"41‘ A. .
A .1.qu
.

.33. 1..

 

‘ u 55x
5.2.5:.
A. . i...

 

ihik‘t:

Date

This is to certify that the
thesis entitled
EFFECTS OF VICINAL FUNCTIONAL GROUPS ON
THE AQUEOUS PHASE HYDROGENATION
OF LACTIC ACID DERIVATIVES
presented by

Norbert Varga

has been accepted towards fulfillment
of the requirements for

MS degree in Chemistry

Q; {Q 41.

Major p fessor

11/12/02

 

0-7639

MS U is an Affirmative Action/Equal Opportunity Institution

 

 

 

 

 

 

 

LIBRARY
Michigan State
University

 

 

 

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

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 cJClRC/DateDuepGS-pJS

 

 

 

 

IIAL 4

 

 

EF

EFFECTS OF VICINAL FUNCTIONAL GROUPS ON
THE AQUEOUS PHASE HYDROGENATION
OF LACTIC ACID DERIVATIVES
By

Norbert Varga

A THESIS

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

MASTER OF SCIENCE
Department of Chemistry

2002

 

 

"31?
hon
hetei

and 1

acid:
1201;
PTOp

prOp

CQn-c

hfrdr

ABSTRACT

EFFECTS OF VICINAL FUNCTIONAL GROUPS ON
THE AQUEOUS PHASE HYDROGENATION
OF LACTIC ACID DERIVATIVES
BY

NORBERT VARGA

Organic acids bearing various functional groups have demonstrated greatly
varying reactivity in catalytic hydrogenations. The goal of this project was to determine
how reactivity and selectivity of a given organic acid or acid ester in aqueous phase
heterogeneous catalytic hydrogenations are affected by different electron-withdrawing
and hydrogen-bonding or sterically hindering vicinal substituents

The Ru/C-catalyzed aqueous-phase hydrogenation reactions of various organic
acids and esters were run in a high pressure Parr batch reactor at 423, 373 and 348 K and
1200 psi pressure. Nine compounds, namely (1) propionic acid (PA) , (2) 2-chloro-
propanoic acid (2CPA), (3) lactic acid (LA), (4) glycolic acid (GA), (5) 2-methoxy-
propanoic acid (2MPA), (6) methoxyacetic acid (MA), (7) 2-acetoxypropionic acid
(2APA), (8) isobutyric acid (IBA), and (9) ethyl lactate (EL), were studied by 'H NMR
spectroscopic and HPLC chromatographic methods. Substrate and alcoholic product
concentrations were monitored with respect to time and the resulting data was used to
determine percent conversion of a given organic acid and percent production of the
corresponding alcohol. The relationship between molecular structure and efficiency of

hydrogenolysis was determined from the selectivity and rate data of each reaction.

Copyrighted by
NORBERT VARGA
2002

 

mm

.m gain... 5 .Xh.

r

Acknowledgements

Great appreciation goes to my advisors, J. E. Jackson and D. J. Miller for providing
scientific and financial support with this thesis. Special thanks to my colleagues who
provided help when it was needed. Most importantly, my deepest gratitude goes to my
friend T. T. O’Kronley who stood by me all along for moral support and helped selflessly

in this thesis.

Vae.’ Hoc diploma Anglice scriptum est!

iv

 

TH;

 

 

US
US
INT

C H;

EX

AI
Bl

TABLE OF CONTENTS

LIST OF TABLES .................................................................
LIST OF FIGURES AND IMA'GES......... ..

LIST OF ABBREVIATIONS........................

CHAPTER 1

1.1 Aqueous Phase Heterogeneous Cataly31s
........16
....31

1.2 Hydrogenation Esters.
1.3 Hydrogenation of Organic Acids

CHAPTER 2

RESULTS AND DISCUSSION”
2.1 Hydrogenation of Lactic Acid
2.2 Hydrogenation of Glycolic Acid

2.3 Hydrogenation of2- -Methoxypropan6ic Ac1d
........54
.......56

....58
.60

2.4 Hydrogenation of Methoxyacetic Acid”
2.5 Hydrogenation of Isobutyric Acid”
2.6 Hydrogenation of Propanoic Acid”

2.7 Hydrogenation of2-Chloropropanoic Acid”
....65

2.8 Hydrogenation of Ethyl Lactate”

2.9 Hydrogenation of2- Acetoxypropanolc Acid..............
....80

2.10 Compiled Results.
Conclusion...

EXPERIMENTAL .................................................................
General Outline for Hydrogenation Reactions. . . . . . . . . . . . . . . . . . ..
Procedure for Handling the Reactor

........86

....90

Analytical Methods...
Preparation of 2-Methoxypropan01c Acid

APPENDICES ......................................................................

....................... vi

vii

ix

....xi

1

.3

.41
.......43
....47

.50

67

...................... 83
........83

84

..................... 91

BIBLIOGRAPHY ...................................................................................... 1 15

 

 

 

 

List 0

Table 1
related

Table 1
catal} It

Table 2
1“ LINE:
H} drogi
generate
a relatix‘
SleOll.

Table 2.
h}dT0ge
H} drogg
generate
relatix'e];
MeOH.

Table 2,
RU C-ca'
The tale
Sfiction,

List of Tables

Table 1.1. Heats and free energies of formation for diethyl oxalate and
related compounds at 1 atm ............................................................................................... 27

Table 1.2. Summary for the hydrogenation of various organic acids,
catalyzed by rhenium black.25 ............................................................................................ 33

Table 2.1. Compiled 1H NMR results of Ru/C-catalyzed aqueous phase
hydrogenations. This table only Shows the main products of reactions.
Hydrogenation of ZAPA also generated HOAc and EL hydrogenation
generated large quantity of EtOH. Values designated with a * indicate

a relatively large degree of uncertainty. Reactions of 2MPA also produced
MeOH, POL and residual PG ........................................................................................... 73

Table 2.2. Compiled HPLC results of Ru/C-catalyzed aqueous phase

hydrogenations. This table only shows the main products of reactions.

Hydrogenation of ZAPA also generated HOAc and EL hydrogenation

generated large quantity of EtOH. Values designated with a * indicate a

relatively large degree of uncertainty. Reactions of 2MPA also produced

MeOH, POL and residual PG. The GA was not analyzed by HPLC .............................. 74

Table 2.3. The calculated 1H NMR results for the aqueous phase

Ru/C-catalyzed hydrogenation of organic acids and esters at 423 K and 1200 psi.

The calculations are described in the Analytical Methods of the Experimental

section. The relative rates of product formation are the ratios of k(product) .................. 79

vi

List of Figures and Images

Figure 1.1. The schematic representation of the addition of hydrogen to
the substrate as the molecule enters the porous catalyst and adsorbs on its
internal surface.6 ................................................................................................................ 7

Figure 1.2. Two modes of adsorption of unsaturated alcohols to the catalyst
surface: (1) Ni interacts with the lone pair of electrons of the hydroxyl
functionality, (2) Pd interacts with carbinol carbon by haptofillic affect.7’8 ...................... 8

Figure 1.3. On top, adsorption is controlled by the relative position of the

large substituent on the chiral center. On the bottom, hydrogen bonding

controls the addition of hydrogen, but just as in the top figure, the less

hindered side adsorbs to the catalyst surface.6 ................................................................... 9

Figure 1.4. The direct attachment of the formyl group to the catalyst
surface.3S ........................................................................................................................... 18

Figure 1.5. Mechanism for the gaseous byproduct formation from diethyl
oxalate.46 ........................................................................................................................... 28

Figure 2.1. A magnified region of the lactic acid mixture showing the doublets
of the corresponding acid and esters ................................................................................. 44

Figure 2.2. Selected 1H NMR spectra of LA hydrogenation in which PG is
produced. For stacked spectra detailing the LA hydrogenation each hour, please
see Appendix 1.1 ............................................................................................................... 45

Figure 2.3. Spectrum (1) is the starting material (GA) and its cyclic ester.
Spectrum (11) shows EG, the final product of GA hydrogenation after Six hours ........... 48

Figure 2.4. Intra-molecular hydrogen bonding of glycolic and lactic acid have a
relative fixed geometry compared to propanoic acid ........................................................ 50

Figure 2.5. 'H NMR spectrum of the hydrogenation of 2MPA. The reaction
produced a mixture of 2MPOL and 2MPA after six hours ............................................. 52

Figure 2.6. lH NMR spectra of 2MPA hydrogenation in which 2MPOL is
produced along with small amount of PG and residual amount of MeOH and POL.

For stacked 1H NMR spectra detailing 2MPA hydrogenation each hour, please see
Appendix 3.1 ..................................................................................................................... 53

vii

THL-z

 

 

 

 

Figure 2.7. 1H NMR spectra of MA hydrogenation. Spectrum (1) shows the starting
material and spectrum (11) corresponds to the reaction mixture containing MA,
2METOH and MeOH after six hours. Stacked spectra detailing MA hydrogenation
each hour are found in Appendix 4.1 ................................................................................ 55

Figure 2.8. 1H NMR spectra of IBA hydrogenation. Spectra (I) shows the starting
material, spectra (11) is the mixture of IBA and IBUOL after six hours. For stacked
1H NMR spectra detailing IBA hydrogenation, see Appendix 5.1 ................................... 57

Figure 2.9. 1H NMR spectrum of PA hydrogenation. Spectrum (1) shows the
starting material, spectrum (11) corresponds to the mixture of PA and POL. For
stacked spectra detailing PA hydrogenation, please see Appendix 6.1 ............................ 59

Figure 2.10. (I) is the 1H NMR spectrum of 2CPA pre-reaction mixture. (II) is
taken after six hours and is the spectrum of PA For stacked 1H NMR spectra
detailing 2CPA hydrogenation, see Appendix 7.1 ........................................................... 61

Figure 2.11. lH NMR spectra of EL hydrogenation. Spectrum I is EL and
spectrum 11 is the mixture of PG and ethanol after a six hour reaction time. For
stacked 1H NMR spectra detailing EL hydrogenation, see Appendix 8.1 ........................ 66

Figure 2.12. Hydrogenation of 2APA at 423 K and 1200 psi produces PG,

ethanol and HOAc. Spectrum I is the starting material (2APA) and Spectrum

11 is the product mixture after six hours. For stacked 1H NMR spectra of 2APA,

see Appendix 9.1 ............................................................................................................... 68

Figure 2.13. Percent conversion of LA, IBA, 2MPA, MA, GA, PA, EL and
2APA in aqueous phase hydrogenation reactions at 423 K and 1200 psi, as
determined by 1H NMR .................................................................................................... 75

Figure 2.14. Percent conversion of LA, IBA, 2MPA, MA, PA and EL in aqueous
phase hydrogenation reactions at 423 K and 1200 psi, as determined by HPLC ............. 76

Figure 2.15. Percent production of LA, IBA, 2MPA, MA, GA, PA EL and
2APA and 2CPA ion aqueous phase hydrogenation reactions at 423 K and 1200
psi, as determined by HPLC ............................................................................................. 77

Figure 2.16. Percent production of LA, IBA, 2MPA, MA, PA, EL and 2APA in
aqueous phase hydrogenation reactions at 423 K an d 1200 psi, as determined by
HPLC ................................................................................................................................ 78

Image 1. The Series 4560 Parr Mini Reactor interfaced with a control panel that
monitors both pressure and temperature ........................................................................... 86

viii

List of Schemes

Scheme 1.1. Chemoselective hydrogenation of an aromatic carboxylic acid to
an aldehyde.6 ...................................................................................................................... 5

Scheme 1.2. Selectivity determined by the catalyst. Platinum hydrogenates
exclusively the olefin moiety of the molecule whereas Zn or Fe selectively
hydrogenates the carbonyl carbon.6 ................................................................................... 6

Scheme 1.3. Chemoselectivity influenced by steric hindrance.6 ....................................... 6

Scheme 1.4. Enantioselective hydrogenation of diketo ester to a B-hydroxy
ester by cinchonidine-modified Pt/Ale3 catalyst ............................................................. 10

Scheme 1.5. Proposed mechanism for the hydrogenolysis of methyl formate to
form two moles of methanol ............................................................................................. 17

Scheme 1.6. Adkins hydrogenated esters with 1.5 eq. of copper-chromium
oxide at 330 atm of H2 and 398 K, producing the corresponding alcohols with
80% yield .......................................................................................................................... 21

Scheme 1.7. The selected reactions by Adkins show that copper-chromium oxide
selectively reduces the phenyl-substituted ester at the carbonyl functionality,

whereas W-6 Raney nickel, or all tested Raney nickels indiscriminately reduced

all unsaturated carbon-carbon bonds, resulting in cylcohexyl functionalities .................. 22

Scheme 1.8. The hydrogenations of esters produced yields in excess of 80%.
Even though the temperatures did not exceed 373 K, reaction conditions required

1.5 equivalent of W-6 Raney nickel catalyst and 330 atm pressure of H2 gas ................. 24

Scheme 1.9. The two step process of hydrogenolysis of diethyl oxalate to
ethylene glycol .................................................................................................................. 26

Scheme 1.10. The hydrogenation at two different pressures, producing two
different results ................................................................................................................. 29

Scheme 1.11. Solvent effects strongly influence the hydrogenation of lactic acid
and succinic acid, using rhenium heptoxide reduced in situ. 5 ......................................... 34

Scheme 1.12. The hydrogenation of lactic acid, 2-chloropropanoic acid, malic acid
and tartaric acid gave good to moderate yields. The stereochemistry of the substrate

was largely retained after hydrogenation resulting in high ee’s ....................................... 38

Scheme 1.13. The proposed mechanism of lactic acid hydrogenation.2 ......................... 39

ix

 

 

 

 

Scheme 2.1. Hydrogenation of lactic acid to generate propylene glycol ........................ 43

Scheme 2.2. Hydrogenation of glycolic acid that generates ethylene glycol.

Scheme 2.3. The hydrogenation of 2MPA to generate 2MPOL .................................... 51
Scheme 2.4. Hydrogenation of methoxyacetic (MA) to generate 2-methoxyethanol
(2METOH) ....................................................................................................................... 54
Scheme 2.5. Summary of 2CPA hydrogenation reactions at varying temperatures and
under controlled conditions ............................................................................................... 64
Scheme 2.6. Hydrogenation of 2CPA at 348 and 423 K and 1200 psi ........................... 67

Scheme 2.7. The reaction mechanism for alcohol formation from (it-substituted
organic acids in aqueous medium under heterogeneous catalytic conditions ................... 71

 

(LAP
(PCP
(lbw
(251E
(GAP
(EC)-
(ZNIP.
(251Pl
CHfl)
{Efl31
UBAI
UBLC
(PAP-
(POL)
QCPA
QCPO
(ELLA
QAPA
(HOAc
(THF )-
IHDO ).
(1)30)“
(HChh.

List of Abbreviations

(LA)—--lactic acid
(PG)---propylene glycol
(MA)---methoxyacetic acid
(2MEOH)---2-methoxyethanol
(GA)---glycolic acid

(EG) ---ethylene glycol
(2MPA)---2-methoxypropanoic acid
(2MPOL)---2-methoxypropanol
(MeOH)---methanol
(EtOH)---ethanol
(IBA)---isobutyric acid
(IBUOL)---isobutanol
(PA)---propanoic acid
(POL)---n-propanol
(2CPA)---2-chloropropanoic acid
(2CPOL)---2-chloropropanol
(EL)—-—ethy1 lactate
(2APA)--—2-acetoxypropanoic acid
(HOAc)---acetic acid
(THF)---tetrahydrofuran
(HDO)---partially deuterated water
(D20)---deuterium oxide (deuterated water)
(HCl)---hydrochloric acid

xi

 

IL I.
.9

V1

 

 

 

 

 
 

Introduction

Organic acids bearing various functional groups have demonstrated greatly
varying reactivity in catalytic hydrogenations. The goal of this project was to determine
how reactivity and selectivity toward hydrogenation of a given organic acid or acid ester
are affected by different electron-withdrawing and hydrogen-bonding or sterically
hindering vicinal substituents. The reactions in this work were run in aqueous medium
under mild conditions, using a carbon-supported ruthenium catalyst, to generate value-
added chemicals, such as propylene glycol which can be produced by lactic acid
hydrogenation. These organic acids and esters are also important feedstocks of the
renewable resource-based chemical industry that uses environmentally friendly chemical
processes also termed “green chemistry”.

The existence and constant evolution of green chemistry is a result of an
environmental consciousness that has impacted all strata of society. Although much
work remains to be done to improve the environment, political attitudes, practices in the
chemical industry and consumer habits have been making tremendous progress. While
the demand for synthetically produced material has increased, the chemical industry has
been finding alternative ways that are not only economically sound but are within the
realm of “benevolent” chemical processes. The renewable resource-based chemical
industry is in the vanguard of these efforts.

Organic acids compose a major class of renewable-resource feedstock chemicals.
They are commonly obtained from fermentation of biomass-derived glucose, which

comes in turn from starch h drol sis.l Thou h corn is currently the most im ortant
Y y g P

rHE'

 

 

 

starch CF

imp‘m‘lr

because
SIECOI' V
block 1“”
Ii
org:mic 5
hydrogen
highl} 9”
Miller 101
and [he Cf;
{I16 much I
because bx‘
published I
both.2

TIF
by Jackson E
h) drogenatit

OCH 3 0r (Mt

substituents. 01

starch crop, other grains as well as non-cereal crops (e.g. potatoes) represent other
important sources of biomass-derived carbohydrate starting materials.l

The hydrogenated derivatives of organic acids and esters have great importance
because they can serve as building blocks of many syntheses. For example, propylene
glycol, which is a hydrogenation product of lactic acid, is a useful molecular building
block for the syntheses of a number of pharmaceutical compounds.

Hydrogenolyses of esters often involving stepwise processes and the use of
organic solvents, have been successfully practiced in industry. While esters undergo
hydrogenolysis more easily than organic acids, lactic acid (LA) has been an exception. Its
highly efficient direct hydrogenolysis, relative to propanoic acid, prompted Jackson and
Miller to probe the relationship between the molecular structure of various organic acids
and the efficiency of hydrogenation under mild conditions.2 Their research investigates
the much preferred aqueous condensed-phase processes over vapor-phase processes
because both the feed purity and temperature requirements are less demanding than many
published hydrogenations that have been run at either high temperatures or pressures or
both.2

The ease of aqueous phase Ru/C-catalyzed hydrogenation of LA , demonstrated
by Jackson and Miller, laid the ground work for further investigation into the
hydrogenation of organic acids bearing a vicinal substituent X (X: H, CH3, Cl, OH,
OCH3 or OAc). The goal of this thesis was to elucidate the relative importance of

substituents on the efficiency of aqueous phase Ru/C-catalyzed hydrogenation.

 

an .-a.2.. 2:51.:

 
   

.L z

.11

Chapter 1

1.1. Aqueous Phase Heterogeneous Catalysis

 

The essential component in the hydrogenation of an organic acid or ester is the
catalyst which can be one of two types: (1) homogeneous or (2) heterogeneous. It is
important to note that a complete review of both types of catalysts even in the context of
ruthenium, the only catalytic species used in this work, is virtually impossible, given the
vast amount of knowledge that has been amassed throughout the years in the science of
catalysis. Hence only a limited review of heterogeneous catalysis directly pertaining to
this work is provided here.

Although stereochemical control remains a great challenge in synthetic organic
chemistry, optimization of reactivity and selectivity to targeted products are of primary
importance, in order to afford an efficient and practical use of these reactions.3
Homogeneous catalysts have served as useful agents for selective hydrogenations. They
must meet two criteria for efficient synthesis: (1) they must possess a high turnover
number (TON), (2) high turnover frequency (TOF).3 Equally important is our ability to
handle them easily and safely that conforms to environmentally friendly protocol.3 With
this in mind, ruthenium metal, which is the cheapest precious metal, is an excellent

candidate as it has demonstrated superb catalytic activity.3

The molecular fine tuning of homogeneous catalysts allow ever increasing stereo
control in selective hydrogenations and maximization of product yields.4 Tuning
heterogeneous catalyst on the other hand is more challenging. The question then arises;
“Why use heterogeneous catalysts?” As opposed to homogeneous catalysts they are easy
to handle and are also easy to separate from the reaction mixture. In addition, they can
also be recovered and regenerated after use which makes them immensely appealing to
industry.4

As opposed to homogeneous catalysis, in which reactants and the catalyst are in
one phase (liquid-phase), heterogeneous catalysis involves reactants and the catalyst that
are in different phases.5 This necessarily means that heat and mass transfer must pass .
through different phases for reaction to occur.5 For instance, regardless of the activity of
the catalyst, a hydrogen molecule must ultimately reach either the external or internal
surface of the catalyst where hydrogenation takes place.5 Following an H2 molecule,
one could see that eight essential steps must happen in the heterogeneous catalytic
hydrogenation reaction. Firstly, the H2 molecule must diffuse from the bulk gas phase
to the gas/liquid interface; (2) then it must be absorbed and diffused in the bulk liquid
phase; (3) in the liquid phase the H2 along with other reactants must transfer to the
external surface of the catalyst particle; (4) the reactants must then diffuse into the porous
catalyst where (5) they adsorb on the catalyst surface; (6) at the surface of the catalyst
hydrogenation takes place; (7) the products desorb and transfer away from the catalyst
and (8) diffuse into the liquid phase or gas phase.5

Selective hydrogenation of a substrate is one of the main challenges in

heterogeneous catalytic reactions. Of particular concern is the development of effective

THE?! '4-

 

 

-
.
r
I'm-l.-' 4
I ‘ l

 

 

 

systems that can control either enantioselectivity or diastereoselectivity.6 In addition,
great interest lies in the chemoselective hydrogenation of unsaturated acids and esters to
the corresponding unsaturated alcohols.6 One example of chemoselectivity is the

heterogeneous catalytic hydrogenation of a carboxylic acid to an aldehyde (Scheme 1.1).6

CT/ZTOZ
0 +7 0
o 350 °C 0
Gas Phase
OH

Scheme 1.1. Chemoselective hydrogenation of an aromatic carboxylic acid to an
aldehyde.6

 

In this reaction two types of selectivity are of interest: (1) the preference toward
reducing the carbonyl carbon instead of the unsaturated carbons in the phenyl rings and
(2) partial hydrogenation of the carboxylic acid to the corresponding aldehyde instead of
an alcohol. The synthetic importance of partial hydrogenation is manifested in
conversion of an alkyne to a cis alkene instead of the generation of saturated species.6
The choice of catalyst is a determining factor in chemoselectivity and the result is
normally the formation of two different products from the same starting material (Scheme
1.2).6

Steric hindrance can also influence the selectivity, normally resulting in
preferential hydrogenation of one site over the other. For example, hydrogenation by
carbon-supported iridium under identical reaction conditions results in higher product
yields for a,B-unsaturated alcohols as the steric bulk close to the olefinic moiety

increases (Scheme 1.3).6

 

H3C / 0 CH3 H
\l/Y H30 OH
CH3 H PtOz, Zn(II), Fe(II) ’
H2 at 30 atm

Scheme 1.2. Selectivity determined by the catalyst Platinum hydrogenates exclusively
the olefin moiety of the molecule whereas Zn or Fe selectively hydrogenates the carbonyl
carbon.6

 

 

 

0 lr/C EtOH OH
M , . N
/ 373 K, 65 atm /

Acrolein 73 %

Mo Ir/C, EtOH > MW

373 K, 65 atm

 

 

   
 

Crotonaldehyde 83 %
\ \O Ir/C, EtOH \ OH
373 K, 65 atm
Cinnamylaldehyde 96 %

Scheme 1.3. Chemoselectivity influenced by steric hindrance.6

The stereochemistry of the product is greatly influenced by the adsorption of the
substrate on the catalyst surface. As shown in Figure 1.1, molecule can be hydrogenated
from the top or the bottom face of the alkene functional group.6 The facial

discrimination is relative to which side of the molecule adsorbs on the catalyst surface.6

Hydrogenation from the "top" face

R ‘n'

H
Trans Product

 
 

Internal catalyst surface

H2

CH2

R ‘1'
H R ‘.'
V

H dro enation from the "bottom" face

 

Cis Product

Figure 1.1. The schematic representation of the addition of hydrogen to the substrate as
the molecule enters the porous catalyst and adsorbs on its internal surface.6

Vicinal bulky substituents and hydrogen bonding functional groups, such as
hydroxyl or amine groups, can determine the direction of adsorption on the catalyst
surface.7’8 For example, the hydroxyl group can interact with the catalyst surface to
determine from which direction the alkene will be adsorbed to the catalyst surface
(Figure 1.2) 6 The mode of interaction between the catalyst’s orbitals and the hydroxy
group is believed to depend on the type of metal as well. For example, it has been

proposed that the orbitals of a nickel catalyst interact with the lone pair of electron of the

hydroxyl group, giving trans product, whereas the palladium’s orbitals interact with the

carbinol carbon, giving cis product (Figure 1.2).9

H30 “ H
pd Pd/C

 

Pd surface

 

Figure 1.2. Two modes of adsorption of unsaturated alcohols to the catalyst surface: (1)
Ni interacts with the lone pair of electrons of the hydroxyl functionality, (2) Pd interacts
with carbinol carbon by haptofillic effect.7‘8

Adsorption to the catalyst surface can influence the diastereoselectivity as well.
The diastereoselective hydrogenation of ketones is influenced by the presence of a chiral
center of the molecule that can direct the hydrogen addition to the substrate.6 It has been
proposed that Cram’s Rule can be applied in heterogeneous catalytic hydrogenations to
explain the stereoselectivity of hydride transfer to the carbonyl functionality. In this
process the hydrogen is transferred to the carbonyl carbon from the less sterically

hindered face of the molecule (Figure 1.3).6‘IO The degree of adsorption to the catalyst

surface is influenced by the relative size of the other two substituents attached to the

chiral center.6 Hydrogen bonding of a carbonyl with a vicinal -—OH or —NH group can
influence the direction of hydrogen addition.6 In principle, the molecule is fixed in one
plane and the adsorption takes place from the less sterically hindered side where the

smaller substituent is situated (Figure 1.3).6

 

Catalyst surface

Ketone with vicinal chiral center

 

 

 

 

 

M. 9H
R E S
Cat. _ s =. H
H2
RL R
Ketone with vicinal chiral center H-bonded through -OH
H‘
o/
0
Cat.
RN“ H2
Rs R

 

 

 

Figure 1.3. On top, adsorption is controlled by the relative position of the large
substituent on the chiral center. On the bottom, hydrogen bonding controls the addition

of hydrogen, but just as in the top figure, the less hindered side adsorbs to the catalyst
surface.

Chiral directing groups that reside on the substrate molecule can be essential for the

diastereoselectivity in such hydrogenation reactions. Although Cram’s Rule has worked

1H:

 

 

 

 

 

 

well for predit
reaction at the=

I11 IIlC I
performed usi-
Altematix‘el);
enantioselecti'.

hithout h} (In

 

Scheme 1.4.
Cinchonidinc-,.'

Althou

 

0ng from the'
too.4 One oft
inside the reac
profoundly aft
h3drogenation
selectivity for

10H .
“rpm.“

well for predicting stereoselectivity in homogeneous catalytic reactions, the details of the
reaction at the catalyst surface in heterogeneous catalysis have not been elucidated.

In the early fifties, enantioselective, heterogeneous catalytic hydrogenation was
performed using metal catalysts with quartz or silk fibroin serving as chiral support.l "'2
Alternatively, Cinchonidine-modified Pt/AI2O3 catalysts can be employed to

enantioselectively hydrogenate a generic diketo ester to the corresponding R ester

without hydrolysis (Scheme 1.4).6‘4

0\ Pt, H2
Cinchonidine

 

R

 

0

Scheme 1.4. Enantioselective hydrogenation of diketo ester to a B-hydroxy ester by
Cinchonidine-modified Pt/AI2O3 catalyst.

Although the choice of a heterogeneous catalyst is vital, mass transfer limitations
of H2 from the gas phase to the liquid phase can severely influence enantioselectivity
too.4 One of the ways to minimize diffusion limitations is by varying the stirring speeds
inside the reactor.4 Blackmond et a1. has demonstrated that increasing the stirring speed
profoundly affects the optical yields for the Pt/Al2O3-catalyzed liquid-phase

hydrogenation of ethyl pyruvate to R and S ethyl lactate.”l6 In their work the

selectivity for (R)-lactic acid increased as the stirring speed was increased from 400 to

1000 rpm.‘3

10

“I
T
q

t. m

 

 

 

 

 

 

 

With an increase of H2 pressure, from 1 to 40 atm, the enantiomeric excess also

17

increased for the R enantiomer. This preference for the R enantiomer was believed to

be due to the difference in the rate determining step for the formation of the R and S

7 An alternative explanation for this selectivity lies in the assumption of a

enantiomers. I
steady-state approximation that allows the determination of the surface concentration of
the ethyl pyruvate that is adsorbed on modified catalytic sites.4‘15

The reaction rate in heterogeneous catalytic hydrogenations is another important
factor. Mass transfer limitations, however, can skew rate determinations. To avoid this
problem, the absence of all transport limitation is assumed to determine the kinetics of
these reactions. Based on this assumption, the rate of reaction increases with an
increasing number of active sites, if the reaction is not sensitive to the structure of the
catalyst.4 Hence, there is a linear relationship between the active site density and the
number of metal atoms on the surface.4 In the hydrogenation of citral, Singh et al.
compared the catalytic activity of Group VIII metals supported on SiO2, A1203 and T102
and found that Pd on silica support shows the greatest catalytic activity at 300 K and
1 atm, when all mass transport limitations are absent.18 Tests were also carried out by
TiO2-supported Pt, and they found that the hydrogenation rate is strongly influenced by
the particle size of the Pt catalyst due to metal-support interaction.18

Just as in organic synthesis, solvent effects have an impact on heterogeneous
catalytic hydrogenation reactions as well. Although these effects are incompletely
understood, the rational behind them is based on systematic correlations of solvent
19.20

polarity or dielectric constant with the rates of reactions and distribution of products.

Because the solvent can interact with the metal support, unraveling the solvent effects in

11

 

 

 
 

these reactions are more difficult as compared to homogeneous catalysis. However, it
was found that polar and nonpolar substrates hydrogenate at different rates in solvents of
varying polarity.

Polar solvents seem to increase the adsorption of non-polar substrates to the
catalyst surface whereas non-polar solvents increase the adsorption of a polar reactant to
the catalyst surface.“’6 In a study of competitive hydrogenation of cyclohexene and
acetone, Scholten et al. demonstrated that the selectivity of Ru-catalyzed liquid-phase
hydrogenation of cyclohexene to cyclohexane was enhanced by addition of ethanol.4
Conversely, the addition of cyclohexane increased the selectivity in the hydrogenation of
acetone to isopropanol.4‘21 Furthermore, addition of ethanol in acetone hydrogenations

produced poor yields.4"21

They proposed that ethanol strongly increased the rate of
desorption from the catalyst surface and simultaneously impeded the rate of
hydrogenation. 2‘ This finding was also confirmed in the gas-phase hydrogenation of
benzene in which ethanol was introduced. Using F TIR spectroscopic methods to
characterize the interaction between the alcohol and the olefin, they observed that the
—OH group of ethanol formed a hydrogen bonding adduct with the alkene functional
group of the substrate.4 Based on this finding, they proposed that the increased
desorption rate was due to a weakened overlap of the olefin 7t electrons with the Ru
catalyst.4‘2"22 In addition, the synergy between H2 concentration and the solvent effect
has profound impact on the concentration of the adsorbed hydrogen on the catalyst
surface. Using 5% PVC at 300 K and 1 atm, Cerveny et al. demonstrated a linear

relationship between the concentration of adsorbed hydrogen on the catalyst and liquid-

phase hydrogen concentration, optimized in various solvents.4

12

THI

 

 

 

 

 

 

 

 

is the tempera

91111101!

it repont’ti ill

  
    

l'sing citral 11>
catalyst. their"
are easier to h"
kinetic studies
temperature ru
product distril
does not at al'
found that Iii;
conrersion at
298 K to 423
also ohsen er
hldrogenatio
benzene also
observed an e
actit'ity,‘
Reacti
Useful to h} dr,
[empt‘ratures (I

Furthemiore. c

P118102 0r Pt ‘I

 

Although a review of the kinetic descriptions of heterogeneous catalysis will not
be reported in this thesis, a few highlights pertinent to this project need to be addressed.
Using citral as the target model for parallel hydrogenation reactions by supported Pt
catalyst, thermodynamic considerations suggest that non-conjugated terminal C=C bonds
are easier to hydrogenate than conjugated C=C followed by C=O bond.4 Based on
kinetic studies of citral hydrogenation by Pt supported on TiO2 or SiO2, in the
temperature range of 298 to 423 K, it was found that increasing temperatures can effect
product distributions and it can also cause a peculiarity in the kinetics of this reaction that
does not at all adhere to simple Arrhenious behaviorf"18 Using Pt/SiO2 at 20 atm, it was
found that the turnover frequency for H2 uptake was significantly lower for citral
conversion at 423 K than at 298 K. This means that as the temperature is increased from
298 K to 423 K, the rate of hydrogenation significantly decreases.l8 Other groups have
also observed a significant decrease in the reactivity in heterogeneous catalytic
hydrogenation of benzene, which was thought to be due to a lower concentration of
benzene adsorbed on the catalyst surface at higher temperatures.4 Because Singh et al.
observed an activity minimum, they argued against the latter proposition of decreased
activity.4

Reaction kinetics can also be altered by the metal-support interactions and can be
useful to hydrogenate C=O bonds selectively.4 Group VIII metals were reduced at high
temperatures (over 700 K), and their interaction with TiO2 support was studied.
Furthermore, citral hydrogenation was also studied under similar conditions, using
Pt/SiO2 or Pt/TiO2 catalysts, and a twofold decrease in turnover frequency was observed

as the temperature was raised from 473 K to 773 K.4 Haller et al. proposed that this

13

unusual kinetic behavior is due a partial loss of oxygen from the support, rendering it a
partially unsaturated cation with its coordination bonds still intact.4 The oxygen vacancy
is the greatest at the metal-support interface that facilitates the partial reduction of the
support. The support can in turn migrate onto the metal catalyst surface and virtually
block the chemisorption of H2 on the metal.

If the support can have such a profound impact, how does the metal catalyst affect
the kinetics? The choice of metal appears to have the greatest affect. Virtually three
orders of magnitude of initial turnover frequency was observed amongst Group VIII
metals in citral hydrogenation.4 Using the silica support for each catalyst, the lowest
initial turnover frequency was observed for cobalt followed by nickel, rhodium,
ruthenium, osmium, iridium, platinum and palladium, respectively, at 300 K and 1 atm
pressure.4 This trend was attributed to percent (1 character of the metal that is defined as
the contribution of d electrons to the spd hybrid orbitals.4'23 The relationship between
the catalytic activity and the percent (1 character of a Group VIII metal is not well
understood.4 In addition, various group VIII metals generated different product
distribution in the hydrogenation of citral.4 Based on extended Huckel calculations,
Delbecq and Sautet proposed that this difference is due to the relative geometry and
strength of adsorption, which in turn depends on two important interactions: the relative
contributions of stabilizing two-electron donation/backdonation and four-electron
destabilizing interactions between the catalyst and the substrate.4 Finally, crystallite
size effects seem to affect the kinetics of some hydrogenation reactions.4 These effects
can be understood from the structure sensitivity of the turnover frequency.4 More

specifically, a change in the size of metal crystallite in the range of 1-10 nm and a change

14

in the available crystal plane can effectively decrease or increase the turnover frequency
of the catalyst.24

Hydrogenation reactions involving C-C, C-0 and C-N scission are sensitive to the
structure of the catalyst, whereas C-H bond formation and scission including
hydrogenation reaction are structure insensitive.4 Apparently, metal atoms crowd the
support surface with varying coordination that lead to inhomogeneity in the potential
energy surface. This phenomenon has rendered the explanation for structure
insensitivity elusive.4

Although heterogeneously catalyzed selective hydrogenations are useful there is
still plenty of work to be done in improving the diastereoselectivity and enantioselectivity
of these reactions. At present there is still a great potential in homogeneous catalysts
from the synthetic point of view as they can be molecularly tuned to improve both
stereoselectivity and product yields. The caveat though is how to remove homogeneous
catalysts effectively from the reaction mixture, minimize environmental hazards that they
may pose, and make their use more cost effective. Because heterogeneous catalysis can
circumvent these problems they can provide an excellent alternative to conventional

stoichiometric hydride reductions.

15

1.2. Hydrogenation of Esters

 

By the middle of the 1930’s numerous heterogeneous catalytic hydrogenations of
esters had been published. The economic feasibility of these processes were quite
questionable as they were done under stringent conditions, requiring high temperatures
and pressures up to 700 K and 350 atm and the use of organic solvents such as dioxane.25

The pioneering work of Christiansen to generate methanol via the hydrogenolysis
of methyl formate provided an alternative route to the common industrial process in

which CO was hydrogenated over Cu/Zn catalyst at 523 K and 99 atm.26

The catalytic
hydrogenation of methyl formate to form methanol, with AH (298K, latm) = -47.5
kJ/mol, is an exothermic process, making this alternative route thermodynamically
feasible.27 Christiansen developed a two step process that initially involved the
generation of a methyl formate intermediate (Equation 1) by reacting carbon monoxide
with methanol in the presence of a dissolved alkali metal such as sodium.28 From one
mole of formate two moles of methanol were produced by catalytic hydrogenation, using

a copper-based catalyst (Equation 2).26'28

CH3OH + C0 —+ CH3OCHO (Eq.1)

CH3OCHO + 2112 —, 2 CH3OH (Eq.2)

The generation of methanol from methyl formate was the focus of several studies.
Evans examined the kinetics of the copper chromite catalyzed hydrogenolysis of methyl

formate in the temperature range of 373 to 500 K at atmospheric pressure. He observed

16

 

met 90 0 0
He also Io
ratio of hj
C e
hydrogen
methanol
pressure:
C4428 10
kl mol u]
undergo:
cleax'agc
then pro(

15 L203:

HCH

Scheme

over 90 % selectivity to methanol while below 90 % conversion of methyl formatezg’30

He also found that the selectivity for methanol was not influenced by temperature or the
ratio of hydrogen to methyl formate.29‘30
Cerveny et al. demonstrated that the carbonylation of methanol and
hydrogenolysis of methyl formate could be carried out in a one pot reaction to generate
methanol in liquid phase in the temperature range of 373 and 500 K and at atmospheric
pressure.”3 ‘ Sorum and Onsager studied this reaction in the narrow temperature range
of 428 to 458 K and at 70 atm pressure and determined that activation energy is 53
kJ/mol under these conditions.32 They proposed a mechanism in which the reaction
undergoes a rate limiting formation of a hemiacetal intermediate followed by a rapid

cleavage to generate formaldehyde and methanol. The hydrogenation of formaldehyde

then produces one mol of methanol which is the net yield of this reaction (Scheme

1.5).2632

Hi:
0 H o
H3C/ T + H2* __> H3C/ \fiH
0 OH“)

O

* * H21: 1
HCHJOH" + CH30H" <—— /“\ + (21130110

H H*

Scheme 1.5. Proposed mechanism for the hydrogenolysis of methyl formate to form two
moles of methanol.

17

Monti et al. found that the very same reaction, carried out in a temperature range of 408
to 473 K and pressure of 17 to 45 atm, has an energy of activation of 62 kJ/mol.32
Although a relatively high selectivity for methanol was demonstrated in the gas phase
reaction, some degree of inhibition by CO was also observed.”33 This was not a great
concern, but they found that CO poisoning can be lowered by increasing the
temperature.34
Trimm et al. studied the hydrogenation of methyl formate by in—situ IR
spectroscopy, using silica-supported copper as catalyst.”35 They found that the rate of

conversion is directly related to the intensity of the IR absorbance at 1666 cm", assigned

to the formyl group directly bound to the copper (Figure 1.4).35

[III/fill!!!

Figure 1.4. The direct attachment of the formyl group to the catalyst surface.35

One of the earliest examples of the hydrogenation of higher esters was
demonstrated by Lazier.36 In the temperature range of 473 to 673 K and pressure range

of 50 to 250 atm, butyl alcohol was obtained from the hydrogenolysis of butyl butyrate

18

with high selectivity, using a copper oxide/zinc oxide catalytic mixture.36 The liquid-
phase hydrogenolysis of butyl acetate was compromised by competing trans-
esterifications in studies with barium-promoted copper chromite catalyst in the

temperature range of 450 to 530 K and pressure from 50 to 200 atm.”37

Grey et al.
described the use of an anionic ruthenium hydride complex for the hydrogenolysis of
esters run neat or using solvents such as toluene or THF. Although decarbonylation and
transesterification reactions compromised his reactions, producing low yields, he was
able to demonstrate the feasibility of mild conditions for hydrogenolysis (T = 373 K, P =
6.2 atm)”39

Ethyl acetate has been a primary target of hydrogenolysis. Two patents and
several studies describe the application of bimetallic rhodium catalyst (Sn/Rh) and Raney
Copper catalyst, claiming 90 % conversion to ethanol in the temperature range of 473 to
560 K and pressures of 1 to 50 atm.40 However, ethanol yields from ethyl acetate
hydrogenolysis (Eq.3) can be compromised under such conditions by the formation of

acetaldehyde which is in equilibrium with ethanol (Eq.4).26‘4O

CH3COOC2H5 + 2H2 ——* 2C2H50H (Eq. 3)

CZHSOH CH3CHO + H2 (Eq. 4)

 

Using Sn/Rh catalyst mixtures, Ferreti demonstrated that an increase of Sn/Rh ratio
increases the rate of hydrogenation of ethyl acetate to ethanol, but the competing

acetaldehyde formation is not affected.4| Furhermore, transesterification reactions were

19

observed during the hydrogenation of higher esters such as methyl propanoate and methyl
butanoate, at temperatures close to 500 K.

Mechanistically similar to the Christiansen hydrogenolysis of methyl formate,
Claus et al. provided an interesting route to ethanol formation. Using a copper catalyst,
methyl acetate was hydrogenated to methanol and ethanol at 40 atm and 500K. They
proposed that the first step involves the formation of ethanol and methanol (Equation 5).
Subsequently the ethanol undergoes a transesterification with another methyl acetate to
form ethyl acetate and methanol (Equation 6). Finally, ethyl acetate is reduced to two

moles of ethanol (Equation 7).42

CH3COOCH3 + 2H2 —_> CH3OH + CH3CH20H (Eq. 5)
CH3COOCH3 + CH3CH20H —> CH3COOCH2CH3 + CH3OH (Eq. 6)

CH3COOC2H5 + 2112 —> 2CH3CH2OH (Eq. 7)

Yan et al. and Evans et al. proposed that the mechanism of the hydrogenolysis of
various acetates involves the dissociative adsorption of acetate to the catalytic
Surface via the acyl fragment?"37 This hypothesis was confirmed by Trimm et al. who
conducted isotopic labeling studies on methyl formate and found that the hydrogenation
of the acyl fragment is a relatively slow step. They proposed that an alkoxy fragment
associated to the catalytic surface quickly forms an alcohol whereas the adsorbed acyl
fragment is slow to release, undergoing rate limiting hydrogenation to the corresponding

alcohol or aldehyde.”37 Though the latter process must have been expected to lower the

20

yield, Agarawal, using Cu suspended on silica, demonstrated that acetaldehyde is
hydrogenated three orders of magnitude faster than ethyl acetate.43

As demonstrated in the 1930’s high temperatures (up to 485 K) were required in
copper-chromium oxide catalyzed hydrogenations of ethyl lactate to generate propylene
glycol and ethanol.”45 These temperatures were high enough to racemize or even
pyrolyze esters. Such high temperatures are impractically expensive for any useful
process; thus lowering temperatures became an important issue. For example, Mozingo
and Folkers successfully hydrogenated malonates, acetoacetates and benzoates at 430 K
for 13 hours, using copper-chromium oxide in the ratio of 20 to 50 weight percent to
esters and reporting 40 % yields.43 Adkins used their findings to demonstrate that
increasing copper-chromium oxide loading up to one or even 1.5 equivalent of the ester
in a temperature range of 621 to 673 K would improve the product yield beyond 40% in

the hydrogenation of esters (Scheme 1.6).

OH OH
3.8K,O. h ._ ~
)\n/O\/ 9 2 rs')\/0H
O .

 

 

Ethyl lactate 1,2-Pr0panediol
O OH OH OH
O 398,9hrs. >
ADM \/
OH O OH OH
Diethyl tartrate l,2,3,4-Butanetetrol

Scheme 1.6. Adkins hydrogenated esters with 1.5 eq. of copper—chromium oxide at 330
atm of H2 and 398 K, producing the corresponding alcohols with 80% yield.

21

While Raney nickel indiscriminately reduces the aromatic ring of phenyl-
substituted esters in all cases, giving cyclohexyl-substituted alcohols, copper-chromium
oxide only reduces the phenyl-substituted esters at the carbonyl functionality (Scheme

1.7). Aside from this limitation on selectivity, Raney nickel is useful in reducing esters

1. 5 eq. copper-chromium oxide:
O,125°C 5000 psi, 0.2 hrs.
Ldry EtOH

Ethyl mandelate (80%) l-Phenyl-1,2-ethanediol

O O

/\O 0/\

1.5 eq. copper-chromium oxide

150°C, 5000 psi, 1 hr.
dry EtOH

 

Diethylphenylmalonate 2-Phenyl-1,3-propane- 2-Phenyl-1-ethanol
diol (50%) (32%)

 

HO
OH
1.5 eq. W-6 Raney nickel
100°C, 5000 psi, 1.3 hrs.
0
O \—
Ethyl mandelate l-cyclohexyl 2-Cyclohexyl- Ethyl cyclohexyl-
1,2-ethanediol l-ethanol acetate

(32-53%) ( 12-24%) (16-45%)

Scheme 1.7. The selected reactions by Adkins show that copper-chromium oxide
selectively reduces the phenyl-substituted ester at the carbonyl functionality, whereas
W—6 Raney nickel, or all tested Raney nickels indiscriminately reduced all unsaturated
carbon-carbon bonds, resulting in cyclohexyl functionalities.

22

because the hydrogenation can proceed to completion at room temperature, producing
80% product yield.44 Depending on the substrate and catalyst, however, the amount of
time can vary. Various types of Raney nickels, designated as W-l through W-8 have
been widely used in industry (W designation refers to the preparation technique of Raney
nickel, varying temperatures, NaOHzalloy ratios, digestion temperatures with respect to

- - 44.45
time and washmg process).

For example, W- 6 Raney nickel requires 14 and 25
hours in the reduction of 5-carbethoxy-2-pyrrolidone and ethyl N-phenylglycinate to the
corresponding amino alcohols, respectively. On the other hand reductions of the same
substrate by W-5 Raney nickel to the corresponding amino alcohols required less time
(0.8-9.0 hours). Adkins did point out that at the time of testing W-6 Raney nickel
catalysts were not completely developed yet. In fact, when tested, W—6 Raney nickel
catalyzed a reduction of N—phenylglycinate that went with extreme violence at 373 K.
Thus temperature control poses a safety concern, making lower temperatures a practical
necessity and safety requirement.

At conditions employing 1.5 equivalent of W-6 Raney nickel to esters at 330 atm
and 373 K, quantitative yields of 1,2 propanediol (1.5 hrs.) and 1,2,3,4-butanetetrol (10
hrs.) were produced from ethyl lactate and diethyl tartrate, respectively (Scheme 1.8).
Under identical conditions, 3,3-dimethyl-1,2-butanediol from ethyl a-hydroxy-B,B-
dimethylbutyrate (373 K for 15 hours) and 2-methyl-1,2-propanediol from ethyl (1-
hydroxy-isobutyrate (3 73 K for 2 hours) form (Scheme 1.8).44‘45

Adkins observed that temperature, pressure and catalyst loading have a profound

effect on the rate of hydrogenation. When investigating the rate of conversion of ethyl

lactate he found that while at room temperature 5.5 hours were required to achieve 75 %

23

conversion, the same reaction required less than an hour at 373 K. It is important to
note that this reaction was carried out using 1.5 equivalent of catalyst (Raney nickel
W-6).44‘45 When the hydrogen pressure was increased from 230 to 330 atm,

conversion of ethyl lactate was accomplished in substantially less time.“45

 

 

 

OH OH OH
O\/ 100°C. 1.5 11% OH < 25°C. 13 hrs/iYOv
o O
Ethyl lactate 1.2-Pr0panediol Ethyl lactate
OH OH OH
/‘\‘(OY 100°C. 5.5 hrs. > /l\/OH + /l\

O

Isopropyl lactate 1,2-Propanediol

100°C. 15 hrs.
__—’

 

OH
3,3-Dimethyl-1,2-butanediol

 

   

Ethyl a-hydroxy-isobutyrate 2-Methyl-1,2-propanediol

Scheme 1.8. The hydrogenations of esters produced yields in excess of 80%. Even
though the temperatures did not exceed 373 K, reaction conditions required 1.5
equivalent of W-6 Raney nickel catalyst and 330 atm pressure of H2 gas.

24

Adkins was the first to demonstrate that dibasic esters can be hydrogenated to the

corresponding diol by the scheme below:

ROOC— (CH2)n— COOR + 4H2 —> HO — (CH2),,,2— OH + ZROH

Dibasic esters that have two carbons separating the carbalkoxy groups produced 80-90 %
yield of diol, whereas less than 40 % yield was observed for malonates to produce
l,3-propanediol.26‘46‘47

An immensely important compound to the chemical industry is ethylene glycol.
It has use not only as antifreeze but as a component in hydraulic fluids, alkyd resins and

46 The industrial production of ethylene glycol involves the process in

polyester fibers.
which ethylene is oxidized to ethylene oxide over a silver catalyst which in turn is treated
with aqueous acid to produce ethylene glycol.46 Trimm et al. suggested an alternative
route for producing ethylene glycol from diethyl oxalate that is cheaply available as a
byproduct from aluminum production. They studied the gas phase hydrogenolysis of
diethyl oxalate over a series of copper-based catalysts. They observed up to 99%
conversion with 85% selectivity to ethylene glycol in the ion-exchanged Cu/SiO2-
catalyzed hydrogenolysis of diethyl oxalate at 513 K and 6 atm. They also observed
some polymerization of the ethyl glycolate intermediate that had formed during the
reaction. These polymerization side products presented a challenge as they caused a
partial deactivation of the catalyst.46

Trimm et al. proposed that the hydrogenolysis of diethyl oxalate involves two

Steps. First ethyl glycolate and one equivalent of ethanol are generated. In the second

25

step the ethyl glycolate is further reduced to ethylene glycol and another equivalent of
ethanol (Scheme 1.9).46 The choice of catalyst was either Raney cOpper or copper
supported on either Silica, alumina, MgO or TiO2. Trimm et al. determined that the
deactivation of catalyst caused by the polymerization of ethyl glycolate can be minimized
by increasing the surface area of the silica-supported copper catalyst .46 They proposed
that diethyl oxalate quickly adsorbs in a dissociative fashion on the copper surface to
generate an ethoxy species, which in turn reacts with hydrogen to form ethanol. The
remaining keto-acyl fragment reacts more slowly to yield ethyl glycolate.46 The
dissociative adsorption of alkoxy and acyl fragments have been previously proposed in

. . 9
stud1es done on acetates and h1gher esters.“4

O

o
0 2H ——> EOH
V \II/JKOA1L 2 Aokw + t
o

O
/H\/ HO OH
' OH + 2H2 ——> + EtOH

Scheme 1.9. The two step process of hydrogenolysis of diethyl oxalate to ethylene
glycol.

Thrimm et al. also examined the thermodynamic feasibility of diethyl oxalate
hydrogenolysis to produce the diol.46 The beats and free energies of formation of
reactants and products were calculated and are reported in Table 1.1.46 From these

values the equilibrium constants of the possible reactions in diethyl oxalate

26

hydrogenolysis were calculated. These calculations demonstrated the feasibility of
diethyl oxalate hydrogenolysis to ethylene glycol with high conversion at relatively low

pressures (1 -6 atm).46

Name Formula AH; (kJ/mol) AG; (kJ/mol)
Diethyl oxalate C2HsOOCCOOC2H5 -768 -562
Ethylene glycol HO(CH2)2OH -3 85 -301
Ethyl glycolate HOCH2COOC2H5 -5 76 -431
Ethanol C2HSOH -235 -168

Table 1.1. Heats and free energies of formation for diethyl oxalate and related
compounds at 1 atm.46 Note that entries for ethanol have been switched, as the Trimm'
paper incorrectly reported these values from the CRC Handbook.

Other than the polymerization of ethyl glycolate on the catalytic surface, gasous
byproducts can also be generated. Trimm et al. proposed a reaction mechanism for the
decarbonylation of ethyl glycolate in which C0, C02 and C2H4 are generated, which
process is proposed to be the same for diethyl oxalate (Figure 1.5). He observed that the
generation of gasous byproducts depends on the partial pressure of the diethyl oxalate or
ethyl glycolate. Contamination from C0, C02 and C2H4 can be minimized and a 99%
conversion with 85% selectivity to ethylene glycol can be achieved at 513K and 6 atm by

increasing the concentration of ethyl glycolate.46

27

 

> C0 + C02 + C2H4

C... .L.
/////////'//

Figure 1.5. Mechanism for the gasous byproduct formation from diethyl oxalate.

 

 

As previously established, hydrogenation of dibasic esters presents a major
challenge in that they can easily form undesired cyclization products at higher
temperatures.47 For example, dimethyl succinate preferentially generates y-butyro-
lactones and methanol at low pressures and high temperature. At 5 atm H2 pressure 7-
butyrolactone further reacts to generate THF and water, whereas at pressures above 300

“‘50 As Trimm et al.

atm it generates 1,4-butanediol, THF and water (Scheme 1.10).
showed that it is possible to develop a process in which controlled quantities of THF and
1,4 butanediol are generated via what is thought to be a 'y-butyrolactone
intermediate.“5 "54 Furthermore, the formation of the desired final product (1,4-
butanediol) at low temperature and higher pressure is thermodynamically favored.47
Various copper catalysts, such as Raney copper, have been successfully used in

7

the hydrogenation of dialkyl succinates.4 Patent applications only demonstrate the use

28

of copper/zinc oxide, silica-supported copper and especially copper chromite. Addition

of zinc oxide to the copper chromite mixture increased the catalytic activity for the

O O
/O 0/ + 2H2 ———> O: 3 + 2MeOH

O

0 2112/1): 5 atm l 2H2/ F: 30247 atm
H20 + V 7

HO——(CH2)4_OH + THF, H20

Scheme 1.10. The hydrogenation at two different pressures, producing two different
results.

conversion of dimethyl succinate to produce y-butyrolactone.47 Coprecipitated
copper/zinc oxide showed a one order of magnitude greater rate of conversion of
dimethyl succinate with respect to per unit surface area of copper than copper chromite or
silica-supported copper catalysts used by themselves.47 Using coprecipitated
copper/zinc oxide in the hydrogenation of dimethyl succinate, y-butyrolactone (GBL)
intermediate is generated along with methanol (Scheme 1.10). Starting with GBL, the
reaction takes different routes as pressures are varied. While at 5 atm, THF and water
are generated from GBL, at higher pressures (30-247 atm) a substantial amount of 1,4-

butanediol is formed along with negligible amounts of THF and water.47

29

The hydrogenation of esters is a non-trivial process that often requires high
temperatures and pressures. In this section a variety of catalysts has been discussed for
such processes. Although esters normally can be hydrogenated more than the
corresponding free organic acids, development of even more effective methodologies for

their reduction is an active research area.

30

 

 

1.3 Hydrogenation of Organic Acids

 

Although there are numerous published examples of the hydrogenation of organic
acids to value—added chemicals, their fundamental chemistry and the economics of their
hydrogenation have been much less discussed. Nevertheless, hydrogenation of organic
acids under mild conditions still has a lot of promise even though they are typically much
more difficult to hydrogenate than esters. Studies by Adkins showed that esters
hydrogenate more poorly in the presence of free acids.55 For example, he obtained well
above 90% yields for the hydrogenation of caproate esters over copper-barium—chromium
oxide between the temperature range of 423 to 473 K and 200-300 atm. pressure in 35
minutes. In contrast, hydrogenation of free caproic acid in ethanol and n-butanol
produced only 10 and 15 % yields, respectively. Hydrogenation of n-butyl caproate in
the presence of caproic acid suppressed the product yield and produced only 75 %
hexanol for the same amount of time.55

In 1959 rhenium “black” was reported as an effective catalyst for the
hydrogenation of organic acids to the corresponding alcohols, under much less stringent
conditions than had previously been done.25 Various forms of copper catalysts
employing chromium, cadmium, cobalt-nickel and ferrous metals coupled with non-
ferrous metals had been used to hydrogenate carboxylic acids in the temperature range of
393 to 673 K and the pressure range of 30 to 400 atm.25‘56‘57 Adams et al. demonstrated
that cadmium-nickel salts of organic acids over copper chromite catalyst produce
alcohols under dry conditions at 513 K and 235 atm.58 When discussing rhenium

catalysts, it is important to note that they show lower activity toward the reduction of

31

alkene functionality than platinum, palladium or nickel.25 However, they must be in a
high ratio to substrate to hydrogenate an organic acid, such as maleic or cinnamic acid,
effectively.25 Of particular interest is the hydrogenation of maleic acid with rhenium
black that exclusively afforded succinic acid in various solvents under varying
temperatures and pressures and with different reaction times (Table 1.2).

Rhenium “black” derived from rhenium heptoxide afforded ethanol in the
hydrogenation of acetic acid with reaction times of 2.5 hours to 60 hours, depending on
the type of solvent used (Table 1.2).25 In general, the catalyst prepared in water produced
very poor yields in the hydrogenation of acetic acid and succinic acid. While the
hydrogenation of maleic acid over water-prepared catalyst produced low yields of
succinic acid, the catalyst prepared in ethanol or dioxane afforded 100 % yield (Table
1.2). The introduction of water for the preparation of the catalyst and the reduction of
substrate is noteworthy because it proved that solvent effects must be taken into account
and that they can be a dominant factor in hydrogenolysis.59

It also became obvious that an increase in temperature or pressures does not
necessarily lead to higher conversion rates and selectivity. Broadbent also demonstrated
with rhenium “black” that higher catalyst loading does not necessarily lead to better
product yields and that the solvent used in the preparation of the catalyst and the
hydrogenation can profoundly affect the product yields of these reactions (Table 1.2).25
For example, hydrogenation of acetic acid produced 68 % yield when acetic acid was
used for the catalyst preparation, but ethanol used in the preparation of the same catalyst

caused a drop in the product yield to 38 %.25

32

Substrates

Acetic acid

Acetic acid

Acetic acid

Acetic acid

Maleic acid

Maleic acid

Maleic acid

Maleic acid

Succinic
acid

Solvents
for
Catalyst
Prep.
Dioxane

Acetic
acid
Ethanol
Water
Dioxane
Acetic
acid
Ethanol
Water

Water

Gram of
Cat./
Mole of

Substrate

1.0
0.2
1.1
1.1
2.6
2.5
1.1
1.0
1.1

Solvent for

Reduction

None

None

None

None

Dioxane

Dioxanc

Water

Water

Water

Avg.

Temp.

(K)
420
433
419
448
421
431
424
473

478

Avg.

Pressure

(atm.)

355
129
258
275
177
163
254
194
272

Time

(hr.)

9
60
15
2.5
2.5
12

4

8
12

%

Yield
40
68
38
27
100
93
100

0

39

Main Product

Ethanol
Ethanol
Ethanol
Ethanol
Succinic Acid
Succinic Acid
Succinic Acid
Succinic Acid

Tetrahydrofu ran

Table 1.2. Summary for the hydrogenation of various organic acids catalyzed by
rhenium black. 25

Rhenium heptoxide with catalyst loading Similar to that of rhenium “black”

afforded good results especially when water was used for hydrogenation. For example,

acetic acid in water was converted to ethanol with 100 % yield in1 0 hours at 423 K and

168 atm.25 Furthermore, the hydrogenation of trifluoroacetic acid, which contains three

very strongly electron-withdrawing atoms, required longer reaction time (18.5 hours)

and higher pressure (300 atm) and temperature (480 K) to generate the corresponding

trifluoroethanol. Under the same conditions but even longer reaction times (27 hours),

trichloroacetic acid did not afford any product.25 Hydrogenolysis of succinic acid

produced 94 % 1,4-butanediol and 6 % n-butyl alcohol when the reaction was run neat.

When water was added the 1, 4 butanediol yield dropped to 59 % and 13 %

tetrahydrofuran were produced. The hydrogenation of lactic acid using rhenium

33

heptoxide produced 84 % propylene glycol in an eight hour reaction time at high pressure
and temperature (Scheme 1.1 1).25

A large array of hydrogenolysis by rhenium heptoxide reduced in situ was
demonstrated by Broadbent. The beauty of these reactions was their simplicity as they
avoided the initial preparation of rhenium black catalyst. However, they often required
very long reaction times and either very high temperatures or pressures, or both.
Broadbent’s work, however, was pivotal in that he was able to demonstrate that various

organic acids, such as acetic acid, can indeed be reduced in good to excellent yields.

 

OH OH OS <0 OH
OH 423 K, 258 atm. _ /‘\/OH0 + + OH
8 hours
(Neat)
O

 

 

 

O
84 0/o 16 "/0
+ n-butyl alcohol
483 K, 245 atm.
4 hours 94 % 6 %
t
O (Nea)
OH
HO
O + Polyester
483 K, 258 atm.>
6 5 hours 59 °/o 13 0/o 3 °/o
(Water)

Scheme 1.11. Solvent effects strongly influence the hysdrogenolysis of lactic acid and
succinic acid using rhenium heptoxide reduced in situ.

34

In a 1988 patent Kitson described heterogeneous catalytic hydrogenation of
various organic acids with 2-12 carbon backbones, using either molybdenum or tungsten
coupled with a second component that could be palladium, rhodium or ruthenium on
carbon support. Just as Broadbent did, Kitson also observed trans-esterification products
from the hydrogenolysis of various organic acids.60 For example, the hydrogenation of
acetic acid produced ethyl acetate and the hydrogenation of propanoic acid produced not
only propanol but propyl propanoate.60 These reactions were carried out at relatively
high temperatures from 180-250 °C, pressures above 10 atm and relatively high catalyst
loading, producing very poor conversions of substrates. The catalyst composition and .
method of preparation determined the degree of conversion as well as the selectivities for
the corresponding alcohols and ester by-products. When the catalyst was tested for
acetic acid hydrogenation, both the temperature variation and catalyst composition
affected the conversions and selectivities of these reactions. For example, a mixture of
2.5 weight percent palladium and 5 weight percent tungsten on carbon support produced
a 16.5% conversion of acetic acid with an 84.3% selectivity toward ethanol at 462 K,
whereas the very same catalyst produced a 44.6 % conversion at 521 K with 82.7 %
selectivity to ethanol. Palladium on carbon support produced marginal conversion (0.3
%) and no products of any kind under the same conditions, and when tungsten was
excluded only 0.6% conversion was observed. Similar results were obtained from
mixtures of molybdenum and palladium using carbon support. Again exclusion of Pd
lead to no reaction, but a mixture of 2.5 weight % molybdenum and 5% palladium

produced 58% conversion and 83% selectivity toward ethanol.”61

35

In the literature to this point, high selectivity was attainable only under the
conditions that Kitson had described, using catalytic mixtures supported on various inert
materials, such as carbon, silicas, aluminas, but the conversions of substrates remained

62 Various combinations of Group VIII metals, such as palladium and ruthenium,

low.
with rhenium seemed promising. Of particular interest was palladium in various mole
percent compositions with rhenium. In the hydrogenation of acetic acid, a mixture of 2.5
% Pd with 5 % Re afforded both lower conversion and selectivity when the temperature
was lowered. Under the same conditions, lowering the Re content of the catalytic
mixture to 2.5 % yielded a dramatic drop in conversion but the selectivity to ethanol
essentially remained the same. The Kitson patent was intended to demonstrate that
heterogeneous catalytic systems composed of two metals and supported on high surface
area inert material such as carbon can produce high selectivities even though the percent
conversions remained low. In hydrogenating acetic acid and propanoic acid,
esterification products such as ethyl acetate and propyl propionate were also formed and
their ratios to the corresponding alcohol from the reductions of carboxylic acids can be
increased by lower conversion per pass in gas phase, whereas at high conversion per pass
or continuous co-feeding of water would generate more of the corresponding alcohol.62
Antons described the preparation of optically active alcohols from optically active
carboxylic acids, employing catalyst either by itself (elemental ruthenium, ruthenium
oxide, hydroxide or halide) or on a support such as carbon, aluminum oxide or silicon
dioxide.63 Of greatest pertinence to this thesis were the hydrogenations of short organic

acids with vicinal substituents such as optically active lactic acid and 2-chloropropanoic

acid to propylene glycol and 2-chloropropanol (Scheme 1.12).63 Although the

36

enantiomeric excess was well above 90 %, relatively poor yields were obtained.
Nevertheless, the quantity of Ru required and the ease of handling these Ru catalysts are
much more cost effective than conventional reducing agents such as lithium aluminum
hydride and sodium borohydride.63 For the hydrogenolysis of lactic acid it was also
observed that with increasing temperatures, 383 to 413 K, the enantiomeric excess of
propylene glycol dropped from 93 % to 71 %.

Optically active propylene glycol and 1,2,3,4-butanetetrol can also be prepared
from optically active lactic and tartaric acids, using ruthenium/rhenium mixtures prepared
from ruthenium oxide and rhenium heptoxide. Although this colloidal mixture
outperformed ruthenium black by producing up to 99 % ee, the yields did not improve .
beyond 80 %.64 Nevertheless, combinations of finely dispersed ruthenium oxide with
metals such osmium, iron, rhodium, copper, zinc/germanium, gallium and trimetallic
species of ruthenium with two other metals such as rhenium/silver, rhenium/copper, and
rhenium/tin seem promising and are under investigation.64

The numerous examples of patent applications do provide various methodologies
for catalytic hydrogenations of organic acids, demonstrating ever-increasing yields and
selectivities, but they do not go into the mechanistic aspects of the chemistry behind all
these processes. Using 1H and 13 C NMR spectroscopic methods for H/D exchange, the
hydrogenation of lactic acid, a particularly important organic acid, was mechanistically
investigated by Kovacs et al.2 The determination of H/D exchange in both the solvent

and the hydrogenation product, propylene glycol, is a prerequisite in order to understand

the hydrogenation mechanism of lactic acid.

37

OH

 

Ru/H2 > a“
80 0C, 200 bar
85 % Yield, 97 % ee

 

L-(+)-l ,2-propanediol

H \CI

9‘9
Ru/HZ > OH
140 °c, 200 bar
66 % Yield, 82 % ee

 

 

L-(+)-2-chloropropanol

OH

H OH
Ru/H2 , ~\“
* '. OH
80 0C, 200 bar

66 % Yield, 97 % ee

 

 

L-malic acid L-l ,2,4-butanetriol

RU/H2
80 0C, 200 bar

 

   

57 % Yield,
L-(-)-tartaric acid L—l .2,3,4-butanetetraol

Scheme 1.12. The hydrogenation of lactic acid, 2-chloropropionic acid, malic acid and
tartaric acid gave good to moderate yields. The stereochemistry of the substrate was
largely retained after hydrogenation resulting in high ee’s.

Control experiments, in which a heterogeneous ruthenium catalyst was used

under the same reaction conditions, when neither lactic acid nor propylene glycol was

38

present, demonstrated rapid H/D exchange between H2 and D20. Propylene glycol, the
hydrogenation product of lactic acid, also underwent catalytic H/D exchange at the C1
and C2 positions, and more slowly at C3. The NMR splitting patterns of I 3 C coupled to
deuterium and the chemical shift differences between the deuterated and the
corresponding protiated carbon sites proved valuable in determining the location and the

extent of H/D exchange in both lactic acid and propylene glycol.2

 

Scheme 1.13. The proposed mechanism of lactic acid hydrogenation.2

Kovacs et al. proposed that hydrogenation only occurs at the C1 position of lactic

acid. This mechanism involves the formation of an acetal intermediate (3) followed by a

39

dehydration to obtain aldehyde (4) that further hydrogenates to the corresponding alcohol
(5) (Scheme 1.13).2 Because the Optical activity at the C2 position of lactic acid was
retained in these reactions, alternative routes involving keto-enol tautomerization were
excluded. However, as deuterium incorporation was observed at the C2 position of
propylene glycol, Kovacs at e1. proposed that H/D exchange takes place after the
propylene glycol is formed and occurs via a surface-bound intermediate.2

The efficient and selective aqueous phase hydrogenation of organic acids in
general presents a greater challenge than that of esters. Ruthenium and copper chromite
are relatively efficient catalysts for organic acid hydrogenations, but they work at
relatively high pressures and temperatures.6 In aqueous-phase hydrogenation, ruthenium
supported on activated carbon produced 95 % conversion of lactic acid and 90 %

selectivity to propylene glycol at optimal reaction conditions (423 K and 99-140 atm).65

40

Chapter 2

Results and Discussion

 

The Ru/C-catalyzed aqueous-phase hydrogenation reactions of ethyl lactate and
various organic acids bearing —OH, -OMe, -OCOCH3, -H or -CH3 substituent, vicinal to
the carboxy functional group, were run in a high pressure Parr batch reactor under mild
conditions (348 to 423 K and 1200 psi). Three-carbon backbone compounds bearing
(it-substituent X (X = -H, -Cl, -OH, -OCH3, -OAc, -CH3) and two- carbon backbone
compounds bearing a-substituent X (X = -OH, -OCH3) were studied by 1H NMR
spectroscopy and HPLC. While 1H NMR analysis of the samples required the use of l
either t-butyl alcohol or dioxane as internal standards, for HPLC analysis ethanol was
used. In the text 1H NMR peaks were designated as follows: singlet (s), doublet (d),
triplet (t), quartet (q), quintet (quint), sextet (sext), septet (sept), multiplet (m), doublet of
a doublet (d/d).

Substrate and alcohol product concentrations were monitored with respect to time
and the resulting data were used to determine the percent conversion of a given organic
acid and the percent production of the corresponding alcohol. The relationship between
molecular structure and the efficiency of hydrogenation of organic acids or esters was
determined from the selectivity and rate data of each reaction. This work has
demonstrated that various vicinal functional groups can influence the reactivity of
organic acids.

While a hydroxy functional group can hydrogen-bond inter- or intra-molecularly,

the methoxy functional group has only electron-withdrawing ability. With this in mind,

41

various electron-withdrawing groups (-OAc, -C1 or —OMe) at the C2 position of a given
organic acid were varied and tested and compared to —OH bearing substrates for their
effect on the reactivity in hydrogenation reactions.

The ester functionality does present a problem as it can easily hydrolyze in
hydrogenation reactions and the resulting rate data contain both hydrolysis of the alkoxy
fragment and the hydrogenation of the acyl fragment. The chlorine substituent at the C2
position was also of concern as it was readily removed from the substrate in Ru/C-
catalyzed aqueous phase hydrogenation reactions. Hence the loss of chlorine rendered
the substrate functionally void at the C2 position.

Finally, steric effects may also play a role in the reactivity of these organic acids
as the substrates must adsorb on the catalyst surface for reduction to take place. In
theory, the relatively bulky -OAc or -OMe group may adversely influence the
adsorption of the substrate on the catalyst surface as free rotation around the sigma bond
between C1 and C2 carbon is evident. The momentary presence of a bulky substituent

close to the sp2 carbon of the carboxyl functionality may hinder the delivery of hydride.

42

2.1. Hydrogenation of Lactic Acid

 

The hydrogenation reaction of lactic acid (LA) was run at 423 K and 1200
psi for six hours. The product from lactic acid hydrogenation, detectable by 1H NMR
and HPLC, was only propylene glycol (PG) (Scheme 2.1). The 1H NMR peaks of
products and reactants from lactic acid were referenced to t-butyl alcohol at 1.22 ppm.
Lactic acid was purchased and used as a mixture consisting of approximately 15 % self-
esterification products (lactyl lactate; and di-lactide which is a cyclic dimer of LA) and
approximately 80 to 85% LA and residual amount of water (see Figure 2.1 for lH NMR
spectrum). Unfortunately, purification by distillation of this mixture, even at low
pressures and temperatures, is non-conducive because its constituents are azeotropic.
Furthermore, polimerizarion of LA during distillation, even at lower temperatures, is

quite significant.

 

OH OH
OH Ru/C /|\/OH
>
H2
O 423 K, 1200 psi

Scheme 2.1. Hydrogenation of lactic acid to generate propylene glycol.

43

Lactyl lactate has two doublets: one at 1.498 and the other at 1.420 ppm. A weak
doublet of the di-lactide is evident at 1.554 ppm (Figure 2.1). In addition, three quartets
of lactic acid and its esterification products are visible on the spectrum. The most
upfield quartet at 3.380 ppm belongs to LA, whereas the more downfield quartets of the
esterification products generate weak signals and one of them is at the HBO peak. These
quartets could not be used for integration due to their close proximity to the HDO,
especially when lH NMR spectra are obtained from a water suppression experiment. In

addition, quartets run across a relatively long base line, thus integration values can easily

 

 

OH
01-1
0
0 01-1 ..
o
0 Hok/
o
o
o
11
0 j,
1.60T 1 1.55 7 A 1.50 1.45 ' V 1.40 ' i ' 1.35

Figure 2.1. A magnified region of the lactic acid mixture showing the doublets of the
corresponding acid and esters.

44

be skewed. It must be noted that one of the main errors for calculating concentrations

can come from long base-line integration that can effectively increase the calculated

concentrations of the samples.

LA (5 4.364 lH(q)/[B]; 1.404 3H(d)/[A]) has a prominent and easily
distinguished doublet at 1.404 ppm that ‘was monitored to determine the lH NMR

concentration of the reactant (Figure 2.2). The esters from the starting material

completely hydrolyzed to LA or directly hydrogenated to form PG within two hours of

 

 

 

OH OH
D
[B] OH Rtl/C [ 1 OH t-butyl alcohol
[A] H2 [C] [E]
O 423 K. 1200 psi
0 hour IA]
1!
181 l
1,. ' 1!.
1 hour -
a C
[D] [E1 1 [.,l
3 hours ll
, 1; ~
6 hours “
i *l’rl'lr' '1' l"""l“"l""’l""1'"‘l""l""’!""'1' 'r" '1‘ 1 '1‘ 1'
4 4 U 4.0 3.8 3.6 3.4 3.2 3.0 2.8 2.6 2.4 2.2 2.0 1.8 1.6 l 4 1.2 1.0 0 8

Figure 2.2. Selected lH NMR spectra of LA hydrogenation in which PG is produced.
For stacked spectra detailing the LA hydrogenation each hour, please see Appendix 1.1.

45

reaction time as their characteristic doublets and quartets were no longer visible in the lH
NMR spectrum. The product PG (5 3.84 lH(m)/[D]; 3.49 2H(d/d)/[E]; 1.12 3H(d)/[C])
has a doublet at 1.12 ppm that was monitored for concentration change (Figure 2.2).

The 1H NMR results for the percent conversion of LA and production of PG
differ significantly from the HPLC results (Figure 213-216). The lH NMR
concentration profiles of the hydrogenation of LA reveal that the starting material is
almost completely used up after five hours (Appendix 1.2 and 1.3). Although the
conversion of the starting material approaches 100i7%, only 72i11% PG was produced
after six hours. On the other hand, the presence of LA was evident from the HPLC
chromatogram even after six hours (for HPLC concentration profile, see Appendix 1.3).
HPLC experiments indicate 87% conversion of LA to 67% PG within this time frame. A
large number of different samples handled by HPLC may cause a discrepancy in the
chromatogram as a different type of sample from a former run may Show up and may
have a retention time that is similar or even identical with that of the analyte.

As patent literature has demonstrated, close to 100 % conversion is possible for
LA under the same conditions} 63'“ While the lH NMR results of this work concur
with published results, the values for percent conversion of production, obtained from
HPLC, should be considered with a hint of skepticism. Ideally the values obtained from
either analytical method should be very close. Furthermore, the concentration versus
time profile of HPLC results (Appendix 1.3) shows an exponential decay that
corresponds well to first order behavior. On the other hand, the concentration profile,

based on 1H NMR experiment, does not closely follow the first order behavior.

46

The agreement between 1H NMR and HPLC with respect to selectivity to PG as
the final product is satisfactory. The selectivity to PG is 82 and 77 % based on IH NMR
and HPLC, respectively. Foreign peaks for side products on the 1H NMR spectrum and
additional humps on the HPLC chromatogram were not observed even after six hours.
However, gaseous byproduct formation, such as CO2, should be considered as
decarboxylation reactions at this temperature and time period are possible.

Two control experiments were run: one under H2 without the catalyst and the
other under He with Ru/C catalyst. There was no evidence of product formation.
Although some degree of decarboxylation may occur, the 1H NMR spectra did not show

any products and the concentration of the LA starting material was invariant.

2.2. Hydrogenation of Glycolic Acid

 

The hydrogenation reactions of glycolic acid (GA) were run at 423 K and 1200
psi. The product from GA hydrogenation, detectable by 1H NMR, was only ethylene
glycol (EG) (Scheme 2.2). The peaks of products and reactants from GA were
referenced to t-butyl alcohol at 1.22 ppm. Two singlets of GA appear in the 1H NMR
spectrum. One of the peaks is at 4.050 ppm accounting for the two ethylene protons of
glycolic acid and a small singlet at 4.181 ppm that belongs to a cyclic di-ester which is

the dimerized derivative of glycolic acid. The singlet of glycolic acid was monitored to

47

determine 1H NMR concentrations. The singlet of ethylene glycol at 3.510 ppm was

monitored to determine the product concentration (Figure 2.3).

 

O
140% HO OH
Ru/C
OH + \__/
H2
GA 423 K, 1200 psi EG

Scheme 2.2. Hydrogenation of glycolic acid that generates ethylene glycol.

 

O O t-butyl alcohol
Hovlb [B]
H
[A] 1 A] O O
(I) l
0 w
1
_ 1 1
1B] 1 O
1 [C] HO OH
(H) 1 \ /
[CI
7 4.4. i I 4.2, 4.0 i ' 3.8 i 3 83:68 I | 13.47 I 3.2 I i 3.0 i I 2.8 2.6187” 2.4. V —2.2 1 III ‘ 1.8V I I 71.6 i 1.4, I I 1.2

Figure 2.3. Spectrum (1) is the starting material containing (GA) and its cyclic ester.
Spectrum (11) shows EG, the final product of GA hydrogenation after six hours.

For stacked spectra detailing the hydrogenation of GA each hour, please see Appendix
2.1.

The concentration profiles for the hydrogenation reaction of GA, determined by

1H NMR, reveal that the starting material is completely used up after five hours (see

48

Appendix 2.2 for ‘H NMR concentration profile). Although the conversion of the
starting material approaches 100i7%, only 60i5% EG is produced.

GA is a one carbon shorter analogue of LA. GA hydrogenation has
demonstrated that the selectivity to EG (62 %) is quite different from that of LA
hydrogenation to PG (82 %). ]H NMR did not show the presence of any hydrogenation
byproducts, such as methanol or ethanol, but gaseous byproducts, such as CO and CO2,
may have formed during the reaction. HPLC analysis was not carried out on GA
because of technical difficulties.

Lactic acid and glycolic acid are important target compounds in this work as their
vicinal hydroxy functional group can intra-molecularly hydrogen-bond to the carboxy
group. The reactivity of LA and GA is dramatically increased by the —OH functional
group as compared to simple unsubstituted propanoic acid. Conceivably, the —OH
substituent induces a fixed intra-molecular geometry, while it simultaneously induces the
sp2 carbon of the adjacent carboxy functionality, which favors its rehybridization to sp3
(Figure 2.4). Considering this increased reactivity, glycolic acid should react at a similar
rate as lactic acid in these reactions.

In addition, the —OH functionality can form an inter-molecular hydrogen bonding
with water that may also play a role in an increased reactivity of these compounds as
compared to propanoic acid that is less capable of hydrogen bonding. As solvent effects
can stabilize the transition state of a reaction, strongly hydrogen bonding water could

facilitate the interaction between the substrate and the active site of the catalyst.

49

H H
o/ 0’70 0/ 070 WW
OH OH O
Glycolic Acid Lactic Acid Propanoic Acid

Figure 2.4. Intra-molecular hydrogen bonding of glycolic and lactic acid have a relative
fixed geometry compared to propanoic acid..

Comparing the rate of conversion of LA to GA from 1H NMR experiments, it is
clear that LA hydrogenates at a marginally faster rate (Data Table 2.3). However, both
compounds are completely reacted after five hours. Similarly to LA, under H2 without

the catalyst, controlled reaction of GA did not produce any products.

2.3. Hydrogenation of 2-Methoxypropanoic Acid

 

The hydrogenation of 2-methoxypropanoic acid (2MPA) was carried out at 423 K
and 1200 psi for six hours (Scheme 2.3). The lH NMR spectrum of the reactant and the
product was referenced to t-butyl alcohol at 1.22 ppm. Three peaks belong to 2-
methoxypropanoic acid (5 1.38 3H(d)/ [A], 3.360 3H(s)/[C] and 4.00 1H(q)/[B]). The
main product of the hydrogenation of 2MPA, detectable by 1H NMR, was 2-methoxy-
propanol (2MPOL) (8 1.092 ppm 3H(d)/[D]; 3.350 3H(s)/[F]; 3.40-3.62 2H(m)/[G];

3.40-3.62 lH(m)/[E]) (see Figure 2.5).

50

OMe

OMe OMe
OH Ru/C OH A/OH
+

H2
mm 0 423 K, 1200 psi 0

ll

2MPOL

Scheme 2.3. The hydrogenation of 2MPA to generate 2MPOL.

The doublet [A] of 2MPA and the doublet [D] of 2MPOL were monitored for
the concentration change in the six hour period (see Appendix 3.1 for lH NMR spectra,
and Appendix 3.2 and 3.3 for concentration profiles). The pre-reaction mixture
contained an estimated 5% LA which was a byproduct of 2MPA synthesis. Because LA
and 2MPA are azeotropes, the removal of the remaining LA from the 2MPA mixture
proved very difficult and severely reduced the product yield after each attempt. Hence,
the pre-reaction mixture contained 5% LA relative to 2MPA. The starting material
(2MPA) was not completely used up in the six hour hydrogenation reaction. Both lH
NMR and HPLC showed that the reaction mixture contained 2MPOL, a substantial
amount of 2MPA, a small amount of PG and a residual amount of methanol (MeOH)
and n-propanol (POL) after six hours (see Figure 2.6). The IH NMR results were non-
conclusive as they showed 432t3% conversion and 55% production of 2MPOL. Percent
error could not be established for 2MPOL production because only one set of lH NMR
data was conclusive for product formation. The discrepancy between the conversion and
production values is likely due to integration error. HPLC analysis gave more

reasonable values with 48 % conversion of 2MPA and 47 % production of 2MPOL.

51

OMe [C]

OH
[B] [E]
[A] [D]

[C]

[F]
/
[E] [G] [H]
[3171 x

[B] [N]

- v fi‘+Yfi fa ._. 7*

.-T,--..,--...- - 1, - ,
4.0 3.8 3.6 3.4 3.2 3.0

OMe [F]

[G]

2.8

t-butyl alcohol

[H] MeOH [Kl

/\/0H
[J] [L]
OH
[N1 OH
[M] [P]

[A] ,[M]
1 , [D]

| ‘ 1‘

1K1 1: "

i I 2.4 V 2.21 20 Y 1.8 i i 1.6 i 1.4 ' . 1:24 1.0 V I 4018‘ i i 10.6

Figure 2.5. ]H NMR spectrum of the hydrogenation of 2MPA. The reaction produced
a mixture of 2MPOL and 2MPA after six hours.

While HPLC shows 98 % selectivity to 2MPOL, 1H NMR results show an

impossibly high selectivity which is greater than 100%.

Hence, only HPLC could be

considered for the selectivity. Although this degree of selectivity is encouraging, it

seems to be higher than expected as some hydrolysis of the methoxy functionality must

take place. Residual amount of MeOH and POL are also formed and are clearly visible

on the ]H NMR spectrum and HPLC. Upfield triplet of POL at 0.82 ppm and MeOH

at 3.32 at ppm are evident. The amount of MeOH and POL formed in this reaction is

estimated at 5 % based 1H NMR results.

52

 

 

OMe [C] OH [E] OH [KI OH
N
B] 0” [v1 0“ Ru/C _ [D] [G] + [J] [L]
[A] + [T] H2 ' OH 11 t-butyl alcohol
0 4123 K. 1200 psi [N] 011+ [Me]OH \
i [C] [Ml [P] I“ l
0 hour .4/ [T] 1’ ,
[VI 1',” w ‘5, ‘1
1 hour i H
1 1
3 hours ii .‘
I ~ ' .
6 W5 “”1“” up] [MllD]
[N] ‘ H . 11 J
{'15:}. )1 2 “‘1 . .11.". [i]
4.68 V V 4.4 4.2 4.0 i i 3.8. i 3.6 I i 3.4 i i 3.2 3.0 7 2.8 V 2.6 2.4: ‘ 22‘ V 82.00 V 1.8. i 1.6 1.4. 1.2 i 11.0

Figure 2.6. 1H NMR spectra of 2MPA hydrogenation in which 2MPOL is produced
along with small amount of PG and residual amount of MeOH and POL. For stacked
1H NMR spectra detailing 2MPA hydrogenation each hour, please see Appendix 3.1.

Two types of control experiments were run for 2MPA at 423 K and 1200 psi for
six hours: one without the catalyst under H2 and another under He with the catalyst.
Product formation was not observed in either of these experiments. Furthermore, the gas
pressure did not change in the control reactions, also confirming that gas was not

consumed nor generated.

53

2.4. Hydrogenation of Methoxyacetic Acid

 

Methoxyacetic acid (MA) was reduced at 423 K and 1200 psi to
2-methoxyethanol (2METOH) (scheme 2.4). The 1H NMR spectrum of MA (5 3.402
3H(s)/[B]; 4.105 2H(s)/[A]) and 2METOH (5 3.402 3H(s)/[C]; 3.706-3553 4H(m)/[D]
and [E]) was referenced to t-butyl alcohol at 1.22 ppm. The methoxy Singlet of the

starting material shifts to 3.405 ppm after reaction has taken place (Figure 2.7).

O

 

MeO OH
OH > \__/
H2
MA 423 K, 1200 psi ZMETOH

Scheme 2.4. Hydrogenation of methoxyacetic acid (MA) to generate 2-methoxyethanol
(2METOH).

A small singlet at 3.322 ppm, corresponding to MeOH singlet, is seen in the lH
NMR spectrum. Although the presence of MeOH was confirmed by a peak
enhancement technique in which a small quantity of MeOH was added to the sample to
produce an intense and sharp singlet at 3.322 ppm, the presence of ethanol cannot be
clearly seen. This is because the small triplet and quartet of ethanol (EtOH) are indeed
overlapped with product and reactant peaks as well as the standard itself. Based on

relative peak intensities, 6 % of MeOH is estimated to be in the reaction mixture.

54

HPLC analysis showed residual quantities of MeOH and EtOH that estimated at less

than 2 % of the generated product mixture.

 

 

 

O
[B] e Ru/C + McOH
OH > in
[A] H: [D] [E]
423 K‘ 1200 psi t-butyl alcohol
181
[A]
(l) i
l
I
(ll) .
[(1
[DIE] " ,l/H
4.6 4.4 4.2 4.0 3.8 3.6 3.4 3.2 3.0 2.8 2.6 2.4 2.2 2.0 1.8 1.6 1.4 l.2 1.0

Figure 2.7. 1H NMR spectra of MA hydrogenation. Spectrum (I) shows the starting
material and spectrum (11) corresponds to the reaction mixture containing MA,
2METOH and MeOH after six hours. Stacked spectra detailing MA hydrogenation
each hour are found in Appendix 4.1.

The concentration profile of this reaction shows that a substantial amount of the
starting material is still present in the reaction mixture after six hours (see Appendix 4.2.
and 4.3 for concentration profiles). Based on 1H NMR , 78:1:7% conversion is calculated
that is in contrast with the HPLC results that demonstrate only 64i11% conversion after

six hours. The percent production of 2METOH was calculated at 46 % by lH NMR and

55

44 % by HPLC. The selectivity to 2METOH was 53 and 69 % as determined 1H NMR
and HPLC, respectively.

The calculated percent conversions of MA and 2MPA compared to LA and GA
reveal that the hydrogenations of LA and GA go at faster rates than do their methoxy-
substituted analogues. Furthermore, comparing MA with 2MPA, it is evident that MA
hydrogenates twice as fast as 2MPA (see Table 2.3).

A control experiment for MA hydrogenation was run under H2 without the
catalyst. During this reaction the hydrogen pressure was stable, indicating that hydrogen
was not being used up. This was confirmed by 1H NMR that showed that the product
did not form and the starting and the final concentration of MA was the same in a six .

hour run.

2.5. Hydrogenation of Isobutyric Acid

 

Hydrogenation of isobutyric acid (IBA) was carried out at 423 K and 1200 psi to
produce isobutyl alcohol (IBUOL). The 1H NMR spectrum of IBA (5 1.134 6H(d)/ [A];
2.607 1H(hept)/[B]) and IBUOL (5 0.861 6H(d)/[C]; 1.718 lH(m)/[D]; 3.334 2H(d)/[E])
was referenced to t-butyl alcohol at 1.22 ppm. The doublet of IBA (1.134 ppm) and the

doublet of IBUOL (0.861 ppm) were monitored for the concentration change (Figure

2.8).

56

The concentration profile of the hydrogenation of IBA reveals that a significant
amount of the starting material is still present after six hours (see Appendix 5.2 and 5.3
for concentration profiles). The results from lH NMR and HPLC are in close agreement
for the values of percent conversion and percent production at each time interval. For
example, after four hours, 1H NMR and HPLC showed 40 and 39 % conversion,
respectively. After six hours, 1H NMR indicated that 523:8 % percent of the starting
material had been converted, but only 25:1:10 % IBUOL had been produced and these
results matched those in HPLC analysis. The relatively low production of IBUOL could
be due to gaseous byproduct formation such as CO2, CO and propane, as decarboxylation
and decarbonylation reactions are quite possible at 423 K for a six hour period. Value of

selectivity to IBUOL was 48% as determined by lH NMR and HPLC, respectively.

 

CH3 [A] CH3 [C] t-butyl alcohol 1
. 1
OH Ru/C )D\/OH 1. 1| [A]
. > H C 3’ ’.
H30 [3] i 112 3 l 1 [El if
1
O 423 K. 1200 psi
(I)
[B]
.1}. T
1
(II) [(171
IE]

[DI

'- a...» ,-..._. .. . '3'“- .-. .1_ . .__:~.. -,. 1. .. ,-.. . .«~....v‘ y~-. ..-..v.... .. ... a _.____.,s. . ,H,._ . ' "'~‘.. ._ .1--. .,.. ............ .....

_.._. A-

‘_f_-..—_,_.

4.6 4.4 42 i 1 84:0 1 i 3.18- 7 3:6 I 3T4“ . (3.2
Figure 2.8. 1H NMR spectra of IBA hydrogenation. Spectrum (1) shows the starting
material, spectrum (11) is the mixture of IBA and IBUOL after six hours. For stacked 1H
NMR spectra detailing IBA hydrogenation, see Appendix 5.1.

3.0 2.8 110 2.4 2.2 2.0 1.8 1.6 1.4 1.2 1.0

57

In a control experiment, run under H2 without catalyst, it was clear that
hydrogenation was not taking place, as H2 pressure was relatively stable all throughout
the reaction. lH NMR spectrum only showed the characteristic peaks of IBA after six
hours. Concentration profiles also demonstrated that under such conditions, reaction did

not take place as the initial and final concentration of IBA was invariant.

2.6 Hydrogenation of Propanoic Acid

The hydrogenation of propanoic acid (PA) was carried out at 423 K and 1200 psi.
The product detectable by lH NMR and HPLC was n-propanol (POL). The spectrum of
PA (5 1.062 3H(t)/[A]; 2.351 2H(q)/[B]) and POL (5 0.864 3H(t)/[C]; 1.531
2H(sext)/ [D]; 3.529 2H(t)/[E]) was referenced to t-butyl alcohol at 1.22 ppm (Figure 2.9).
The triplet of PA at 1.062 ppm and the triplet of POL were monitored for concentration
change. Concentration profile of this reaction shows a slow decrease of the
concentration of the starting material and a correspondingly slow increase of the
concentration of POL (see Appendix 6.2 and 6.3 for concentration profile).

Conversion values, determined by lH NMR and HPLC, were in good agreement;
355% and 33 %, respectively. While 1H NMR showed 1816% production of POL
after six hours, HPLC indicated that this value is only 14 %. The discrepancy may be
due to the integration error of the triplet in the 1H NMR spectrum. The selectivity was

51 and 42 %, as determined by 1H NMR and HPLC, respectively. Similarly to IBA

58

hydrogenation, decarboxylation and decarbonylation could be side reactions that may
occur to a great degree in which CO2, CO and ethane are produced.

In a control experiment, run under H2 without catalyst, product formation was not
observed. This result is expected as PA is relatively slow to reduce to POL compared to
other functionalized organic acids with very poor yields in catalytic hydrogenation.
Furthermore, the invariant concentration of PA during the control experiment indicated

that gaseous byproduct formations are unlikely.

 

181 "’1
OH Ru/C 4 /\/OH
O 423 K. 1200 psi
1
(I) I [A]
181 ‘ 1

(11) [D] 1

[El 1 1' [C]

4.6 4.4 4.2 4.0 3.8: 3:6 3.4 3.2 3.0 2.8 2.6 V 2.41 2.2 21.0. i i 118777 8 "116 i 1.41 I i 1.2 i 1.0.

Figure 2.9. ]H NMR spectrum of PA hydrogenation. Spectrum (1) shows the starting
material; spectrum (11) corresponds to the mixture of PA and POL. For stacked spectra
detailing PA hydrogenation, please see Appendix 6.].

It is important to contrast the hydrogenation of PA with IBA. Clearly, the rate of
conversion of IBA to IBOL is substantially faster than the conversion of PA to POL.
However, the selectivity to the corresponding alcohol is higher for PA hydrogenation

than that of [BA (selectivities are 51% and 48% for PA and IBA, respectively). Based

59

on HPLC analysis, the rate of product (POL) formation from PA was slower than

IBUOL formation from IBA and these results concur with that of 1H NMR.

2.7. Hydrogenation of 2-Chloropropanoic Acid

Antons showed that 2-chloropropionic acid (2CPA) hydrogenation generates 2-
chloropropanol when Mohr’s salt of the ruthenium catalyst is used.63 In this work carbon
supported ruthenium (Ru/ C) was used for the hydrogenation of 2CPA in aqueous
medium. Although the initial expectation was to see the generation of 2-chloropropanol
(2CPOL), the hydrogenation of 2CPA only generated propanoic acid (PA) at 423 K and
1200 psi.

As 2CPA was introduced in the reactor, the reactor pressure rapidly increased by
150 psi from the initial pressure (1200 psi), indicating that gas was being generated at
423 K. Furthermore, the collected samples were turning dark green with the progression
of time. This suggested that HCl was produced in this reaction that in turn was leaching
the metal components from the inside wall of the reactor (pH = 3.5 of the pre-reaction
mixture; pH= l of the collected reaction mixture). Therefore, these reactions were
repeated at 373 K to minimize the corrosion in the reactor. (Leaching of the reactor
interior can be avoided by protective Teflon coating, but this measure can also
significantly increase the time required for cooling of the reactor after the reaction.)

Although the degree of leaching was lowered at 373 K, this reaction only produced PA

60

after six hours at both temperatures. 2CPA rapidly reacted under these conditions to
form PA in less than one hour at 423 K and less than two hours at 373 K.

The 1H NMR spectrum of 2CPA (5 1.656 3H(d)/ [A]; 4.552 1H(q)/[B]) and PA (5
1.064 3H(t)/ [D]; 2.393 2H(q)/[E]) was referenced to t-butyl alcohol. The doublet of
2CPA was monitored for concentration change. It is important to note that
approximately 5 % of the starting material contained 2,2-dichloropropanoic acid (5 2.223

3H(s)/[C]). The sample collected after one hour of reaction time, run at 423 K,

 

«“C' 1151
OH Rmc _ //A\Wr/OH ‘
[D] 1

423 KH 21200 psi 0 l
(I) [A] ‘
1| ‘
[B] K] ” I
("l
W] m 1?]
4.8. 1 74.6 V 74.4: I 4.2 410' 33.8 I i 36 7‘ V 2.4 i I 3.2 l I 3.0. V 2.8 i 2.6 l I 2.4 A I 2.2 t V 2.0 V i 11.8 1.6 i i 1.4 i i 1.2 t i 1.01

Figure 2.10. (I) is the lH NMR spectrum of 2CPA pre-reaction mixture. (II) is taken
after six hours and is the spectrum of PA. For stacked 1H NMR spectra detailing 2CPA
hydrogenation, see Appendix 7.].

61

contained only PA as confirmed by 1H NMR. The doublet of 2CPA and the singlet of
2,2-dichloropropanoic acid were not observed on the spectrum, but characteristic PA
peaks were evident (Figure 2.10). Furthermore, the concentration of PA did not change
in six hours (for 1H NMR concentration profile, see Appendix 7.2). Analyzing the
reaction samples of 2CPA by HPLC presented a challenge as negative peaks were
produced in the chromatogram, thus HPLC was not used for concentration determination.

The 2CPA did not hydrogenate at the carboxy functionality. The generation of
PA , under these reaction conditions, demonstrated that reaction only takes place at the
C2 position of 2CPA. At 423 K the conversion of 2CPA is complete in 20 minutes and
the selectivity to PA approaches 100%. At 373 K the conversion of 2CPA is 73 % in‘
one hour and in two hours it is completed with selectivity similar to that of the former
reaction. Uncertainties were not determined in these reactions as they were run only
once per temperature and concentration variation. At these pressures and temperatures a
possible decarboxylation producing C02 should also be considered, but PA

concentrations were invariant in each sample.

Even more interesting is the behavior of this reaction in a control experiment run
at 373 K without any catalyst. After two hours the presence of PA was evident but
2CPA never completely reacted to form only PA. In addition, after two hours, a residual
LA doublet and quartet peaks were observed. The characteristic LA peaks were no
longer present in the 1H NMR spectrum after six hours.

One of the main questions is why 2CPA reacts so quickly at the C2 position
losing a chlorine atom, but it does not further react to hydrogenate the resulting PA to

form POL. To address this problem, catalyst poisoning had to be considered.

62

Qualitatively, it could be determined that HCl was generated because the samples that
were removed from the reactor during the hydrogenation of 2CPA were dark green.
Then it was assumed that either the chloride or the proton of the acid poisons the catalyst.

To test this assumption, a hydrogenation reaction LA was run with added NaCl
and a PA hydrogenation was carried out with added HCl. Because the presence of Cl'
ion could inhibit the hydrogenation reaction, LA, an easily reducible species under
normal hydrogenation conditions, should become difficult to reduce, if at all. In one
control experiment LA was hydrogenated in the presence of one weight equivalent of Cl'
(10 g of LA and 10 g of NaCl). In this reaction LA hydrogenation was not affected by
the presence of chloride anion.

Testing for the effect of acid on the reaction, PA was selected instead of LA to
avoid a possible dehydration reaction. Concentrated HCI (1 mL) was added to the
aqueous reaction mixture of PA and was allowed to react in the reactor under conditions
previously described for other substrates. PA hydrogenation was not changed by the
addition of acid as the rates of conversion and production of POL were virtually the same
as in the PA reduction by the regular protocol. Based on these results, a conclusion can
be drawn that neither Cl' nor H+ play a direct role in catalyst poisoning of these reactions.

The reactor itself may provide an ostensible explanation for the metal leaching.
The stainless steel wall of the reactor is composed of an alloy that contains a variety of
metals such as chromium that could be “washed out” of the reactor by HCl that is
generated in these reactions. Conceivably, these metals could block the active sites on
the catalyst surface and thus prevent the hydrogenation process at the carboxy functional

group.

63

The formation of PA was evident in reactions at 423 and 373 K, but in a control
experiment at 373 K and 1200 psi in which the catalyst was omitted, an appreciable
amount of PA was still produced. Furthermore, LA formation was observed in the 1H
NMR spectrum within two hours, but at the end of the six hour period it was no longer
present in the reaction mixture. To explain the generation of LA under these conditions,
water must attack 2CPA at the C2 position where the chlorine is attached. Given that
the chlorine can be easily removed in an SN2 mechanism from a substrate, water can
carry out a nucleophillic attack at the C2 position. This attack may be facilitated by
increased nucleophilicity of water at these pressures and temperatures. Scheme 2.5
provides a summary of 2CPA hydrogenation reaction at various temperatures and in

control conditions.

 

 

 

h OH
Cl 6 ours > /\I(
OH 423, 373 K 0
CI 0H
0 Control _
' OH
373 K OH OH
2 hours
0 3 % 0 8 % 0
CI
4 hours
OH OH
+
0 15 % 0

Scheme 2.5. Summary of 2CPA hydrogenation reactions at varying temperatures and
under controlled conditions.

64

2.8. Hydrogenation of Ethyl Lactate

As previously demonstrated by Adkins et al. the hydrogenation of ethyl lactate

45 In this project the

(EL) to propylene glycol (PG) goes with surprising ease.“
hydrogenation of EL was carried out at 423 K and 1200 psi. In the 1H NMR analysis of
reaction samples, dioxane (6 3.606 ppm) was used as the internal standard because the
prominent singlet of t-butyl alcohol at 1.22 ppm, used for other reaction samples,

overlaps with the doublet of EL. Unfortunately, this doublet was the only acceptable set
of peaks for the concentration determination of EL.

EL has four easily distinguishable peaks on the lH NMR spectrum (6 4.247
lH(q)/[B]; 4.084 2H(q)/[C]; 1.263 3H(d)/[A]; 1.117 3H(t)/[D]). The end products of EL
hydrogenolysis are PG (6 3.762-3.66l 1H(m)/[F]; 3.414-3.242 2H(d/d)/[G]; 0.962
3H(d)/[E]) and ETOH (8 3.500 2H(q)/[J]; 1.014 3H(t)/[H]) (Figure 2.11).

Doublet A of EL, multiplet G of PG and quartet J of ethanol were monitored for
the concentration change (see Appendix 8.2 and 8.3 for concentration profile). As has
been discussed in other sections, long base-line integration should be avoided to
minimize error in concentration determination. In the case of PG, only peak G, which is
doublet of doublet spanning a large base line, could be used for concentration

determination. This is why there is a discrepancy between the HPLC and 1H NMR
results for the percent of product formation.

While 1H NMR showed a 100i10% conversion and over 100i16% production of

PG, HPLC results indicated only 82% conversion and 77% production of PG for the

65

same reaction. Selectivity from IH NMR could not be determined due to unrealistic

percentage values for PG formation.

 

 

OH OH OH
[3] o\/[D] Ru/C _ [F] OH + _J
H ' , 1H1 1J1
[A] [C] 2 [L] [G]
O I 423 K, 1200 psi
dioxane
[Al
[D]
1
l ) [(31 ll 1.
181 .1 ‘ H
l
l
l
(11) l l [Elthl
J , l
. (. 1
m 1 ‘.. ,'. 1
4.6V 4.4. 4.2 4.0 3.8' i 73.6 3.4 3.21 V i 3.0 2.8. 2.6 2.4 V V 12.24 I 1 2.0 - 1.8V v i 71.6. i 91.4 4 ~I.Z V GI.0 V V 10.8

Figure 2.11. lH NMR spectrum of EL hydrogenation. Spectrum I is EL and spectrum
11 is the mixture of PG and ethanol after a six hour reaction time. For stacked lH NMR
spectra detailing EL hydrogenation, see Appendix 8.].

However, HPLC showed 92 % and 97 % selectivity for PG and ethanol, respectively.
These are realistic values as excellent PG yields have been demonstrated by others.”45
Furthermore, a control experiment, under H2 not employing catalyst in aqueous medium,

showed that only the hydrolysis product (LA) and ethanol are produced with 100 %

estimated selectivity.

66

2.9. Hydrogenation of 2-Acetoxypropanoic Acid

The hydrogenation of 2-acetoxypropanoic (2APA) acid was run at two different
temperatures (423, 348 K) and 1200 psi pressure. This reaction produced different
results at different temperatures. While at 423 K, PG, ethanol and acetic acid (HOAc)
were produced, reactions at 348 K only afforded LA and HOAc (Scheme 2.6).

The 1H NMR peaks of products and reactants of these reactions were referenced

to t-butyl alcohol at 1.22 ppm. The starting material 2APA, has two easily

OAc OH

OH OH
Ru/C ; + J + HOAc
H2

0 423 K, 1200 psi

   

OH

348 K, 1200 psi
OH
+ HOAc

Scheme 2.6. Hydrogenation of 2CPA at 348 and 423 K and 1200 psi.

67

distinguishable peaks upfield and a weak-signal quartet close to the water peak (6 5.011

1H(q)/[B]; 2.131 3H(s)/[C]; 1.481 3H(d)/[A]). The characteristic PG peaks (6 3.905-

3.789(m); 3.544-3.383(d/d) and l.124(d)), ethanol quartet at 3.544 ppm and HOAc

singlet at 2.058 ppm are prominent in the final sample collected after six hours (Figure

2.12).

[B

[A]

(U

I?”

(II)

5.2 5.0

Figure 2.12. Hydrogenation of 2APA at 423 K and 1200 psi produces PG, ethanol and

HOAc. Spectrum I is the starting material (2APA) and Spectrum 11 is the product

 

 

OAc [C] OH
OH Ru/C t [E] + +
H2 [1)] [F] 161 1H1
0 423 K, 1200 psi [C]
HDO

 

. F
I [In] [H], I _]

4.8 4.6 4.4 4.2 4.0 3.8 3.6 3.4

t-butyl alcohol I

 

HOAc
[A]
1; ,
. .i'
HOAc
[GI
LID]
; /
2.0. V 11.87 ' I 1.6 i V 1.4 T [2va 91.0

mixture after six hours. For stacked 1H NMR spectra of 2APA, see Appendix 9.1.

two hours and that the hydrolysis of the acetoxy group and the formation of ethanol is
significantly faster than PG generation (see Appendix 9.2 and 9.3 for concentration

profile). Previous work has demonstrated that acyl fragments are attached to the catalyst

The concentration profile reveals an increase of concentration of LA in the first

68

26.29.35

surface longer than alkoxy fragments. As hydrolysis of the ethoxy fragment of

2APA takes place, followed by LA and PG formation, a mixed kinetics can be expected
for this reaction.

The hydrogenation of 2APA is also a relatively fast reaction compared to other
acids such as propanoic acid, but 2APA’s overall reactivity is also greater than those of
LA or GA. lHNMR results of 2APA reduction at 423 K showed a 100i9% conversion
of 2-acetoxypropionic acid to form PG (74:1:11%). Although the weak signal of the
EtOH quartet at 3.544 ppm is evident, the determination of its concentration came from
HPLC analysis. An estimated 5 % of EtOH and 90 % HOAc are produced. The
selectivity to PG is 70 % which was also confirmed by HPLC. A weak ethanol peak is
present in the chromatogram that indicates an estimated 1%. It is reasonable to suggest
that the rate of hydrolysis of the acetoxy functionality is significantly faster than the rate
of its hydrogenation. This is consistent with prior findings in which the hydrogenation
of acetic acid to ethanol was difficult, requiring stringent conditions.3S

One of the problems with the hydrogenation of 2APA is its hydrolyzing acetoxy
group to form HOAc at relatively high temperatures. The acetoxy group hydrolyzed
within two hours as it was confirmed by lHNMR, and the integration value of the
developing singlet of HOAc after two hours did not change. Because the initial goal
was to observe whether the highly electron-withdrawing acetoxy group is strongly
activating in hydrogenation reactions, this temperature (423 K) had to be omitted.
Hoping that the hydrolysis of the acetoxy group can be prevented or at least minimized at

lower temperature, the hydrogenation of 2APA was carried out at 348 K. It is important

69

to retain the highly electron-withdrawing acetoxy group at the C2 position long enough to
see its effect on the hydrogenation at the carboxy group.

If one was to assume that the a more electron-withdrawing group would increase
reactivity, then one should see an increased rate of conversion of 2APA to 2-acetoxy
propanol or PG relative to PA or LA reduction. Unfortunately, even at 348 K, though
not as fast as at 423 K, the hydrolysis of the acetoxy group was evident. Results of
lHNMR clearly show a slower rate of acetic acid formation and the generation of LA as
the other product of this reaction. But because this temperature is below the optimal
range, no propylene glycol was formed in this reaction (Scheme 2.6)

A control experiment, under H2 and without catalyst, showed that hydrolysis
products (LA and HOAc) are generated exclusively from 2APA in two hours. This

result can be used to conclude that the hydrolysis of 2APA is unaffected by the presence

of the catalyst.

70

2.10. Compiled Results

A general mechanism describing the hydrogenation reaction of organic acids is
described in the scheme below. The organic chemical process in these reactions can be

understood by hydride delivery at the carboxy carbon (1) to generate the hemiacetal.

X H- x x H' x

( OH H ( H k/OH

—-> —.> -—>
< 1 OH -1120

O OH (9 ‘

1 2 3 4
X=-H,

-OH,

-CH3

-OMe

Scheme 2.7. The reaction mechanism for alcohol formation from a-substituted organic
acids in aqueous medium under heterogeneous catalytic conditions.

intermediate (2). Hemiacetal forms aldehyde (3) that in turn undergoes another hydride
delivery to form alcohol (4). Hydrogenation at the C2 carbon of 2CPA and 2APA by
hydride delivery may be an SN2 process. However, arguments can be made that
insertion of hydrogen followed by B elimination can be an alternative mechanism. The
hydrogenation of 2APA is more easily understood as it is hydrolyzed by water to form

LA under these conditions. The LA in turn can hydrogenate, as described in scheme 2.].

71

In the general procedure 10 weight % aqueous mixtures were described to react
with 1 weight % catalyst mixture (2.16 g/ 100 g of pre-reaction mixture). The activated
carbon-supported ruthenium contained only 5% ruthenium. Thus in these reactions, only
0.05 gram of ruthenium was used as the active catalyst. This naturally translates into a
non-uniform molar ratio of catalyst to substrate. For example, the results indicate that
the percent conversion 2MPA was slower than that of IBA, but one must also remember
that the catalyst ratio to 2MPA was also greater than the ratio between the catalyst and
IBA. However, even a higher catalyst loading for 2MPA hydrogenation did not produce
better conversion rates compared to IBA reductions.

The results of carbon-supported ruthenium-catalyzed hydrogenation reactions are
summarized in table 2.1 ('HN MR) and 2.2 (HPLC) and these results are plotted in Figure
13-16. Note that some of the values in the table are designated with *, meaning that they
incurred integration or instrumental error. Instrumental error was only determined for
HPLC analysis from a series of identical concentration ethanol internal standards. The
relative error, however, was only 0.15% which is negligible. The ethanol standards
produced an average response factor of 33701422t6534. The magnitude of the standard
deviation is negligible relative to the average counts of the ethanol standard. One can
conclude that instrumental error did not play a significant role in any discrepancy that
may have arisen from HPLC analysis. Only one HPLC analysis was carried out per type
of reaction except for GA which was not analyzed by HPLC. Therefore, uncertainties
from HPLC, other than instrumental error, could not be determined. Table 2.3
summarizes the selectivity values and rate data and relative rates of main product

formation.

72

.11-Wg951dw i I

 

1 Substrate

LA-

GA

EL

2MPA

MA

IBA

PA

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2CPA 2APA
" Main PAM PG '- EG PG ‘ 2MPOL IZMETVOHA PG IBOL POL
Product ‘ . ‘ .. . _. 7 .
_ % Conv. 100* 32 w 29 41 12 48 65 20 5
3 (1 hr.) _ ‘ , .
% Conv. 100* 61 41 68 21 66 93 27 20
' (2 hr.)
% Conv. 100* 73 45 78 29 78 100 35 22
(3 hr.) . , g . . ., ..
% Conv. 100* 89 ' 76 84 41 84 100 40 32
(4 hr.) 2 ,
% Conv. 100* 100 100 90 2.3.- 39 88 100 48 35
(5 hr.) _
% Conv. 100* 100 100 100 43 86 100 52 35
(6 hr.) ‘ -
% Prod. 121* 30 22 63* 16 19 32 5 _ 12
(1 hr.)
% Prod. 106* 55 32 81*“ 34 33 53 10 13
' (2 hr.)
% Prod. 94* 77 _ 44 89* 31 37 60 14 19
(3 hr.)
% Prod. 98* 74 52 99* 47 41 73 18 13
. (4 hr.)
% Prod. 96* 83 - 61 108* 50 46 66 21 15
(5 hr.) .
% Prod. 1 108'*"””82“. 62 105*" 5‘5““ 46 74 25 18
(6 hr.)

 

Table 2.1. Compiled 1H NMR results of Ru/C-catalyzed aqueous phase hydrogenations.

This table only shows the main products of reactions.

Hydrogenation of 2APA also

generated HOAc and EL hydrogenation generated large quantity of EtOH. Values
designated with a * indicate a relatively large degree of uncertainty. Reactions of 2MPA
also produced MeOH, POL and residual PG.

73

 

 

" Substrate

LA

EL

2MPA

MA

2APA

IBA

PA

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1 Main PG PG 2MPOL 2METOH ‘ 1 PG IBUOL POL
Product 1 g
% Conv. 7* 33 3 21 * 15 1
(1 hr.) 1
% Conv. 32 55 15 35 * 24 8
(“I“) _ __ _.
% Conv. 64 69 30 44 * 32 20
(3 lift)... - .
% Conv. 75 76 3O 51 * 39 21
(4 hr.)
% Conv. 83 75 38 57 * 46 23
(5 hr.)
% Conv. 87 82 48 64 * 52 33
(6 hr.)
°/o Prod. 1* 33 7 12 17 6 4
(1 hr.) .
°/o Prod. ‘ 21 59 17 24 31 11 8
(2 hr?) . .
°/o Prod. 46 62 24 31 38 15 9
(3 hr.) 1
% Prod. 60 72 36 36 46 18 12
(4 hr.) 1
1 °/o Prod. 1 60 97 43 41 46 19 14
(5 hr.)
1' % Prod. 1 67 7‘7’*’ 47 44 51 25 14
1 (6 hr.) 1

 

Table 2.2. Compiled HPLC results of Ru/C-catalyzed aqueous phase hydrogenations.
Hydrogenation of 2APA also
generated HOAc and EL hydrogenation generated large quantity of EtOH. Values
designated with a * indicate a relatively large degree of uncertainty. Reactions of 2MPA
also produced MeOH, POL and residual PG. The GA was not analyzed by HPLC.

This table only shows the main products of reactions.

74

 

3:22 m. 3 6656586 8 do 82 6:6 M
mmv 8 macros»: cosmcowoho»: 32938:? E <m<N use .5— {A JG £2 .<m2~ 563 {1:0 "888280 :5ch .m~.~ charm

 

 

‘ Eon o_ocmaoa>x23<u~lll :Qflom. 35ml Eon 29.396101“ - nWmi w__oo>_ol§,lw
- .28 266126562le ownomocoooaroeoEAIil 28 2538. III 28 ozoSIOL

2&5on 0.5...
o m v m N F o

2&5

11 . $3

o\ooN

 

$0”

 

$94

 

$8

UOQSJSAUOO °/°

l: <on

 

$05

 

o\oow

fir

i o\oom

 

 

 

£609

 

75

0:: 3 6655666 8 .23 82 Ba

x mmv S 30:88 :oumcowobin 82938ch 5 Am— vaw «5 £2 .<m2~ £02 {A we 565280 885m .v—.N e..:wmh

 

 

 

 

 

. 2302 _>.i=im.-dll.l Eula £8391” Eon ozmoflmaofioEIXl Eon. mammaegxoLW—zWLtl Eon oiSdflLll you owomulfilJ

 

23.565 2.5

. o\oo

 

 

 

 

 

 

 

i l. A$0..

o\oom

o\oom

o\oov

£60m

uogmanuoa 1y.

.- - 1 $00

. 11 . $05

1- o\oow

ll 1 I . o\oom

 

oboo F

 

76

822 z. 3 0658860
00 an 89 080 M SE 0.0 805008 cosacuwohga 000890500300 8 «S $805008 <mUN 0:0 <m<~ 80¢ Um AH 80¢ Um <n— 80¢
40m <0 80¢ OH <2 80¢ IOHEN «SEN 80¢ AOAEN 3%: 80¢ ADD“: «5 80¢ Um ¢0 8260008 808$ ..m_ N 0.59m

 

 

 

 

 

0.60 0mco_00..090_51.~l £00 0Wm0005>x900$~ Em... Buzmwnvaaeml . . 0.800. 3.5m 80:. 3020 0203090 l,
i 200 0_ocmao._d 80¢ _o:mno._n_-c+ Eon. 0:820 80: .0020 0c0_>£m+ 200 300208505. 80.: _ocm£0>xo£02-Nle.
200 0_0cmnoa>xo£02-~ E0: _ocmao.a>xo£0§.m+ E00 0:333. Ea: 6200.0 Esoofllll 200 0:03 Ea: .0020 0c0_>aoi+,
23.50.: 25h
0 n v m N F o
, $0
nX00 _.
exoom
$8
$9.
$8 /o
$8 m
D.
n
£005 m.
o
u
$8
£00m
$00—
$0:
600$
.1 1 $09

 

 

77

. .043...
.3 008.8800 m0 .60 com. 00.0 M mg “0 8000000 800080800»: 000290300000 8 <m<~ 80¢ Um 0:0 An.— 80¢ Un— J...—

80¢ 40m .<—>. 80¢ low—2N JEEN 80¢ ADA—SN J4m— 80¢ AODm— {A 80¢ Um m0 00003008 8080a dud PEEK

7 0W0 0.0c0ao.m>x200<.~ Ea... .00».m10c0_>ao._n.il.ll.
, 0.080. 35m 80¢ .0020 oceaaoil 0.00 0.9.0021 8050;398:101
0.00 0000008505. 80.. _oc0£o>xo£0s.-wl*l 0.00 0.0:000.n>x0502-m Eat .0cmaoa>xo£0.2-m+.
0.00 0.52.00. 80: 8500.0 _>.:nom.+ 0.00 0:00.. 80.. .0020 0c0.>00.n.+

 

260.59.. 65:.
0 m. 4 n N F o

 

. $0
$8
$00
$8
.‘\\1 $04

£60m

 

£60m

(1/lou1) uouenueauoo

$05

$3

nxbm

 

 

 

 

 

o\ooo —

78

 

 

Substrate Main Product °/o Selectivity to k’(acid) k(prod uct) Relative Rate of
Main Product (l/hr) (l/hr) Product Formation
2-Chloropropanoic Acid Propanoic Acid 100 1.0EI 1.0E1 250
(2CPA) (PA)
2-Acetoxypropanoic Acid Propylene Glycol 74 1.4 9.91-L-1 25
(2APA) (PG)
Ethyl Lactate Propylene Glycol l00 4.6E-l 4.6E-l 12
(EL) (PG)
Lactic Acid Propylene Glycol 82 5.4E-l 4,413.] 11
(LA) (PG)
Glycolic Acid Ethylene Glycol 62 3.1E-1 1.981 5
(GA) (GA)
Methoxyacetic Acid 2-Methoxy Ethanol 53 2.6E-l 1.4E-1 4
(MA)
Isobutyric Acid lsobutyl Alcohol 48 1.2E-l 5.6E-2 1
(IBA) (IBUOL)
2~Methoxypropanoic Acid 2-Methoxypropanol 100* 9.6E-2 9.6E-2 2
(2MPA) (2MPA)
Propanoic Acid n-Propanol 51 7813-2 405-2 1
(PA) (POL)

 

 

Table 2.3. The calculated 1H NMR results for the aqueous-phase Ru/-catalyzed
hydrogenation of organic acids and esters at 423 K and 1200 psi. The calculations are
described in the Analytical Methods of the Experimental section. The relative rates of
product formation are the ratios of k(product).

79

Conclusion

 

The effects of vicinal substituents on the aqueous-phase hydrogenations of lactic
acid and ester derivatives were studied. Electron-withdrawing and/or hydrogen bonding
substituent are thought to activate the substrate upon hydrogenation. Based on the
results of this study, it is clear that hydrogen bonding has a profound effect on the rate of
hydrogenation, but the electron-withdrawing ability of a vicinal functional group of an
organic acid also seem to have a significant influence on the selectivity to the
corresponding alcohol (see results in Table 2.3). As observed for lactic acid, glycolic
acid and ethyl lactate, their reactivity and selectivity must be due to the vicinal hydroxy .
functional group. One can theorize that this behavior may be because the fixed geometry
of these compounds, due H-bonding, facilitates the adsorption to the catalyst surface.

Substrates bearing a methoxy functional group vicinal to the carboxy functionality
have demonstrated lower activity than their H-bonding analogues. To answer the initial
question of H-bonding versus electron-withdrawing, it is clear that the non-hydrogen-
bonding but electron-withdrawing vicinal methoxy substituents did not accelerate the
reaction rate as much as the —OH groups do. However, these methoxy-substituted acids
still showed good to excellent selectivities to the corresponding alcohol products (see
Table 2.3). In addition, the appreciably lower reactivity of 2-methoxypropanoic acid
compared methoxyacetic acid is suggestive that effects other than H-bonding or electron-
withdrawal may effect reactivity toward hydrogenation.

Steric hindrance has been an important issue in organic chemistry and it should

also be considered in catalytic hydrogenations. For example, the methoxy group of

80

2MPA can be thought of simply as a sterically bulky group near the carboxy
functionality. Having both methoxy and methyl groups at the C2 position, 2MPA might
be expected to react slower than MA which lacks a C2 methyl group. Indeed, MA reacts
at three-fold greater rate as compared to 2MPA.

The steric argument is weakened, however, by the case of IBA that reacts twice as

fast as PA or even 2MPA. The IBA has 2 methyl groups at the C2 position, it lacks
electron-withdrawing and vicinal H-bonding functional group, yet the rate of conversion
to IBUOL is much faster than that of PA. This phenomenon needs to be studied and
pyvalic acid should the next candidate for hydrogenation to see how the addition of one
more methyl group affects this reaction. It is conceivable that there is yet an unknown
mechanism that may take place at the catalyst surface that could explain such
discrepancy. The relative rate of alcohol product formation is only marginally greater
for IBA hydrogenation. Taking the uncertainty values into account, IBA shows virtually
identical product formation rate with PA.
Based on both HPLC and lHNMR methods, it is clear that the increasing order of
reactivity with respect to hydrogenation at the carboxyl functional group to the
corresponding alcohol is as follows: PA < 2MPA < IBA < MA < GA < LA < EL. The
rate of conversion of substrates, however, showed a somewhat different order: PA < IBA
< 2MPA < MA < APA < LA < EL.

The conversion of 2CPA to PA cannot be ranked in the same category with the
rest of the compounds listed because it did not undergo hydrogenation at the carboxyl
functional group. Furthermore, 2APA hydrogenation is also inconclusive as competing

hydrolysis and hydrogenation take place.

81

The percent production of main products in the results table (Table 2.3) was not
entirely conclusive for a few compounds. The analysis for the hydrogenation of ethyl
lactate, for example, automatically incurred an error by having to integrate a PG
multiplet near the internal standard. Integration across long base line substantially added
to the integral value of the peaks. The 2CPA also presented a challenge as an
appreciable amount of chromium metal had been dissolved from the inside of the reactor.
However, Teflon coating of the reactor may completely eliminate leaching, thus
hydrogenation may progress and the formation of n-propanol can be expected. HPLC was
not useful for 2CPA because negative peaks in the chromatogram could not be integrated
into meaningful counts to calculate the concentrations. In addition, the acidic HPLC
column quickly hydrolyzed 2APA it showed up at the LA spot in the chromatogram and
had very similar response factor. Therefore, percent production of PG from 2APA had

to be calculated from the LA that was the hydrolysis product of 2APA.

82

Experimental

 

General Outline for the Hydrogenation Reactions

 

Except for 2-methoxy propanoic acid all compounds were purchased from
Aldrich and used as is. The hydrogenation reactions of 10 weight percent aqueous
solutions of substrates were run in high pressure Model 4560 Parr Mini Reactor (Image
1) with one weight percent activated carbon-supported ruthenium catalyst under 1200 psi
H2 pressure and at 423, 373 or 348 K. As described in the previous section, the type of
substrate used in a given hydrogenation reaction determined the choice of temperature.
For example, a highly acidic substrate, such as 2-chlroropropanoic acid, was reacted at
373 K to minimize damage to the inside metal surface of the reactor. Prior to each
hydrogenation experiment, about a 1-2 mL sample was saved from the 10 weight percent
aqueous pre-reaction mixture and designated as a “prerun” sample. The prerun sample
was used to determine the initial concentration of the reaction mixture by 1H NMR or
HPLC before it was introduced in the reactor. 1H NMR and HPLC concentration of a
prerun sample was recalculated, using simple dilution calculation (M; x VI = M; x V;
where M = concentration of compound, V = volume of compound), to obtain the actual
concentration of the stock solution. Ideally, the absolute concentration and the 1H NMR

or HPLC-determined concentration of the stock solution ought to be identical.

83

Procedure for Handling the Reactor

 

Before the hydrogenation reaction of a substrate, the internal parts of the reactor
were scrupulously cleaned and dried to avoid side reactions resulting from residual
compounds from previous experiments. The Ru/C catalyst (one weight percent relative
to the reaction mixture) was placed in the reaction chamber. After the reactor was
assembled, it was flushed three times with H2 or He gas at approximately 200 psi
pressure to remove any impurities from the reactor chamber and valves or any parts that
would otherwise be in contact with the reaction mixture. It is important to note that the
flushing must be done at a slow rate to avoid blowing the catalyst out the valves of the '
reactor. After reducing the catalyst at 423 K and 300 psi pressure over a three hour
period, the reactor was allowed to cool down to RT and depressurized to 0 psi. The
reactor was heated up again to the desired temperature (348, 373 or 423 K). The
reaction mixture was poured into a charge vessel previously cleaned and flushed three
times with H2 gas at 200 psi. The charge vessel containing the reaction mixture was
flushed twice with H2 at approximately 200 psi to remove all air. The charge vessel was
then secured to the reactor and charged up to 250 psi. The control valve between the
charge vessel and the reactor was opened to allow the transfer of aqueous substrate into
the reactor at this pressure. The pressure of the connected charge vessel and reactor was
increased up to 300 psi. The control valve between the reactor and the charge vessel was
then turned off and the charge vessel was vented before it was disassembled from the
reactor. The digital monitor interfaced with the reactor indicated the accurate pressure

and temperature inside the reactor. A metal-jacketed condenser was secured to the

84

reactor through the valve used for the charging the vessel. The condenser was quickly
flushed with N2 gas and used for collecting 1-2 mL of sample from the reactor under
pressure. This sample was designated as “prerunpot” and its concentration should be
very close to the initial (prerun) concentration of the stock solution. Prerunpot sample
must be collected as quickly as possible to minimize product formation. Finally, the
reactor pressure was raised to 1200 psi and the reaction was “officially” started.
Approximately 1-2 mL of sample was collected periodically using the condenser
(normally samples were collected each hour). After each sample collection, the
condenser was flushed with N2 gas for a few minutes and the pressure of the reactor was
readjusted to 1200 psi. The collected samples were filtered by Millex sterile syringe
driven filter units and were labeled accordingly. The inside pressure of the reactor was
recorded as the reaction proceeded. When the hydrogenation reaction was ended, the
reactor was slowly cooled down by removing its heating mantle and adjusting the
temperature to RT. The reactor was slowly vented from the gas and carefully opened
and washed for later use.

There are important considerations in operating a high pressure reactor as it poses
potential danger to its operator and surroundings. As a precaution, after charging the
reactor, the main control valve and the main gas tank should be turned off. When
introducing high pressure gas through a gas line to the reactor, gas must pass through
several control valves which should be opened one at a time. The gas tank and the
reactor should be checked periodically for possible leaking and the metal stirring impeller
must be properly installed to avoid damage to the reactor. The Teflon seals in the reactor

head should be periodically checked and replaced if needed. Finally, the reactor should

85

be periodically cleaned with 20 % aqueous acetone solution under pressure and higher

than room temperature to remove all contaminants.

 

Image 1. The Series 4560 Parr Mini Reactor interfaced with a control panel that
monitors both pressure and temperature.

Analytical methods

 

HPLC and 1H NMR were used to determine the absolute concentrations of each
sample. Each 1H NMR sample was prepared from 10 uL of dioxane or t-butyl alcohol
internal standard and 20 uL of a reaction sample and diluted to one mL by D20. The
stock solutions of the internal standards were prepared such that their 10 uL aliquots
diluted to one mL would make exactly 0.020 M concentration. After identifying the
composition of each reaction mixture by 1H NMR, the well-defined resonances,

preferably with lowest degree of splitting, were used to calculate the concentration of

86

each sample. It was also important to use an internal standard that does not overlap with
any reactant and product peaks and has only a singlet. In general one should integrate a
set of the lowest number of peaks of a given compound to avoid long base-line
integration and thus minimize error in the concentration calculation. In fact, when
possible, singlets should be used for concentration determination. To determine the
concentrations of the reactants and products in a reaction mixture, all peaks were
integrated and their values were reset, based on the integral value of the internal standard
(t-butyl alcohol set at 900; dioxane set at 800). The concentration of the standard was
normalized to 100 hydrogen atoms, and this value was inversely related to the normalized
integration value of a given product or reactant to obtain its lH NMR concentration. For
example, the integration value of a non-overlapping doublet belonging to the methyl
group of lactic acid is found at 1.41 ppm and can be used to calculate the lactic acid

concentration with respect to the reaction time by the equation below:

 

 

 

 

0.020 M t-BuOH x M of LA
900 t-BuOH singlet = 300 LA doublet x = 0-020 M
9 3

The resulting IH NMR concentration was recalculated by dilution method to obtain the
actual concentration of the reactant or product in the reaction mixture.

The HPLC samples were prepared from 100 uL of solutions of reaction mixtures
or stock solutions of standards diluted to 1.00 mL using HPLC grade water. The stock

solutions were prepared as approximately one weight percent HPLC grade aqueous

87

solutions (0.1g of substrate/ 10 g of solution). These standards were all products and
reactants that were identified by lHNMR in the various hydrogenation reactions. Each
standard having different HPLC response factor gave different number of counts (for
standard concentration and response factor see Appendix 10). The absolute
concentration of a standard was divided by the corresponding number of peak area
(counts) giving a conversion factor that in turn was multiplied by the peak area (counts)

of the corresponding compound from a reaction sample, as shown below:

 

 

C(standard)i Ci
= (CF of standard)i =
(PA of standard), (PA)i
Ci = Concentration of Species i CF = Conversion Factor

C(standard)i = Concentration of standard, PA = Peak Area

Ethanol was used as the internal standard and was prepared the same way as previously
described for reaction standards. Ethanol samples were inserted in the HPLC tray
between sets of reaction samples to monitor the concentration variation between
measurements. The peak areas of the ethanol internal standard showed negligible
standard deviation (see previous section).

The percent conversion of the substrate was calculated from the concentration of

the substrate as described below:

C(substrate)t
Conversion % = 1 — x 100

C(substrate)0

 

88

The percent production of the product derived from the starting material was calculated

as described below:

C(product)t

Product % = X 100
C(substrate)0

 

The selectivity for a reaction product was calculated from the concentrations of the

starting material and the concentration of the reaction product as:

. . C(product)t
Selectlv1ty % = x 100 only if C(product)o = 0
C(substrate)o— C(substrate)t

 

The concentration profiles of the reactions were plotted as concentration of
reactant or product versus time. The data points were fitted with an exponential decay
curve, using the Kaleidograph program. This program is designed to calculate the rate
constant of a reaction, thus the rate constant (k’(,ub,umc)) was obtained. The k’(subsm)

value was used to calculate k(product) of a corresponding product of a reaction as:

k(product) = k,(substrate) X (SCICCtiVity)

Hence the calculated k(pmducl, allowed the determination of relative rates of the reactions.

89

 

Preparation of 2-methoxypropjonic acid

 

The preparation of 2-methoxypropanoic acid (2MPA) was carried out as
described in the literature, with a couple of minor modifications.66 These modifications
involved washing and drying NaH in excess of 100% over the stoichiometric amount of
the starting material (ethyl lactate) by hexanes and dry THF in the open. Approximately
1.1 equivalent of pure and dry NaH was collected and used for the deprotonation of ethyl
lactate in dry THF. Although the literature procedure described the use of a continuous
extractor for final extraction, ordinary extraction technique was applied 30 times. The

yield was slightly above the literature value.

90

ad

 

 

 

e.. N._
L
_ z
c.— _ Q:
,4. l./. ,. _ ,. . ......\.
i , ~
{ _ W
e.. N.—
_ ll
. l). .1 .31..
: _ .
: .
c.— N._
A fit 1 r
. ._
L,
_ _
c.— N.
:
a; N.
.. f.
3 N.

v; e.. a; o."
h k ll-
1. o; a; A:
_ . Pl . l l l
.2— 9. a; A:
. r . h - _
v; e.. a; ed
1 - L
..._ 9. «A A:
v; e._ a; =.~
v; e.. a; c.~
Tl z -

Nd Ya ed QN 9n N6. v.n
- . . - L .1 z - .2. .

N.~ «an 0.” ad 9.... N n 9....

Fl 1.1 a a Lil _ tlk - L l , pl 1

«a Va 0." ad ad fin TM

1 y in Ill H A Ell l l k L I

fin tn o.~ ad ad Nd v.m

1 11 L 111. 1L 1 l

N.~ Ya 9." ad ad fin wan
- l . W . 1.- .. L -1- .

N." TN 0." QN =6. ~.n 9.».

~.~ V." o.~ a.~ 5.... ~.n «an

.2 82 use a ms. a 28; a... a 6.:

:, 393w 2.2.805 9 Ads: Eon 26m— mo couscowegn 03%
38:3 cowbfiwodxzm .8 «50on m2 Z w: AA ice—25¢.

”V Y?
Ego: e
mi v...
mSo; m
Nd. v...
950; v
~.v V...
£30; m
N6 96
230: N
~.v «.6
So; _
Ni vi
So: 0

i»

91

Beam mco._.>ao.n_lll Eon 96.3 IOI

 

2352: as:
m n v m N F

 

 

:wmood

mmonood
gmnoo .o

11 11 . 1 1 . 11 11 1 .o.w0.m . .1.

N356

 

rum—.06

 

Nvtod

T

mood

5.0

. 03.0

1 No.0

mmlwmhbb

 

 

«:22 885 3 $26.28 an oofi Ba x 31m 28 289 do
cozomm. cozmcomoagc 9: Co 2.85 cozmzcmocoo .N... 56:22?

mmod

mod

('l/low) uonmuaouoo

92

vammwm Food

momma Rood

 

Nvmmomnood

vnnmm 30.0

i .3020 05.39.". l-lh . 0.00 00010.. IT-

 

20:85 25:
m v m N
85880.0
2 5880.0
228390
825380
38028.0
1 . 1 1 l . 1086880-

O._n_I >9 005::an 7d com? Dam v. mmv am 20.4 0:03 *0

 

_. 0
103.03% 1‘ 10. L o

l. L wood

1 l .- 1. V 30.0

1- 1 1 M wood

1 . wood

1 . 8.0
$8823 08:25.0

1 1 1 1 30.0

53009 5:20.86»: 05 “—o 0:65 cozmzcmocoo .m... 50:09:

(mow) uogmuaouoo

93

 

n6 :4 N.— v.—
:

a... 3 2, I

W . ,0. l .P:

0..
_

u e e.. N. 9.—
V

a... A: N. w.—

-- . 0 . _ ,1
0.0 0.. ~ . ..._

 

“E
G
°.
.-
“1
”H
V.
v-

1
1
1
1 1
1
1 1
1

04— . a; 11

041 - a..—

o.— w—
.011.

e; m.—

e; n.—

c.N N N YN 9N n.N 9n N.n Y».
L L z z r L k V
9N N.N v.N 9N w.N 9n Nd «an
L1 . . . . t 1 1 W 1 . 1 . 0
e.N N.N YN e.N w.N 9n N.n «an
H 0 r 1.1 . . ,1. 11 t . l
9N N.N YN 9N ”N :6 NM Yn
. . -l - , L Ll H . 1 - . , . 1-

:.N N.N v.N 9N a N c n N n v m
L.
c.N N.N v.N e.N «N QM NJ. I
=.N N.N v.N 9N «N :6 N.m Y...
1 1» l 1 1 L

08 82 0% 00 ms. a was: 0% 5

A65 Bobw 000350 9 Addy 200 0:00bw mo soggwecx:
0 00020 0000500 BmefiodBm 00 0000000 M22 F ._.N £659?

00:0: 0
. ad a... Ni v...
230: m
0.». _ ad. a... N... vi
00:0: .0
ad 9v N.v v...
050: m
_
QM 9v N... vi
0030; N
_.
ad .3. N... v...
50: 0
a6 a... N.v vi
,, So: o

94

.0020 0:0.»9wllll. 200 2.0351911

 

 

 

20.52: as:
0 m 4 N N F
9
o o
1 - 1 1 1 00% -
008.0
588.0
088.0
N033
1 -1 1 1 0430.0 1 - 1 -1
00:0.n1 3:0 o 8080
1 1 1 088.. 1

mmomoo

 

 

E22 865 3 855.00% 50 8NF 0:0 v. 0N4 a .28 2.820 do
5:000. cozmcomegc 05 oo 0:005 cos—02:00:00 N.N x5592

mood

Foo

mwoo

No.0

mmoo

moo

("I/10w) uonenuaouoo

95

N.—

 

 

 

 

L

Y. 1 a; a; 01: 1.1rN11.N.11 11 v.,N c.N 1 1n.N1111L1 own N1.1n
0.. _.. 1 1e.— .11 1 a. _1 1 11 1o1N 1111 1N:N 1 . va _ L c N 1¢.N1. ._ 1- 1“.le 111 1 N »
L02— 11 Lo.1_ W- 111 11 o.1N1 NLN L v..N L 9N QN =6 1 NHN
L...- cfl— J— L Lo.N , L N..N L «N o.N 1 11 .nHN 1-1.-WW1 L N.N

Y. c; a; o.N N.N 1N 9N a.N oi. N.N
L; c; m. o N N.N v N o N n.N oi. N n
._ o; a; o.N N.N v.N 9N n.N ed N.N

1 1 . 1 1111 1 1 111 1 111.11111111111 11 1 11 1 1 . 1 1 1 1 1. l 11 1 1 11.

.me oofi 0:0 M va 00 00:0..— 50 E 300:
. L. 3:000:01: 0:0 3002 6:523 _o=0aoac0xo£0E-N 9

 

;
L.
L

L 2E8 200 0_o:00€%xo£0E-N .00 0200:0353 000:0
0:00:00 0050—0000103m .00 0000000 522 I_ .2” 05:250.

.._......1...1m..,0.....(..l... if}. a). 1 l 1 11 l 1 l 1 . 1 . 1 1
00:0; 0
. 0.0 0.0 .00 0... N0 0...
\ .....<...\J.z. v s. - 1 1 1 . 1 1 1
0.50: m
L 0.0 0.0 0.0 0 0 N... 0.0
L 050; v
L.
L 0.0 0.0 0.0 0.0 0.0 0.0
1.1.111 11111111111111.111 .1 .1 . . 1 . 1 .L
L 00:0: m
. 0.0 0.0 0.0 0.0 N4 0...
. . 0 1 1 1111 1 1 1 11 1
L
. 0030: N
0.0 0.0 0.0 0... N... 0.0
L
. So: _
L0 0.0 0.0 0... N... 0.0
L L 0 L
50: o

96

L .000090000502Nlil 000 0_00Lm000>xo£021.N.l¢I

20:85 2::

 

mum—10.0

 

.1 000 0.0.9

00000.0

000000 .0

05000.0

   
 

moo—0.0

      
   

000000.

002.00

1. 0008.01

    
      

0_.NN00

 

 

«:22 00390 3 000 00.0.0 0:0 x mm». 30 200 0L0cmaoa>xo£0E1N
00 $3000.. 003000020? 05 00 0.390 00303000000 .Nd 0000:0090

000.0

00.0

000.0

mwoo

(“I/10w) uouenuaouog

97

 

L0cmnoa>xo£1021NII1|1 1%100Lcoaegxom31021N lOl L
23505 as:
m m v m N F O

 

 

        
  
            

Nwmwmmoooo

Commmrooo

m0 Emwmooo

000030000
0mm 0000000

0080 «.0000
0000000000

000200000 500000000

N 0 09000000

omwvmtooo

0000000000

 

 

01.0: 3 005E000 in 8.9 000 x 0N0 30 200 0.000000%xoc30E1N
00 0030000 0030:0090? 0.3 00 9305 00303000000 .0...” 5.2.0030

000.0

N000

000.0

voo .0

000.0

000.0

000.0

000.0

000.0

(1/10ul) uonwuaauoo

98

06 N6

 

vi a...

v.— o.—
, . . . . ‘ 7 ,. . 0 V 7 ,V , v y . , s ‘ . , . . A . .
{N , _ r ,7, 7.“ ~ . 77.,"0‘4‘7

 

vi 04.

ad 0d

0d N.N
, _-H ..

.me 00S 05“ M mmv 8 E50: x6 5 EOE was $002 Amp—EEG.
_o§£o0xo£ofi-m 8 ?:>0 Eon ouoom0x0£2= no 2235080057

 

 

 

03:0 9505.3 030038-950 mo 88on m2 Z P .3“ 50:22:.

99

m _ocm£m0xo£m2-~.lll.f Bow 2580x262 IOI.

 

 

 

 

 

23:85 as:
o w. v m N P o
7 I . _ . _ o
o
582.
338.0 838d 338.0
IIII\ OII IVII I IIIIIIIIIIIII II I I\II\II: II. .
8886 «:83 mooo
‘Il.\\
858.0
T . 328.0
583 8&8 o
1.31% III’ III! I , III :3
585.0
I| I |I 0,031.0 I _ I I I 05.0
\I I I JiillI _ IIfIl, II\II - 8.0
283
33.

 

£22 565 3 35.59% 7d cow? new x mmv am you 2680x238
.o 8:89 5.65093; 9: .o 2.35 5392850 .3. 59522

(mow) uogmeuaouoa
1 OO

_ocmc.«mwxo£m2-m+ 200 usummxxocamzlol

 

 

23.59: 25.0
o m e m N F o
82850.0 8&853
\| 53330...

. Emmmmood

mt F380 0 88883
II\I ' I»§D.D I‘ I \‘\ \\(\“ ‘\ ‘ “‘
$58886 an 5386

89883

£3883

£8583
$838.0
89385.0

 

 

oI_n=I_ >0 cmEEEEu an 003 05 x mmv 0m Eon 2880x0508 S
.8089 cozmcmmeuE 05 .6 2005 5092850 .né 505%:

I N000

30.0

000.0

000.0

00.0

N000

('mow) uonenuaouoa

101

0.0 0.. N._ na— 0._ 0.. 0.” N.N v.~ 0." 0.n 0.0 «.0

 

 

7 I 7 7 7 7 7 L I» I I
.7 ,7
7 7
0.0 0.— ~.— 01— 0._ a; 0.N N.N YN 0.~ 0d 0.n Nd
.7II77 7 . 7 7 I7... . 7 -ILLLIIIFL‘II. I 7 . 7 7 I 7 . II PI 7 I7
7 7
0.0 0.— N. v; 0; a; 0.N N.N YN ed 0d 0.0 Nd
. . 7 7 I- 7I. 7 7 7 I 7 I0 II ILIIIII I [I I
7.77.7.7 . 7... 7,. \I-..
77 7. 7
0.0 0.— N.7_7 v4 0; 04 0." N.N van ed Wm 0.». NJ.
7.. :7 . :07. 77 I7II|7.III7 III 7- 7 7. 7 7
7 7
. 7. 7
0.0 0._ N. 7 v4 0; a; 0.~ N.N v.~ 0.~ a.~ 0.0. N.n
7 r I I I :7I II. .7\I LIL III I I I I 7 I . 7 7
0.0 0.— N. 7 v.— 04 a; 0.~ N.N TN 0.~ 0d 0.m N.m
777
_ 7 7
w 0 0._ ”H v._ 0; a; 0.N N.N YN 0d wé 0.0 Nd
.. _ 7 7 . 7 7 . .:I..I I7I I7I . . 7II III7I0 7

.7... .78. .77... v. 3.. .7. .52. x7. .7. 30.5.7 72.8.7W
00:09.7 9 2.50 Eon 0:00.587 00 57850800: 820

38:3 82.28-0730 0o 27020 02 Z :. ém 7.6.5.5....

102

_ocmSnom_ + How 2705002 + 7

.3735 as:
7. m N . o

 

IImmmmoPplll

 

00—. 00.0

   
 

Nmmmood

00000.0 000 wood

 

  
   
  
 

00000.0
00000.0

rmmmod

 

 

«:22 8.9.. 7... 85572.... 78 SN. 2... x 8.1.. .778 277.7387
0c .8008. cozmcomegc m5 770 man cozmzcoocoo .~.m 00:00.2

000.0

I _.0.0

9.0.0

I. No.0

mmod

(mow) uoueuuaouoa

103

_ocmSmom_Il7IP .EoImetbgofl IT 7

0.0520 25.0

 

$N®0500

I1I . I r . .IIIIIImmNBNood

038.030...

      

 

memmmmood

 

0VN000®000

.30000080I1

 

  

  

mmmmmmr 00.0

N000

I 30.0

000.0

IIIII I. moo...

_.0.0

 

 

07.... 7... 8.777753% .8 8N. Em v7 NNv a 2% 0777.738.
.6 20:39 20.650900; $5 76 man coamzcmocoo .n.m 50:20.?

N000

('mow) uonenuaouog

104

0.—

0.~ «.N v.~

0HN II ~.~ v..~
.0.N N.N ed
0 N I ~..~ v m
0.~ N.N v N

fin

~.n

.me 00m. 05 M mg a mason xmm E 36.0 3.8089: 8 2.0 065080 00
80050800; 0800 982.3 030788-030 mo «08% 022 I. 4.0 $050.3.

0....

0.0

0.0

0.0

06

$397 w

~.v

1?

vi

0...

v6

v6

105

Bamaoié III Eon 2559.". IOI

2332: as;
w m N F o

 

 

NNmmoo .o

       

omFNOd mmFNod

  
      

mummood

mommoo .o

mmmmod

Nwomod

mnwwod

350.0
gamma?

 

 

mEz c906 E vwEEEEU an 009 95 v. mmv Hm Eom 29.805
be 8:89 coamcmmegc 9: ho man cougcoocoo .N.@ 523%:

mood

m—od

No.0

mmod

mmod

(mow) uonuzuaouoo

106

_. _ocmqoi-clll Boa oanoEIOI

 

23.35 «E:-
o m c m N v o

 

 

 

||-‘\\OII
33353 11.1 8838.0

 

 

 

 

 

   
   

1+1 .
,T .T. -- $03303 Ego @5888 o ...... -. - -
Noommowood
mnmnwnood
28800.0 i . 080E1.11 888098 :I ,

000005050
111|1- 1: 1|.:. - - - . :. . . Nov-Nmmwéqol

Nmmmoom V0.0

 

 

od: 3 85.52% 7d 83 Em x g a. 20m 2880;.
8 scream. 885093; 9: __o 2005 8388050 .md x559?

N000

30.0

- 000.0

000.0

«0.0

N80

3.0.0

(mow) uonmuaouoa

107

0.0 0..

0.0 0..

 

N.—

 

 

 

. 1 P . . . . w . - . l - 1 _ , _ k . .
i x \ :1 )K11. - 1 I -111 1 111 11 - 1-1 1 1- 1- J. 1 1 - 1 11 11 1 -
.... e.. 0.. a." N.~ v.“ ed fin 0.». a... 1m 9....
1.3.1111 _ 0 0 - 0 _ - - : k 1 . .. 0 1 1
\.-\u\1 \11 ,1 x 1 II 1111 I .- \ 11 1-x... .. \11 1 I a 11.. | 1 1 n 1 11 . 1 11 1111 1111111111
\.\ g ._ .... a...
_._ _
1. e.. 0.. 0.N ~.~ vd 9n .3 06 fl... 2.... o..-
_- : _ , . . r - ». 1 h :1 1.1 - 1111.11111 W 1

7..

0.. 0.. 0d N.N YN ed ad 0.0 Ni. Ya o...

 

0.. 0.~ NA v.~ ed a.” 0.... N.” ...n o...

0.. 0.~ ad TN ed a.” 06 ”.0 w.»- 0.».

0.. 0.~ N.N in ed 0d 0.x. ".0 TM 0....

 

_
._
__
.
_

_ .23: Bow 20:80.5 83080 8 .8 00m. 0cm v. mmv 3 B20;
x5 5 $508 Eon 2888005200 .0 808880»; 820
: 38:3 38:83.93. 00 86on v.2 Z I. .6 5050.7.

0.».

06 N... v...

0... N... v...

06 ~.v 9...
9:5: m

0... N... v...
£30; .

0... N... v4

.30: _ .0

04 N... 96

108

0. m5 2%. .2839“. lxl a. 2m... 28 oanoaeozoldl o. 83 28 2853+ Q8: 28 28.859251.

 

 

 

    

 

 

.352: as.»
m m w m N . 3 no to
? fi Le a a .
o
- waged - 1 -
8386
22.8...
8:3. .
mmmad ,
, 81o... . 1
. 088 o .886 . «55.0
1MB“: , 1,- 155.31 mm§1$m$ ; 1 -1-
80.0 Bowed £08... .5861 .1) 8.80
mmmmod :80 o
.830

mmmmod

Stod-

1

 

 

v. mum cam mmv 6 20m o_ocmaoa9o_no.m
.0 585023; m5 .0 28:. 8:928:00 .NK 59.2.3.

000.0

5.0

900

No.0

mmod

00.0

('mom) uonenuaouog

109

0.0 0. . ~.. 0.. 0.. 0. . 0.N N.N v.~ 0.~ 0.0 0.0 N.0 0.0 0.0 0.0 0.v ~.v v.9

 

: _:..
t 0 _ 3:0: 0
0.0 0. _ n.. v... 0.. 0.. 0." N.N 0.0 0.0 0.0 0.0 «.0 v.0 0.0 0.0 0.0 0.0 v...
._- ,. ., -, ,0
v _ . . .s I .I. ._:,\. _ \. ... .
t 0 3:0: 0
0.0 0.. a. n... 0.. 0.. 0.~ N.N v.~ ed 0.~ 0.0 0.0 0.0 0.0 0.0 0.0 N... v.9
0 0, 1‘ ml , .01--..r“\\\ 0, I. . . - .xrx - 0L. ,. g . W M .. .
._ £00; 0
0.0 0.. N.— v.. 0.. 0.. 0.0 0.0 in 0.~ 0.~ 0.0 0.0 v.0 0 0 0.0 0.0 N... v...
. . § £30; 0
M
0.0 0.. 0.. v. 0.. 0. 0... - v" 00 0.~ 0.0 0.0 V0 00 00 0.0 Na 0..
_ 350: m
0.0 0. _ 0.. 9... 0.. 0.. 0.~ n.” v.0 0.0 0.~ 0.0 «.0 0.0 0&0 0.0 0.0 N... v...
. _. So; _
f
0.0 0.. 0.. 0.. 0.. 0.. 0.~ N.N 0.0 0.0 0.~ 0.0 ~.0 0.0 0.0 0.0 0.9 N... v...
- _.
4 g A Son 0
.8 82 Ea M MS. a 28; x... 5 E88 6525 Ea 8.0
.820 002.380 8 3.00 898. .050 00 0283080.... 89.0 J

.3823 03388.93. 00 «58% .022 F .—.0 x...=2.0<

0.0

0.0.

110

 

 

0&6 000..»000mlll 0.0.00- _.>£m+

 

 

00035 as:
0 m w n N F
0
528.0
0580.0
; I -iLli ; 30300;? . - \
0808.0
08 50
£20..
[ - - mvoadlx a!
605.0
0080.0 -
58.0 ¢ 4 0880
T
028.0

 

 

0.22 0290 3 850028 00 8E 00.... x 3. a 298. 350
.0 0000000000? 05 .0 20000 00000000000 0.0 5002:?

000.0

. _.0.0

05.0

No.0

000.0

(mow) uonmuoouog

111

 

7 Bobmffliaoilll 0.0.09 .000 +

 

23505 20:.

 

VN0000000

 

 

/ $000280

 

 

- 0000000000

05.000 :000

   

v 0000NN000

 

.00va0000

         

 

. . 500,330.
0080030 0

 

0.5..VON000.

         

vwow ..0n00.0

 

 

0:: 3 8200200 00 000. 20 x 3. .0 2000 350
.0 00000000002 00. .0 00.000 00000000000 0.0 500023

.000

1 N000

000.0

30.0

000.0

000.0

mood

. 000.0

('I/Iow) uoumuaouoo

112

 

 

113

0. . N. . 0.. 0.. 0.. 0.N N.N v.N 0.N 0.N 0.0 N.0 v.0 0.0 0.0 0.v N.0. vi 0...
_ 1 1 . 0 . _ 0
.\ 1.1/34:. ._... 1... 00...... ,1 1 .3 x. 1 1.1 1 111 1111 1 1 1 1 11 1 1 2., ....1 ,0... 1 5...? 1 1 1 1 11 1 11
, _ .
I. 0 m_
0.. N.- 0.. 0.. 0 . 0.N W N.N 0.N 0.N 0.N 0.0 N.0 0.0 0.0 0.0 0.0 N... 0.0 06
. .4. 1 _.. .111 1 1 111 1 1 0 11 1.11 1 .
1 1 _. 0).... _. 011 1 111 11 1111 1111 * 1 1 1 1 1 1111 11 1 11111 1 - 1 no... 0...... 1 1 1.1.. .1 1 1 1 111 1 11111 1
, g ,. . 0 “_
.1 g _
1 g u
0.. N.. v . 0.. 0.. 0.N _. N.N v.N 0.N 0.N 0.0 N.0 v.0 0 0 0.0 0.0 N.0 0.0 0.0
G: .3). . . .
w 0 1, ._
0 ._
0.. N._ v.— 0.. 0.. 0.N .__ N.N 1N 0.N 0.N 0.0 N.0 0.0 0.0 0.0 0... H... v... 0.0
0 1 _ L1F1 1 11111. 1 a 0 . V 0 . 0 . 1 M 11111 1" L 11 111111 1111111,. 1, 0 . _. _ _ . ._ 0
1 1 1 1 .1 01.241. \10. 1 111.< 11 l 1 1 . (1 0... l 1 1 1 1 1111 1 1 1 R ...r. 11 1 1 1 11.. 1111 1 11 1| 1 1
: . w
0.. N _ 0.. 0.. 0.. 0.N . N.N 0.N 0.N 0.N 0.0 N.0 v.0 0.0 0.0 0... N.» w... 0.0
,< 1 . 1.....1 \111 1 1 1.1 1 1 11 1 1 1 1 1 1 1.. 111111 I 1 1 ..... 1
.,_,
0.. N. _ w... 0.. 0.. 0.N __ N.N 0.N 0.N 0.N 0.0 N.0 0.0 0 0 0.0 0... N... v... 0.0
1 1111 1 1......_\. .101 1 1 1 11 1 l 1 1 11111 1 1111111111 1 1 1 11
0.. N Y. 0.. 0.. 0.N N.N 0.N 0.N 0.N 0.0 N.0 0.0 0.0 0 0 0.0 N... v... 0...
1 1 1 1 1 .1 1 0 11 1.1 1 0 _
1 1 1. .11., .a 1 )1 1 1 11111111111 )1 1 1 1 1 1 11 11 1 1 11 1 1 1 1111 1 1 111 1
. 0
. 00 82 0.. 00 ms. 0 .152 E 5 :00 05. 200 6.0 :20
00030.0 00 A<m<Nv 0.00 0.0000000000000010 .0 0000000000»...

 

00000 0:00:00 000.00.00-93. .0 00.00% 0.22 :_ ...0 0.000000.

.8210 mamaaoi If Rom 0:00.. III 200 o?01.m®.axxoumow-ml§l

 

   

 

 

  
   
  

20...:on 25...
m n v m N P
9 O 9
o
11wmm1~8d $50.? 1
111 1 £88.10
0300.0 0308.0 vmomood
1 13300.01 1 111111 .1 1 1
850.0 505 o .
820.0 .
i 4 QFMFO o
1 1 1 .1393 1 1 11

$5..

 

 

E22 865 B $58.28 an 89 new v. a? a 20m

20:30.0?280130 53080062 05 00 man cozmscwocoo .«d x_u:oqn<

wood

wood

wood

wood

5.0

N56

1 30.0

90.0

1 30.0

No.0

(mow) uonlmuaouoo

114

 

 

_ .820 000.3105 oodowflEwocoWIdl. u_om1o_8M._ IMI 13.153 Rom o_co_109a1-w28«1~|16|

 

Aarsos 25»
o n v m N F
11111.11 1 1 1 1 1 1 1 1 1 1 1, 1 1 11E0mflqqo 1.1
8338 o $53280
11111/ . 1.111111 1.1 .11 111- 1. 1
memoveoo o 63883 .gmomoo 0 888800 mmmwmvmoo o1
éooo
111111 1.11 1.3330311 . A. 1 1 1 1
88830.0 . «Bounced .
R8933
11. .111 198008010. 1 1 1 1 1.1 1 1 1. 1
51088010 mommmmvoo;
«80880.0
1 1 1 888890 1 1
1 1 1 11 1 1 1 1 1 1 1 1111888390.
98583

 

 

  

 

 

 

9a: 3 855.2% 2 82 Ba x m? a 28
0_0cmaoa>xoumom-w n_0 cozmcmmegc 05 *0 man cozmbcwocoo .nd x_u:aan<

Food

1 wood

1 mood

vooo

mood

oooo

woo .o

wood

wood

(mow) uouwuaouoo

115

Appendix 10. Table of HPLC standards (prepared as ~0.1 g of standard/ 10

.3 of solution) and their corresponding peak area (counts).

'A'WX-‘W'S'ahfl‘i MV’h-wéa ATM?“

va-vamaofi-F «watt.

wflTWJmWWWm‘ETfiil-fi rWt‘

 

' HPLC Standard

Cone. of Stock
3 _S‘01Htion.(mol/Lg)p

HPLC Peak Area
(Counts) . -.

 

Lactic Acid (LA) _

 

 

 

 

i 11599-2 _ ' 4599399
f: 9:012ka 6119,9111???) : 17813-2..- 7651,4924
: Ethyl Lactate__<EL>_ 9149-3 1846357
:9 Ethanol 119.1091.) . 2795-2. 9984916 A
._ Pr.9999.9i.9.é9191_(w.A1.2174132 .; {8395.9}.

 

n-Pr999991_(P_0L)

3 1.72132

i 4218308

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Methanol <MEOH> 9749-9 w ' 1929536,
2-Acetoxypropanoic f '
Acid (2APA) 8.73E-3 3218327
Ace.ti.c,.Ac.id.<H0Ac>. 29919-2 : 3562999,,_,,,_
2-Chloropropanoic § ;
.ACldaCPA) . 9 1.79.1.5‘2..__.__.. 1 114.975.4517- .. ______
f 2- -Methoxypropanoic ' f
;; Acid(2MPA) , 9.85E-3 1; 37.38495
Chloropropano‘l 1.41E-2‘ : 6299451 . _______
‘ Glycolic Acid ;
2 (GA). .. . 18413-2 ..... _.3Z_2.2?}9..__
EtherneGchol (EG) 91719-2 , 9590556
gi Methoxyacetic Acid
(MA) W 1319-2 4.674.999
2- -Methoxyethano1
(2METOH) 1.5.45-2 : 4424119 .,
Isobutyric Acid '
: (IBA1 19919-2 14567477
; 19999999911113110191 n 9 1579-2 539.9779

 

116

 

 

Bibliography

(1) Association, N. C. G.: St Louis, Missouri, 2000.

(2) Jackson, J. 13.; Miller, D. J .; Kovacs, D. G. Publication being submitted
2002.

(3) Noyori, R.; Okhuma, T. Angew. Chem. Int. Ed. 2001, 40, 41-67.

(4) Singh, U. K.; Vannice, A. M. Appl. Catal. A: General 2001, 213, 1-24.
(5) Baiker, A. Chem. Rev. 1999, 99, 453-473.

(6) Augustine, R. Catalysis Today 1997, 37, 419-440.

(7) Thompson, H. W.; Naipawer, R. E. J. Am. Chem. Soc 1973, 95, 6379-
6382.

(8) Thompson, H. W. J. Org. Chem. 1971, 36, 2577-2582.

(9) Fried, J .; Edwards, A. Organic Reactions in Steroid Chemistry; Van
Nostrand Reinhold, 1976; Vol. 1.

(10) Cram, D. J.; Abd Elhafez, F. A. J. Am. Chem. Soc 1952, 74, 5828-5851.

(1 l) Noyori, R. Asymmetric Catalysis in Organic Synthesis; Wiley: New York,
NY, 1994.

(12) Akabori, S.; Izumi, Y.; Fuji, Y.; Sakurai, S. Nature 1956, 178, 323-340.

(13) Singh, U. K.; Landau, R. N.; Sim, Y.; Leblond, C.; Blackmond, D. G.;
Tanielyan, S.; Augustine, R. L. J. Catal. 1995, 154, 91-104.

(14) Sun, Y.; Wang, J.; Leblond, C.; Landau, R. N.; Blackmond, D. G. J.
Catal. 1996, 161, 759-771.

(15) Sun, Y.; Landau, R. N.; Wang, J.; Leblond, C.; Blackmond, D. G. J. Am.
Chem. Soc 1996, 118, 1348-1353.

(16) Sun, Y.; Wang, J .; Leblond, C.; Reamer, R. A.; Laquidara, J.; Sowa, J. R.
J.; Blackmond, D. G. J. Org. Chem. 1997, 548, 65-72.

(1 7) Blaser, H.-U.; Jalett, H.-P.; Garland, M.; Studer, 11.; Hhies, A.; Wirth-
Tijani, J. J. Catal. 1998, 173, 282-292.

117

(18)
(19)
(20>
<21)
(22)
(23)
(24)

(25)

(26)
(27)
(28)

(29)

(30)

(31)

(32)

(33)

(34)

(35)

(36)

Singh, U. K.; Vannice, A. M. J. Catal. 2000, 191, 165-180.

Wehrli, J. T.; Baiker, A. J. J. Mol. Catal. 1989, 57, 1989-1998.
Blaser, H.-U.; Jalett, H.-P.; Wiehl, J. J. Mo]. Catal. 1991, 68, 215-228.
Struyk, J.; Scholten, J. J. F. Appl. Catal. 1990, 62, 151-173.

West, R. J. Am. Chem. Soc 1959, 81, 1615-1625.

Pauling, L. Proc. R. Soc. London Ser. A 1949, 196, 343-348.

Boudart, M. Adv. Catal. 1969, 20, 153-159.

Broadbent, H. S.; Campbell, G. C.; Bartley, W. J .; Johnson, J. H. J. Org.
Chem. 1959, 24, 1847-1854.

Turek, T.; Trimm, D. L.; Cant, N. W. Catal. Rev. Sci. Eng. 1994, 36, 645.
CRC Handbook 1992.
Christiansen J. A. In US Patent 1,302,011, 1919.

Evans, J. W.; Casey, P. S.; Mainwright, M. S.; Trimm, D. L.; Cant, N. W.
App]. Catal. 1983, 7, 31-41.

Evans, J. W.; Tonner, S. P.; Mainwright, M. S.; Trimm, D. L. Chem. Eng.
Today 1983, 509-516.

Cerveny, L.; Havlicek, V. Sb. Vys. Sk. Chem-Techno]. Praze, C: Org.
Chem. Technol. 1988, C30, 125-141.

Sorum, P. A.; Onsager, O. T. Proc. Int. Congr. Catal. 1984, 8, 233-244.

Kim, K. M.; Woo, H. C.; Cheong, M.; Kim, J. C.; Lee, K. H.; Lee, J. S.;
Kim, Y. G. App]. Catal. A: Generalo 1992, 83, 15-30.

Gormley, R. J .; Rao, V. U. S.; Soong, Y.; Micheli, E. App]. Catal. A:
Generalo 1992, 87, 81-101.

Monti, D. M.; Cant, N. W.; Trimm, D. L.; Mainwright, M. S. Journal of
Catalysis 1986, 100, 17-27.

Lazier, W. A. In US Patent 2079414, 1937.

118

 

(37)

(38)

(39)

(40)

(41)

(42)

(43)

(44)
(45)

(46)

(47)

(48)

(49)

(50)

(51)

(52)

(53)

Yan, T. Y.; Albright, L. F.; Case, L. C. Ind. Eng. Chem. Res. Prod. Res.
Dev. 1965, 4, 101-110.

Grey, R. A.; Pez, G. P.; Wallo, A.; Corsi, J. J. Chem. Soc. Chem.
Commun. 1980, 783-784.

Grey, R. A.; Pez, G. P. In US Patent 4,232,170, 1979.

Evans, J. W.; Mainwright, M. S.; Cant, N. W.; Trimm, D. L. J.
Catal. 1984, 88, 203-213.

Ferreti, O. A.; Boumonville, J. P.; Mabillon, G.; Martino, G.; Candy, J. P.;
Basset, J.-M. J. Mol. Catal. 1991, 67, 283-294.

Claus, P.; Lucas, M.; Lucke, B. App]. Catal. A: General 1991, 79, 1-18.

Agarwal, A. K.; Mainwright, M. S.; Trimm, D. L.; Cant, N. W. J. Mol.
Catal. 1988, 45, 247-254.

Adkins, H.; Billica, H. R. J. Am. Chem. Soc 1948, 70, 3121-3125.
Adkins, H.; Bowden, E. J. Am. Chem. Soc 1934, 56, 689-691.

Thomas, D. J .; Wehrli, J. T.; Mainwright, M. S.; Trimm, D. L. App].
Catal. A: General 1992, 86, 101-114.

Turek, T.; Trimm, D. L.; Black, D. S.; Cant, N. W. App]. Catal. A:
General 1994, 116, 137-150.

Kohler, M. A.; Cant, N. W.; Mainwright, M. S.; Trimm, D. L. In Proc. 9th
International Congress on Catalysis; The Chemical Institute of Canada,

Ottawa: Calgary, 1988, p 1043.

Monti, D. M.; Cant, N. W.; Trimm, D. L.; Mainwright, M. S. J. Catal.,
100, 28.

Thomas, D. J .; Stammbach, M. R.; Cant, N. W.; Mainwright, M. S.;
Trimm, D. L. Ind. Eng. Chem. Res. 1990, 29, 204.

C outeau, W.; Dungelhoeff, M. L.; Hendrickx, A.; Verbeest, R. In Ger.
Patent 2 553 959: Germany, 1976.

Bradley, M. W. Belg. Patent 892 958 1982.

Sharif, M.; Turner, K. In Eur. Patent App]. 143 634, 1986.

119

 

(54)
(55)
(56)

(57)

(59)

(60)

(61)

(62)

(63)
(64)

(65)

(66)

Kouba, J. A.; Zletz, A. In US Patent 4 613 707, 1986.

Adkins, H.; Folkers, K. J. Am. Chem. Soc 1932, 54, 1145-1154.

Arnold, F. R.; Lazier, W. A. In US Patent 2,116,552: USA, 1938.

Guyer, A.; Bieler, A.; Jaberg, K. Helv. Chim. Acta 1947, 30, 39.

(58) Adams, P. T.; Selff, R. E.; Tolbert, B. M. J. Org. Chem. 1952, 74,
2416-2423

Rayadhyaksha, R. A.; Karwa, S. L. Chem. Eng. Sci. 1984, 41, 1765-1770.

Kitson, M.; Williams, P. S. In US Patent 4, 777,303; BP Chemicals Ltd.:
USA, 1988.

Kitson, M.; Williams, P. S. In US Patent 4,826, 795; BP Chemicals Ltd.:
USA, 1989.

Kitson, M.; Williams, P. S. In US Patent 4,804,791; BP Chemicals Ltd.:
USA, 1989. -

Antons, S. In US Patent 5, 731,479; Bayer Aktiengesellschaft: USA, 1998.
Antons, S. In US Patent; Bayer Aktiengesellschaft: USA, 2002.

Zhang, 2.; Jackson, J. E.; Miller, D. J. App]. Catal. A: General 2001, 219,
89-98.

Yonezava, Y.; Hino, T.; Tokita, Y.; Matsuda, K.; Ikeda, T. Tetrahedron.
1997, 53(24), 14287-14296.

120

 

IIIIIIIIIIIIIIIIIIIIIIIIIIIIIII

mull1134131111111Will