working electrode. Evidently, the
hydrogenation of lactic acid to propylene glycol depends on specific characteristics of the
Ru/C catalyst such as the nature and distribution of pores and of the Ru nanoparticles in

these pores.

4.3.2 Anion adsorption on electrode surfaces

The adsorption of anions on electrode surfaces and their effects on hydrogen adsorption
on metal surfaces have been extensively studied. The relative strengths of adsorption of
the anions were found by Horanyi to decrease in the order C1‘ > HSO4' > ClO4‘ on Pt

electrode.75 In studies of chemical catalytic hydrogenation of lactic acid, Cl' was found to

125

inhibit hydrogenation, whereas an opposite effect is observed in ECH. This discrepancy
reflects the fundamental difference in the initial step of the two processes. In catalytic
hydrogenation, the first step in the process is the dissociation of the H2 molecule into
atoms, which is equivalent to the electrocatalytic oxidation of H2. In the case of ECH, the
initial step involves the cathodic reduction of protons in solution. Adsorption of anions
was found to inhibit the hydrogen oxidation process on Pt, but not the electroreduction
step.73 Therefore, the effect of anions depends on the source of H atoms, i.e. molecular H2
or protons. This difference was attributed to the involvement of overpotential deposited
OPD H atoms in ECH and not underpotential deposited UPD H atoms, which are the
ones whose adsorption is influenced by anion adsorption. OPD hydrogen atoms are
weakly bound on the cathode surface and have a certain degree of mobility, thus enabling
them to reach the active sites. However, in the case of ECH of lactic acid on Ru, the
adsorption of anions seems to be strong and accompanied by some amount of charge
transfer since they are not easily desorbed even in the hydrogen evolution region. The
increase in ECH activity at more negative potentials may also be attributed to the increase
in the mobility of OPD H atoms and a change in the electronic properties of the metal
surface due to cathodic polarization.73 In several studies, the adsorption of Cl' on the
electrode surface was found to shift the metal oxide formation peak in the cyclic
voltammogram to more positive potentials and the hydrogen adsorption peak to more
negative potentials.76 Therefore, Cl' adsorption was found to inhibit the formation of
metal oxides such as RuOH, resulting in increased hydrogen adsorption. Some Cl'
adsorption was assumed to occur even at more negative potentials of hydrogen

evolution. The enhancement effect of HCl towards ECH of lactic acid observed in our

126

studies may thus be attributed to the inhibition of Ru oxide formation. The adsorption of
Cl' does inhibit lactic acid conversion initially, but the effect is overcome at longer times
by the desorption of Cl' resulting from anodic oxidation of Cl' and reduced solution

concentration.

Using FTIR spectroscopy, HSO4' and C104' were found to adsorb on metals through three
of the O atoms in a C3v configuration.”79 Even though C104' is considered to be weakly
adsorbed on electrode surfaces in many cases, Colom and coworkers80 observed
electroreduction of ClO4' to Cl' on Ru electrodes as shown in scheme 4.2. The extent of
reduction was found to be dependent on the concentration of ClO4' , hydrogen adsorption
and temperature. The perchlorate reduction process requires the presence of adsorbed
hydrogen and is enhanced at higher temperatures. Therefore, despite the similar modes of
adsorption of perchlorate and sulfate, the enhancement of lactaldehyde formation with
perchlorate may to some extent be attributed to its decomposition to Cl' on the surface
followed by the desorption and anodic oxidation of Cl'. C12 evolution at the anode was

observed when HC104 was used as the electrolyte.

(0104')ad + 8Had———> 01' + 4H20

Scheme 4.2: Electrocatalytic reduction of perchlorate to chloride on Ru proposed by

Colom80

However, it is interesting to note that there is significant concentration dependent

inhibition of lactaldehyde and propylene glycol formation with 0.01 M and 0.1 M H2SO4

127

and HC104. Comparison of rates of product formation between the different electrolytes
reveals that 8042' has a significant detrimental effect towards lactaldehyde formation
whereas perchlorate significantly inhibits propylene glycol formation only at high
concentration. These results indicate a general adsorption of Cl' on all types of active
sites, whereas perchlorate and sulfate show preferential adsorption on specific active
sites. On the basis of in situ FTIR studies, Marinkovic and coworkers81 have reported that
bisulfate ions adsorb strongly on Ru(0001) surface due to the right geometry provided by
the hexagonal symmetry at potentials about 300 mV more negative than for adsorption on
Pt, whereas its adsorption on polycrystalline Ru is inhibited by surface oxide formation.
The potential at which anion adsorption on a metal surface occurs is determined by its
work function.82 A more detailed picture of the effect of anion adsorption on lactic acid
conversion can be obtained using in situ spectroscopic studies with Ru single crystal
electrodes in order to determine the exact mode of adsorption and the nature of surface

sites occupied by each type of anion.

4.3.3 Study of anion hydrogen adsorption on Ru using cyclic voltammetry

Shown in figures 4.4a and b are the cyclic voltammograms of a polycrystalline Ru
electrode (0.2 cm2) in the three electrolytes at 0.01 M concentration in the absence and
presence of lactic acid (11.1 mM) recorded at a scan rate of 10 mV/s. In all cases, the
cathodic hydrogen adsorption and anodic hydrogen oxidation/desorption peaks can be
observed. In figure 4.4a, in the case of HC104 and HCl, the H adsorption peaks overlap
with the hydrogen evolution current. Also, in the presence of lactic acid, the oxygen

evolution occurs at much lower potentials. However, in the case of H2SO4, in the absence

128

 

and presence of lactic acid, there is significant hysteresis, which can be attributed to the
differences in the properties of the surface caused by anion adsorption on Ru.83 The peak
currents are also much lower in the case of H2SO4. Unlike HC104 and HCl, there is a
delay in the onset of hydrogen desorption in H2SO4, indicating the formation of a stable
adlayer of H which is resistant to oxidation. All these results are consistent with
adsorption of sulfate on Ru at the H adsorption/desorption potential. There is no cathodic
peak after the hydrogen desorption peak during the oxidation sweep in 0.01 M HC104,
indicating that perchlorate reduction to chloride, which is known to be a slow and
irreversible process at room temperature80 does not occur at this scan rate and at this

concentration.

In the presence of lactic acid (figure 4.4b), the hydrogen adsorption peak currents
decrease in the case of H2SO4 and HC104 due to partial blocking of surface sites by lactic
acid adsorption, but remain constant in the case of HCl. This indicates that in HCl, in
spite of lactic acid adsorption, the surface coverage of adsorbed hydrogen remains
constant due to an enhancement of hydrogen adsorption by adsorbed Cl. Using
radiotracer analysis, Horanyi and coworkers84 reported strong adsorption of Cl on Ru in
the hydrogen adsorption/evolution region at negative potentials even with high surface
coverage of H and stated the possibility of Cl' ions influencing the adsorption of H and O
on Ru. They also showed much stronger adsorption of Cl' ions compared to C104' and
8042'. Though the cyclic voltammograms yield some information about the extent of
anion adsorption in the hydrogen adsorption region, they don’t reveal much about anion

adsorption/desorption in the more negative hydrogen evolution region, where ECH

129

actually occurs. ATR-SEIRAS studies of the ECH of lactic acid, described in chapter 6

give more detailed information about the effect of anion adsorption on ECH.

 

.
“—-u-;.£__

 

i(X10E-5 A)

‘ —— HC104 1 1
‘ ------- H2804 ‘1
: ': ';"J_C_.'_ J

 

 

 

 

 

E(mV)

Figure 4.4a: Cyclic voltammogram of a Ru polycrystalline electrode in various

electrolytes (0.01 M) (A= 0.2 cm2, 0= 10 mV/s)

130

 

 

 

 

7
'\
/ -.
. I
1 1‘? ,.-
. 5 . ' o
’ ' I
1 1 4‘. '
1 ‘2 1 ‘1 1’; __7rc‘1647 1
i 9 l 3" O, I
1 13 1 . I ,l 1 ----- H2304.
’ " I 2 "- ' / - - H01
1 § 1 *\ /.’I' — 4
* ,' ~ .. .{fu'x
____ 4:1..9L-i‘. -‘on __ ___T_.,:_fi . __f_._7_m__ _
-530 . _ 0 500 1000
,/
-2

 

Figure 4.4b: Cyclic voltammogram of a Ru polycrystalline electrode in various

electrolytes (0.01 M) in the presence of lactic acid (0.01 M) (A= 0.2 cm2, 0= 10 mV/s)

131

4.3.4 Hydrogen adsorption on metal surfaces

The activity of the metal in catalytic hydrogenation is thought to depend on the heat of
adsorption of H2 on the metal and the dissociation energy of the adsorbed molecular
hydrogen to atomic hydrogen.7| It also depends on the relative rates of hydrogen
evolution and hydrogenation. The supporting matrix can also modify the adsorption
energy of atomic hydrogen on the metal through the effect of the empty (1 orbital of the
matrix. These orbitals can increase or decrease the strength of the M-H bond and change
the activity of the catalyst. Since ECH occurs at the adlineation point, a higher
concentration of the organic molecule at the adlineation point, which can be achieved by
using a stronger adsorbing support, increases the rate of the reaction. The selectivity of
the metals and their supports towards the hydrogenation of specific functional groups
such as ketones, aldehydes, carboxylic acids, aromatic rings, nitro groups, etc. has been
extensively studied, but there is no physical theory to explain the selectivity on the basis
of molecular hydrogen energy of adsorption, organic molecule adsorption energy or it-
bond energy. Though Ru is known to be one of the least active metals for hydrogen
evolution and is thus expected to be efficient for hydrogenation, Menard and coworkers
in the study of the ECH of cyclohexanone71 found Ru to be less efficient for
hydrogenation compared to Rh and Pt, thus indicating that hydrogen transfer from the
metal to adsorbed organic molecules is the rate limiting step and an increase in the
organic molecule concentration at the catalyst-solution interface should increase the rate
of hydrogenation. This was proved by the improvement of the ECH rate on using a more

strongly adsorbing carbon support instead of alumina.

132

Electrocatalysis in the hydrogen evolution region is affected by the states and binding

energies of chemisorbed H. The kinetics of HER at any electrode are characterized by
specific values of exchange current density, i0 and fractional hydrogen coverage (0" in
the steady state of HER). The exchange current density is given by the following

equation.85

i0 = k0,. (1-9”) exp(a AGHO/RT) exp(-a Fn/RT)

where k is a constant, C A is the concentration of protons in solution, AG"0 is the standard
Gibbs free energy of chemisorption of H, a is the activation barrier symmetry factor, F is

the Faraday constant and n is the overpotential.

The exchange current density is determined by the work function of the metal, which in
turn is correlated with the MH chemisorption energy. The MH chemisorption energy is

derived from the Eley-Pauling relation shown below.86

DHM = (DMM + DHH)/2 + (XM — )CH)2

where D represents the respective diatomic bond dissociation energies in kcal mol'1 and x
the electronegativities of M and H. The (W — 3002 is an important factor relating log io to
the properties of the metal, since the repulsion between M-H dipoles leads to a coverage

dependent work firnction that determines the polarity of the M-H bond.

Gerisher87 showed that when log i0 values were plotted versus the standard Gibbs energy
of chemisorption of H, AGH0 for a series of metals, a volcano curve was obtained with the
active noble metals lying around the vertex and the other metals lying on either side of
the maximum, as shown in figure 4.5. The maximum corresponds to AG”0 = 0 and weak

or strong H chemisorption corresponds to positive and negative values of AGHo

133

respectively. The maximum in the volcano curve results from the enhancing effect of an

increase in the free energy of chemisorption, AGH0 and the retarding effect of an increase
in the surface coverage, 0H as AGHO increases. Though not shown in the figure, Ru might
be expected to appear near the maximum of the volcano curve, similar to the other Pt

group metals.

The appearance of Pt group metals at the top of the volcano curve is not consistent with
the fact that these metals have high negative values of free energies of H chemisorption.
It was proposed by Conway88 that the relevant AG”0 for Pt group metals is not the A630
for adsorption of H on the bare metal, but the lower AG“O for the OPD H, which is the
adsorbed intermediate in HER and is deposited among the UPD H atoms, which in turn
are deposited on the metal surface, thus proving the involvement of OPD H atoms in

HER and hydrogenation at the potential of HER.

134

 

5.1

-log i0

0

 

125

transition metals
(Ni,Co,Fe,Cu,Au

._
Kfivfi WW . f .1" v- v -2.-- r .

  
   

Platinum metals
(Re,Rh,lr)

Nb and Vb
o metals (Sn,Ga,Bi)

 

sp'metals
(Pb,Cd,Ti,ln)

_- 203 203 3 5
M-H bond strengthlkJ mor1

Figure 4.5: Volcano curve for electrocatalysis in the hydrogen evolution region (HER) in

terms of log i0 plotted as a function of heat of adsorption of H (from ref. 86)

4.3.5 ECH of other a-functionalized carboxylic acids

To further confirm the unexpected formation of lactaldehyde via ECH of lactic acid,

glycolic (2-hydroxyacetic) acid, the C2 analogue of lactic acid, was subjected to

electrohydrogenation in 10 mM HC104 at 70°C and 100 mA current for 21 h. The

products of electrohydrogenation were analyzed by HPLC-RI, ESl-MS and 1H NMR (see

Appendix). Almost complete conversion of glycolic acid to glycolic aldehyde was

observed (scheme 4.3). Similarly, glyceric acid was converted to glyceraldehyde. Both of

these products are well known and their identification and quantification is

135

straightforward. The amino acid, alanine which is an analog of lactic acid with an a—NH2

group in place of -—OH was also converted to the corresponding aldehyde.

 

 

 

0 ECH p
5 HO
HO\/U\OH \/‘
RuIC/RVC
glycolic acid Q'YCOHC aldehyde
0
O ECH I
HO 4 HO
1A0}, , j}
H O RuIC/RVC H O
glyceric acid glyceraldehyde
NH2 ECH ”“2
OH > I
Ru/C/RVC O
O
alanine 2-amin0 propanal

Scheme 4.3: ECH of lactic acid analogs using 5% Ru/C agglomerated in RVC (T= 70°C,
i= 100 mA, electrolyte: 0.01 M HCl). Complete conversion of the reactants was obtained,

but actual yields of the aldehydes were not determined.

4.3.6 Reason for lactaldehyde formation

The unexpected finding that lactaldehyde (2-hydroxypropanal), a proposed intermediate
in the path of lactic acid hydrogenation, is the major product instead of the expected
propylene glycol, which only appears as a minor part of the product mixture is a
thermodynamically unusual result which reflects the relative surface adsorption and

aqueous hydration properties of the starting materials and the two products.

136

Why does electrohydrogenation of lactic acid stop at lactaldehyde, rather than
progressing on to 1,2-propanediol as in the chemical reduction? The most likely reason is
that the lactic acid reduction leads directly to the hydrated form of lactaldehyde. Like
other a-hydroxy aldehydes, lactaldehyde strongly favors the hydrated form (scheme 4.1).
Under the relatively low temperature and acidic conditions, the dehydration required to
enable further hydrogenation to the diol would be slow, thus leading to the formation of
small quantities of propylene glycol. This notion is also supported by the fact that when
mandelic acid, an aromatic analog of lactic acid was subject to ECH under similar
conditions (scheme 4.3), the completely hydrogenated product, namely cyclohexyl
ethylene glycol was formed to an extent of approximately 50%, which is much higher
than the lactic acid to propylene glycol conversion (scheme 4.4). The intermediate,
cyclohexyl glycolic aldehyde, which was also present in the product mixture is expected
to be less hydrated than lactaldehyde due to the higher electron donating effect of the

cyclohexyl substituent.

 

OH 0H OH
OH ECH O/i\\ O/K/OH
> I
o Ru/C/RVC 0
mandelic acid 50% 50%

Scheme 4.4: ECH of mandelic acid

A similar selectivity to the corresponding aldehyde was observed by Yokoyarna and

coworkers89 in the vapor phase hydrogenation of benzoic acid using a Cr/Zr02 catalyst. In

137

this case, the carboxylic acid interacts strongly with the catalyst surface, displacing the
weakly bound aldehyde as soon as it is formed, preventing successive hydrogenation to
the alcohol and resulting in high aldehyde selectivity (96% at 98% conversion in the
Yokoyarna case). Studies of lactic acid, propylene glycol and glycerol binding on Ru
surfaces,90 show preferential lactic acid binding by at least 2 orders of magnitude. In
terms of chemical functionality, the hydrated lactaldehyde (1,1 ,2-pr0panetriol) is much
more like propylene glycol and glycerol than lactic acid, and its expected affinity for the
surface should be similarly weak. Thus, like Yokoyarna, we interpret the lactaldehyde
preference as evidence for the desorption rate exceeding the dehydration/hydrogenation

rate at these low temperatures.

It is also interesting to note that although no lactaldehyde formation was observed in the
Ru/C catalyzed aqueous phase chemical hydrogenation of lactic acid, Dumesic and
coworkers91 did observe lactaldehyde formation in the vapor phase hydrogenation of
lactic acid using a 10% Cu/SiO2 catalyst. The formation of the aldehyde was found to be
dependent on the equilibrium between the aldehyde and diol at a given hydrogen partial
pressure. The highest yield of lactaldehyde (22.2%) was obtained at a hydrogen partial
pressure of 0. 10 MPa at 473 K, and the yield was found to decrease with decreasing
temperature and increasing hydrogen partial pressure. Acetaldehyde formation by the
vapor phase hydrogenation of acetic acid over silica-supported copper catalyst was also
described by Dumesic and coworkers in an earlier report.92 In these cases, the formation
of aldehyde was attributed to the hydrogenation of intermediate surface acyl species.
Falomi and coworkers93 reported the Pd/C catalyzed hydrogenation of substituted amino

acids to aldehydes at atmospheric pressure in ethanol. The conversion involved the

138

preparation of an ester with 2-chloro-4,6-dimethoxyl [1,3,5] triazine followed by
hydrogenation. Reduction of the aldehyde to the alcohol was reported only at higher
pressures and longer reaction times. However, as with the mandelide studies, alcohol was
formed as the major product when aromatic carboxylic acids were used. These studies
reflect a dependence of the relative rates of product formation on the surface hydrogen

concentration, reaction conditions and the mode of adsorption of the substrate.

4.4 Conclusions

Electrocatalytic hydrogenation of lactic acid has been accomplished using 5% Ru/C
agglomerated in Reticulated Vitreous Carbon (RVC) under mild conditions of
temperature and pressure as compared to chemical catalytic hydrogenation. The major
product of electrocatalytic hydrogenation is lactaldehyde with small quantities of 1,2-
propanediol. This unexpected process can provide a clean, convenient and efficient
method to prepare lactaldehyde, a compound that is not commercially available and that
is difficult to obtain in pure form.

In the previous preparative and mechanistic studies25 of the Ru/C catalyzed
hydrogenation of lactic acid to 1,2-propanediol, lactaldehyde was proposed as an
intermediate, but was never detected. Using electrohydrogenation, lactaldehyde is the
primary lactic acid conversion product, supporting its proposed interrnediacy en route to
1,2-propanediol. Though this project was begun as a way to carry out lactic acid to
propylene glycol reduction under mild conditions, the discovery of a pathway that
directly reduces aqueous carboxylic acid to aldehyde and then stops is substantially more

interesting, in terms of both fundamental chemistry and practical process. The conversion

139

of amino acids to amino aldehydes, achieved in the case of alanine is an important
application, as amino aldehydes are used as chiral building blocks in the synthesis of

more complex molecules used in pharmaceuticals.

The fact that the product ratio of ECH depends on the electrolyte used is also intriguing
and can be exploited in future applications to control the selectivity of the reaction
towards specific products by using specific electrolytes or a combination of various
electrolytes. Study of ECH of lactic acid under various conditions showed trends similar
to those observed with Ru/carbon felt, i.e. increase in product yields with increase in
temperature and ECH current. A significant enhancement of the ECH rate was observed
in HCl electrolyte compared to that in H2SO4 and HC104, especially during the later
stages of the reaction. The results suggest that this effect is due to the desorption of Cl
from the electrode surface at longer times (as a result of C12 evolution at the anode and
decrease in the solution concentration of Cl') and also an electronic effect of Cl partially
covering the Ru electrode surface, which facilitates adsorption of reactants.

Using cyclic voltammetric studies on Ru polycrystalline electrode in various electrolytes,
it was found that sulfate is strongly adsorbed on the electrode surface at the potential of
hydrogen adsorption/desorption. Furthermore, lactic acid adsorption suppresses hydrogen
adsorption in H2SO4 and HC104, but not in HCl. Although these studies give us some
insights into the effect of electrolyte anion adsorption on the ECH of lactic acid, the
infrared spectroelectrochemical studies described in chapter 6 allowed us to study these

effects in more detail.

140

CHAPTER 5
BULK ELECTROCATALYTIC HYDROGENATION OF LACTIC ACID
USING RETICULATED VITREOUS CARBON AGGLOMERATED WITH
RUTHENIUM ON NON-CARBON SUPPORTS

5.1 Introduction
In chapter 4, ECH of lactic acid was studied using RVC agglomerated with 5% Ru/C, the
same catalyst that was used for the catalytic hydrogenation of lactic acid in the batch
reactor. Using ECH, a greener and milder process was achieved for the hydrogenation of
lactic acid with the products being lactaldehyde and small quantities of propylene glycol.
It is intriguing that conversion of lactic acid to propylene glycol was achieved only with
the 5% Ru/C powder catalyst. In earlier studies with electrodeposited Ru on carbon felt
as the cathode, lactaldehyde was the only product and the rate of ECH was much slower.
The reason for the better performance of the former method and the difference in product
distribution is not yet clear. Due to the dynamic nature of the catalyst agglomeration
process during the course of ECH, it is difficult to determine the exact quantity and
surface area of active catalyst at given time. However, the particle size (<2 nm) of Ru in
the 5% Ru/C catalyst is much smaller than that of the electrodeposited Ru on carbon felt,
so the dispersion of Ru on the carbon support is much higher, thus implying much higher
active surface area per gram of Ru in the commercial catalyst. Also nanometer sized
metal particles are known to have properties vastly different from bulk metal.
Furthermore, in the case of Ru/C catalyzed lactic acid hydrogenation, the role of carbon
as a support is believed to be merely to provide a high surface area for dispersion of Ru,

since the carbon support is highly porous with a BET surface area of ~600 mZ/g. Since

141

hydrogenation of lactic acid to propylene glycol was also achieved with Ru sponge in the
batch process, it was concluded that carbon does not play a chemically active role in the
catalytic reaction, though it is certainly capable of adsorbing considerable product and
feed, complicating carbon balance analyses.

In order to explore in more detail the role of the support material, ECH of lactic acid was
attempted with a similar RVC cathode, but using powder catalysts containing Ru
dispersed on non-carbon supports. The two support materials chosen were SiO2 and
BaSO4, both of which are relatively inert materials and are common catalyst supports for
various metals. Alumina is known to have a high concentration of basic sites, resulting in
a strong affinity for acids such as lactic acid and was therefore not used. The study used a
commercial 5% Ru/SiO2 catalyst and a 5% Ru/BaSO4 (not commercially available)
catalyst prepared in the lab by the wet impregnation method. Additionally, 5% Ru/SiO2
catalyst was also prepared by the wet impregnation method and the activity of the catalyst
prepared in-house was compared to that of the commercial catalyst.

Menard and coworkers in their study of the ECH of phenol67 and cyclohexanone71 to
cyclohexanol found that the rate of ECH was dependent on the nature of the support and
that a higher rate of ECH was obtained when a strongly adsorbing support such as carbon
was used, as opposed to A1203 or BaSO4. This was attributed to the support playing a
major role in the adsorption of the organic substrate and the hydrogenation occurring at
the adlineation point between the support and the hydrogen covered metal. The present
study of the ECH of lactic acid using various supported Ru catalysts offers similar insight

into the role of the support in lactic acid hydrogenation.

142

5.2 Experimental

5.2.1 Catalyst preparation

The support material,BaSO4 and the Ru precursor, RuCl3.xH2O were purchased from
Aldrich and were used without further purification. SiO2 was purchased from Silicycle
(Quebec, Canada) and 5% Ru/SiO2 obtained from Kaida Chemicals (Shaanxi, China). 5%
Ru/Si02 and 5% Ru/BaSO4 were also prepared using the wet impregnation method.192
To measure the incipient wetness, the support was first dried for 5 h at 100°C. Then
HPLC grade water was added to 5 g of dried support until the appearance of liquid phase.
The incipient wetness of the support is given by the maximum weight of water added just
before the appearance of the liquid phase. The incipient wetness of SiO2 and BaSO4 were
determined to be 1.56 g and 0.26 g of water per gram of catalyst, respectively.

For the preparation of 5% Ru/SiO2, 0.513 g. of RuC13.xH2O was dissolved in 7.8 g of
HPLC grade water and 5 g of SiO2 was added to the solution. The mixture was stirred for
5 min to ensure uniform distribution of Ru and then dried in a rotary evaporator for 2 h at
80°C. The catalyst was dried at room temperature overnight and then transferred to the
quartz tube reactor for reduction. The reactor was purged with argon for 2 h at room
temperature. Hydrogen gas was then passed through the catalyst at 30 mL/min, and the
temperature was ramped to 400°C at 2°C per minute. The reduction was carried out for 16
h and the catalyst was then cooled in a flow of argon.

For the preparation of 5% Ru/BaSO4, 1.026 g of RuCl3.xH2O was dissolved in minimum
quantity of HPLC grade water, and 10 g of BaSO4 was added to the solution. The rest of

the procedure was similar to that for the preparation of 5% Ru/SiO2 described above.

143

The SEM images of the catalysts are shown in figures 5.1, 5.2 and 5.4. The presence of

Ru was confirmed by the presence of the Ru L- peak in the EDS (figures 5.3 and 5.5).

5.2.2 Electrochemical cell setup

Electrohydrogenation studies of lactic acid in several aqueous electrolyte solutions were
carried out in the glass electrochemical cell described in chapter 3. It is a two
compartment cell with the anode and the cathode compartments separated by a glass frit
in order to prevent the diffusion of gaseous products. The reference electrode is Ag/AgCl
(E0: -0.35 V vs. SCE). The counter electrode is Pt wire. The Reticulated Vitreous Carbon
(RVC) working electrode (dimensions 4.0 X 3.0 X 0.5 cm), with a pore size of 100 pores
per inch (ppi) is connected to the external circuit via a copper wire, positioned to
minimize the surface area of Cu exposed to solution; electrochernistry on the copper
surface can thus be assumed to be negligible. A voltmeter is connected between the
reference and working electrode to measure the potential of the working electrode.
Controlled current was applied using a model 273 potentiostat/galvanostat from Princeton
Applied Research. The temperature of the system was controlled using a heating tape
wrapped around the cell and connected to a temperature controller from Omega. Samples
were withdrawn for analysis immediately after the cell was switched on and at intervals
during the electrohydrogenation.

The electrolytes were prepared by the dilution of the concentrated acids, namely 98%
sulfuric acid (CCI), 36.5% hydrochloric acid (CCI), and 70% perchloric acid (Spectrum)
with HPLC grade water. S-Lactic acid and propylene glycol were purchased from

Aldrich and were used without further purification.

144

 

Figure 5.1: SEM images of 5% Ru/SiO2 (commercial catalyst) (A) Secondary electron

image (B) Backscattered electron image

 

500m 500m

Figure 5.2: SEM images of 5% Ru/SiO2 prepared by wet impregnation method. Arrows

indicate Ru particles

145

3000-

2000-1

1000~

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I l
0 2 3 4
keV
25004
2000—
1500—
1000— o
500- . _
1' N M i 'L"
0 1 1 ‘11”' . k 1
0 1 2 3 4
keV

Figure 5.3: Energy Dispersive Spectra (EDS) of 5% Ru/SiO2: commercial catalyst (top);

in-house prepared catalyst (bottom)

146

 

2pm 211m

Figure 5.4: SEM images of 5% Ru/BaSO4 prepared by wet impregnation method,
showing both low magnification (top) and high magnification (bottom) images of two

different regions.

147

300‘“ "

200-

1001

 

 

 

 

 

 

 

 

 

 

*‘1

keV

Figure 5.5: Energy Dispersive Spectra (EDS) of 5%Ru/BaSO4

5.2.3 Electrocatalytic hydrogenation

Before each experiment, 1.0 g (dry weight) of the catalyst (5% Ru/SiO2 or 5%
Ru/BaSO4) was added to the electrolyte solution, and the solution was stirred for 2 hours
under cathodic polarization (I= 20-40 mA) in order to ensure proper agglomeration of the
catalyst in the pores of the RVC. The potential of the working electrode during this
process was in the range of -100 to -400 mV (vs. Ag/AgCl) depending on the electrolyte
and the temperature of the cell. The stirring rate was kept low in order to prevent the
agglomerated catalyst from being washed off the pores. After the agglomeration, 1 mL of
lactic acid solution in the respective electrolyte was added so that the final concentration

of lactic acid in solution was 0.1 wt% or 11.1 mM, and the total volume of the solution

148

was 75 mL. Samples were withdrawn for analysis immediately after addition of lactic

acid and at intervals during the electrohydrogenation.

5.2.4 Characterization of the electrode surface

The electrode surface was characterized using a J SM-6400 V scanning electron
microscope (J EOL, Ltd., Tokyo, Japan). Micrographs were recorded using secondary
electron mode generated with an accelerating voltage of 20 kV. The presence of Ru was
verified by energy dispersive X-ray (EDX) analysis (Noran Instruments, Inc.,
Middletown, WI) connected to the ISM-6400 V. SEM images were obtained using the

AnalySIS software (Software Imaging System Corp., Lakewood, CO).

5.2.5 Analysis of products

Reaction mixtures were filtered using Millipore syringe filters (0.25 pm) in order to
remove the catalyst, and then analyzed by HPLC with refractive index detection, using a
Biorad HPX87H ion exchange column at 65°C. The mobile phase used was 5 mM
H2SO4, run isocratically at a flow rate of 0.6 mL./min.

The method used for the analysis and quantification of reactant, lactic acid and products,

lactaldehyde and propylene glycol is similar to the one described in chapter 3.

5.3 Results and Discussion
5.3.1 Catalyst structure
Figure 5.1 shows the SEM images in the secondary electron (A) and backscattered

electron mode (B) of the commercial 5% Ru/SiO2 catalyst. In the backscattered electron

149

mode, the areas with higher atomic number elements appear brighter. The backscattered
electron image is therefore an atomic number image rather than a topographical image.
Here the Ru particles can be seen as bright spots, which correspond to the brighter areas
in the secondary electron image.

The characteristics of the silica support used in the commercial catalyst and the in-house

prepared catalyst are shown in table 5.1 (data provided by the manufacturer).

 

 

 

 

 

Commercial 5%Ru/Si02 In-house prepared 5%
Ru/SiO2
Particle size (pm) 100 40-63
Pore diameter (A) 100 60
Specific surface area 300 500
(mz/g)

 

 

 

 

Table 5.1: Comparison of the characteristics of the silica support in 5% Ru/SiO2 catalyst

Figure 5.2 shows the SEM of the 5% Ru/Si02 catalyst prepared by the wet impregnation
method. Some Ru particles can be seen as bright spots on the silica surface and are
indicated by arrows.

Figure 5.4 shows the SEM of the 5%Ru/BaSO4 catalyst prepared by the wet impregnation
method. In this case, the morphology of the deposits is very different from that of
Ru/SiO2. The Ru appears to be deposited in the form of islands or patches, which are
about an order of magnitude larger than the Ru particles on silica. The reason for this is
that due to the low porosity and specific surface area of BaSO4, all the Ru3+ from the

precursor solution is deposited only on the outer surface of the support particles, and this

150

results in much higher concentration of Ru in a smaller area. This leads to aggregation of
the particles and island formation. On the other hand, on the highly porous silica support,
the same metal loading leads to the formation of more highly dispersed and smaller sized

Ru particles.

5.3.2 Factors governing lactic acid conversion (5% Ru/SiO2)

ECH of lactic acid (LA) yields lactaldehyde (LAL) and/or propylene glycol (PG) under
all conditions studied. The reductive catalyst agglomeration process before substrate
addition was used to ensure that all Ru in contact with the RVC was in the metallic state

and saturated with hydrogen.

5.3.2.1 Temperature

Electrohydrogenation experiments with 11.1 mM lactic acid in 0.01 M HCl were carried
out at 25, 50 and 70°C for 21 h using a constant current of 40 mA. Lactaldehyde was the
only product detected at lower temperatures. At 70°C, a small amount of acrylic acid was
detected. The formation of acrylic acid may be explained by the dehydration of part of
the lactic acid on the acidic sites of the silica support at high temperature. The yields of
the product, lactaldehyde obtained at various temperatures are shown in table 5.2.

The trend here is quite different from those observed with Ru/carbon felt and with 5%
Ru/C/RVC. Since the material balances are low, it is clear that other volatile or
polymerized products may be formed. Therefore, the actual current efficiencies at various
temperatures cannot be compared, as other reactions such as dehydration, etc. are

occurring apart from hydrogenation, especially at higher temperatures. Some amount of

151

the reactant and product may be adsorbed on silica. No attempt was made to desorb and
quantify these materials. Unlike the case with Ru/carbon felt and 5% Ru/C/RVC,
significant conversion occurs even at room temperature, but at 50°C, the conversion to
lactaldehyde appears to be minimum. No specific conclusions can be drawn about the
relative rates of hydrogenation and hydrogen evolution at various temperatures with the

available data.

 

 

 

 

Temperature LA (% LAL (% yield) Material
(°C) remaining) balance(%)
25 29.3 28.8 58.1

50 59.1 1 .5 60.6

70 34.4 21 .8 56.2

 

 

 

 

 

 

Table 5.2: ECH of lactic acid using 5% Ru/SiO2/RVC at various temperatures (11.1 mM
lactic acid in 0.01 M HCl for 21 h, i= 40 mA). The error in the HPLC is estimated to be

i3%.

5.3.2.2 ECH Current

Electrohydrogenation of 11.1 mM lactic acid was carried out in 0.01 M HCl for 21 h at
70°C at various currents. The results are shown in table 5.3. The material balances are
quite low suggesting the formation of other products and adsorption on silica. However,

the highest conversion and yield of lactaldehyde was obtained at 100 mA.

152

 

 

 

 

Current (mA) LA (% LAL (% yield) Material
remaining) balance(%)

20 56.2 11.3 67.5

60 60.8 6.7 67.5

100 0.4 54.5 54.9

 

 

 

 

 

 

Table 5.3: ECH of lactic acid using 5% Ru/SiO2/RVC at various currents (11.1 mM
lactic acid in 0.01 M HCl for 21h at 70°C). The error in the HPLC is estimated to be

28%.

5.3.2.3 Nature of electrolyte

ECH of 11.1 mM lactic acid was carried out using 5% Ru/SiO2 agglomerated in RVC in
various electrolytes at 0.01 M and 0.1 M concentrations. The results are shown in figure
5.6. In both cases, lactaldehyde is the major product formed and no propylene glycol
formation was observed. Small amounts of acrylic acid were also formed (not shown in
the figure) due to dehydration of lactic acid on the acidic sites of silica. In the case‘of
both 0.01 M and 0.1 M electrolyte concentrations, the relative rate of formation of LAL
was found to decrease in the order, HCl>HClO4>H2SO4. In H2SO4, there is negligible
conversion of LA. In HC104, the rate of formation of lactaldehyde was somewhat slower
than the corresponding rate with 5% Ru/C/RVC.

Similar to the previous case, the initial induction period in HCl, followed by a rate

increase at longer times is also seen in this case.

153

 

120

 

 

 

 

 

 

 

 

 

 

 

 

100 ’ ’ ’_—_‘“
+LAL(0.01MH2804) 1
80 +LAL (0.01M H0104) 9.
1 % 60 - - = +LAL (0.01M H01) '
1 3 40 -...._.LA(0.01M H2804) 1
1 . 20 - +LA(0.01M HC104) ’.
_._ LA (0.01M H01)
0 LA war/w —- ’41
-20 30
120 ‘
100
80 _.:LAT'(611TH2_35477,
2 60 _ +LAL(0.1M H0104) 1]
i +LAL(O.1M H01) 1
a: 40 - +LA(0.1MH2$O4) ]
20 - + LA (0.1M HCIO4)
_._LA (0.1M H01)
0 L. L L- L LL
‘ I 61’
-20

Time (h)

Figure 5.6: Comparison of rates of ECH of lactic acid (11.1 mM) with 5% Ru/SiO2/RVC
in various electrolytes at a concentration of 0.01 M (top) and 0.1 M (bottom) (T= 70°C,

i= 100 mA). The error in HPLC is estimated to be :t3%.

154

For the purpose of comparison, ECH of lactic acid was carried out in 0.01 M and 0.1 M
HCl using 5% Ru/SiO2, prepared in-house by the wet impregnation method. The results
are shown in figure 5.7. Quite surprisingly, the results were significantly different from
those obtained with the commercial catalyst. Lactaldehyde formation was much slower
and in 0.1 M HCI, propylene glycol formation was observed at longer times (45 h) with
the yield being similar to that obtained with 5% Ru/C/RVC. The slower rate of
lactaldehyde formation may be attributed to the lower metal loading on the external silica
surface, in case of the in-house prepared catalyst, as seen in the EDS. Some Ru particles
may be distributed in the pores of the catalyst and are not accounted for in the EDS.
From table 5.1, it can be seen that the silica support used for the in-house prepared
catalyst has a smaller pore volume and higher specific surface area compared to the silica
support in the commercial catalyst. It is possible that this allows higher dispersion of the
deposited Ru nanoparticles, thus providing more surface area for hydrogen adsorption,

which in turn favors the formation of propylene glycol.

155

 

 

%yhm

 

 

 

10 20 30 4o 50 6F
.20

 

Time (h)

 

_.;"1_A (0.01M HCI) :
—-— LAL (0.01 M H01) '

+LA(0.1M HCI) 11

~ -4... LAL (0.1M H01) 1
+1=G(0.1M H01) "

 

 

Figure 5.7: ECH of lactic acid (11.1 mM, T= 70°C, i= 100 mA) in BC], with 5%

Ru/SiO2/RVC prepared in-house. The error in HPLC is estimated to be 28%.

5.3.3 Factors governing lactic acid conversion (5% Ru/BaSO4)

5.3.3.1 Temperature

Electrohydrogenation experiments with 11.1 mM lactic acid in 0.01 M H2SO4 were

carried out at 25, 50 and 70°C for 21 h using a constant current of 40 mA. The results are

shown in table 5.4. Lactaldehyde was the only product detected. The increase in the yield

of lactaldehyde was not much when the temperature was raised from 25 to 50°C, but on

increasing the temperature to 70°C, a much larger increase in yield was obtained.

The data obtained in this case reflect a greater increase in the rate of hydrogenation

compared to hydrogen evolution with increase in temperature.

156

 

 

 

 

Temperature LA (% LAL (% yield) Material
(°C) remaining) balance(%)
25 75 .9 1 .2 77.1

50 72.3 2.9 75.2

70 64.4 1 0. 8 75 .2

 

 

 

 

 

 

Table 5.4: ECH of lactic acid using 5% Ru/BaSO4/RVC at various currents (11.1 mM
lactic acid in 0.01 M H2SO4 for 21 h at 70°C). The error in the HPLC is estimated to be

:l:3%.

5.3.3.2 ECH Current

Electrohydrogenation of 11.1 mM lactic acid was carried out in 0.01 M H2804 for 21 h at
70°C at various currents. The results are shown in table 5.5. As expected, there is an
increase in the yield of lactaldehyde with increase in the ECH current, similar to the trend
observed with Ru/carbon felt and 5% Ru/C/RVC.

These results suggest that the rate of hydrogenation is dependent on the concentration of
hydrogen atoms on the catalyst surface at a given time, which increases with increasing
current. Thus, in spite of the increased rate of hydrogen evolution, the rate of

lactaldehyde formation is also higher at higher currents.

157

 

 

 

 

 

 

Current (mA) LA (% LAL (% yield) Material
remaining) balance(%)

20 94.4 0.9 95.3

60 75.8 11.4 87.2

100 54.0 27.0 81.0

 

 

 

 

 

 

Table 5.5: ECH of lactic acid using 5% Ru/BaSO4/RVC at various currents (11.1 mM
lactic acid in 0.01 M H2SO4 for 21 h at 70°C). The error in the HPLC is estimated to be

i3°/o.

5.3.3.3 Nature of electrolyte

ECH of 11.1 mM lactic acid was carried out using 5% Ru/BaSO4 agglomerated in RVC
in various electrolytes at 0.01 M and 0.1 M concentrations. The results are shown in
figure 5.8. In both cases, lactaldehyde is the major product formed and no propylene
glycol formation was observed.

In this case, the results are quite different from the ones obtained with other systems.
Similar rates of ECH were obtained in all the electrolytes at 0.01 M concentration.
However, at 0.1 M electrolyte concentration, ECH was strongly inhibited in H2804 and
HC104. In 0.1 M HCl, after an initial induction period, there is a significant increase in

the rate and complete conversion of LA is obtained after 48 h.

158

 

120

 

 

 

 

 

 

 

100
80 l + LAL—(0 OTMTI-ESTOI); 1.
1‘ § 60 - -—-— LAL (0.01M 110104) 1 g
. '3 + LAL (0.01M HCI) '
g 40 ~
——*~— LA (0.01M 112504)
20 r + LA (0.01M HCIO4) ,
0 * 1+LAQ-91MHE'1L-
-20 30 ‘
Time (h) '1
120 i
1 u ’_._ 010111 #125011)?
E +LAL(0.1MH0104) .1
1 g .+LAL(0.1MH01) :
«4»— LA (0.1M H2504) ;1
+ LA (0.1M H0104) ‘
+ LAG-110-1191. -
60

 

 

 

Time (11)

Figure 5.8: Comparison of rates of ECH of lactic acid (11.1 mM) with 5%
Ru/BaSO4/RVC in various electrolytes at a concentration of 0.01 M (top) and 0.1 M

(bottom) (T= 70°C, i= 100 mA). The error in HPLC is estimated to be 13%.

159

These results indicate that the nature of the support plays an important role in
determining the nature and distribution of products formed during ECH. It also supports
the conclusion that propylene glycol formation depends on certain specific characteristics
of the Ru/C catalyst such as the nature and distribution of pores and of the Ru
nanoparticles in these pores and not just on the nature of the metal electrocatalyst. The
supports used in this study differ both chemically and physically and therefore they differ

in the relative amounts of organic substrates and electrolyte anions adsorbed on the

 

surface. Silica has a porosity comparable to that of carbon, but barium sulfate has a very 1:
low porosity. This can be concluded from the far lower incipient wetness value of BaSO4,
compared to SiO2, indicating lower uptake of water by BaSO4. From the SEM images of
these catalysts, it is clear that the Ru particles are deposited mostly on the surface of the
support and are a few hundred nanometers in diameter. This is very different from the
case of 5% Ru/C where many of the Ru particles are present inside the pores and are less
than 10 nm in diameter.

From the results on 5% Ru/BaSO4, it is clear that there is a limiting electrolyte
concentration between 0.01 and 0.1 M at which most of the catalytic sites are blocked by
the anions and there is significant inhibition of ECH. Once again the desorption of Cl'
ions at longer times, as a result of C12 evolution at the anode and depletion of C1' in
solution, results in an increase in the rate of ECH. However, at 0.01 M concentration,
sulfate does not exert the inhibitory effect that is observed in the case of LAL formation
on Ru/C and Ru/SiO2. Therefore, support nature and porosity plays an important role in
determining the nature and amount of surface area available for the adsorption of the

various species, namely organic acid, hydrogen and anions.

160

5.4 Conclusions

ECH of lactic acid was carried out using RVC cathode agglomerated with 5% Ru/Si02
and 5% Ru/BaSO4, in order to study the effect of support on the rate and product
distribution of ECH. The results obtained in various electrolytes showed some general
trends similar to 5% Ru/C/RVC, but a few differences were observed. Lactaldehyde was
the only product in all cases except with 5% Ru/Si02 prepared in-house by wet
impregnation, where in spite of the low rate of ECH, some propylene glycol formation
was observed at longer times.

The rate of ECH was highest in HCl in all cases, whereas much stronger inhibition of
ECH was observed with 5% Ru/SiO2 in H2SO4 and HC104, compared to 5% Ru/C/RVC.
However, with 5% Ru/BaSO4, in 0.01 M electrolytes, the rate of ECH was similar in all
three electrolytes, indicating a limiting electrolyte concentration between 0.01 and 0.1 M
at which most of the catalytic sites are blocked by the anions. While the data reported
here are not sufficient to exactly determine the role of the support and the extent of
adsorption of various species (reactants, products, electrolyte anions, etc.) on the support,
it does indicate a dependence of the ECH of lactic acid on the chemical nature and

porosity of the support and the nature and distribution of Ru particles on the support.

161

CHAPTER 6
IN-SITU VIBRATIONAL SPECTROSCOPIC STUDIES OF

ELECTROCATALYTIC HYDROGENATION OF LACTIC ACID
6.1 Introduction
To date no studies have been carried out to probe the nature of the adsorbed species on
the surface of the Ru/C catalyst during lactic acid hydrogenation. Attenuated total
reflection (ATR)-IR, coupled with ECH is one of the best ways to study the mechanism
on the surface of the heterogeneous catalytic reaction described above.
Most of the surface analytical techniques are suited for ex situ analysis in which the solid
surface to be studied has to be removed from the solution, washed and subjected to ultra
high vacuum before analysis. Though such methods can provide some insight into the
morphology and the nature and distribution of chemical species on the surface, they may
completely change the surface characteristics of the solid. Examples of ex situ techniques
are X-ray Photoelectron spectroscopy (XPS), ultra violet photoelectron spectroscpopy
(UV PES), Scanning tunneling microscopy (STM), Atomic force microscopy (AF M),
Fourier transform infra red transmission and emission spectroscopies and Diffuse
reflectance Fourier transform infrared spectroscopy (DRIFT). Transmission spectroscopy
generally involves dispersal of the sample in an infrared transparent medium like KBr,
leading to difficulties due to absorption of water and undesired reactions with the infrared
transparent medium.

A number of studies of aromatic compounds like benzene,94 benzoquinonefis'g7

98-108 110

hydroquinone, pyridine carboxylic acids,'09 hydroxy pyridines, etc. coordinated to

electrode surfaces have been carried out during the past two decades by Soriaga and

162

coworkers. They used ultra high vacuum (UHV) techniques such as Electron Energy 1055
spectroscopy (EELS) and Auger electron spectroscopy to determine the intensities of the
vibrational modes, orientations and packing densities of the adsorbed molecules at
various applied potentials in order to determine potential dependent changes in the
structure of the adsorbed molecules. However in all these techniques, the electrode had to
be transferred after the electrochemical reaction from the electrolyte solution to the UHV
chamber for analysis of the interface. Shorter lived intermediates could therefore not be
detected. Moreover, in many cases adsorbed species on surfaces behave differently in
solution and in UHV.

Spectroelectrochemistry has been used for the past few decades to study electrochemical
interfaces in situ to get information on molecular structure which can be difficult to
obtain using other techniques. In situ IR-RAS (Infrared Reflection Absorption
Spectroscopy)l l "'34 and SNIFTIRS (Subtractively Normalized Interfacial Fourier

135-139

Transform Infrared Spectroscopy) studies have been reported for CO oxidation,1 1 "

113 1 119-122 1 123-125,130 126
9 9

electrooxidation of formic acid,1 '4" '8 methano ethano acetonitrile,

‘28 and amines'29 and the orientation of interfacial water

ethylene glycol,127 acetaldehyde
and electrolyte anions on electrode surfaces. '3 "'34 In SNIFTIRS, the solution background
is removed by minimizing the spectral drift, which involves co-adding sets of
interferograms acquired at a suitable pair of electrode materials and alternating the
potential periodically.140 But these techniques have inherent problems such as low
sensitivity and the inability of simple background subtraction to correct for changes in the

concentration of ionic species in the diffusion layer with change in potential. Therefore,

they have limited applications in the study of adsorbed species because in spite of the thin

163

layer configuration used, the spectra contain artifacts and features from solution species.
The higher surface sensitivity of ATR allows much higher signal to noise ratio by
averaging only a few interferograms aquired in relatively shorter time, which is
advantageous for real-time monitoring of surface reactions. However, in the external
reflection mode, the working electrode has to be pressed against the prism in order to
minimize the thickness of the electrolyte layer. This limits mass transport of chemical
species to and from the electrode surface and also increases the resistance in the cell.
Further, when used in the study of ECH which occurs in the hydrogen
adsorption/evolution region, the evolved H2 gas is trapped between the working electrode
and the cell window and distorts the spectral and electrochemical measurements.

In situ Raman spectroscopy can be used for monitoring electrogenerated reactive
intermediates and adsorbates on electrode surfaces as all molecules show some amount of
Raman scattering at characteristic wavelengths. This is especially useful for carbon
electrodes which do not support electromagnetic field enhancement that forms the basis
of SERS (Surface Enhanced Raman Spectroscopy). With the advent of highly advanced
spectrometers using sensitive detectors, Raman spectra of molecules can be obtained on
carbon electrode interfaces without the surface enhancement. This technique has been
used to follow molecular structural changes associated with electrochemical charge
transfer and to study the structure and behaviour of monolayers on carbon surfaces in air,

vacuum and electrolyte solutions.Ml

164

6.2 ATR-IR and the SEIRAS effect

ATR-IR (Attenuated total reflection- infrared) spectroscopy was developed by Harrick
and Fahrenfort. In this technique, the sample is placed in contact with an internal
reflection element (IRE) of high refractive index, either in the form of a thin fihn, liquid
or powder. Infrared radiation is focused onto the edge of the IRE, reflected through the
IRE and then directed to a suitable detector as shown in figure 6.1.142 The material of the
IRE should be chosen according to the refractive index of the sample to be studied.
Although complete internal reflection occurs at the sample/IRE interface, the IR beam 1..
penetrates a short distance into the sample, in the form of an evanescent wave. This wave
is perpendicular to the reflecting surface and its electric field amplitude decays
exponentially with distance from the surface. The IR radiation can be absorbed by
molecules close to the surface of the IRE, and an absorption spectrum characteristic of
the sample in contact with the IRE can be obtained. The penetration depth, tip of the wave
is defined as the distance required for the amplitude of the electric field to fall to e’1 of its

value at the surface. It is given by the following equation.

A
dp = 21m1[sin20-(n2/n1)2] ”2 (I)
where n; and n2 are the refractive indices of the IRE and sample respectively, )1 is the
wavelength of the IR beam and 0 is the angle of incidence with respect to the surface
normal.
The incident light passes through the higher refractive index IRE and is reflected at the

surface of the sample. The propagating light passing through the IRE forms a stationary

wave perpendicular to the reflecting surface. ATR can be carried out in single as well as

165

multiple reflection modes. The intensity of the propagating wave is attenuated with each
reflection due to absorption by the sample. The energy loss in the reflected wave is
termed as attenuated total reflectance. For N reflections, the total reflectance or the
reflected power, RN is expressed as follows.

RN=(1-ade)N (2)
where (1c is the effective layer thickness and 01 is the absorption co-efficient of the layer.

The reflection loss is proportional to the number of reflections.

sample

 

 

 

 

 

 

1 R
Figure 6.1: Schematic diagram of Attenuated Total Reflection (ATR) in the horizontal

configuration showing the important parameters for spectral acquisition and calculation

of penetration depth

Several examples exist in literature where ATR-IR has been used to study solid/liquid
interfaces both in organic and aqueous solvents, and valuable information has been

obtained not only about the nature of the bonds present on the surface, but also on the
exact mode of binding and the orientation of the species. This has been made possible

due to the development of Surface Enhanced Infrared Absorption Spectroscopy

166

(SEIRAS). This is based on the principle that there is an enhanced absorption of IR
radiation on surfaces by certain types of bonds, which leads to an increase in the intensity
of the bands in the IR spectrum. The signal enhancement is believed to arise due to an
enhanced electric field within the metal film, caused by the excitation of surface
plasmons associated with the metal and depends on the metal morphology, i.e size and
shape of particles and interparticle distance. The enhancement is restricted to molecules
very near the metal film, specifically adsorbed species, thus enabling the selective
characterization of adsorbed species. The enhancement factors found in SEIRAS are
usually in the range of 10-100, but in some cases can range up to 103.

The SEIRAS enhancement has been explained using both electromagnetic and chemical
theories and is believed to be analogous to SERS (Surface Enhanced Raman
Spectrometry).”°’143 Models described by Osawa140 and Chew144 have been used to
calculate enhancement factors. The island film has been modeled as a set of rotating
ellipsoids with the axis of rotation normal to the substrate surface, as shown in figure 6.2.
The enhancement factor is found to be proportional to the background absorption by the
metal, which in turn depends on the thickness of the metal filmm’145

The local electric field is normal to the local surfaces of the island. Therefore only
molecular vibrations that give dipole changes normal to the surface can be excited,
yielding the surface selection rule. This surface selection rule has been used in numerous
studies to determine the orientation of adsorbed molecules because the enhancement
factors and hence the intensities of the vibrational bands are proportional to cos20, where
0 is the angle of the oscillating dipole with the surface normal. Therefore, bands due to

vibrations parallel to the surface will not be observed in the SEIRAS spectrum. For

167

example, in the ATR-FTIR studies of adsorption of anthraquinone 2-carboxylic acid
(AQ-COOH)I40 on Ag electrode surfaces, compared with the transmission IR spectrum in
KBr pellet of the Na salt of AQ-COOH, it was observed that the COOsyrn mode and the
b2u modes of the aromatic ring are selectively enhanced, whereas the COOasym and the b3u
modes are absent in the surface spectrum. On the basis of the surface selection rule, it
was concluded that the molecule is oriented with the aromatic ring perpendicular to the
surface. On the other hand, the reduced species, 9,10-dihydroxy anthracene-Z-
carboxylate, shows only two weak bands at 780 and 757 cm"l due to the OH out-of-plane

bending modes, suggesting a flat orientation on the surface.

Infrared radiation

Adsorbed molecule

(Ed)

Electric

v...;._,.-,$Flb$llate ..

     

 

 

Figure 6.2: Electromagnetic enhancement effect of Surface enhanced infrared
absorption spectroscopy (SEIRAS) showing the metal particles of the island film
modeled as a set of rotating ellipsoids with the axis of rotation normal to the substrate

surface (from ref. 140) (E.- is the electric field of the incident radiation).

168

The vibrations of the adsorbed molecules excited by the enhanced local field also
polarize the metal island, resulting in the perturbation of the polarizability of the islands
at the frequencies of molecular vibrations. The image dipole induced in the island by
oscillating dipoles of adsorbed molecules may also contribute to the perturbations. This
phenomenon also explains the appearance or enhancement in some cases of totally
symmetric modes compared to antisymmetric modes in the surface spectra, for example,
in the case of p-nitro benzoic acid (PNBA) adsorption on Ag.I45
The chemical model of SEIRAS was proposed to explain the fact that chemisorbed
molecules exhibit larger enhancement than physisorbed molecules.I40 One possible
reason for this is that chemisorption aligns the surface molecules and preferentially
oriented molecules give larger intensities than randomly oriented molecules for vibrations
that give dipole changes normal to the surface. The increase of absorption co-efficient or
dipole moment of molecules upon chemisorption also contributes to the enhancement.
For example, the absorption co-efficient of CO increases by a factor of 4 by

chemisorption on Ag.I46 Griffiths147

suggested that vibrations of strongly polarizable
groups within a molecule give larger enhancements than other groups due to donor-
acceptor interactions with the metal surface, eg. the interactions between the nitro group
and metal in adsorbed nitrophenols on Cu and Ag films and O-H stretch enhancement on
metal bonding. However, the contribution of chemical effects is believed to be much
smaller than that of the electromagnetic effects. The chemical enhancement effect in
SEIRAS operates over a shorter range, usually one monolayer from the surface,

compared to the electromagnetic enhancement effect, which operates over several

monolayers, depending on the conditions. Devlin and ConstaniI48 suggested an

169

enhancement of IR absorption resulting from the coupling of vibrational modes with
charge transfer between the metal and adsorbate, eg. the charge transfer enhancement
observed for thiocyanateI49 and tetracyano ethylene anion radical.150 The total
enhancement is found to be roughly equal to the product of the enhancements due to each
of the three contributions, namely electromagnetic effect, chemical effect and molecular
orientation.

Hahn and coworkers!“ used SEIRAS to study the anodic oxidation of methane at noble
metal electrodes (Pt, Pd, Au, Ru and Rh) in order to explore the possibility of using
methane as a fuel in fuel cells. Using deuterated water solutions, adsorbed intermediates
such as CO and CH0 was detected with the final product being CO2. A mechanism
involving dissociative adsorption of CH4, followed by reaction with adsorbed water or —
OH was proposed.

The most remarkable application of SEIRAS is the study of methanol electrooxidation on
Pt and Pt-Ru electrodes in order to determine the role of Ru in reducing anode poisoning
by adsorbed CO. In 2001 , Watanabe and coworkers'52 proposed a mechanism for
methanol electrooxidation from results of ATR-SEIRAS. Adsorbed COL (linearly
adsorbed), COB (bridged) and formate were identified as intermediates of methanol
electrooxidation to CO2 at intermediate potentials, and CO and COOH were detected at
high potentials. In 2003, Osawa et al.‘53 detected using SEIRAS the fonnate species,
HCOO, giving rise to the C00 vibration around 1300 cm", which increased in intensity
on Pt when CO was oxidized, leading to increased methanol oxidation current. Based on
this finding, the dual path mechanism of methanol oxidation, involving both CO and non-

CO paths was established. In 2004, Yajima et al. '54 in their study of methanol

170

electrooxidation on Pt, Pt/Ru and Ru, detected water molecules adsorbed on Ru sites
giving rise to an O-H stretching vibration at 3610 cm’l and assigned water molecules as
the oxygen donor for CO oxidation, which resulted in removal of the poison from the
electrode surface and led to enhanced methanol oxidation current. Recently Okada and
coworkers15 5 observed water molecules on the surface of Pt in 0.1 M HC104, with the O-
H stretching frequency having two components at 3500 and 3000-3100 cm". These
characteristics changed by interaction with neighbouring adsorbed CO in the presence of
methanol giving rise to a new O-H stretching band at 3658 cm". In both cases, a linear
quantitative relationship was observed, with an increase in the formate band intensity at
1330 cm'1 being correlated with a simultaneous decrease in the C-0 and water band
intensities. This served as a proof that the water molecules adsorbed along with methanol
and co-existing with adsorbed CO intermediate are the species responsible for the
oxidation of the catalytic poison.

'56 studied ethanol electrooxidation on a Pt film electrode

Recently Adzic and coworkers
using ATR-SEIRAS. By following the behavior of acetic acid, one of the products of
ethanol oxidation, as a function of potential, it was found that acetate was strongly
adsorbed even at high potentials, thus blocking Pt sites and inhibiting ethanol oxidation.
The acetate was found to be derived mainly by direct oxidation of ethanol and partially
from adsorbed acetaldehyde at low potentials. The main reaction intermediates are
adsorbed acetaldehyde and acetyl, which formed adsorbed CO and acetic acid. The first
and second layer adsorbates on Au electrodes with increase in potential were elucidated

by Futamata157 with respect to the peak frequency of the electrolyte ions such as sulfate

and perochlorate. The constant peak frequency of the second layer ions were explained

171

by the fact that first layer adsorbates suppress the specific interaction of the second layer
adsorbates with the Au surface. Watanabe and coworkers158 used ATR-SEIRAS to study
the oxidation of adsorbed CO on a Pt-Fe alloy, the surface of which was covered by Pt
skin, and found that the linear CO had a much lower saturation coverage and higher
mobility on the Pt skin compared to pure Pt due to increased (1 vacancy on the Pt skin and
weakening of the Pt-C bond. An adsorbed HOOH species, which could not be observed
on pure Pt, was observed on the alloy above 0.6 V in the electrolyte solution in the
absence of CO. However, unlike the case with Pt/Ru, the Pt skin did not show high
activity for direct CO oxidation and hence it was concluded that the CO tolerance is due
to lower CO coverage and bond strength and not due to enhanced oxidation. Using ATR-
SEIRAS, Osawa and coworkers15 9"60 also elucidated the mechanism of galvanostatic
formic acid electrooxidation on Pt and the cause of the observed potential oscillations. By
following the band intensities of COL, formate and COB as a function of potential and by
using 13C labeled formic acid, adsorbed fonnate was identified as the intermediate of
oxidation of formic acid to CO2. In a different set of experiments,I61 the oxidation current
was found to correlate well with the forrnate band intensity in the electrooxidation of
formic acid on Pt, pointing to the involvement of formate as the intermediate and the
oxidation of formate to CO2 as the rate determining step.

Osawa and coworkers reported time-resolved ATR-SEIRAS studies of fumarate
adsorption on a Au(111) surface in 0.1 M HC104 solution.I62 By correlating the
appearance of the C00 symmetric stretch and the disappearance of the C=O stretching
bands with the change in potential, it was concluded that fumaric acid adsorbed on the

electrode surface through one of its two COOH groups. The absence of the C00

172

asymmetric stretching band indicates that the carboxylate group is symmetrically bonded
to the surface with a perpendicular orientation. Finally, surface enhanced IR with ATR
configuration was also used to study hydrogen adsorption on Pt at potentials under which
hydrogen gas evolution occurs.“”’'64 The kinetics and mechanism of the HER were
determined from the IR and electrochemical data, and it was determined that adsorbed H
atoms giving a Pt-H vibration around 2100 cm'I i.e. HQPD and not Hum), are the reaction

intermediates of HER. These studies demonstrate the monitoring of fast dynamics at

 

 

electrochemical interfaces using the ATR-SEIRAS technique at the same timescale as
that used in electrochemical measurements, but providing much more detailed
information about kinetics, dynamics and structural details.

In the hydrogenation reaction, our aim is to look at the surface coordinated species
without interference from other IR absorbing species like water which is present in large
excess. ATR is a technique where the passage of the IR radiation through bulk water can
be effectively avoided as the evanescent wave penetrates only small distances from the
surface of the internal reflection element allowing the detection of mainly adsorbed
species and species very close to the surface. This technique also provides high

sensitivity, no mass transport limitations and no distortion from hydrogen gas evolution.

6.3 Experimental

6.3.1 ATR cell configuration

The spectroelectrochemical cell used in the present work is shown in figure 6.3. It
consists of a hemispherical Si window, 25 mm in diameter, and a teflon holder, screwed

onto a stainless steel base plate. The glass ATR cell (about 20mL in volume) was

173

mounted on the base plate and sealed with an o-ring to prevent leaks. The surface area of
the IRE exposed to the solution was 3.14 cmz. The 0011 consists of a water jacket through
which water was circulated using a water circulator in order to maintain a constant
temperature. The counter electrode was a Pt wire, which was attached to the cell and the
Ag/AgCl reference electrode is connected through the salt bridge. The Si window coated
with a thin film of Ru on Au was used as the working electrode. Electrical contact with 1

the window was attained through a gold wire sealed to a teflon holder, which was

 

mounted in a glass tube. The gold wire was sealed to the glass tube at the top and the
entire assembly was screwed onto the cell. A constant potential was controlled using a

PAR 173 potentiostat.

6.3.2 Preparation of Au thin film

The Si window was polished with 0.01 um alumina powder for 10 min., rinsed with
milli-Q water, sonicated with acetone for 40 min. and in water for 40 min. The window
was then heated in piranha solution (1:1 H202 and H2804) at 80°C for 1 h in order to
remove surface oxides, cooled and rinsed with water.

The chemical deposition of Au on the Si Internal Reflection Element (IRE) was carried
out using the method described by Osawa and coworkers.‘65 The total reflecting plane
was immersed in 40% NH4F solution for 3 min to terminate the Si surface with hydrogen.
Deposition of Au was done at 60°C by adding a mixture of plating solution

and 2% HF onto the hydrogen terminated Si surface. The composition of the plating
solution was 0.015 M NaAuCl4.2H2O + 0.15 M Na2803 + 0.05 M Na2S2O3.5H2O + 0.05

M NH4C1.

174

W
R

Ag/AgCl 5'1: Au wrre b
reference electrode .11 glass t” e

teflon ring

Ill

 

<— water

salt bridge ~‘L ‘— Ar

 

solution jacketed cell

 

 

 

 

water <—
teflon holder

A (Pt counter electrode)
Au ring

' a“ \A base plate
Si window

 

 

1 1' ‘-j ;~. _.
H.111 '.‘. , ——u—<r
_,1:-.' ' - . . . .... A. or“ '
~73“ .. .'.'.-‘,—v].'f.‘ -, - . ,. ~-
. ,.'." '_ v‘ <
. “ml-mem

 

 

 

 

= W
'~ 33:?
I

 

   

evanescent wave

solution
\ Ru layer

Au layer
Si window

 

 

source detector
Figure 6.3: Schematic diagram of the cell used in the in-situ ATR-FTIR studies of the

ECH of lactic acid

175

 

The thickness of the Au deposit can be controlled by varying the deposition time. The
purpose of HF is to reduce the metal complex i.e. Au3+ to AuO. The reaction for Au
deposition is shown below.

Si°(s) + 6F'(aq) -—> SiF62'(aq) + 40'

AuC14' (aq) + 3e' —+ Au0(s) + 4Cl'(aq)

6.3.3 Deposition of Ru

Once theAu film was formed, the window was rinsed with milli-Q water and then
immersed in a solution of 1 mM RuCl3.xH2O in order to deposit an adlayer of Ru3+ on the
Au surface. Prior to the ECH experiments, the electroreduction of Ru3+ to Ru0 was

carried out in the spectroelectrochemical cell by applying a potential of 0.0 mV (vs.

Ag/AgCl).

6.3.4 Experimental setup

After the cell was assembled, the water circulator was turned on. Once the temperature
was stabilized at 30°C, about 19 mL of water was added to the cell and an IR spectrum
was recorded. Then the required amount of electrolyte was added to the water so that the
final electrolyte concentration was about 0.01 M. Argon gas was bubbled through the
solution for about 30 min. in order to remove oxygen and other dissolved gases. A
constant potential of 0.0 V (vs. Ag/AgCl) was applied for 10 min. to reduce the Ru3+
adlayer to metallic Ru. The working electrode was then polarized at -400 mV and an IR
spectrum was recorded. Then the required amount of lactic acid solution was added so

that the final concentration of lactic acid in the system was 0.01 M and this point marked

176

the beginning of ECH. The working electrode was maintained at a constant potential of -
400 mV and IR spectra were recorded every 8 min. for 20 h.

FTIR spectra were taken with a Nicolet 20SX/C FTIR spectrometer, equipped with a
liquid nitrogen cooled linearized narrow band MCT-B detector. The sample compartment
was purged using CO2 and H20 free air from a Puregas Heatless Dryer. The IR source
(Everglow) was focused at the electrode/electrolyte interface by passing it through the Si
window. The incident angle was 30° from the surface normal, the beam diameter was 8
mm and the area of the IRE exposed to the solution was 3.14 cmz. Each spectrum was
acquired by integrating 1000 interferograms with a resolution of 4 cm'l, to achieve the
required sensitivity. The acquisition time was 8 min. per spectrum. The spectra were
referenced against the spectrum of water or the spectrum of the electrolyte alone taken at
the same potential under identical conditions. All spectra are shown in absorbance units,
defined as A= -log (I/Io), where I and 10 are the intensities of the reflected radiation at the

sample and reference solutions respectively.

6.4 Results and Discussion

6.4.1 Optimum concentration of lactic acid

For the IR spectra to be selective to species adsorbed on or close to the surface, the
concentration of the lactic acid and electrolytes in solution should be optimum, so that
interference from solution species can be minimized. In order to calculate the optimum
concentration, the penetration depth, tip of the IR beam needs to be determined from
equation 1. Assuming 0= 30°, X= 5.88 pm (1700 cm"), n1= 3.4 and n2: 1.33, the

penetration depth was calculated to be about 800 nm. Assuming the lactic acid molecular

177

diameter to be 0.5 nm and the molecular cross section to be 0.2 nmz, the monolayer
coverage of lactic acid was calculated to be 5 X 1014 molecules/cm2 or 8.3 X 10-10
moles/cmz. Considering the adsorbed lactic acid to be the only lactic acid present in the
cylindrical volume enclosed by the penetration depth, the required solution concentration
of lactic acid needs to be approximately 0.01 M. This would ensure that the IR spectra
recorded are characteristic only of the adsorbed layer or species present in the electrical

double layer.

6.4.2 IR spectra of electrolytes
The IR spectra of the three electrolytes at 0.01 M concentration, recorded at -400 mV are
shown in figure 6.4. The IR spectrum of water taken under similar conditions (without
the cathodic polarization) was used as reference. In all cases, the bands below 900 cm'1
cannot be clearly distinguished due to noise caused by absorption of IR radiation by Si in
that region. Using cyclic voltammetry and ATR-SEIRAS studies of CO adsorption and a
metal deposition technique similar to that used in this study, Osawa193 demonstrated the
the formation of metal overlayers that are almost pinhole free and having electrochemical
and surface properties similar to those of the corresponding bulk metals with apparent
surface enhancement factors relative to the smooth metal surfaces, in the range of 20-70.
Therefore, the observed spectral features may be attributed to adsorbed species on Ru and
adsorption on Au may be assumed to be negligible.
The spectrum of 0.01 M H2SO4 consists of bands at 1110 (not shown), 1617 and 3190
183,190 -

cm". The 11100m'l peak is due to the S-O stretching vibration, rndicating the

presence of adsorbed sulfate on the electrode surface. The 1617 and 3190 cm'1 peaks are

178

 

due to the H-O-H bending and O-H stretching vibrations of adsorbed water molecules.
These bands indicate the presence of water molecules co-adsorbed with sulfate ions on
the electrode surface. Isolated or non-hydrogen bonded water molecules show O-H
stretching frequencies around 3600 cm".184 The shift in the O-H stretch frequency from
this value is indicative of the extent of hydrogen bonding. The H-O-H bending fi'equency
of water adsorbed on Ru(001) surface in UHV has been reported to be in the range of
1560-1520 cm'l. '66’167 Therefore, a downward shift of the O-H stretch frequency to 3190
cm’1 and an upward shift of the 6H-O-H mode to 1617 cm'1 is indicative of strongly (
hydrogen bonded water.183 A similar phenomenon was observed by Osawa and
coworkers in the study of interfacial water molecules on Au electrodes!“183 Since
sulfate is an effective hydrogen bond acceptor, it is usually co-adsorbed with water
molecules on the electrode surface. A 6H-O-H frequency of 1617 cm'1 is also indicative
of interaction of water with the metal surface via an oxygen lone pair orbital.190 The
frequency of this mode is higher in this case compared to that obtained under UHV due to
the decrease in the strength of interaction with the metal in the presence of hydrogen
bonding.

The spectrum of 0.01 M HCl consists of strong negative bands at 1647 and 3380 cm".
This indicates displacement of the adsorbed water molecules from the electrode due to
chloride adsorption. The spectrum of 0.01 M HC104 also consists of negative bands at
1600 and 3350 cm’l due to partial displacement of adsorbed water by ClO4' ions. These
results are consistent with those of Osawa and coworkers,'68 who reported stronger
adsorption of chloride anion, compared to perchlorate anion on a Au(111) surface. This

was confirmed in their studies by a shift in the peak intensity versus potential profile to

179

the negative direction on adding 1 mM HCl to 0.5 M HC104 due to the displacement of
perchlorate ions from the electrode surface by chloride ions. Since perchlorate and
chloride ions are poor hydrogen bond acceptors, the co-adsorbed ion disrupts the
hydrogen bonding between interfacial water molecules and either partially or completely
displaces them.“59’”0’188 It is believed that in the case of perchlorate ions, the observed IR
bands are not due to water adsorbed directly on the electrode surface, but are due to the
water molecules present in the hydration shell of the adsorbed perchlorate ions,
associated with the anions through electrostatic interactions, thus resulting in non-
hydrogen bonded or asymmetrically hydrogen bonded water molecules!“189 In our case,
we only observe negative bands due to displacement of adsorbed water from the surface.
No bands are observed due to non-hydro gen bonded or asymmetrically hydrogen-bonded
water molecules. These observations serve as evidence that the IR spectra recorded are
mainly due to adsorbed molecules and ions or species present very close to the electrode
surface in the electrical double layer and are not due to solution species as in the latter
case, the spectra would be dominated by strong water bands even in the presence of
electrolytes.

The band at 1110 cm'1 due to the Cl-O stretch indicates the adsorption of ClO4' on the
electrode surface in the C3v or lower configuration, i.e. with only the Cl-O vibration
perpendicular to the surface being IR active. There exists some evidence in literature
showing that the perchlorate anion adsorbs on the Au(1 l l) electrode surface via three 0
atoms on a threefold hollow site, due to the symmetry of arrangement of the surface
atoms on the Au(1 1 1) surface. Sawatari and coworkers77 studied perchlorate adsorption

on Pt(111) surface using IR-RAS and concluded that perchlorate is adsorbed through one

180

or three 0 atoms with C3V symmetry. The coordination mode of adsorbed perchlorate
may be deduced from the vibrational spectra on the basis of group theory, though the
process involves some amount of ambiguity. The free perchlorate ion has Td symmetry,
which makes it IR inactive because there is no change in the overall dipole moment of the
ion on vibration of the Cl-O bonds. However, upon adsorption on a surface, the
perchlorate anion assumes C3v or C2v or lower symmetry, depending on the orientation,
adsorption site and the nature of the surface. Due to the reduction in symmetry, the triply
degenerate Cl-O stretching mode splits into two for C 3v symmetry and three for C2" or
lower symmetry. According to the surface selection rule, in case of the C3V and C2V
symmetries, only the mode which gives a dipole change normal to the surface is IR
active. Therefore, the presence of a single band at 1110 cm", assigned to the symmetric
Cl-O stretching mode indicates a C2., or C 3v symmetry for the adsorbed perchlorate.
However, infrared spectra alone do not enable us to distinguish between the C2" and C3v
modes. The Cl-O stretching frequency observed here is very close to that of the
symmetric Cl-O stretching modes of perchlorate ions with monodentate or bidentate co-

ordination with metal ions (1060-1030 cm")

181

—— Electrolyte
Abs-[02 --------- 8 min.

1200 min.

.. 0.01MH2SO4

  

 

.......... ,1)
In lily-M
0.01M HCI
‘ «M 0_01M HCIO4

. U
- . "_~‘~‘I‘."\,

......

1"Mk/yfftrrt M

  

 

I 3600 ' 3200 1 2500 . 2400 ' 2000 T 1600 Y 1200 '
-1
Wavenumbers (cm )

4000

Figure 6.4: ATR-IR spectra of the electrolyte solutions, obtained using the spectrum of
water as reference (T= 30°C, E= -400 mV) (solid lines) and ATR-IR spectra taken during
the ECH of lactic acid with the water spectrum as the reference (T= 30°C, E= -400 mV)

(dotted lines)

182

.n

The mode of adsorption of water and electrolyte anions as a function of potential have
been studied by a number of researchers using spectroscopy168 and theoretical models.”"
“9’187 It is believed that at potentials below the pzc (potential of zero charge) of the metal,
water molecules are adsorbed on the surface via an oxygen lone pair and are oriented
with the hydrogen atoms closer to the surface than the oxygen atom (mode A in figure
6.5). Around the pzc, the water molecules are oriented flat on the electrode surface and
no hands are observed in the IR spectrum due to the surface selection rule (mode B). At
potentials above the pzc, the hydrogen atoms of water are oriented towards the solution
and are hydrogen bonded with the second water layer, forming an ice-like structure
(mode C).178 This ice-like structure is disrupted at more positive potential due to anion
adsorption. The various possible modes of adsorption of water are shown in figure 6.5.
By comparison with the IR spectra of adsorbed water at various potentials reported by
Osawa,168 the ice-like structure and parallel orientations may be ruled out. On the basis of
similarity in band shapes and the cathodic polarization used, the spectra can be attributed
to water molecules adsorbed on the Ru surface through the oxygen atom with the

hydrogen atoms oriented towards the surface (mode A).'9°"9'

183

 

electrode

............
............
A B C surface
0 t
. a om l

Hamm
covalent bond

 

................... hydrogen bond

Figure 6.5: Schematic diagram showing different modes of adsorption of water

molecules on the electrode surface at various potentials

6.4.3 In-situ FTIR study of the ECH of lactic acid

6.4.3.1 ECH of lactic acid in 0.01 M H2SO4

The infrared spectra of lactic acid solutions recorded at various times at 30°C and -400
mV are shown in figure 6.6. The spectrum below 1200 cm'l is not shown due to strong
absorption by Si in the range of 1150 to 1100 cm".

In order to obtain useful information from infrared spectra, a suitable reference spectrum
should be chosen. In many cases, particularly in in situ FTIR studies of methanol
electrooxidation, the reference spectrum is the one taken at a high positive potential
where most of the surface bound species (CH3OH, C O and intermediates) are completely
oxidized to gaseous CO2. In case of ECH of lactic acid, the organic acid is known to

adsorb strongly on the electrode surface and is very stable to oxidation. Therefore, the

184

reference spectrum cannot be taken in the presence of lactic acid, as this would not allow

us to distinguish features from surface bound lactic acid and the changes over time.

Therefore, all spectra were referenced against the spectra of the respective electrolyte

solutions taken under identical conditions, unless otherwise mentioned.

The initial spectrum taken immediately after addition of lactic acid to the electrolyte

shows bands at 3610, 1630, 1435, 1357 and 1283 cm'l and a negative band at 3168 cm". .1

The weak band at 1712 cm"l has been assigned to the doubly degenerate H-O-H bending

 

mode of adsorbed H3O+ ions.'8°"33'I90 The O—H stretching bands of H3O+ are obscured by
the O-H bands of water. Osawa and coworkers observed a similar peak in their study of
interfacial water at various potentials. '68 On the basis of this 1712 cm'1 band and the
absence of the band due to the symmetric H-O-H bending mode, a structure for adsorbed
H3O+ was proposed in which the C3v axis is nearly parallel to the surface, as shown in
figure 6.7. Since at potentials below the pzc, water molecules are adsorbed with the
oxygen lone pair directed towards the solution, the hydronium ion with this orientation is
readily formed by addition of a proton to the oxygen lone pair without disturbing the

orientation of the other two H atoms.

1200 min.

Ab
{[002 1000 min.

800 min.
600 min.
400 min.

200 min.
32 min.
- - __ 24 min.

16min.
8 min.

 

 

 

 

 

V

f I

T ' I I ' l ' l ' I ' l
4000 3600 3200 2800 2400 2000 1600 1200

Wavenumbers (cm'1)

Figure 6.6: Real time ATR-IR spectra of the EC H of lactic acid in 0.01 M H2SO4

(T=30°C, E= -400 mV) using the spectrum of 0.01 M H2SO4 taken under identical

conditions as the reference

186

. O atom

 

 

Hamm
electrode _ covalent bond
surface hydrogen bond

Figure 6.7: Schematic diagram showing the mode of adsorption of hydronium ion on the

electrode surface

The 1283, 1357 and 1435 cm" bands are due to the alcoholic 00 stretch, O-H
deformation and the carboxylate symmetric stretch respectively.185 The spectra clearly
show strong bands due to the C00 symmetric stretch and the alcoholic C-O stretch,
indicating that lactic acid is adsorbed on Ru in the chelating bidentate mode. Therefore,
the (ll-hydroxy group is not involved in bonding to the metal catalyst. Shown in figure 6.8.
are some possible modes of adsorption of lactic acid on the Ru surface. The monodentate
and the dissociative adsorption modes may be ruled out as these modes would result in
much stronger C=O stretching bands. The weak band observed around 1710 cm'l may be
attributed to the C=O stretch of solution phase lactic acid.

The spectra show strong symmetric stretching (1435 cm") and weak asymmetric
stretching bands. These observations can be explained by the surface selection rule and
strongly support the chelating bidentate mode of adsorption and the fact that spectra show
predominantly adsorbed lactic acid species, as SEIRAS is sensitive to surface species and
less sensitive to solution species. The transition dipole of the symmetric COO vibration is
parallel to the diagonal of the carboxylate group and when lactic acid is co-ordinated to

the metal surface through the carboxylate group, has a strong component in the direction

187

normal to the surface as shown in figure 6.9, which explains the high intensity of this
band. In contrast, the direction of the transition dipole of the C00 asymmetric stretch is
along the line joining the two oxygen atoms, which in the adsorbed molecule is parallel to
the surface and hence this vibration is IR inactive. The O-H deformation mode also gives
a weak band at 1357 cm'1 due to this bond being parallel to the surface when lactic acid is
adsorbed in the chelating bidentate orientation.

The negative band at 3168 cm'1 indicates the displacement of strongly hydrogen bonded
water molecules from the surface by lactic acid. However, the bands at 1630 and 3610
cm" are due to the H-O-H bending and O-H stretching vibrations of non-hydrogen
bonded water molecules. Therefore, it can be concluded that the adsorption of lactic acid
on the electrode surface is associated with the presence of relatively more isolated water
molecules on the surface. The O-H stretching vibrations of lactic acid cannot be
distinguished from those of water and this band is believed to overlap with the 3610 cm'1
peak from water. The close-packed monolayer of lactic acid adsorbed on Ru through the
carboxylate groups is also expected to have -—O-H groups, which are isolated from one-
another and from water molecules in the vicinity. forming very few hydrogen bonds, thus
giving the high frequency band around 3610 cm".

As the reaction progresses, several changes are observed in the spectra. The peak at 3168
cm'1 due to the O-H stretching vibration of strongly hydrogen bonded water reappears
and the peaks at 1435, 1357 and 1283 cm'1 from adsorbed lactic acid decrease in
intensity. This indicates that as the reaction progresses, the Ru surface is poisoned and
this results in the decrease in the adsorption of lactic acid. The poison may be associated

with sulfate or S containing species derived from the sulfate ions of the electrolyte, as the

188

decrease in adsorption of lactic acid is associated with the appearance of strongly
hydrogen bonded water near the electrode surface, which has been shown to be
associated with the hydrophilic sulfate anions. This explains the lower rate of ECH of

lactic acid in H2SO4 compared to HCl and HC104.

HOH OIOH HO (13H

\
0 o
(1) 9 RM ' R"
RU Ru RU
monodentate chelating bidentate dissociative adsorption

Figure 6.8: Possible modes of adsorption of lactic acid on a metal surface

COO symmetric COO asymmetric

 

 

 

 

stretch stretch
2
R- Resultant dipole
o—H o—H
33‘ R \
o\-—~ o 03- 0
RM 1/
V ‘ R
’ x

Plane of electrode surface

3’

Figure 6.9: Symmetric and asymmetric stretching modes of the carboxylate group and

their resultant transition dipoles

189

 

6.4.3.2 ECH of lactic acid in 0.01 M HCl

The infrared spectra of lactic acid recorded at various times at 30°C and -400 mV are
shown in figure 6.10. The bands due to lactic acid are similar in position and intensity to
those observed with H2804. The initial spectrum taken immediately after addition of
lactic acid to the electrolyte shows bands at 1435, 1357 and 1283 cm". The weak band at
1712 cm'I has been assigned to the doubly degenerate H-O-H bending mode of adsorbed
hydronium (H3OI) ions. The 1283, 1357 and 1435 cm'1 bands are due to the alcoholic C-
O stretch, O-H deformation and the carboxylate symmetric stretch respectively. The
presence of strong bands due to the C00 symmetric stretch and the alcoholic 00 stretch
and a weak band due to O-H deformation indicate that lactic acid is adsorbed on Ru in
the chelating bidentate mode.

The O-H stretching bands from water and lactic acid and H-O-H bending bands from
water appear at 3120 and 1595 cm'1 respectively. Therefore, the lactic acid may be
associated with strongly hydrogen-bonded water. However, in this case there is no
decrease in intensity with time of bands due to lactic acid. Therefore, the Ru surface is

apparently not poisoned by electrolyte anions in this case.

190

Abs 0.02 1200 min.
M 1000 min.
800 min.
600 min.

400 min.

200 min.

40 min.

32 min.

24 min.

16min.

8 min.

 

U V I I U I V U

l T l l l
2800 2400 2000 1600 1200

Wavenumbers (cm")

1 l l
4000 3600 3200

Figure 6.10: Real time ATR-IR spectra of the EC H of lactic acid in 0.01 M HCl
(T=30°C, E: -400 mV) using the spectrum of 0.01 M HCl taken under identical

conditions as the reference

Figure 6.4 shows IR spectra taken during the ECH of lactic acid using the water spectrum
as the reference. Therefore, the spectra are characteristic of both lactic acid and
electrolytes. It can be seen that the spectra are dominated by peaks associated with
electrolytes, indicating strong adsorption of electrolyte even in the presence of lactic acid.
However, in the case of 0.01 M HCl, a comparison of the relative intensities of the

negative going water bands at the beginning and end of the reaction indicates a slight

191

decrease in the intensity of the bands probably associated with desorption of some
amount of chloride from the surface, followed by re-adsorption of water molecules, as a
result of a decrease in the solution concentration of chloride as C12 is evolved at the
anode.

It is also interesting to note the appearance of new bands at 3400 and 1647 cm’1 at around
200 min, associated with O-H stretching and H-O-H bending vibrations of water in figure
6.9. The appearance of these bands corresponds approximately to the increase in the rate
of ECH of lactic acid after an initial slower phase, observed while using HCl as the
electrolyte with 5% Ru/C agglomerated in RVC.. This suggests that the partial desorption
of chloride frees up sites allowing the adsorption of water molecules, which are required
for ECH. From the frequencies of the water bands, they appear to be relatively isolated or
poorly hydrogen bonded, which may be a result of association with the Ru surface and
the hydrophobic nature of the co-adsorbed chloride. These observations are consistent
with the conclusion that adsorbed water plays an important role in lactic acid
hydrogenation, derived from the experiments indicating a decrease in the rate of
hydrogenation of lactic acid in organic solvent/water mixtures”. From ab initio studies of
the water layer on Pt(111) and Rh(l 11) surfaces, Vassilev and coworkers186 discovered
that water molecules adsorbed on a metal surface possess a small metal-like conductivity
resulting from the mixing of local states of the oxygen atom with the electron bands of
the metal substrate. The extent of mixing of the atomic states from oxygen and metal was
found to depend on the orientation of the water molecules, the distance from the surface
and the polarization of the surface. The analysis of the electronic states indicated an

electron charge transfer from the water molecules towards the surface, resulting in a

 

reduction of the work function of the system. This phenomenon may be responsible for
facilitating the electron transfer to protons or H 30+ to form H atoms, which are required
for ECH. In the case of H2S04 and HCl electrolytes, a weak feature due to H3O+ was
observed in the IR spectra. However, the peak corresponding to the M-H bond, known to

'1 163’!“ was not observed in any of the spectra, indicating that

appear around 21000m
either the surface concentration of adsorbed H is too low to be detected by IR or it is a
very short-lived species, with lifetime much shorter than the timescale of the scans.

The cyclic voltammetry results described in chapter 4 indicate that adsorbed Cl on the
electrode surface when HCl is used as the electrolyte is favorable for H adsorption.
Therefore, the enhancement of the ECH rate in HCl is probably a result of the electronic
effects of both adsorbed chloride and water. The specific sites occupied by the water

molecules and the exact configuration required for involvement in the hydrogenation rate

enhancement is not yet clear.

6.4.3.3 ECH of lactic acid in 0.01 M HC104

The infrared spectra of lactic acid recorded at various times at 30°C and -400 mV are
shown in figure 6.11. The bands due to lactic acid are similar in position and intensity to
those observed with H2S04 and HCl. The initial spectrum taken immediately after
addition of lactic acid to the electrolyte shows bands at 1435, 1340 and 1275 cm".

The 1275, 1340 and 1435 cm" bands can be assigned to the alcoholic 00 stretch, O-H
deformation and the carboxylate symmetric stretch respectively. The presence of strong

bands due to the C00 symmetric stretch and the alcoholic C-O stretch and a weak band

 

due to O-H deformation indicate that lactic acid is adsorbed on Ru in the chelating
bidentate mode.

The O-H stretching bands from water and lactic acid and H-O-H bending bands fi'om
water appear at 3400 and 1625 cm‘| respectively. There is a slight decrease in the
intensity of the lactic acid bands with time at 1000 min., indicating poisoning of the
surface by perchlorate or chloride, but the decrease in intensity is not as pronounced as in
the case of H2SO4 and occurs at much longer time compared to the H2SO4 case.

Since in all the three electrolytes, lactic acid was found to adsorb on the Ru surface in the

 

chelating bidentate mode without involving the a-hydroxy groups, it can be concluded
that the (ll-hydroxy group does not facilitate the hydrogenation of lactic acid (and other a-
hydroxy acids) by forming a more stable adsorbed intermediate or by influencing the
initial adsorption of lactic acid, but it may indirectly increase the hydrogenation rate by
facilitating the adsorption of water on the metal surface through the formation of
hydrogen bonds. This might be expected from the fact that in the chelating bidentate

mode, the lactic acid is adsorbed with the —O-H groups exposed to the solution.

194

Abs[0.04

1200 min.

1000 min.
800 min.

600 min.
400 min.
200 min.
40 min.

32 min.
24 min.
16min.
8 min.

 

fi I 1 I I l I

l ' l l l
2800 2400 2000 1600 1200

Wavenumbers (cm'1)

r l ' l
4000 3600 3200

Figure 6.11: Real time ATR-IR spectra of the ECH of lactic acid in 0.01 M HC104

(T=30°C, E= -400 mV) using the spectrum of 0.01 M HC104 taken under identical

conditions as the reference

6.5 Conclusions

The ECH of lactic acid on a Ru electrode surface has been studied in real time using
ATR-FTIR spectroscopy. The conditions used for the FTIR studies were close to the ones
used for bulk ECH studies, 50 the spectra can be correlated with changes occurring on or

near the electrode surface during the course of the experiment. The optimum

195

I

concentration of lactic acid was calculated to obtain maximum surface selectivity. From
the spectra, the ATR technique was found to be very sensitive to species adsorbed on the
surface or present within the electrical double layer. The in-situ infrared spectra of lactic
acid were obtained in three electrolytes- H2804, HCl and HC104. Background subtraction
and analysis of the spectra allowed us to detect an electrode surface poisoning effect,
probably due to species derived from sulfate and perchlorate anions associated with the
electrolyte, which blocks lactic acid adsorption. The difference in the rate of ECH

observed in the three electrolytes can be correlated with the extent of surface poisoning

--'_

by the electrolyte anions or species derived from them. It was also found that lactic acid
adsorbs on Ru through the chelating bidentate mode without involving the a-OH group.
Therefore, the til-hydroxy group does not facilitate the hydrogenation of lactic acid (and
other a-hydroxy acids) by forming a more stable adsorbed intermediate or by influencing
the initial adsorption of lactic acid, but it may indirectly increase the hydrogenation rate
by facilitating the adsorption of water on the metal surface through the formation of
hydrogen bonds.

It was also determined that chloride does not block adsorption of lactic acid, and the
enhancement of the ECH rate in HCl may be due to partial desorption of chloride from
the electrode surface at longer times and subsequent adsorption of water molecules,

which play an important role in ECH.

196

CHAPTER 7

CONCLUSIONS AND FUTURE WORK
7.1 Conclusions
Electrocatalytic hydrogenation provides a greener, milder and more convenient
alternative to the chemical catalytic hydrogenation of lactic acid, which requires 150°C
and 1200 psi H2 gas, mainly due to the poor solubility of hydrogen gas in water. In ECH,
the hydrogen required for hydrogenation is produced close to the reaction site on the
catalyst surface, thus eliminating the kinetic barrier to the dissolution and splitting of H2.
As a practical matter, ECH eliminates the requirement for external supply of hydrogen,
thus improving safety. The milder conditions and easy controllability of the ECH method
may also improve stereoselectivity of the reduction. Finally, it is amenable to in~situ
vibrational spectroscopic studies, which give structural information about adsorbed
intermediates and enable us to probe into mechanistic details.
In this study, the ECH of various (ll-functionalized carboxylic acids has been attempted
and optimized. To the best of our knowledge. this is the first study of the ECH of
carboxylic acids. Initial studies using the model compound benzoyl formic acid showed
that the ECH reaction is dependent more on the availability of surface sites, rather than
the rate of mass transport, consistent with the results obtained with the chemical catalytic
hydrogenation of lactic acid. ECH of lactic acid using electrodeposited Ru on carbon felt,
though giving a poor yield of product, enabled the determination of kinetic parameters,
such as order of reaction, fractional surface coverages of lactic acid and hydrogen, rate

constants, adsorption constants and activation energy using the Langmuir-Hinshelwood

197

 

model. The apparent activation energy determined to be 76.0 kJmol'I is approximately
half the value of 138 kJmol", determined for the batch hydrogenation of lactic acid.

ECH of lactic acid using 5% Ru/C agglomerated in RVC resulted in much higher
conversion. The effects of various parameters such as temperature, ECH current and
nature of electrolyte on the yields and distribution of products were studied. In all cases,
lactaldehyde was obtained as a major product with small quantities of propylene glycol
obtained only with 5% Ru/C/RVC. This unexpected result is attributed to the existence of
lactaldehyde mostly in the hydrated form under the mild conditions used. Dehydration of
the hydrated lactaldehyde and subsequent hydrogenation to propylene glycol is
presumably slower than desorption into solution. The formation of propylene glycol is
also favored by higher surface concentration of H. In the chemical catalytic
hydrogenation of lactic acid, lactaldehyde was proposed to be an intermediate, but was
never isolated. The formation of lactaldehyde as the major product of ECH supports its
proposed intermediacy en route to propylene glycol. Although this project was begun as a
way to carry out lactic acid to propylene glycol reduction under mild conditions, the
discovery of a pathway that directly reduces an aqueous carboxylic acid to an aldehyde
and then stops is substantially more interesting. in terms of both fundamental chemistry
and practical process. This unexpected process can provide a clean, convenient and
efficient method to prepare lactaldehyde, a compound that is not commercially available

and that is difficult to obtain in pure form.

ECH of lactic acid is also influenced by the nature of the electrolyte used, indicating
significant anion adsorption and modification of the electrode surface under the

conditions used. The most intriguing findin g is the much higher enhancement of the rate

198

 

 

 

of ECH in I-ICl, as compared to H2804 and HC104, especially during the later stages of
the reaction. From control experiments in the absence of electrolyte and from the in situ
ATR-IR studies, this has been attributed to the desorption of Cl from the electrode
surface at longer times due to C12 evolution at the anode, thus leaving sites for the
adsorption of relevant species such as water molecules, which are required for
hydrogenation. The chloride ions partially covering the Ru surface may also have an
electronic effect that favors adsorption of lactic acid and hydrogen. The fact that the
product ratio of ECH depends on the electrolyte used is also intriguing and can be
exploited in future applications to control the selectivity of the reaction towards specific
products by using specific electrolytes or a combination of different electrolytes.

Using in-situ FTIR, the initial mode of adsorption of lactic acid on the electrode surface
was determined to be chelating bidentate, without involving the a-OH group. Therefore,
the a-hydroxy group does not facilitate the hydrogenation of lactic acid (and other a-
hydroxy acids) by forming a more stable adsorbed intermediate or by influencing the
initial adsorption of lactic acid, but it may indirectly increase the hydrogenation rate by
facilitating the adsorption of water on the metal surface through the formation of
hydrogen bonds. The study of the ECH of lactic acid using RVC cathode and Ru
dispersed on non-carbon supports shows that the chemical nature and porosity of the
support and the characteristics of the Ru particles deposited on the support play an
important role in determining the rate and product distribution in ECH.

This work is a step towards ultimately using electrocatalytic hydrogenation as a
convenient laboratory process and possibly a commercial process for the conversion of

lactic acid to value added chemicals. It demonstrates the feasibility of the process at the

199

 

laboratory scale and the feasibility of using in-siru spectroelectrochemical studies for
understanding the mechanism. This will lead to the development of more efficient
catalysts in the future. Lactic acid is an important feedstock due to its ready availability
from the fermentation of biomass and therefore. its conversion to value-added chemicals
using various catalytic processes is being extensively studied. Though electrocatalytic
reactions are not widely used in commercial processes due to the difficulty involved in
scaling up, it is a viable alternative that may be used in medium scale conversions, where

a high product selectivity is desired.

7.2 Suggestions for future work

In this work, the ECH of lactic acid has been achieved under mild conditions of
temperature and pressure compared to the chemical catalytic method. However, much
more work needs to be carried out in order to improve the process both with respect to
application and fundamental understanding. From the calculations reported in chapter 4,
it was seen that the current efficiency of the process is too low for practical applications,
and this can be attributed to the slow kinetics of the ECH reaction. In the 5% Ru/C RVC
system, the amount of active electrocatalyst at any given time is very low as the
technique relies on efficient agglomeration of the catalyst, which is affected by various
factors such as particle size, fluid dynamics. etc. The kinetics can be improved by
developing better electrode materials based on Ru and C, which will consist of well-
dispersed Ru nanoparticles on a solid phase carbon of similar porosity.

The process needs to be improved to achieve higher yields of propylene glycol, which is

the desired product. The exact conditions which favor the formation of propylene glycol,

200

both in terms of reaction conditions and electrode microenvironment need to be
determined. Toward this aim, in-situ Raman spectroelectrochemistry using single crystal
Ru electrodes appears to be the technique of choice. From initial studies on Raman
spectroscopy, it was found that lactic acid does give reasonably strong Raman signals in
aqueous solution, which are slightly attenuated on Ru. Raman spectroscopy is desirable
for the study of electrode reactions in aqueous phase for several reasons. Water, being a
poor Raman scatterer does not give strong signals and hence offers minimum
interference. Also Raman spectroscopy is very useful for obtaining information in the
lower frequency region (400-9000m"), which could not be studied with infrared
spectroscopy, due to strong absorption by the Si window in this region. The bands in this
region will provide more detailed information about the modes of adsorption of
electrolyte anions and the presence of M-O or M-C bonds, giving further insight into the
manner in which anion adsorption affects lactic acid and hydrogen adsorption and the
ECH reaction. Spectroelectrochemistry experiments with single crystal Ru electrodes
should afford more information about the specific sites on which the various reactants
(lactic acid, hydrogen and water) and intermediates adsorb and would enable us finally to
propose a reaction mechanism for lactic acid hydrogenation involving surface bound
intermediates. The use of single crystal electrodes is expected to simplify the spectra
considerably, compared to polycrystalline electrodes. thus making spectral interpretation
easier.

Though this study deals mainly with the EC” of lactic acid, the process has been
extended to other (Jr-functionalized carboxylic acids, which are essentially analogs of

lactic acid, namely, glycolic acid, glyceric acid and alanine. Since this process was found

201

J ...._

to give the aldehyde as the major product. it would be interesting and useful to extend the
ECH method to other water soluble unsubstituted organic acids such as propanoic acid
and isobutyric acid to produce the corresponding aldehydes, which have commercial
value and are more difficult to obtain from other sources. However, attempts at ECH of
these substrates have been unsuccessful with no conversion achieved to date. One reason
for this may be that the adsorption of electrolyte anions makes the electrode surface
hydrophilic, thus inhibiting the adsorption of the organic acid, which are rendered
hydrophobic by the hydrocarbon chain. This problem may be overcome by including a
non-micelle forming cationic surfactant such as didodecyl dimethyl ammonium bromide
(DDAB) in the reaction mixture, which would bind to the electrode surface through
electrostatic interactions by displacing the anions, and its long hydrocarbon tail would
provide a hydrophobic surface which would enhance the adsorption of the organic acid,
as shown in figure 7.1. A similar method of enhancing the rate of ECH of hydrophobic
substrates such as alkyl phenols using cationic surfactants has been reported by
Thomalla.38'39

However, in spite of the absence of electrolytes and harsher conditions, the
hydrogenation of these acids was found to be slow in the batch process, compared to
lactic acid and amino acids. This finding has been attributed to a kinetic effect of the 01-
hydroxy or amino group, which facilitates adsorption of water molecules essential for
hydrogenation. Therefore, even if the substrate adsorption is enhanced, other challenges

may have to be overcome to achieve hydrogenation of these substrates.

202

 

 

 

)\‘(OH /A‘ \A/VWN’/ Cl-
/ \
O

Isobutyric acid D DAB
S- Substrate
l---l- Adsorbed hydrogen

Ru particle
H S

S H 8 811
Main 5218

‘L

 

 

 

Figure 7.1: Use of a cationic surfactant to modify the electrode surface

 

APPENDIX

204

 

 

 

 

 

, .
L-JtlelL W1) 1er 01 La

EFT T f7 I I I T I I T I I TYfiiT T I T “H"-I T T T 1' r7 T I I I 1

 

 

I ~ I I I l T
9 8 7 6 5 4 3 2 1

._J LJ l_lt_J {—49

1.53 0.18 0.46 1.11 0.99 0.45 3.40
0.78 1.60 0.59 0.62

Figure Al: NMR spectrum of the D20 extract obtained from PCC oxidation of 1,2-

propanediol (‘H, D2O).

Lactaldehyde, pyruvaldehyde and 1,2-propanediol are present in the mixture; 8 0.9 ppm

[d, -CH3 (1,2-propanediol)], 8 1.1 ppm ((1, -CH3 (lactaldehyde)], 8 1.9 ppm [5, -CH3

(pyruvaldehyde)], 8 3.3 ppm [m, -CHOH (1 ,2-propanediol)], 8 3.6 ppm [d, -CH2OH (1,2-

propanediol)], 8 3.8 ppm [m, -CHOH (lactaldehyde)], 8 8.0 ppm [d, -CHO

(lactaldehyde)], 8 8.3 ppm [5, -CH0, (pyruvaldehyde)], 8 7.7 and 8.5 ppm [d, C6H5N-

(pyridine)]. Peak splitting is obscured in some cases due to the presence of chromate

from FCC.

205

 

 

 

 

 

‘ 1 1 I 1 I ' i ‘ ‘ ‘ ‘. '1 . 1 1.
I- . ..........lt tmtethltfllthiWMIh‘“Ltllflllllll.rel 01.111. ”-1.... .11.“ |.J:l..‘1u. arm. 1.}.

220 200 180 160 140 120 100 80 60 40 20 0 pp

 

Figure A2: NMR spectrum of the D20 extract obtained from PCC oxidation of 1,2-

propanediol (13C, D20).

Lactaldehyde, pyruvaldehyde and 1,2-propanediol are present; 8 19 ppm {-CH3 (1,2-
propanediol)], 8 20 ppm {-CH3 (lactaldehyde)], 8 66 ppm {-CHOH (l ,2-propanediol)], 8
67 ppm [-CH2OH (1,2-propanediol)], 8 69 ppm {-CH3 (pyruvaldehyde)], 8 72 ppm [-
CHOH (lactaldehyde)], 8 126, 143 and 144 ppm [C6H5N- (pyridine)], 8 164 ppm

(unknown peak), 8 174 ppm {-CHO (pyruvaldehyde)], 8 175 ppm [-CHO (lactaldehyde)],

206

.a
[I

 

 

 

 

 

 

 

 

l' r l
r I I l I T] T T U U I I I I I I I l I r I I j ' T1
4.5 4.1 1.5 3 0 at. 2.1 r s 1.0 o 5 pp-
‘1‘ 1-1-1 ‘1’ '-t-' H-l
0.72 070 4.56512 242

Figure A3: NMR spectrum of lactaldehyde dimethyl acetal ('H, D20)

Peak assignments: 8 1.0 ppm (d, 3H, -CH3), 8 3.2 ppm (5, 3H, -OCH3), 8 3.3 ppm (s, 3H,

-OCH3), 5 3.6 ppm (m, 1H, -CHOH), 5 4.1 ppm [d, 1H, -CH(OCH3)2]

207

 

 

 

 

1

i;

1.

'1’

l.
i" \J\‘ + L M
,.._-.-.- . r:- . ~. - ~ “T""T."'
4.5 4.0 3.5 3.0 2.5 2.0 1.5 ppm

3.58 1.00

Figure A4: NMR spectrum of lactaldehyde obtained by the hydrolysis of lactaldehyde

dimethyl acetal in 0.05 M 112504 for 2 h (‘11, 020)

Peak assignments: 8 1.3 ppm (d, 3H, -CH3), 8 4.25 ppm (q, 1H, -CHOH) (aldehyde peak

not visible)

208

a-
O

 

I, .
ll

 

 

 

Figure A5: NMR spectrum of lactaldehyde obtained by hydrolysis of lactaldehyde

dimethyl acetal in 0.05 M st04 for 2 h ('30. 030)

Peak assignments: 8 20ppm (-CH3), 8 66 ppm (-CHOH), 8 179 ppm (-CHO hydrated)

209

70

 

1400
1025 60

 

 

2800
3450
2900 56

 

 

 

4000 3500 3000 2500 2000 1500 1000 500 ' 0
Wavenumber (cm‘)

Figure A6: FTIR spectrum of lactaldehyde obtained by hydrolysis of lactaldehyde

dimethyl acetal in 0.05 M H2804 for 2 h (NaCl plate)

Peak assignments: 1025 cm": uC-O (alcohol), 1400 cm": 000 (hydrated aldehyde),

2800 and 2900 cm": DC-H (sp3), 3450 cm": uO-H

210

 

 

my. - .0

 

 

 

 

 

 

 

 

 

 

 

 

YTTTTI—ITj—ITII'

9 a 7 6 5 4 3 2 1 -0 on

Figure A7: NMR spectrum of product mixture obtained after electrocatalytic

hydrogenation of lactic acid in 0.01 M HClo.1 for 21 h ('H, 020)

Both lactaldehyde and 1,2-propanediol are present in the mixture; 8 1.0 ppm [d, -CH3
(1,2-propanediol)], 8 1.2 ppm (d, -CH3 (lactaldehyde)], 8 1.75 ppm (unknown peak,
visible in all electrohydrogenation product mixtures, probably associated with mineral
acid used as electrolyte), 8 3.3 ppm [m, —CHOH (1,2-propanediol)], 8 3.7 ppm [d, -
CH2OH (1,2-propanediol)], 8 4.0 ppm [q, -CHOH (lactaldehyde)], 8 8.3 ppm [d, -CHO

(lactaldehyde)]

211

 

 

 

 

 

 

 

 

 

 

 

 

 

2.2 2.0 1.8 1.6 1.1 1.2 1.0 0.0 0.: pp.
H-1 1H '14 ‘14

2.09 0.88 0.47 0.55

Figure A8.1: Expansion of the peaks in figure A7 (8 0.5-3.0 ppm).

 

Figure A8.2: Expansion of the peaks in figure A7 (8 2.6-4.4 ppm)

 

IIIIIIIIIIIWIFITIllllIIIIIIIIIIIIIIIIIIIIIIIIIIIIITTIIITTIIITTIIITI'IIIIIITTIII

8.38 8.36 8.34 8.32 8.30 8.28 8.26 pp.
l__'_J
0.18

Figure A8.3: Expansion of the peaks in figure A7 (8 8.24-8.40 ppm)

__'_ '
l

Splitting of the —CH0 peak into a doublet is not very clear and may be obscured by the
presence of residual acid or C104' ions from the electrolyte. The —CHOH peak is a quartet
instead of a multiplet in both the chemically synthesized and the electrocatalytically
synthesized lactaldehyde. This may be a result of the acidic environment and the lability

of the aldehyde proton.

213

 

 

l enumllllki-lll1hrIMIIJIHIIhnldJnhuL‘l‘JIdullhlnl l. 1111.“... l

200 180 160 140 120 100 80 60 40 20 0

 

Figure A9: NMR spectrum of product mixture obtained after electrohydrogenation of

lactic acid in 0.01 M HC104 for 21 h (‘30, 020)

Both lactaldehyde and 1,2-propanediol are present; 8 19 ppm [~CH3 (l,2-propanediol)], 8

20 ppm {-CH3 (lactaldehyde)], 8 66 ppm {-CHOH (l.2-propanediol)], 8 67 ppm [-CH2OH

(l,2-propanediol)], 8 72 ppm {-CHOH (lactaldehyde)], 8 178 ppm [-CHO (lactaldehyde)]

214

1 74.03

.i

l$llll§lll

‘1 ‘1 O a
$ (71 0
111111 lllllLlrlllllilrll

G
01

1J%l 11

Relative Ablmdance
0|
l$l l l 1T1 l

N 01 b b
17.11131 1 1 1T1 r 1 8111 lTlLl 81 1 13111131 l'lclnl 11

0

AM

-I

56.06

 

 

§6’73 730.091

0

{75°05

 

IITIITIIIII

70

8

I I

l
[II‘IFTfIIT

80 90

Figure A10: ESI-MS of chemically synthesized lactaldehyde

Peak assignments: m/z 74.03 (M: ion peak), m/z 56.06 (M-H2O)+

215

 

 

 

 

73.8

 

 

1
3
3
i
m
64.9
1 56.0 82.8
.1 51.0 664 57.1 69.161164;1 66.0 69.4 72.172,-7 ".3 76.0 79.0 620 83.8 81.0 88.8 90111117 915 m 90.7
vlrIfi—rII—r7T117fr7r—fi77f7117771111fi71117771T17f111771]~
50 55 60 as 70 75 60 as 911 115 >100

mlz

Figure A11: ESI-MS of lactaldehyde obtained after electrohydrogenation of lactic acid in

0.01 M HC104

Peak assignments: m/z 82.8 (unknown peak), m/z 73.8 (M+ ion peak), m/z 64.0 (formed

by loss of 20 from HC104), m/z 56.0 (M-HZO)+

216

a...
O

 

 

 

 

 

Figure A12: NMR spectrum of product mixture obtained after electrocatalytic

hydrogenation of glycolic acid in 0.01 M HCl for 21 h ('H, D20)

Only glycolic aldehyde was detected in the mixture; 6 3.2 ppm [d, -CH2- (glycolic
aldehyde)], 8 1.75 ppm (unknown peak, visible in all electrohydrogenation product
mixtures, probably associated with mineral acid used as electrolyte), 8 8.3 ppm [d, -CHO

(glycolic aldehyde)]

 

A-r
I

 

 

l
'(t

 

 

 

 

_A__ J IR; A
T—TWTI TIYTIITIII lt—Tr—r ITIY! 7T TY—T ‘T—TITY r1rrTfY—T—rlrtlr r111]
3.28 3.27 3.26 3.25 3.23 3.23 3.22 3.21 3.20 3.19 3.18 3.17 3.16 pp-
L__J
2.00

Figure A13.l: Expansion of the peaks in figure A12 (5 3.16-3.28 ppm)

Iq ‘ ‘
NM M / V‘flt‘A-H‘w ‘4 5" “IV/448:“ . W W VAWMM J‘wfi‘l‘NJ It, “Viv-”’51 J."&\‘-\V‘V'W~MJ‘WW-’WJJ'WNWV N “MP" W U " “I H H

‘lllll [ill FYITTT ITTYTIIITY 'lll Til, YTIIY TITTl Illl I

8.39 8.38 8.37 8.36 8.35 8.34 8.33 8.32 8.31 8.30 8.29 8.28

 

012

Figure A13.2: Expansion of the peaks in figure A12 (5 8.28-8.39 ppm)

218

TT‘F]

pp-

1

3"
a

 

 

Splitting of the —CH0 peak into a triplet is not very clear and may be obscured by the

presence of residual acid or Cl' ions from the electrolyte.

 

219

 

 

755
70 1

65.--:

 

 

Relative Abundance
8

35'
3o}
25 ‘
20:?

59.29
15 1

I——Ma—L—.-é- - 1“

10- «_‘ 1 75.32
5 -
" .50-15 76.22
0""? 1' ? -,-1;.'.-2T~..:_ ‘ t r ' ' ' “r {1'
50 60 70 80
Figure A14: ESI-MS of sample obtained after electrohydrogenation of glycolic acid in

0.01 M HCl

 

220

Peak assignments: m/z 76.22 (M+ ion peak, glycolic acid), m/z 75.32 [(M-H)+, glycolic

acid], m/z 60.15 (M+ ion peak, glycolic aldehyde), m/z 59.29 [(M-H)+, glycolic aldehyde]

 

 

 

 

1 1 j ifi 1 j * 5* 1 T V 1

50 45 40 35 30 2.5 2.0 15 10 pp-
1+ u H4
I963 3.08 7729

Figure A15: NMR spectrum of alanine ('H, D20)

Peak assignments: 5 1.3 ppm (d, 3H, -CH3), 6 2.1 ppm (d, 2H, -NH2), 8 3.6 ppm (q, 1H, -

CHOH)

221

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure A16: NMR spectrum of product mixture obtained after electrocatalytic

hydrogenation of alanine in 0.01 M HCl for 21 h ('H, D20)

The product was identified as 2-amino propanal; 8 1.2 ppm (d, 2H, -CH3), 8 1.75 ppm
(unknown peak, visible in all electrohydrogenation product mixtures, probably associated
with mineral acid used as electrolyte), 8 3.6 ppm (q, 1H, -CHOH), 8 8.3 ppm (d, 1H, -

CHO).

222

 

" Ll

 

 

 

 

 

Figure A17.1: Expansion of the peaks in figure A16 (8 1.2-3.8 ppm)

 

 

 

1,1. .. ,
WWW W11 ”MM 1111/11» 111”!” “11'11”111111141",V111" M11111" 1‘AMWMW

 

1f111v11fr111TT1 ttfrtr11fitrtlllT1
8.37 8.36 8.35 8.34 8.33 8.32 8.31 pm
l__J
0.04

Figure Al7.2: Expansion of the peaks in figure A16 (8 8.30-8.37 ppm)

Splitting of the —CH0 peak into a doublet is quite clear in this case. The —CHOH peak is

a quartet instead of a multiplet. This may be a result of the acidic environment and the

lability of the aldehyde proton.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

d
c OH
e b a
f d
e
1
1
1
‘x 1 .
F1 L11 11
n_.1L 4.411; \_____,_____/ \z
Y—fi I I T — 1 ' _T' T 7‘ T'_"T T T -' T 1 I l T 7 T f
7 6 5 4 3 2 1 on

Figure A18: NMR spectrum of the product of batch reduction of l-phenyl 1,2-ethanediol

(PED) ('H, CDCl3).

Cyclohexyl ethylene glycol (CEG) (structure shown above) was the major product with
small amount of 2-cyclohexyl ethanol (CE) (from GC analysis); (C atoms are labeled as
shown in the structure) 6 1.04 ppm [q, -Cf-H (cyclohexyl)], 6 1.20 ppm [m, Ce-H and

Cd-H (cyclohexyl)], 8 1.42 ppm [m, Cc-H (cyclohexyl)], 5 1.66 ppm [t, CZ-H (CE)], 1.76

225

 

 

 

ppm [t, Ca-OH (CEG)], 5 1.88 ppm [d, Cb-OH (CEG)], 5 2.12 ppm [s (broad), O—H

(water)], 5 3.46 ppm [dt, Cb-H (CEG)], 5 3.54 ppm [t, Cl-H (CE)], 5 3.72 ppm [dd, Ca-

 

 

 

 

 

 

 

H(CEG)]
1 l
1 I I
l1 ‘
11 1, .
11 1 .‘
I 1‘ 111
1 1 I11 11 111'
I 1! 111I 1111 111
l
I 1 111 I 11 1
JAKJ 1 1/ 11/11
11 TI IrTIT' ”111 ”ll‘HlIl 1 1"” 11111117 T1111'11111‘1111 11”” ””1 1' I'" I‘m
2.4 2.2 2.0 1.3 1.5 1.4 1.2 1.0 0.3 pp.
5—4114"_1 T171 $3: 1l07 331 215

Figure Al9.1: Expansion of the peaks in figure A18 (8 0.6-2.6 ppm)

226

 

 

 

 

 

 

 

1 i111 1; 1 1 11.
1 1l 1 ,1 1 I'

 

U

 

 

 

 

1
1 1
1 1 ~.
(111 11111 1111
1
I‘V' J.;/'1’ 1
I U 1\ .151 1V1 11‘: 11/11-
L. A A/ k'v'uw-x'/ uk/ W
1 TFVr ‘T' . uI 1
4.0 3.9 3.8 3.7 3.6 3.5 3.4 3.3 3.2 pp
‘r—J—r—J L____J __.T.__J
0.42 0.52 1.00 1.00

Figure A19.2: Expansion of the peaks in figure A18 (5 3.1-4.0 ppm)

227

 

 

 

 

 

 

 

80 70 60 50 4O 30 20 “I

Figure A20: NMR spectrum of the product of batch reduction of l-phenyl 1,2-ethanediol

(PED) (”(3, CDCI3)

Cyclohexyl ethylene glycol (CEG) was the major product with small amount of 2-
cyclohexyl ethanol (CE); (C atoms are labeled as shown in the structure) 8 26.2 ppm [Cf
(cyclohexyl), 6 26.3 ppm [Ce (cyclohexyl)], 5 26.6 ppm [Cd (cyclohexyl)], 6 28.9 ppm
[Cc (cyclohexyl) (CE)], 8 29.2 ppm [Cc (cyclohexyl) (CEG)], 5 41.0 ppm [C2 (CE)], 8

65.0 ppm [c1 (CE)], 6 77.3 ppm [Cb (CEG)], 6 77.5 ppm [Ca ((3130))

228

 

in

 

 

 

 

 

, .1 5.!
WWWM1MwwWWWWWMWW

l Ill—TIT ll11lT—l171‘rTl ITIIII llTlIl ll) Tlllll’ [I ll] Ill lTrrrTrrTTTIlT llTl]

34 32 30 28 25 24 22 20 pp-

 

Figure A21.l: Expansion of the peaks in figure A20 (5 20-34 ppm)

229

 

 

 

1.1111 1

 

l
.5 ‘ " 'i' ' .1
. , . w. v 1» ~. 1 4
W‘memmmmumwm W 1w,”M'~LM«,W~)M\|1~LMW
I I I I T I I I I I fir I T h ' I I I T T T I T I I l l I r I I
so 79 '78 77 76 75 7‘ pl-

Figure A21.2: Expansion of the peaks in figure A20 (6 73-81 ppm)

230

 

‘ 2920

‘ 3397
-' 2849

 

 

1.40 -1 1 1451 1064

 

 

 

 

 

4000 3000 2000 1000

Wavenumbers (cm")

Figure A22: FTIR spectrum of cyclohexyl ethylene glycol (CEG) synthesized by batch

reduction of phenyl ethanediol (PED) (KBr pellet)

Peak assignments: 1064 cm": oC-O (alcohol), 1451 cm": SCHz-C, 2849 and 2920 cm":

uC-H (sp3), 3397 cm": uO-H

231

 

73.1

 

 

 

 

 

 

 

 

 

 

 

 

 

273.3
129.2
170.2
11 1 183.2
812 1472 i 1
1 f
115.2 ~ f 217.3 2883
551 1°12 .' 155.2 1 ‘992 L232: 260.3 | .-
' ‘ ‘ ' 1 ' 1 1 1 1
0191.11...1111.11'1...7-.,T.1..I1T. L 11 11.1.11 1.1.41 111 111..., .4 343: 1.11.1-1 .
m/z 50 100 150 200 250 300

Figure A23: Electron ionization mass spectrum (El-MS) of trimethyl silyl (TMS)
derivative of cyclohexyl ethylene glycol (CEG) synthesized by batch reduction of phenyl

ethanediol (PED) derivatized using BSTFA and separated by GC (DBl column)

Peak assignments: m/z 288.3 (M+ ion peak of disilylated CEG); m/z 273.3 (M-15 peak of

disilylated CEG); m/z 147.2 (2TMS); m/z 101.2 (CHZOTMS); m/z 73.1 (TMS)

232

 

 

 

 

 

 

 

 

 

 

273.3
147.2
129.2 1
1 17021832
' 1
116.2 1 1 1 1
1°12 1 1 1 2112 2323 1
. ”55.2 . 198.2 1 - 260.3 , 288.3
"-47 1th11‘: 11111110111I 1hr 111 -111 1111. .11; 11 '245;? L.1L 111‘ .1; l r r
100 150 200 250 300

Figure A24: Electron ionization mass spectrum (El-MS) of CEG obtained by ECH of
mandelic acid, afier precipitation of sulfate, evaporation of water, derivatization using

BSTFA and separation by GC (DB1 column)

Peak assignments: m/z 288.3 (M+ ion peak of disilylated CEG); m/z 273.3 (M-lS peak of

disilylated CEG); m/z 147.2 (2TMS); m/z 101.2 (CH20TMS); m/z 73.1 (TMS)

233

 

75.1 117.2

1

132.2

 

 

 

 

 

 

1
1
4 55,1 1
“1:61 1

 

12 99.2 1 .
1.1.1 .1111111.11.....1.111.1.°17'2 11 1' "172552 171-217.92 11 214-3 247.3
.fi 1 l ' 1 1 1 1 1 r . Y Y 1 1,. ._2..._

mlz 1110 150 260 250

 

Figure A25: Electron ionization mass spectrum (El-MS) of 2-cyclohexyl 2-hydroxy
acetaldehyde (CHA) obtained by ECH of mandelic acid, after precipitation of sulfate,

evaporation of water, derivatization using BSTFA and separation by GC (DB1 column)

Peak assignments: m/z 214.3 (M+ ion peak of monosilylated CHA); m/z 199.2 (M-lS

peak of monosilylated CHA); m/z 147.2 (2TMS); m/z 73.1 (TMS)

234

 

10.

11.

12.

13.

14.

15.

16.

17.

18.

REFERENCES

(a) Adkins, H.; Wojcik, B.; Covert, L.W., J. Am. Chem. Soc. 1932, 54, 1145-1154
(b) Adkins, H.; Wojcik, B.; Covert, L.W., J. Am. Chem. Soc. 1933, 55, 1669-1676

Bowden, E.; Adkins, H., J. Am. Chem. Soc. 1934, 56, 689-691
Um, S.; Han, S.; Jung, K.; Baek, J. Repub. Korea, Patent no. KR 9601911, 1996

Roesky, R.; Borchert, H.; Dingerdissen, U. PCT Int. Appl., 1998, 16, Patent no. W0
9841492

Luo, G.; Yan, S.; Qiao, M.; Zhuang, J .; Fan, K., App]. Catal. A: General, 2004, 275,
95-102

 

Schurch, M.; Heinz, T.; Aeschimann, R.; Mallat, T.; Pfaltz, A.; Baiker, A., J. Catal,
1998, 173, 187-195

Orito, Y.; Imaj, S.; Niwa, S., J. Chem. Soc. Jpn., 1979, 8, 1118
Adkins, H.; Billica, H.R., J. Am. Chem. Soc., 1948, 70, 3121-3125

Carnahan, B.; Ford, T.; Gresham, W.; Grisby, W.; Hager, G., J. Am. Chem. Soc.,
1955, 77, 3766

Broadbent, H.; Campbell, G.; Bartley, W.; Johnson, J ., J. Org. Chem, 1959, 24,
1847

Mendes, M.J.; Santos, O.A.A.; Jordao, E., App]. Catal. A: General, 2001, 21 7(1-2),
253-262

Toba, M.; Tanaka, S.; Niwa, S., Appl. Catal. A: General, 1999, 189(2), 243-250
Rachmady, W. Vannice, M.A., J. Catal, 2000, 192, 322-334

Zhang, Z.; Jackson, J .E.; Miller, D.J., Appl. Catal. A: General, 2001, 219, 89-98
Zhang, 2.; Jackson, J .E.; Miller, D.J., Ind. Eng. Chem. Res. 2002, 41, 691

Jere, F.T.; Miller, D.J.; Jackson, J .E., Org. Lett., 2003, 5(4), 527-530

Sorum, P.A.; Onsager, O.T., Proceedings- 8'” International Congress on Catalysis,
Berlin, 1984, 8, 11-23

Monti, D.M.; Cant, N.W.; Trimm, D.L.; Wainwright, M.S., J. Catal., 1986, 100, 17-
27

235

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

Kohler, M.A.; Cant, N.W.; Wainwright, M.S.; Trimm, D.L., Proceedings, 9’”
International Congress on Catalysis, 1988, 9, 1043-1050

Natal-Santiago, M.A.; Sanchez-Castillo, M.A.; Cortright, R.D.; Dumesic, J .A., J.
Catal, 2000, 193, 16-28

Pallassana, V.; Neurock, M., J. Catal., 2002, 209, 289-305

Luo, G.; Yan, S.; Qiao, M.; Fan, K., J Mol. Catal. A: Chemical, 2005, 230, 69-77
Deshpande, V.M.; Ramnarayan, K.; Narasimhan, C.S., J. Catal., 1990, 121, 174
Pouilloux, Y. Autin, F.; Guimon, C.; Barrault, J ., J. Catal., 1998, 176, 215-224
Kovacs, D.G.; Peereboom, L.; Jackson, J .E.; Miller, D.J., Unpublished results
Varga, N., MS thesis

Farrugia, J .; Jackson, J .E.; Miller, D.J., Unpublished results

Dabo, P.; Mahdavi, B.; Menard, H.; Lessard, J ., Electrochim. Acta, 1997, 42(9),
1457-1459

Quiroz, M.A.; Cordova, F.; Lamy-Pitara, E.; Barbier, J ., Electrochim. Acta, 2000,
45, 4291-4298

Mahdavi, B.; Chambrion, P.; Binette, J .; Martel, E. Lessard, J ., Can. J. Chem, 1995,
73, 846-852

Bryan, J.; Grimshaw, J., Electrochim. Acta, 1997, 42(13-14), 2101-2107

Senda, Y.; Tateoka, M.; Itoh, H.; Ishiyama. J .. Bull. Chem. Soc. Jpn., 1991, 64(11),
3302-3305

DaSilva, J .G.; Goulart, M.O.F.; Navarro, M., Tetrahedron, 1999, 55, 7405-7410
Cheong, A.K.; Bolduc, Y.; Lessard, J ., Can. J. Chem, 1993, 71 , 1850-1856
Romulus, A.M.; Savall, A., J. Appl. Electrochem, 2000, 30, 967-972

Hu, X.E.; Yang, H.W.; Wang, X.J.; Bai, RS, .1. App]. Electrochem, 2002, 32, 321-
328

Polcaro, A.M.; Palmas, S.; Dernini. S., Ind. Eng. Chem. Res., 1993, 32, 1315-1322

Ilikti, H.; Rekik, N.; Thomalla, M., J. Appl. Electrochem, 2002, 32, 603

[\J
Lu
0\

 

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

55.

56.

57.

Ilikti, H.; Rekik, N.; Thomalla, M., J. App]. Electrochem, 2004, 34, 127-136
Beraud, V.; Lessard, J.; Thomalla, M., Can. J. Chem, 1997, 75, 1529-1535
Roessler, A.; Dossenbach, 0.; Rys. P., J. Electrochem. Soc., 2003, 150(1), D1-D5

(a) Amouzegar, K.; Savadogo, 0., J. Appl. Electrochem, 1997, 27, 539-542 (b)
Amouzegar, K.; Savadogo, 0., Electrochim. Acta, 1994, 39(4). 557-559

Lofrano, R.C.Z.; Madurro, J .M.; Abrantes, L.M.; Romero, J .R., J. Mol. Catal. A:
Chemical, 2004, 218, 73-79

Takano, N.; Nakade, A.; Takeno, N., Bull. Chem. Soc. Jpn., 1997, 70, 837-840
Takano, N.; Seki, C.; Shimono, 1., Bull. Chem. Soc. Jpn., 2000, 73(3), 745-746

Santana, D.S.; Lima, M.V.F.; Daniel, J R.R.; Navaro, M., Tet. Lett., 2003, 44, 4725-
4727

Roessler, A.; Dossenbach, 0.; Marte, W.; Rys, P., Dyes and Pigments, 2002, 54,
141-146

Raju, R.R.; Mohan, S.K.; Reddy, S.J., .1. Sci. 1nd. Res, 2003, 62(4), 334-338

Wang, J .; Swain, G.M.; Tachibana, T.; Kobashi, K. Electrochemical and Solid
State Letters, 2000, 3(6), 286-289

Wang, J.; Swain, G.M. J. Electrochem. Soc., 2003, 150(1), E24-E32

Wang, J .; Swain, G.M.; Tachibana, T.; Kobashi, K. Proceedings- Electrochemical
Society, 2000, 99-32 (Diamond Materials), 428-439

Dube, P.; Kerdouss, F.; Laplante, F.; Proulx. P.; Brossard, L.; Menard, H. J. App].
Electrochem, 2003, 33, 541-547

Polcaro, A.M.; Demini, S.; Palmas, S. Electrochim. Acta, 1992, 37, 365-367

Inoue, H.; Yoshida, Y.; Ogata, S.; Shimamune, T.; Iwakura, C., J. Electrochem.
Soc., 1998, 145(1), 138-141

Iwakura, C., J. Mater. Res., 1998, 13(4). 821-824
Amouzegar, K.; Savadogo, 0., Electrochim. A eta, 1998, 43(5-6), 503-508

Amouzegar, K.; Savadogo, 0., J. App]. Electrochem, 1997, 27(5), 539-542

237

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

68.

69.

70.

71.

72.

73.

74.

75.

Vedenyapin, A.A.; Baturova, M.D.; Ioseliani, G.I.; Areshidze, G.K., Russ. Chem.
Bull, 1997, 46(1), 76-80

George, G., J. Chem. Edu., 1978, 55(6), 410

Nambiar, K.P.; Stauffer, D.M.; Kolodziej, P.A.; Benner, SA, .1. Am. Chem. Soc.,
1983, 105, 5886

Wang, J .; Fabrication, characterization and electrocatalytic activity of
metal/diamond composite electrodes, PhD. Thesis

Cyr, A.; Chiltz, F.; Jeanson, P.; Martel, A.; Brossard, L.; Lessard, J.; Menard, H.
Can. J. Chem, 2000, 78, 307.

Dabo, P.; Cyr, A.; Lessard, J .; Brossard, L.; Menard, H., Can. J. Chem. 1999, 77,
1225

Giles, R.D.; Harrison, J .A.; Thirsk, HR, .1. Electroanal. Chem, 1969, 20, 47-60
Breiter, M.W., J. Electroana]. Chem, 1984, 178, 53-59

F leischmann, M.; Grenness, M., J. Chem. Soc.- Faraday Trans. 1, 1972, 68(12),
2305

Laplante, F.; Brossard, L.; Menard, H., Can. J. Chem, 2003, 81(3), 258-264

Cyr, A.; Chiltz, F.; Jeanson, P.; Martel. A.; Brossard, L.; Lessard, J.; Menard, H.,
Can. J.Chem., 2000, 78, 307-315.

Dabo, P.; Cyr, A.; Lessard, J .; Brossard, L.; Menard, H., Can. J. Chem, 1999, 77,
1225-1229

Ruest, L.; Menard, H.; Moreau, V.; Laplante, F ., Can. J. Chem. 2002, 80, 1662-
1667

Dube, P.; Kerdouss, F.; Proulx. P.; Brossard. L.; Menard, H., .1. App]. Electrochem.
2003, 33, 541-547

Robin, D.; Comtois, M.; Made], A.; Lemieux, R.; Cheong, A.K.; Belot, G.;
Lessard, J., CanJ. Chem, 1990, 68, 1218.

Lamy-Pitara, E.; El Mouahid, S.; Barbier, J ., Electrochim. Acta. 2000, 43, 4299
Polcaro, A.M.; Palmas, S.; Demini, S., Electrochim. Acta, 1993, 38(2-3), 199-203

Horanyi, G., Electrochim. Acta, 1991, 36, 1453

238

 

76.

77.

78.
79.

80.

81.

82.
83.

84.

85.

86.

87.

88.

89.

90.

91.

92.

93.

94.

95.

Hadzi-Jordanov, S.; Angerstein-Koziowska, H.; Yukovic, M.; Conway, HE, J.
Phys. Chem. 1977, 81 (24), 2271

Sawatari, Y; Inukai, J.; Ito, M., J. Electron Spec. and Related Phenomena, 1993,
64/65, 515

Hoshi, N.; Kuroda, M.; Ogawa, T.; Koga, 0.; Hori, Y., Langmuir, 2004, 20, 5066
Shingaya, Y.; Ito, M., Electrochim. Acta, 1992, 3 7, 365
Colom, F .; Gonzalez-Tejera, M.J., J. Electroanal. Chem, 1985, 190, 243-255

Marinkovic, N.S.; Wang, J .X.; Zajonz, H.; Adzic, R.R., J. Electroanal. Chem. 2001,
500 (1-2), 388-394

Shingaya, Y.; Ito, M., J. Electroanal. Chem, 1999, 467, 299
Ahmadi, A.; Evans, R.W.; Attard, G., J. Electroanal. Chem, 1993, 350, 279-295
Horanyi, G.; Rizmayer, E.M., J. Electroanal. Chem, 1984, 181, 199-208

Wendt, H., ‘Electrochemical Hydrogen Technologies’, (Elsevier, Amsterdam,
1990), p.8

Wiekowski, A., ‘Interfacial Electrochemistry: Theory, Experiment and
Applications’, Marcel Dekker, Inc., 1999, p. 133

Gerischer, H., Bull. Chem. Soc. Belg, 1958, 67, 506-512.

Conway, B.E.; MacDougall, B.R.; Kozlowska, H.A., J. Chem. Soc.- Faraday
Trans. 1, 1972, 68, 1566-1574

Yokoyarna, T.; Yamagata, N., Applied Catalysis A: General, 2001, 221, 227-239
Peereboom, L.; Miller, D.J.; Jackson, J .E., Unpublished results

Cortright, R.D.; Sanchez-Castillo, M.; Dumesic, J .A., Appl. Catal. B, 2002, 39, 353
Cortright, R.D.; Dumesic, J .A., Adv. Catal., 2001, 46, 161

Falomi, M.; Giacomelli, G.; Porcheddu, A.; Taddei, M., J. Org. Chem, 1999, 64,
8962

Kim, Y.G.; Soto, J .E.; Chen, X.; Park, Y.S.; Soriaga, M.P., J. Electroanal. Chem,
2003, 554-555, 167-174

Kim, Y.G.; Baricuatro, J .H.; Soriaga. M.P.; Wayne, S.D., J. Electroanal. Chem,

239

2001, 236(1), 197-198

96. Soto, J.E.; Kim, Y.G.; Soriaga, M.P., Electrochemistry Communications, 1999,
1(3,4), 135-138

97. Pang, K.P.; Benziger, J .B.; Soriaga, M.P.; Hubbard, A.T., J. Phys. Chem, 1984,
88(20), 4583-4586

98. Mebrahtu, T.; Berry, G.M.; Soriaga, M.P.. .1. Electroanal. Chem, 1988, 239(1-2),
375-386

99. Chia, V.K.F.; White, J.H.; Soriaga. M.P.. Hubbard, A.T., J. Electroanal. Chem,
1987, 217(1), 121-128

100. Chia, V.K.F.; Soriaga, M.P.; Hubbard, A.T., J. Phys. Chem, 1987, 91(1), 78-82

10]. Song, D.; Soriaga, M.P.; Hubbard, A.T., J. Electroanal. Chem, 1985, 193(1-2),
255-264

102. Song, D.; Soriaga, M.P.; Vjeira, K.L.; Zapien, D.C.; Hubbard, A.T., J. Phys.
Chem, 1985, 89(19), 3999-4002

103. Soriaga, M.P.; Song, D.; Hubbard, A.T., .1. Phys. Chem, 1985, 89(2), 285-289

104. Soriaga, M.P.; Song, D.; Zapien, D.C.; Hubbard. A.T., Langmuir, 1985, 1(1), 123-
127

105. White, J.E.; Soriaga, M.P.; Hubbard, A.T., J. Electroanal. Chem, 1984, 177(1-2),
89-96

106. Soriaga, M.P.; White, J .E.; Song, D.; Hubbard, A.T., J. Electroanal. Chem, 1984,
1 71(1-2), 359-363

107. Chia, V.K.F.; Soriaga, M.P.; Hubbard. A.T., J. Electroanal. Chem, 1984, 167(1-
2), 97-106

108. Chia, V.K.F.; Stickney, J.L.; Soriaga, M.P.; Rosasco, S.D.; Salaita, G.N.; Hubbard,
A.T.; Benziger, J .B.; Pang, K.W.P., J. Electroanal. Chem, 1984, 163(1-2), 407-
413

109. Stern, D.A.; Laguren-Davidson, L.; Frank. D.G.; Gui, J .Y.; Lin, C.H.; Lu, F .;
Salaita, G.N.; Walton, N.; Zapien, D.C.; Hubbard, A.T. J. Am. Chem. Soc., 1989,
111, 877-891

110. Gui, J.Y.; Stern, D.A.; Lin, C.H.; Gao, F.; Hubbard, A.T. Langmuir, 1991, 7,
3183-3189

240

111.

112.

113.

114.

115.

116.

117.

118.

119.

120.

121.

122.

123.

124.

125.

126.

127.

128.

129.

Sungho, P.; Tong, Y.Y.; Wieckowski, A.; Weaver, M.J., Langmuir, 2002, 18,
3233-3240

Kardash, D.; Huang, J.; Korzeniewski, C .. Langmuir, 2000, 16(4), 2019-2023

Kim, K.S.; Korzeniewski, C.; Tornquist. W.J., J. Chem. Phys., 1994, 100(1), 628-
630

Lin, W.F.; Christiansen, P.A., Faraday Discussions, 2002, 121, 267-284

Lin, W.F.; Christiansen, P.A.; Hamnett. A., Phys. Chem. Chem. Phys., 2001, 3,
3312-3319

Arenz, M.; Stamenkovic, V.; Schmidt, T.J.; Wandelt, K.; Ross, P.N.; Markovic,
N.M., Phys. Chem. Chem. Phys., 2003, 5, 4242-4251

Iwasita, T.; Xia, X.H.; Liess, H.D.; Vielstich. W., J. Phys. Chem. B, 1997, 101,
7542-7547

Iwasita, T.; Xia, X.; Herrero, 13.; Liess, H.D., Langmuir, 1996, 12, 4260-4265
Kardash, D.; Korzeniewski, C., Langmuir, 2000, 16, 8419-8425
Sungho, P.; Xie, Y.; Weaver, M.J., Langmuir, 2002, 18, 5792-5798

Iwasita, T.; Hoster, H.; John-Anacker, A.; Lin, W.F.; Vielstich, W., Langmuir,
2000, 16, 522-529

Gomez, R.; Weaver, M.J., J. Electroanal. C hem, 1997, 435, 205-215

Caram, J .A.; Gutierrez, C., .l. Electroanal. Chem, 1992, 336, 309

Gootzen, J.F.E.; Visscher, W.; Van Veen, J .A.R., Langmuir, 1996, 12, 5076-5082
Xia, X.H.; Liess, H.D.; Iwasita, T., J. Electroanal. Chem, 1997, 43 7, 233-240

Morin, S.; Conway, B.E.; Edens, G.J.; Weaver. M.J., J. Electroanal. Chem, 1997,
421, 213-220

Chang, S.C.; Ho, Y.; Weaver, M.J., J. Am. Chem. Soc., 1991, 113, 9506-9513
Rodriguez, J .L.; Pastor, E.; Xia, X.H.; Iwasita, T., Langmuir, 2000, 16, 5479-5486

Huerta, F .; Morallon, E.; Perez. J.M.; Vazquez, J .L.; Aldaz, A., J. Electroanal.
Chem, 1999, 469, 159-169

241

 

130.

131.

132.

133.

134.

135.

136.

137.

138.

139.

140.

141.

142.

143.

144.

145.

146.

147.

148.

Camara, G.A.; De Lima, R.B.; Iwasita, T., J. Electroanal. Chem, 2005, 585(1),
128-131

Kunimatsu, K.; Bewick, A., Indian. .1. Technol, 1986, 24, 407

Parry, D.B.; Samant, M.G.; Seki, H.; Philpott, M.R.; Ashley, K., Langmuir, 1993,
9, 1878

Osawa, M.; Ataka, K.; Yoshii, K.; Yotsuyanigi, T., J. Electron Spectrosc. Relat.
Phenom, 1993, 64/65, 371

Osawa, M.; Ataka, K.; Yoshii, K.; Yotsuyanigi, T., Langmuir, 1994, 10, 640

Nicholas, R.J.; Burgess, 1.; Young, K.L.; Zamlynny, V.; Lipkowski, J ., J.
Electroanal. Chem, 2004, 563(1), 33-39

Hong-Qiang, L.; Roscoe, S.G.; Lipkowski. J .. J. Electroanal. Chem, 1999,
478(1 ,2), 67-75

Hong-Qiang, L.; Chen, A.; Roscoe, S.G.; Lipkowski, J ., Proc. Electrochem. Soc.,
2000, 15, 145-148

Hoon-Khosla, M.; Fawcett, W.R.; Goddard, J .D.; Lian, W.Q.; Lipkowski, J .,
Langmuir, 2000, 16(5), 2356-23 62

Hoon-Khosla, M.; Fawcett, W.R.; Chen, A.; Lipkowski, J .; Pettinger, B.,
Electrochim. Acta, 1999, 45(4-5), 611-621

Osawa, M., Bull. Chem. Soc. Jpn., 1997, 70, 2861-2880
Itoh, T.; McCreery, R.L. J. Am. Chem. Soc., 2002, 124, 10894-10902

Hind, A.R.; Bhargava, S.K.; McKinnon. A. Advances in Colloid and Interface
Science, 2001, 93, 91-114

Aroca, R.F.; Ross, D.J., App]. Spec., 2004, 58(11), 324A-338A
Chew, H.; Kerker, M., J. Opt. Soc. Am, 1985, B2, 1025

Osawa, M.; Ikeda, M., J. Phys. Chem, 1991, 95, 9914-9919
Dumas, P.; Tobin, R.G.; Richards, P.L.. Surf Sci, 1986, 1 71, 555
Merkin, G.T.; Griffiths, P.R, Langmuir. 1997, 13(23), 6159-6163

Devlin, P.J., Consani, K., J. Phys. Chem. 1981, 85, 2597

149.

150.

151.

152.

153.

154.

155.

156.

157.

158.

159.

160.

161.

162.

163.

164.

165.

166.

167.

Pons, S.; Khoo, S.B.; Bewick, A.; Datta, M.; Smith, J .J .; Hinman, A.S.;
Zachrnann, G., J. Phys. Chem, 1984, 88, 3575

Wadayama, T.; Sakurai, T.; Ichikawa, S.; Suetaka, W., Surf, Sci., 1988, 198, L359
Hahn, F.; Melendres, C.A., Electrochim. Acta, 2001, 46, 3525-3534
Zhu, Y.; Uchida, H.; Yajima, T.; Watanabe, M., Langmuir, 2001, I 7, 146-154

Chen, Y.X.; Miki, A.; Ye, S.; Sakai, H.; Osawa, M., J. Am. Chem. Soc., 2003,
125(13), 3680-3681

Yajima, T.; Uchida, H., Watanabe, M., J. Phys. Chem. B, 2004, 108, 2654-2659

Shiroishi, H.; Ayato, Y.; Kunimatsu, K.; Okada. T., J. Electroanal. Chem, 2005,
581(1), 132-138

Shao, M.H.; Adzic, R.R., Electrochim. A eta, 2005, 50(12), 2415-2422
Futamata, M.; J. Electroanal. Chem, 2003, 55 0-55 1, 93-103
Watanabe, M.; Zhu, Y.; Uchida, H., J. Phys. Chem. B, 2000, 104, 1762-1768

Samjeske, G.; Miki, A.; Ye, S.; Yamakata, A.; Mukouyama, Y.; Okamoto, H.;
Osawa, M., J. Phys. Chem. B, 2005, 109, 23509-23516

Samjeske, G.; Osawa, M., Angew. Chem. Int. Ed., 2005, 44, 5694-5698

Miki, A.; Ye, S.; Senzaki, T.; Osawa. M.. J. Electroanal. Chem, 2004, 563(1), 23-
31

Noda, H.; Ataka, K.; Wan, L.J.; Osawa. M.. Surf Sci., 1999, 427-428, 190-194

Kunimatsu, K.; Uchida, H.; Osawa, M.; Watanabe, M., J. Electroanal. Chem,
2006, 587, 299-307

Kunimatsu, K.; Senzaki, T.; Tsushima, M.; Osawa, M., Chem. Phys. Lett., 2005,
401, 451-454

Miyake, H., Ye, S.; Osawa, M.. Electrochemistry Communications, 2002, 4,
973-977

Thiel, P.A.; Madey, T.E., Surf. Sci. Rep., 1987, 7, 211

Thiel, P.A.; DePaola, R.A.; Hol‘l‘mann. F.M.. J. Chem. Phys., 1984, 80, 5326

243

168.

169.

170.

171.

172.

173.

174.

175.

176.

177.

178.

179.

180.

181.

182.

183.

184.

185.

186.

187.

Ataka, K.; Yotsuyanigi, T.; Osawa, M., J. Phys. Chem, 1996, 100, 10664-10672
Brink, G.; Falk, M., Can. J. Chem, 1970, 48, 2096

(a) Walrafen, G.E., J. Chem. Phys., 1962, 36, 1035; (b) Walrafen, GE, .1. Chem.
Phys., 1971, 55, 768

Jonsson, B.; Chem. Phys. Lett., 1981, 82. 520
Parsonage, N.G.; Nicholson, D., .1. Chem. Soc., Faraday Trans. 2, 1987, 83, 663
Lee, C.Y.; McCammon, J .A.; Rossky, P.J., J. Chem. Phys., 1984, 80, 4448

(a) Nagy, G.; Heinzinger, K., J. Electroanal. Chem, 1990, 296, 549; (b) Nagy, G.;
Heinzinger, K., J. Electroanal. Chem, 1992, 32 7, 25

Nagy, G.; Heinzinger, K.; Spohr. E., J Chem. Soc. Faraday Discuss., 1992, 94,
307

Glosli, J .N.; Philpott, M.R., .1. Chem. Phys., 1992, 96, 6962
Xia, X.; Perera, L.; Essemann, E.; Berkowitz, M.L., Surf Sci., 1995, 335, 401
Doering, D.; Madey, T.E., Surf Sci., 1982, 123, 305

Russell, A.E., Lin, A.S.; O’Grady, W.E., J. Chem. Soc. Faraday Trans, 1993, 89,
195

Nakamoto, K., ‘Infrared and Raman spectra of Inorganic and Co-ordination
Compounds’, 4th ed., Wiley, New York. 1986

Wagner, F.T.; Moylan, T.E., Surf. Sci., 1987, 182, 125
Wagner, F.T.; Moylan, T.E., Surf Sci., 1989, 216, 361
Ataka, K.; Osawa, M., Langmuir, 1998. 14. 951-959

Pretsch, E.; Seibl, J .; Clerc, D.; Simon, W., ‘Tables of Spectral Data for Structure
Determination of Organic Compounds (2"d ed.)’, Springer-Verlag, 1985, pp 185

Cassanas, G.; Morssli, M.; Fabregue, E.; Bardet, L., Journal of Raman
Spectroscopy, 1991, 22, 409-413 ‘

Vassilev, P.; Van Santen, R.A.; Koper. M .T.M., J. Chem. Phys., 2005, 122, 054701

Akiyama, R.; Hirata, F., J. Chem. Phys., 1998, 1 08(12), 4904-4911

188.

189.

190.

191.

192.

193.

Kitamura, F.; Nanbu, N.; Ohsaka. T.; Tokuda, K.. J. Electroanal. Chem, 1998,
452, 241-249

Zheng, W.; Tadjeddine, A., J. Chem. Phys., 2003, 119(24), 13096-13099

Wandlowski, T.; Ataka, K.; Pronkin. S.; Dicsing, D., Electrochim. Acta, 2004,
49(8), 1233-1247

F utamata, M.; Luo, L.; Nishihara, C.. Surf Sci., 2005, 590, 196-211

Zhang, Z. ‘Aqueous-phase hydrogenation of biomass derived lactic acid to
propylene glycol’, Ph.D. Thesis

Yan, Y.G.; Li, Q.X.; Huo, S.J.; Ma. M.; Cai. W.B.; Osawa, M., J. Phys. Chem. B,
2005, 109, 7900-7906

 

IIIIIIIIIIIIIIIIIIIIIIIIIIIIII

llllllllllllllllllllllllllllllllllllllllllllllllllll