. 1:. .

anflmnm
by»:
g

Ea“. ,1.

., it . « >
33,252 a
3 :4
n}...
z.
. ,

 

‘3
{21. J. .

.fll ‘

 

 

 

 

V (xx :1 '
k." ‘ \ .
at, v

UV
\1“ i.‘

\ \.
'Xs

 

 

LBRARY
Michigan State
L_‘niversity

 

 

This is to certify that the
dissertation entitled

CATALYTIC HYDROGENATION OF AMINO ACIDS AND
POLYHYDROXYALKANOATES

presented by

Frank Thomas Jere

has been accepted towards fulfillment
of the requirements for the

PhD. degree in Chemical Engineenng

Baa-w 0 Mat

Major Prflessor‘s Signature
8/1l /o 3

Date

MSU is an Afiirmative Action/Equal Opportunity Institution

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

 

DATE DUE DATE DUE DATE DUE

 

Jail Mme

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 cJClRCIDatoDuopesopJS

CATALYTIC HYDROGENATION OF AMINO ACIDS AND
POLYHYDROXYALKANOATES

By

Frank Thomas Jere

A DISSERTATION

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

DOCTOR OF PHILOSPHY

Department of Chemical Engineering and Materials Science

2003

ABSTRACT

CATALYTIC HYDROGENATION OF AMINO ACIDS AND
POLYHYDROXYALKANOATES

By

Frank Thomas Jere

Aqueous phase hydrogenation of L-alanine (L-(+)-2-amino-propanoic acid)
to L-alaninol (L-(+)-2-amino-propanol) was performed in a three-phase, stirred
batch reactor. Under optimal conditions selectivities over 97% with 93% alanine
conversion and > 99 ee% was achieved. Observed side products included
ammonia, ethylamine, and isopropylamine. The catalyst used for this
hydrogenation was 5 wt% ruthenium on carbon.

Calculations using standard engineering correlations showed that mass
transfer through the gas-liquid and liquid-solid films and intra-particle mass
transfer did not limit the rate of conversion. The absence of mass transfer
limitations allowed for the derivation of a kinetic model. Conditions within the
following experimental matrix were used for kinetic model parameter estimation:
alanine feed concentration from 0.22 to 0.46 M, phosphoric acid feed
concentration from 0 to 1.2 M, temperatures from 90 to 150°C, hydrogen
pressures from 5 to 13 MPa H2, and 1 to 2 grams catalyst per 100 grams feed. A
complete kinetic model was developed that predicts reaction trajectories to an

8% error between predicted and experimental data. Modeling results showed

showed that the activation energy for the surface reaction step was 81.5 kJ/mol.
The model predicted the observed dependences on hydrogen pressure,
temperature, and ionic solution structure.

Hydrogenations have also been performed on biobased polymers PHB
(poly-3-hydroxybutanoic acid) and PLA (polylactic acid; 2-hydroxypropanoic
acid). These reactions occurred through a two step process of hydrolysis of
polymer chains to monomer and soluble short chained oligomers followed by
hydrogenation to alcohol product. PHB was found to be only slightly reactive
toward hydrogenation. PLA, on the other hand, was easily converted to

propylene glycol (PG) at 150°C and 1000 psi H2 over a 5% Ru/C catalyst.

Dedicated to my love, my life, my wife.

iv

ACKNOWLEDGEMENTS

I would like to express my sincerest gratitude to my advisor, Dr. Dennis
Miller, for enthusiastic support during this work. Thanks to Dr. James Jackson
for guidance and furthering my development as a scientist. Thanks to Dr. John
Frost and Dr. Ramani Narayan for their service as members of my doctoral
committee.

Special thanks to fellow group members Mike Shafer, Atul Dhale,
Dushyant Shekawat, Shubbam Chopade, Zhizang Zhang, Dalila Kovacs, Lars
Peerboom, Prashant Srivastava, Brian Hogle, Abe Schuitman, and Brad Rokosz
for their suggestions and help during this project.

I also want to offer my heartfelt thanks to my parents for their support

throughout my education.

TABLE OF CONTENTS

Chapter 1. Introduction

1.1 Background

1.2 Literature Review
1.2.1 Amino acid background
1.2.2 Amino acid production
1.2.3 Amino alcohols

1.2.3.1 Amino alcohol synthesis techniques

1.2.4 Organic acid hydrogenation

Chapter 2. Experimental Methods
2.1 Amino acid hydrogenation studies
2.1.1 Materials
2.1.2 Batch reactor system
2.1.3 Amino acid analysis
2.2.3.1 HPLC analysis
2.2.3.2 Use of ‘H NMR for analysis of amino
acids and amino alcohols
2.2.3.3 Amino acid hydrogenation conditions
matrix for kinetic parameter determination
2.2 Experimental methods for PLA and PHB hydrogenation
2.2.1 Materials
2.2.2 Batch reactor system
2.2.3 Analysis for PLA and PHB reaction products
2.2.3.1 HPLC analysis of monomeric materials
2.2.3.2 HPLC analysis of oligomeric materials
2.2.3.3 NMR analysis of oligomeric lactic acid

Chapter 3. Amino acid hydrogenation results
3.1 Exploratory studies
3.2 Experimental results with alanine
3.2.1 Control experiments with alanine
3.2.2 Effect of acid addition
3.2.2.1 Sub-stoichiometric phosphoric acid
addition
3.2.2.2 Effect of molar excess phosphoric acid
on conversion
3.2.3 Other acid solutions
3.2.4 Effect of pressure
3.2.5 Temperature effect on amino acid hydrogenation
3.2.6 Effect of increasing alanine concentration on
molar conversion rate
3.3 Alaninol degradation
3.3.1 Control experiment with alaninol

vi

ANNU'IOJOOA-i

15

15
15
16
16
16
16
17

18
18
21
21
21
22

25

27
28
30
34

34
34

3.4

Chapter 4.
4.1

4.2
4.3
4.4
4.5

Chapter 5.
5.1

5.2

Chapter 6.
6.1
6.2
6.3

6.4

6.5

Chapter 7.

3.3.2 Temperature effect on alaninol degradation
3.3.3 Effect of phosphoric acid on alaninol
degradation rate
Hydrogenation of B-alanine

Kinetics and mass transfer
Mass transfer calculations
4.1.1 Catalyst suspension
4.1.2 Mass transfer coefficients
4 1 3 Reaction rate and mass transfer
4.1.4 Gas-liquid mass transfer coefficient
4.1.5 Liquid-solid mass transfer coefficient
4 1 6 lntraparticle mass transport
4.1.7 Results from mass transfer calculations
4.1.8 Limiting reactant
Langmuir-Hinshelwood Model Development
Solution charge balance
Determination of kinetic parameters
Confidence intervals

Deuterium incorporation
Deuterium incorporation in amino acid hydrogenation-
mechanistic analysis
5.1.1 Methods
5.1.2 Results
Method for deuterium labeling
5.2.1 Background on amino acid labeling
5.2.2 Deuteration method
5.2.3 Deuteration results

Amino acid hydrogenation economic analysis
Purpose of economic study
Production rate
Alaninol production via hydrogenation
6.3.1 Hydrogenation process description
6.3.2 Material costs for hydrogenation
Alaninol production via borohydride reduction
6.4.1 Reduction process description
6.4.2 Material costs for reduction
Economics conclusion

Amino acid hydrogenation conclusions
and recommendations

Hydrogenation of alanine

Kinetic modeling

Deuterium incorporation

vii

36
36

37

38
38
38
39
4O
42
44
45
46
47
48
51
54
62

63
63

63
63
66
67
68
69

71
71
71
71
71
72
73
74
74
75

77
77

78
78

7.4 Recommendations

Chapter 8. PHA hydrogenation
8.1 Sustainable chemical processes
8.2 Background on feedstocks
8.2.1 PLA
8.2.2 PHB
8.3 Polylactic acid results
8.3.1 Initial results
8.3.2 Use of tetrahydrofuran as a hydrogenation solvent
8.3.3 PLA oligomer hydrogenation and hydrolysis
8.3.4 Temperature effect on PLA hydrogenation
8.3.4.1 Hydrogenation of PLA pellets at 130°C
8.3.4.2 Hydrogenation of PLA pellets at 150°C
8.3.4.3 Hydrogenation of PLA pellets at 175°C
8.3.5 Effect of PLA surface area on hydrogenation rate
8.3.6 Addition of surface wetting agent
8.3.7 Conclusion for PLA
8.4 Results for PHB hydrogenation
8.5 Recommendation for PHA hydrogenation studies

Appendix A. Alanine hydrogenation results

Appendix B. Acid-Base calculations

Appendix C. Excel program for kinetic modeling
C.1 Set up of spreadsheet
0.2 Description of calculation procedure
0.3 Visual basic code

References

viii

79

80
80
8O
80
81
82
83
87
89
91
91
92
93
95
95
97
98
99

100
122
124
124
126
127

149

Table 1.

Table 2.
Table 3.

Table 4.

Table 5.

Table 6.
Table 7.
Table 8.

Table 9.

LIST OF TABLES

World production, preferred production method, and
price of selected amino acids

Amino alcohols available at commercial scale

Measured pH and calculated fraction of protonated alanine
for data points in Figure 4. Reactions performed with 100
grams 0.22 M alanine feed solution under 1000 psi H2 with
1 g 5% Ru/C powder catalyst at 100°C.

Measured pH and calculated fraction of protonated alanine
for reactions with 0.15, 0.29, and 0.58 M phosphoric acid.
Reactions performed with 100 grams 0.22 M alanine solution,
under 1000 psi H2 with 1 g 5% Ru/C powder catalyst
Summary of mass transfer calculations. Reactions
performed with 0.22 M alanine in 100 grams 0.29 M
phosphoric acid solution 1 g Ru/C catalyst under 1000 psi
H2.

Conditions matrix used for parameter determination

Rate Constants for L—H Rate Expression

Summary of Incorporation studies with alanine and alaninol

Deuterium labeling summary

Table 10. Summary of economic evaluation

Table 11. Alanine hydrogenation: 100 grams solution, 0.22 M alanine,

Table 12.

Table 13.

0.29 M phosphoric acid, 100°C, 1000 psi H2, 1 g 5% Ru/C
catalyst

Alanine hydrogenation: 100 grams solution, 0.22 M alanine,
0.59 M phosphoric acid, 100°C, 1000 psi H2, 1 g 5% Ru/C
catalyst

Alanine hydrogenation: 100 grams solution, 0.22 M alanine,

1.20 M phosphoric acid, 100°C, 1000 psi H2, 1 g 5% Ru/C
catalyst

ix

22

27

47

55

57

70

76

102

103

104

Table 14.

Table 15.

Table 16.

Table 17.

Table 18.

Table 19.

Table 20.

Table 21.

Table 22.

Table 23.

Table 24.

Alanine hydrogenation: 100 grams solution, 0.46 M alanine,
0.59 M phosphoric acid, 100°C, 1000 psi H2, 1 g 5% Ru/C
catalyst

Alanine hydrogenation: 100 grams solution, 0.22 M alanine,
0.29 M phosphoric acid, 100°C, 2000 psi H2, 1 g 5% RulC
catalyst

Alanine hydrogenation: 100 grams solution, 0.22 M alanine,
0.29 M phosphoric acid, 100°C, 1500 psi H2, 1 g 5% Ru/C
catalyst

Alanine hydrogenation: 100 grams solution, 0.22 M alanine,
0.29 M phosphoric acid, 100°C, 500 psi H2, 1 g 5% Ru/C
catalyst

Alanine hydrogenation: 100 grams solution, 0.22 M alanine,
0.29 M phosphoric acid, 100°C, 250 psi H2, 1 g 5% Ru/C
catalyst

Alanine hydrogenation: 100 grams solution, 0.22 M alanine,
0.19 M phosphoric acid, 100°C, 1000 psi H2, 1 g 5% RulC
catalyst

Alanine hydrogenation: 100 grams solution, 0.22 M alanine,
0.15 M phosphoric acid, 100°C, 1000 psi H2, 1 g 5% Ru/C
catalyst

Alanine hydrogenation: 100 grams solution, 0.22 M alanine,
0.08 M phosphoric acid, 100°C, 1000 psi H2, 1 g 5% Ru/C
catalyst

Alanine hydrogenation: 100 grams solution, 0.22 M alanine,
0.29 M phosphoric acid, 125°C, 1000 psi H2, 1 g 5% Ru/C
catalyst

Alanine hydrogenation: 100 grams solution, 0.22 M alanine,
0.29 M phosphoric acid, 90°C, 1000 psi H2, 1 g 5% Ru/C
catalyst

Alanine hydrogenation: 100 grams solution, 0.22 M alanine,
0.29 M phosphoric acid, 110°C, 1000 psi H2, 1 g 5% RulC
catalyst

105

106

107

108

109

110

111

112

113

114

115

Table 25.

Table 26.

Table 27.

Table 28.

Table 29.

Table 30.

Table 31.

Alanine hydrogenation: 100 grams solution, 0.22 M alanine,
0.29 M phosphoric acid, 125°C, 1800 psi H2, 1 g 5% RuIC
catalyst

Alanine hydrogenation: 100 grams solution, 0.22 M alanine,
0.29 M phosphoric acid, 125°C, 500 psi H2, 1 g 5% Ru/C
catalyst

Alanine hydrogenation: 100 grams solution, 0.22 M alanine,
0.29 M phosphoric acid, 125°C, 250 psi H2, 1 g 5% Ru/C
catalyst

Alanine hydrogenation: 100 grams solution, 0.22 M alanine,
0.29 M phosphoric acid, 125°C, 100 psi H2, 1 g 5% Ru/C
catalyst

Alanine hydrogenation: 100 grams solution, 0.22 M alanine,
0.29 M phosphoric acid, 100°C, 100 psi H2, 1 g 5% RulC
catalyst

Alanine hydrogenation: 100 grams solution, 0.22 M alanine,
0.29 M phosphoric acid, 100°C, 100 psi H2, 1 g 5% RulC
catalyst

Comparison of Henderson-Hasselbach and charge
balance method for determining ionic solution conditions

xi

116

117

118

119

120

121

122

LIST OF FIGURES

Figure 1. Naturally occurring a-amino acids 4
Figure 2. Amino acids in solution 5
Figure 3. Amino acids in solution 19
Figure 4. Effect of a deficiency of phosphoric acid on alanine 23

conversion: (+) 0.29 M, (-X--) 0.15 M, (—i—) 0.07 M, and (-A--)
0.0 M phosphoric acid. Reaction performed with 100 grams of

0.22 M alanine solution, under hydrogen pressure of 1000 psi with

1 g 5% Ru/C powder catalyst in a 300 mL stirred reactor at 100°C.

Figure 5. Effect of excess phosphoric acid on alanine conversion: 26
(—><—) 0.29 M, (-+-) 0.58 M, and (+) 1.12 M phosphoric acid.
Reactions performed with 100 grams of 0.22 M alanine solution,

under hydrogen pressure of 1000 psi with 1 g 5% Ru/C powder catalyst

in a 300 mL stirred reactor at 100°C.

Figure 6. Dependence of conversion rate on acid type. Reactions 29
performed at 100°C with 100 grams solution of 0.22 M L-alanine, 1 g

of 5% Ru/C catalyst, 1000 psi hydrogen pressure. (0) 0.29 M Phosphoric
acid, (A) 0.29 M triflic acid, (I) 0.29 M sulfuric acid.

Figure 7. Dependence of alanine conversion rate on H2 pressure. 31
Reactions performed at 125°C with 100 grams solution of 0.22 M

L-alanine, 0.29 M phosphoric acid, and 1 g of 5% Ru/C catalyst.

Absolute bomb pressure: (—+—) 1800 psi, (—l—) 1000 psi, (--A-)

250 psi, (--x..) 100 psi.

Figure 8. Secondary racemization of L-alaninol. Upper graph: (+), 32
L-alanine concentration; (—i—), L-alaninol concentration, (-A-),

D-alanine. Lower graph shows enantiomeric excess of alaninol in

solution. Reaction performed with 100 grams of a 0.22 M L-alanine

solution, under H2 pressure of 1000 psi at 150°C with 1 g of 5% Ru/C
powder catalyst. Phosphoric acid in solution was 0.29 M.

xii

Figure 9. Formation of enatiomerically pure L-alaninol. Upper graph:
(+),L-alanine concentration; (—i—), L-alaninol concentration,
(-A-), D-alanine. Lower graph shows enantiomeric excess of alaninol
in solution. Reaction performed with 100 grams of a 0.22 M L-alanine
solution, under H2 pressure of 1000 psi at 100°C with 1 g of 5% Ru/C
powder catalyst. Phosphoric acid in solution was 0.29 M.

Figure 10. Effect of alanine concentration on molar conversion rate.
Reactions performed at 100°C with 100 grams L-alanine solution,

1 g of 5% Ru/C catalyst, 1000 psi H2 pressure. Feed concentrations:
(0) 0.22 M alanine and 0.29 M phosphoric acid, (I) 0.46 M alanine
and 0.59 M phosphoric acid, (A) 1.2 M alanine and 1.5 M phosphoric
acid.

Figure 11. Three phase mass transfer in catalytic system

Figure 12. Experimental versus predicted alanine conversion rate at
various temperatures. Reactions performed under 1000 psi H2, with
0.29 M phosphoric acid and 0.22 M alanine. (o)—Experimental Data,
125°C; (- - - -), Model Prediction, 125°C; (A)—Experimental Data,

1 10°C;( ----- )—Model Prediction, 1 10°C; (or-Experimental Data,
100°C; (- — —)-Mode| Prediction, 100°C; (I)—Experimenta| Data,
90°C; (— - — -), Model Prediction, 90°C.

Figure 13. Predicted versus experimental alanine conversion rate

with varying hydrogen pressure. Reactions performed at 125°C, with
0.29 M phosphoric acid and 0.22 M alanine. (o)—Experimental Data,
1800 psi, (- - - -)—Model Prediction, 1800 psi; (A)—Experimental Data,
1000 psi; ( ----- )—Mode| Prediction, 1000 psi; (o)—Experimental Data,
500 psi, (— — —)—Mode| Prediction, 500 psi; (I)—Experimenta| Data,
250 psi; (— - - -)—Model Prediction 250 psi.

Figure 14. Experimental versus predicted alanine conversion rate

with an excess of phosphoric acid. Reactions performed at 100°C,
under 1000 psi H2 and 0.22 M alanine. (o)—Experimental Data, 0.29 M
phosphoric acid; (- - - -)—Mode| Prediction, 0.29 M phosphoric acid;
(A)—Experimental Data, 0.6 M phosphoric acid; ( ----- )—

Model Prediction, 0.6 M phosphoric acid; (or-Experimental Data, 1.2 M
phosphoric acid; (— —- —)—Mode| Prediction, 1.2 M phosphoric acid.

xiii

33

35

40

58

59

60

Figure 15. Experimental versus predicted alanine conversion rate with
a deficiency of phosphoric acid. Reactions performed at 100°C,

under 1000 psi H2 and 0.22 M alanine. (o)—Experimental Data, 0.29 M
phosphoric acid; (- - - -)—Mode| Prediction, 0.29 M phosphoric acid;
(I)-—Experimental Data, 0.19 M phosphoric acid; ( ----- )—Model
Prediction, 0.19 M phosphoric acid; (A)—Experimental Data, 0.14 M
phosphoric acid; (— — —)—Model Prediction, 0.14 M phosphoric acid;
(o)—Experimental Data, 0.07 M phosphoric acid; (— - — -)-—Model
Prediction, 0.07 M phosphoric acid.

Figure 16. General incorporation scheme for amino acids
Figure 17. Alanine interaction with the catalyst in D2/D2O
Figure 18. Hydrogenation process diagram

Figure 19. Reduction process diagram

Figure 20. PLA hydrogenation versus hydrolysis. 100 grams 1 wt%
PLA at 130°C. Hydrogenation carried out under 1000 psi H2 with 1 g
5% Ru/C catalyst. Hydrolysis carried out under 1000 psi helium.
Hydrogenation results-—(A)-PG (M), (-)—lactic acid. Hydrolysis results
—-(o)-lactic acid (M).

Figure 21. Comparison in the rate of formation of propylene glycol
measured from PLA and predicted from lactic acid. Hydrogenation
carried out with 100 grams of 1 wt% solution with 1% Ru/C catalyst at
130°C, and 1000 psi H2. Hydrogenation results-—(-)-PG yield.
Predicted PG yield from lactic acid (— - - -).

Figure 22. Comparison in the rate of formation of propylene glycol
measured from high and low chain length PLA. Hydrogenations carried
out with 100 grams of 1 wt% PLA solution with 1% Ru/C catalyst at
130°C, and 1000 psi H2. PG formation from low molecular weight
PLA—(o)-PG(M). PG formation from high molecular weight PLA

—-(-)-PG(M).

Figure 23. Comparison in the rate of formation of propylene glycol
measured from oligomeric PLA and predicted from lactic acid.
Hydrogenation carried out with 200 grams of 2.5 wt% solution with
1% Ru/C catalyst at 130°C, and 1000 psi H2. Hydrogenation results
—(-)-PG(M). Predicted PG yield from lactic acid (- - - -).

xiv

61

65
66
72
74

85

86

88

90

Figure 24. Comparison in the rate of formation of propylene glycol
measured from PLA and predicted from lactic acid. Hydrogenation
carried out with 200 grams of 2.5 wt% solution with 1% Ru/C catalyst
at 150°C, and 1000 psi H2. Hydrogenation results—(I)-PG(M).
Predicted PG yield from lactic acid (- - - -).

Figure 25. Comparison in the rate of formation of propylene glycol
from PLA pellet and powderized PLA. Hydrogenation carried out with
200 grams of 2.5 wt% solution with 1 g 5% RulC catalyst at 150°C,
and 1000 psi H2. PG yield from PLA pellet—(-)-PG. PG yield from
PLA powder—(o)-PG.

Figure 26. Kinetic parameter Set Up page

XV

94

96

125

LIST OF SYMBOLS

A= deprotonated alanine (M)

a= mass transfer area per unit volume of reactor (m2/2m3 Z,
= ratio of catalyst surface area to volume of fluid (m 2/)m

A= zwitterionic alanine (M)

A p-rotonated alanine (M)

a3= gas-liquid interfacial area per unit volume of reactor (mz’lm3 ), (cmzlcm3 )

Al= alanine (M)

AL = alanine concentration in bulk solution (M)

Alol = alaninol (M)

Alol“ = protonated alaninol (M)

Am = Ammonia (M)

Am" = protonated ammonia (M)

A3 = concentraction of alanine at catalyst surface (M)

A7 = total alanine concentration (M)

memm. = experimental alanine concentration (M)

0”,“.an = predicted alanine concentration (M)

CCOOH = concentration of alanine protonated (M)

0,, = species surface concentration (moi/L)

CTOTAL= total concentration of alanine (M)

03: diffusion rate of species a (m 2ls)

De= effective diffusivity (m2/s), De- - 8 Da

d.= impeller diameter (m), (cm)

dp= particle diameter (m)

d1 = reactor diameter (m)

Ea= Ethylamine (M)

EA = protonated ethylamine2 (M)

g= -gravitation constant (m/s2 )2

g= -gravitational constant (m/s2 )

H= atomic hydrogen

H” = the hydronium cation (M)

H2 = molecular hydrogen (M)

H2 L = concentration of hydrogen in the liquid phase (M)

H2 3 = concentration of hydrogen at the catalyst surface (M)

H2 = concentration of hydrogen In the gas phase (M)

IPA = protonated isopropyl amine (M)

KM = first acidity constant of alanine

KA2 = second acidity constants of alanine

kL = liquid film mass transfer coefficient (m/s), (cm/s)

kLa = the gas-liquid mass transfer coefficient (1/s)

Kp1 = first acidity constant of phosphoric acid

Kp2 = second acidity constant of phosphoric acid

Kpa = third acidity constant of phosphoric acid

xvi

k3 = the liquid solid mass transfer coefficient (m/s)

L = characteristic length of the catalyst (m), (dp/6)

L = ratio of catalyst surface area per unit volume fluid (m2/m3)
N = stirring speed (rotations/s)

Nmm = minimum stirring speed (rotations/sec)

OH' = anionic hydronium ion (M)

P = H3PO4 (M)
P' = H2PO4' (M)
P2' = HPO42' (M)

P3' Is PO43‘ (M)

PT = the total phosphoric acid concentration

r = reaction rate (mol/L-s)

-RG = observed reaction rate of the species (kmol/m3-s)
S1 = a site occupied by alanine (-COOH form) or phosphoric acid (H3PO4 form)
(M)

S2 = a site occupied by hydrogen (M)

Sh = Sherwood number (dimensionless)

ST = fluid surface tension (N/m)

u9 = gas velocity (cm/s)

u9 = gas velocity (m/s)

VL = liquid volume (m ) .

VL = volume of the solution (cm3)

W' = catalyst loading (9 cat/100 9 feed)

w' = catalyst loading (kg/100 kg solution)

w‘ = catalyst loading (g/100 9 solution)

Greek

mp2 = observable modulus

B = impeller geometry constant

e = support porosity (0.6 for the catalyst used)
pL = fluid density (kg/m3), (glcma)

(IL = liquid viscosity (kg/m-s)

pp = catalyst density (kg/m3)

xvii

Chapter 1. Introduction
1.1 Background

Currently the primary feedstock for organic chemical production is
petroleum. Petroleum, while plentiful in 2003, is a finite resource, and as such it
is prudent to seek chemical production routes that use renewable resources as
feedstocks to replace traditional synthesis techniques as they become more
cosfly.

As the US. government and manufacturers have sought to increase
chemical production from renewables, considerable effort has been made to
improve fermentation technologies which use corn-derived glucose, and other
renewables, as feedstocks for organic chemical production. Several organic
acids including citric acid,"2 gluconic acid,3 and itaconic acid3 have been
traditionally produced via fermentation at the commercial scale. Today new
technologies are targeting propanoic acid,“'5'6 butyric acid,7'8 and fumaric acid9
for industrial scale production via fermentation as well. Two fermentation-derived
organic acids have received considerable attention for use as platforms to
produce multiple products. Lactic acid is a versatile feedstock that can be used

to make acetaldehyde, propylene glycol, acrylic acid, 2,3-pentanedione, and

 

'Grewal, H.; Kalra, K. Biotechnol. Adv. 1995, 13, 209.

:Pazouki, M.; Panda, T. Bioprocess Eng. 1998, 19, 435.

3Milson, P. E. Meers, J. L. Gluconic and ltaconic Acids. Comprehensive Biotechnology, Moo-
Young, M, Blanch, H. W., Drew, H. W. ,Wang, D. l. C., Eds.; Pergamon Press: New York, 1985;
Vol. 3, Chapter 35

:WBoyaval P.; Corre, C. Lari 1995 75, 453.

:Jin,Z.W.;Yang, S.T. Biofechnol.Prog.1998, 14,457.

:Rehberger, J. L.; Glatz, B. A. J. Food Prof. 1998, 61, 211.

:“Vandak D.; Zigova, J.; Sturdik, E.; Schlosser, 8. Process Biochem. 1997, 32, 245.

:Hwang, S.; Hansen, C. L. Biotechnol. Bioeng. 1997, 54, 451.

9,Du J. X. Cao, N. J.; Gong, C. S.; Tsao, G. T.; Yuan, N. J. Appl. Biochem. Biotechnol. 1997, 54,
451.

polylactic acid.1o Succinic acid has also received considerable interest and high
efficiency fermentations are under development.”12

Another class of organic acids that are part of the “renewables” category
are amino acids because they are primarily synthesized via fermentation with
glucose or sucrose as feed materials. Their multiple functional groups and
chirality make amino acids an attractive feedstock for chemical synthesis.

Hydrogenation of amino acids to amino alcohols is one route to utilize
these materials. Addition of small scale, high-value chemical products to a
biorefinery’s product slate, such as amino alcohols from amino acids, could
improve the overall economic viability of these plants.

In addition to economic incentives, the method for amino acid
hydrogenation described here meets several of “T he Twelve Green Chemistry
Principles” outlined by Dr. Paul Anastas13 and criteria outlined by the US.
Department of Energy in their Chemical Vision 2020 technical roadmap seriesz“
high atom economy, benign co-product (water), environmentally friendly solvent
(water), high selectivities, and use of heterogeneous catalysts.

The focus of this study was to examine the stereoretentive hydrogenation
of amino acids to amino alcohols, with focusing on L-alanine ((S)-2-
aminopropanoic acid) to L-alaninol ((S)-2-aminopropanol). In this work we

determined the best catalyst and operating conditions for amino acid

 

‘° Datta, H.; Tsai, s. P. ACS. Symp. Ser. #666, 1997, 224.

‘1 Gorkam, R. FL; Eitemann, M. A., Sridhar, J. ACS Symposium Ser. #6661997, 237.

‘2 Nghiem, N. P.; Davison, B. H.; Richardson, G. R. Appl. Biochem. Biotechnol. 1997, 63-65, 565.
‘3 Anastas, P. T.; Warner, J. C. Green chemistry: theory and practice; Oxford University Press:
New York, NY, 1998.

1‘ httpvlwwwchemicalvision2020.org/techroadmaps.html

2

hydrogenation and derived a mechanism-based, kinetic model to describe
reaction trajectories over a wide range of conditions. Deuterium exchange
experiments were used to gain understanding of reactant and product interaction
with the catalyst. Work in this area has also prompted the development of a
simple method for deuterium labeling of amine bearing compounds.

In addition to the focus on amino acid hydrogenation, preliminary studies
were devoted to polylactic acid (PLA) and poly-3-hydroxybutanoic acid (PHB)
hydrogenation. One of the major advantages of biobased polymers is
biodegradability. The preferred method of disposal is composting, where PLA
polymer hydrolyzes to lactic acid and is then consumed by bacteria which
release carbon dioxide. Alternatively, PLA could be converted to propylene
glycol (PG) through hydrogenation. Hydrogenation could also be used in
manufacturing facilities as an alternative to disposal of batches that do not meet
product specifications.

1.2 Literature Review
1.2.1 Amino acid background

Amino acids are carboxylic acids which bear an amine group. In nature
amino acids are most commonly found as 2-amino acids, or or-amino acids. The
general formula of amino acids is RCH(NH2)COOH, where R is an alkyl or aryl
side chain, and may contain hydroxyl, amino, mercapto, sulfide, and carboxy
groups, as shown in Figure 1. Amino acids contain at least two functional groups
and 19 of the 20 naturally occurring a—amino acids are chiral compounds, which

make them attractive synthetic building blocks.

Figure 1. Naturally occurring oI-amino acids

0
10H

NH2

Alanine

O
OH
“OM
NH2 0

Aspartic acid

NHz
H2NM0H
O O

Glutamine

OH
0
NH2

Isoleucine

O
HOJK/VS\

I‘~IH2

Methionine

Tyrosine

Arginine
O

OH
HzN' ' "
SH

Cysteine

o
”wk/[km

Glycine

O

H2N OH
Phenylalanine
OH
""NH2
HO
O
Threonine
""NH2

HO
O

Valine

O

HOMNHZ

H20

2

Asparagine
NHz

HOMOH

O O
Glutamic acid

O

N
HN NH2

Histidine

Tryptophan

Amino acids are unique chemically in that they are both acidic and basic
because of their carboxylic acid and amine functionalities. In aqueous solution,
the structure is determined by pH: under acidic conditions amino acids exist
primarily as the cationic ammonium carboxylic acid 1, under neutral conditions as
the zwitterionic form 2, and under basic conditions as the deprotonated 2-

aminocarboxylate ion 3, as shown in Figure 2.

afgflH‘T‘T R}O¥TR‘O?_O-

NH3” NH3+ NH2
Predominates Predominates Predominates
pH<2 pH~5-9 pH>10
1 2 3

Figure 2: Amino acids in solution

Solution behavior, as discussed in Section 3.1, is a very important variable in
amino acid hydrogenation.
1.2.2 Amino acid production

Amino acids are produced through four different processes: fermentation,
extraction, chemical synthesis, and enzymatic routes. The most economical
means of producing bulk amino acids is via fermentation of sucrose or glucose
via overproducing microbial strains.

In 1999, the estimated worldwide production of amino acids was 1.6 X 106
tons/year. Amino acids are used primarily for human and animal nutrition with

only 0.6% of the annual production utilized for pharmaceuticals, cosmetics,

agrochemicals, and industrial uses. Table 1 summarizes the production rate,

preferred synthesis method, and price of several amino acids.15

In this study, L-alanine was the feed material in most reactions.

lndustrially, L-alanine is produced from L-aspartic acid in a pressurized

bioreactor by immobilized Pseudomonas dacunhae cells.16

reports of L-alanine production via fermentation with Arthrobacter oxydans from

However, there are

glucose with a yield of 52% and 95% ee.17 Alanine is also still produced by

isolation from protein hydrolysates.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Amino Acid World Production Preferred Production Method Price, $lkg
in 1999, tons/yr
L-Alanine 1 .200 Enzymatic catalysis, fermentation 12-29
L-Arginine 1,500 Fermentation, extraction 44-51
L-Asparagine 150 Extraction
L-Aspartic acid 12,000 10-12
L-Cysteine 3,500 Reduction of L-cystine (extraction), 59-88
enzymatic synthesis
L-Glutamic acid 700,000 Fermentation 3
L-Glutamine 1,000 Fermentation, extraction 44-52
Glycine 15,000 Chemical synthsis 7-13
L-Histidine 300 Fermentation, extraction 1 10-147
L-Hydroxyproline 100 Fermentation, extraction
L-Isoleucine 550 Fermentation, extraction 294-331
L-Leucine 800 Fermentation, extraction 74-147
L-Lysine 430,000 fermentation 5-6
D,L-Methionine 450,000 Chemical synthesis and enzymatic 5-6
resolution
L—Methionine 400 Enmnatic resolution 74-88
L-Phenylalanine 1 1 ,000 Fermentation 59-88
L-Proline 800 Fermentation, extraction 147-294
L—Serine 300 Fermentation, extraction 74-132
L-Threonine 30,000 Fermentation 125-147
L-Tryptophan 1,200 Fermentation 74-147
L-Tyrosine 1 50 Extraction 74-1 10
L-Valine 1,100 Enzymatic catalysis, fermentation 110-147

 

Table 1. World production, preferred production method, and price of selected amino acids

 

'5 Amino Acids. Ulmann’ s Encylopedia of Industrial Chemistry, Karlheinz, D.; Berd, H.; Kleeman,

A.; Krimmer, H.; Leuchtenberger, W, Weckbecker, C., 7‘". ed.; Wiley. New York, 2002.
”Furui, M. Yamashita, K. J. Ferment. Technol. 1983, 61, 587.
' Hashimoto, S.; Katsumata, R. Proc. Ann. Meet. (Agric. Chem. Soc. Jpn.) 1994, 341.

 

1.2.3 Amino alcohols
Amino alcohols are important intermediates in pharmaceutical and peptide

iii-192°” as insecticidal intermediates,22 and chiral auxiliaries.23

chemistry,
Because of the growing emphasis on chiral molecules in these applications,24 the
worldwide market for fine chemicals and pharmaceuticals sold as single
enantiomers in 2002 was $7 billion and $159 billion, respectively.25 Economical
formation of enantiomerically pure amino alcohols may provide a route to lower
raw material costs for many therapeutics.
1.2.3.1 Amino alcohol synthesis techniques

The earliest report of amino alcohol production was from Karrer in 1921
via the reduction of amino acid esters with sodium in ethanol.26 Since then,
many such reductive techniques have been performed using both amino acids
and amino esters as starting materials. Reducing agents include sodium

borohydride (NaBH4)-sulfuric acid,27 NaBH4-iodine,28 lithium aluminum hydride

(LiAlH4),29'3°'3"32'33'34 borane-methyl sulfide in the presence of boron trifluoride

 

TenBrInk“ R. E. J. Org. Chem. 1987, 52,418.
19Nicollaides, E. D.; Tinney, F. J.; Kaltenbronn, J. S; Repine, J. T.; DeJohn, D. A. Lunney, E..A;
Roark, W. H., Marriot, J. G.; Davis, R. E.; Voigtmen, R. E. J. Med. Chem. 1986, 29, 959.
2°Fincham, C. l.; Higginbottom, M; Hill, D. R.; HonIvell, D. C.; O'Toole, J. C.; Ratcliffe, G. 8.;
Rees, D. C.; Roberts, E. J. Med. Chem. 1992, 35, 1472.
:Auvin-Guette, C; Rebuffat, S.; Prigent, Y; Bodo, B. J. Am. Chem. Soc. 1992, 114, 2170.
Z,Wu S.; Takeya, R.; Eto, M; Tomizawa, C. J. Pestic. Sci. 1987, 12, 221.
23Coppola, G. M.; Schuster, H. F. Asymmetric Synthesis of Chiral Molecules using Amino Acids;
Wiley-lnterscience: New York, 1987.
22;.Ager D. J. Prakash, l; Schaad, D. R. Chem. Rev. 1996, 96, 835.
25Rouhi, A. M. Chemical and Engineering News 2003, 81 (18), 45-55.
as 27Karr,er P.; Karrer, W.; Thomann, H, Harlacher, E. Mader, W. Helv. Chim. Acta. 1921, 4, 76.
:Abiko, A.; Masamune, S. Tetrahedron Lett. 1992, 33, 5517.
:McKennon, M. J.; Meyers, A. l. J. Org. Chem. 1993, 58, 3568.
WMeyers A. l.; Dickman, D. A.; Bailey,T. R. J. Am. Chem. Soc. 1985, 107, 7974.
:°Dickman, D. A.; Meyers, A. l.; Smith, G. A.; Gawley, R. E. Org. Synth. 1990, Col! Vol. 7, 530.
“,Evans D. A.; Takacs, J. M.; McGee, L. R.; Ennis, M. D.; Mathre, D. J.; Bartroli, J. Pure Appl.
Chem. 1981, 53, 1109.
32 Evans, D. A.; Bartroli, J.; Shih, T. L. J. Am. Chem. Soc. 1981, 103, 2127.

etherate,3‘5""3'37'3‘"39

and lithium borohydride in the presence of trimethylsilyl
chloride.“’o Amino ester hydrochlorides have also been reduced to the
corresponding amino alcohols with LiAlH4“"2 or NaBH...43

The primary advantage of these reductive techniques is that they are, for
the most part, stereo-retentive. However, they suffer from one or a combination
of the following: high cost, flammable and toxic reagents, laborious product
separation procedures, and large waste streams. Thus, production on the multi-
kilogram scale of amino alcohols is costly. Despite this, there are many amino
alcohols produced at commercial scale listed in the Handbook of Chiral
Chemicals,“ as shown in Table 2. These compounds may be produced using a
technique related to those listed above (or bio-catalytically).

Another route that has been examined is catalytic hydrogenation of amino
esters. In the 1940’s, Adkins and co-workers produced amino alcohols from
amino acid esters over Raney nickel and reported optical rotation close to that of
pure samples; therefore enantiomeric excess was near 100%.“5 In addition to

these studies performed more than half a century ago, in 2001 Studer, et. al,46

studied hydrogenation of amino acid esters as well. In their study, Nishimura

 

:Evans, D. A.; Ennis, M. D.; Mathre, D. J. J. Am. Chem. Soc. 1982, 104, 1737.
:‘bEvans D. A.; Nelson, J. V.; Taber, T. R. Top. Stereochem. 1982, 13, 1.
3“,Lane C. F, Myatt, H. L.; Daniels, J.; Hopps, H. B. J. Org. Chem. 1974, 40, 3527.
3" 37Poindexter, G. S, Meyers, A. l. Tetrahedron Lett. 1977,3527.
33;,Smith G. A.; Gawley, R. E. Org. Synth. 1985, 63, 136.
:Brown, H. C.; Choi, Y. M.; Narasimhan, S. J. Org. Chem. 1982, 47, 3153.
:HGage J. R.; Evans, D. A. Org. Synth. 1990, 68, 77.
:Nicolas, 5.; Russell, K. C.; Hruby, V. J. J. Org. Chem. 1993, 58, 766.
1;,Karrer P.; Portmann, P.; Suter, M. Helv. Chim Acta1949,32, 1156.
:Karrer, P.; Naik, A. R. Helv. Chim.Acta1948,31, 1617.
:Seki, H, Koga, K.; Matsuo, H., Yamada, S. Chem. Pharrn. Bull, 1965, 13, 995
:Ager, 0., ed., Handbook of Chiral Chemicals, Marcel Dekker, Inc., New York, N. Y. ..1999
:Adkins, H.; Pavlic, A. A. J. Am. Chem. Soc. 1947, 69, 3040.
“S,tuder M., Burkhardt, S., Blaser, H. U. Adv. Synth. Catal. 2001, 343(8), 802.

28m ESoEEoo “a 23.55 30:85 oc_E< .N 228.

 

wwmamco\E_xom

65:821.;

 

mmwsmmggxem €59ch .LcmEEmom

65652.9

 

mnmcmh .comumfigm .mmmzmoo\E_xmm 5062852. owz .tonzoz ..mmctccom

_e:_ce_a_>cecd-$.@

 

 

mESED .mwmauoD\E_xom 66620208. owz .toazez .LcmEEmom

_ec_ca_e_>ee;d.$-Ev

 

 

 

mmmsch§x6m _0c_o=m._-t9-._
89:58 9853.31; . acexec-wesesoauiév

 

£86596 €8st .LomEEoom

3552-5 .8829 3553-329

 

«.8262 .meEEwom

29:82-9 aceaea. weEEeeiAE

 

 

mwmsmmggxmm .meEEmom

2888.73-33 .aEEea- F._>£eE-e-ec_E<-N-A+V.AmIV.

 

mmwsmoQEimm

2868.73-53 . .8989F..seeeeesE<-N-A+V-EV

 

 

SwameExot .LemEEcom

2863287324: . sesame-F-_EaE-m.ec_E<-m-E-Ath

 

mwmsmoD\E_xmm .toazoz comEEoom

A_ec__a>.uv ..eceaa. :26E-m-ec_Ee-N.A-v-Amv

 

mmmzchExmm €8ch ..emEEcom

A_ec__e>-e .8523-wseeeaeseywmvérv

 

 

2:85. .ocmSngEméging
2E9: _eaesaec_Ea-w-A-v-Ev

 

 

Lc__aasm

 

ccaoquo

 

 

catalyst (mixed Rh/Pt oxide) was loaded to 10 wt% at room temperature with 10
MPa hydrogen and amino acid ester. Conversions and yields in excess of 90%
were achieved without racemization of the amino alcohol product. In all of these
studies, amino acid esters were used as a starting material, which requires an
esterification step prior to hydrogenation.

Recently, two patents by Antons‘w'8 reported production of optically active
amino alcohols via aqueous phase hydrogenation of the free amino acid in
aqueous solution over supported ruthenium catalysts. However, these patents
do not discuss either mechanism or kinetics.

1.2.4 Organic acid hydrogenation

Hydrogenation of organic acids and esters over metal catalysts has been
the subject of numerous studies over the last several decades. In the 1930’s
Adkins studied the hydrogenation of esters and ketones to alcohols over copper-
chromium oxide49 and nickel catalysts”5 under 2200-2900 psi H2 at 250°C. In the
1950’s, Camahan and Ford50 used ruthenium catalysts to hydrogenate organic
acids such as acetic acid, oxalic, adipic, succinic, and hydroxyacetic acid in a
temperature range of 94 to 192°C under pressures in excess of 7000 psi H2 with

yields ranging from 47 to 88%. Later, Broadbent51

studied hydrogenations of
organic acids catalyzed by rhenium that was prepared in situ in a temperature

range of 137-286°C and pressures ranging from 2200 to 4800 psi H2. In their

 

‘7 Antons, s.; Beitzke, B. 0.3. Patent 5,536,879, 1996.

4" Antons, S., Tilling, A., Wolters, E. US. Patent 6,310,254, 2001.

‘9 Bowden, E., Adkins, H. J. Am. Chem. Soc. 1934, 56, 689.

5° Carhnahan, J. E.; Ford, T. A.; Gresham, W. F.; Grigsby, W. E.; Hager, G. F. J. Am. Chem. Soc.
1955, 77, 3766.

5' Broadbent, H. 5.; Campbell, G. C.; Bartley, W. J.; Johnson, J. H. J. Am. Chem. Soc. 1959, 24,
1847.

10

study, formic, acetic, propionic, butyric, capric, lauric, stearic, lactic, maleic,
succinic, and glutan’c acid were all reduced to their corresponding alcohol.
In this research group, lactic acid has been reduced to propylene glycol

5253 under mild conditions with ruthenium on carbon as the

(1 ,2-propanediol)
preferred catalyst. Reactions were performed in batch and trickle bed reactors at
temperatures ranging from 100 to 170°C and pressures ranging from 1000 to
2000 psi. At 95% conversion of lactic acid, selectivities in excess of 90% were
reported.

Antons"54 also studied hydrogenation of lactic and malic acid over
ruthenium catalysts and reported that the addition of a second metal, preferably
rhenium, enhances reaction rates. At temperatures of 60-70°C and 1450 psi H2,
they produced alcohols with enantiomeric excess greater than 97%.

Recently, Dumesic55 has performed vapor phase lactic acid hydrogenation

over silica-supported copper. At 473 K and a hydrogen partial pressure of 100

psi H2, they achieved propylene glycol yields of 88% with high selectivities.

 

52zmamg,z. G. “Jackson J. E., Miller, D. J. Appl. Catal. A- Gen. 2001, 219 89.
53Zhang,Z. G, Ph. D. Dissertation, Michigan State University, 2000.
ZAntons, S.; TIlling, A. S.; Wolters, E. US Patent6, 355, 848, 2002.
55Cortright, R. D.; Sanchez-Castillo, M, DumesicJ. A. Applied Cate/B. Env. 2002, 39, 353.

11

Chapter 2. Experimental Methods
2.1 Amino acid hydrogenation studies
2.1.1 Materials

Feeds (L(+)-a|anine, glycine, L(+)-serine, L(+)-phenylalanine) were
purchased from Aldrich Chemical Co. in enantiomerically pure forms with optical
purities exceeding 99 ee%. The acid used in the reaction was ACS grade 85%
phosphoric acid (J. T. Baker). Water used was of HPLC grade purity (J.T. Baker,
USA). Ultra high purity hydrogen (AGA Gas, 99.999%) was used for all runs.
For deuterium labeling experiments Dz gas (AGA Gas, 99.5%) and deuterium
oxide (Sigma Aldrich, 99.9%) were used in place of hydrogen and water.

The catalyst used was a 5% ruthenium on carbon powder obtained from
PMC, Inc. (Severville, TN). The particles had a mean diameter of 150 microns,
BET surface area of 860 m2/kg, porosity of 0.6, and bulk density of 800 kg/m3.
2.1.2 Batch reactor system

Reactions were carried out in a 300 mL batch reactor (Parr Instrument Co,
USA) equipped with a gas entraining impeller and a temperature controller that
maintained temperature to :l:1°C. The reactor was constructed of 316 stainless
steel and rated to a maximum pressure of 3000 psi at 350°C. The reactor was
equipped with a liquid sampling system and a gas sampling vent so that analyses
could be performed during the course of an experiment. A Teflon liner was
employed in the reactor to minimize corrosion from the acidic reaction media.

Reactions were typically conducted using the following procedure.

Catalyst (usually 1 g dry basis per 100 9 feed solution) was placed in the reactor.

12

The vessel was purged with nitrogen to flush out oxygen, heated to 150°C, and
charged with hydrogen to 250 psi. The catalyst was reduced for 1 hr at these
conditions with the gas valve slightly open to regulate pressure and allow
impurities to escape. The reactor was then cooled to room temperature and the
hydrogen vented.

The reactor vessel was next charged with 100 grams of a 2 wt% amino
acid solution containing inorganic acid and again purged with hydrogen. Heating
and stirring (1200 rpm) were then initiated. The reaction began when the reactor
reached the desired temperature, and was pressurized with hydrogen.

Reactions were performed at temperatures from 80 to 150°C and at pressures
from 100 to 2000 psi H2. Throughout the reaction liquid samples were taken and
analyzed. Reaction times ranged from four to 16 hours with 6 hours being the
preferred time. At the conclusion of the reaction the vessel was cooled and the
pressure reduced.

2.1.3 Amino acid analysis

2.1.3.1 HPLC analysis

Liquid products were analyzed using high pressure liquid chromatography
(HPLC). The system consisted of a computer-controlled pump Meters 600), a
UV detector (Waters 610), and differential refractometer (Waters 490). Waters
600 Maxima software was used to monitor both detectors. The separation was
carried out on a Waters NovaPak 013 (3.9 X 150 mm) column operating at a flow
rate of 1 mL/min at 40°C, with a 70:30 mobile phase of 0.2 M acetate buffer (pH

5) solution and methanol.

13

To separate enantiomers of amino acids and amino alcohols, the following
derivatization procedure was modified from an existing method:58 1 mL of 1 wt%
TAGIT (2,3,4,6-Tetra-O-acetyl-B—D-glucopyranosyl isothiocyanate) in acetonitrile
was added to 1 mL of 1% (v/v) triethylamine in water solution. To the above, 40
pL of reaction sample were added, shaken vigorously, and allowed to derivatize
for 1 hour. Internal standard, 40 ILL 3 wt% citraconic acid in water, was then
added and a 5 ML aliquot injected into the HPLC. Components were detected at
a UV wavelength of 250 nm. This method separated both the enantiomers of
alanine and alaninol and gave distinct peaks for ammonia, ethylamine, and
isopropylamine.
2.1.3.2 Use of 1H NMR for analysis of amino acids and amino alcohols

Proton NMR (‘H) was used for experiments performed in 020 under
deuterium or hydrogen pressure. Catalyst was removed from samples using a
cellulose-based syringe filter. One mL of sample was added to a round bottom
flask and dried by rotary evaporation to remove water. The sample was
dissolved in 0.7 mL of deuterated water, which contained 1% acetic acid as an
internal standard, before transfer to an NMR tube. NMR spectra were recorded
on a 300 MHz Varian spectrometer. Extent of deuteration was determined by
comparison of the integration value of the internal standard peak to relevant
peaks of the product and reactant coupled with species concentration data from

HPLC.

 

5" Gal, J. J. Liquid Chromatography 1986 9, (2.3), 673.

14

2.1.4 Amino acid hydrogenation conditions matrix for kinetic parameter
determination

Hydrogenation of amino acids was performed in a Parr 300 mL mini-
reactor. Amino acids used as starting materials include glycine, alanine, serine,
proline, and phenylalanine. However, only alanine was studied in detail and
therefore, the kinetic analysis was limited to it. Reactions were performed within
the following variable ranges:

1. Feed concentration: 0.22 to 1.10 M

2. Temperature: 80 to 150°C

3. Pressure: 0 to 2000 psi H2

4. Phosphoric acid concentration: 0 to 1.16 M

Appendix A contains details of selected reactions (concentrations,
conversion, etc. versus time data) used for the kinetic analysis in this study.
2.2 Experimental methods for PLA and PHB hydrogenation
2.2.1 Materials

Poly(3-hydroxybutyric acid) PHB was obtained from Sigma Aldrich with a
weight-average molecular weight of ~437,000. Samples of PLA were obtained
from Polysciences, Inc. with weight-average molecular weights of ~300,000 and
~16,000. Poly-L-Iactic acid was supplied by the Composites Center at Michigan
State University with a number average molecular weight of ~140,000. Lactic
acid, 85% in aqueous solution, was obtained from J. T. Baker. Organic solvents

methanol and THF were obtained from J.T Baker and Aldrich, respectively.

15

Hydrogen, Ru/C catalyst, and water were obtained from the same sources as
described in Section 2.1.1.
2.2.2 Batch reactor system

Hydrogenation of PLA and PHB were carried out in the same system as
described in Section 2.1.2 using the same catalyst preparation. Feed materials
were introduced to the reactor directly by opening the reactor head, instead of
transferring the feed from a pressurized vessel. Before experiments were
initiated, the reactor was purged with helium or hydrogen if a hydrolysis or
hydrogenation experiment were run, respectively.
2.2.3 Analysis for PLA and PHB reaction products
2.2.3.1 HPLC analysis of monomeric materials

Analysis of lactic acid, 3-hydroxybutanoic acid, propylene glycol, and 1,3-
butanediol were performed by HPLC using a HPX-87H Aminex carbohydrates
column (Biorad) at 50°C with 5 mM sulfuric acid flowing at 0.45 mL/min.
Samples were diluted to 5% by volume in water and a 10 pL aliquot was injected.
Samples were detected using a RI detector.
2.2.3.2 HPLC analysis of oligomeric materials

Acetylated oligomers of lactic acid were detected using High Pressure
Liquid Chromatography. The system consisted of a computer-controlled pump
(Waters), a UV detector (Waters 610), and differential refractometer (Waters
490). Waters 600 Maxima software was used to monitor both detectors. A
Waters NovaPak C18 (3.9 X 150 mm) column operating at a flowrate of 1 mL/min

at 40°C was used for the separation. The mobile phase was delivered at a flow

16

rate of 1 mL/min and operated along a 40 minute gradient from 95:5 (v/v)
waterzacetonitrile with 0.1% volume trifluoroacetic acid to 95:5 (v/v)
acetonitrilezwater with 0.1% volume trifluoroacetic acid. This method separated
oligomers of lactic acid up to degree of polymerization 14.

The samples of lactic acid oligomers were derivatized according to the
following procedure. The sample (5 g) was dried by rotary evaporation to
remove water or solvent. To this, 5 mL of acetic anhydride was added and
heated at 80°C for 4 hours. The contents were dried using a rotary evaporator to
remove excess acetic acid, water, and acetic anhydride. To the dried sample
one ml of acetonitrile and 1 mL of water was added and a 20 ul aliquot was
injected.
2.2.3.3 NMR analysis of oligomeric lactic acid

In principle, proton NMR can be used to estimate the average degree of
polymerization of PLA in reaction samples by comparison to the methyl peak in
the acetyl group with the methyl groups in PLA. To prepare samples for this, 1
mL of the sample left over from HPLC analysis, Section 2.2.3.2, was dried to
remove acetonitrile and water. Deuterated chloroform was added to the solution
to dilute the solution to 1 wt% and transferred to an NMR tube. NMR spectra
were recorded on a 300 MHz Varian spectrometer. This method did not prove to

be reliable and was not used extensively for analysis purposes.

17

Chapter 3. Amino acid hydrogenation results
3.1 Exploratory studies

To determine the feasibility of amino acid hydrogenation, several sets of
reaction conditions were investigated with a variety of amino acids and catalysts
at different reaction conditions. The results presented in this section are meant
to provide a qualitative description of observations from these studies. These
reactions were performed before the development of an analytical method by
HPLC, and thus 1H-NMR spectra were used to identify product and reactant.

Glycine, as the simplest amino acid, was the first material examined to
explore the feasibility of aqueous phase hydrogenation. Initial studies were
carried out using nickel, palladium, and ruthenium catalysts. Ruthenium was the
only catalyst to yield glycinol (ethanolamine) in the temperature range of 100 to
150°C under 1000 psi H2.

A ramping experiment with glycine was carried out under hydrogenation
conditions from 100 to 210°C under 1000 psi H2 with 100 grams of a 2 wt%
glycine solution and 1 g 5% Ru/C catalyst. A glycinol yield of less than 10% was
observed at temperatures less than 150°C. Samples collected after reaction
temperatures exceeded 150°C were pink and not suitable for NMR analysis.

Comparatively, lactic acid hydrogenates to propylene glycol rapidly under
similar experimental conditions. Although they are similar compounds, there is at
least one important difference that sets amino acids apart from a—hydroxy

carboxylic acids with respect to aqueous phase hydrogenation.

18

As discussed in Section 1.2.1, and shown in Figure 3, the ionic state of
amino acids is dependent on pH. From prior work with lactic acid and its salts,53
it is known that for hydrogenation to occur, the carboxylic acid functionality may
not be deprotonated form [2, 3]. In dilute solutions, amino acids exist in the
zwitterionic form 2 where the carboxylic moiety is deprotonated and
hydrogenation is not thermodynamically favorable. Amino acids, because of the
amino group at the alpha position, have a lower pKa (1.8-2.6) than do normal

alkanoic acids (pKaz4.8) and lactic acid (pKa=3.8).

O O O
H-A -_h.. -
R " R ‘ R

NH3+ NH3" NH2
Predominates Predominates Predominates
pH<2 pH~5-9 pH>10
1 2 3

Figure 3. Amino acids in solution

To ensure that a high fraction of amino acid in solution was in the
protonated state 1 (Figure 3) acid was added to the feed until an initial pH of ~1-2
was reached. To verify that the protonated form was present, the measured pH
was entered into a rearranged form of the Hendersen-Hasselbach equation to

estimate the fraction of amino acid that was protonated.

 

— H
base] CCOOH 10 p
H: K +l [ —> = 1
p p a 09[lacicIlJ ctota. 10-PKa +10-pH ( )

An experiment was carried out at 150°C with 100 grams of 2 wt%
aqueous glycine under 1000 psi H2 with 1 g 5% Ru/C. To acidify the feed

solution, 0.3 M HCI was added to the feed mixture, which brought the solution pH

19

to ~1. At this temperature, a yield of roughly 40% was observed after 2 hours;
this was an improvement over the previous case.

Phenylalanine, proline, alanine, and serine were also included in the initial
investigation for hydrogenation feasibility. They were reacted under
hydrogenation conditions of 150°C in 100 grams of 2 wt% amino acid solutions
under 1000 psi H2 with 1 g 5% Ru/C catalyst and 0.3 M HCI added. In each
case, some product alcohol was formed, 20-40% as indicated by NMR analysis.
In the case of phenylalanine, the ring was rapidly hydrogenated also to give
cyclohexylalanine and cyclohexylalaninol.

The use of hydrochloric acid as an acidulant was not an appropriate
choice for this reaction system because the reactor was constructed from
stainless steel which was corroded by HCI. Frequently samples were colored
pink, indicating metal leaching. The resulting dissolved metal ions made the
NMR sample spectra collected very noisy.

The completion of these studies coincided with two very important
developments in this work. First, the HPLC method described in Section 2.1.3.1
was developed and, secondly, phosphoric acid became the acidulant of choice.

The studies that pertain to amino acid kinetics were performed exclusively
with alanine as the reactant and all cases, unless othenIvise specified, contained

phosphoric acid.

20

3.2 Experimental results with alanine
3.2.1 Control experiments with alanine

To determine the stability of alanine in solution, two experiments were
performed. In the first experiment, 100 grams of 0.22 M alanine and 0.29 M
phosphoric acid were fed to reactor with no catalyst present under 1000 psi H2.
The reactor operated sequentially at 100, 125, and 150°C for 2 hours at each
temperature. Throughout this experiment a 1% yield of alaninol ((S)-2-
aminopropanol) was observed and 5% yield of ammonia was observed after the
150°C interval. The presence of alaninol was probably due to a small amount of
catalyst present in the reactor. In the second experiment, 100 grams of 0.22 M
alanine and 0.29 M phosphoric acid were fed to reactor with 1 g of 5% Ru/C
catalyst present under 1000 psi He. The reactor operated sequentially at 100,
125, and 150°C for 2 hours at each temperature. In this case, no side products
were observed until after the 150°C interval. In this case, a 2% yield of ammonia
and 5% yield D-alanine was observed. Amino acids are well known to racemize,
particularly at elevated temperatures, a property that can be used for determining
the relative or absolute age of biological materials.57
3.2.2 Effect of acid addition

The amount of phosphoric acid relative to alanine in the reaction mixture
determines the maximum achievable yield. This ratio is important because the
product formed is basic which causes the solution pH to increase as alanine is
converted. With sub-stochiometric amounts of phosphoric acid added to the

feed, alanine conversion ceased when the molar concentration of alaninol

 

5’ Bada, J. L. Annu. Fiev. Earth Planet Sci. 1995, 13, 241.

21

approximately equaled the molar concentration of phosphoric acid added.
Therefore, the effect of acid fed to the reactor was studied extensively, both with
sub-stoichiometric and excess phosphoric acid relative to alanine.
3.2.2.1 Sub-stoichiometic phosphoric acid addition

In the studies presented in this section, 100 grams of a 0.22 M alanine
solution was hydrogenated with varying phosphoric acid concentrations as
specified in the text. In all cases 1000 psi H2, 100°C, and 1 g of 5% Ru/C

catalyst were used.

 

o M 0.07M 0.15 M 0.29 M

W") PH X CCOOH/Ct PH X CCOOH/Ct PH X CCOOH/Ct PH X CCOOH/Ct
6.9 0.00 0.00 2.8 0.00 0.26 2.4 0.00 0.46 2.0 0.00 0.69
7.0 0.12 0.00 3.2 0.28 0.13 2.5 0.14 0.40 2.1 0.31 0.64
7.9 0.12 0.00 3.8 0.36 0.04 2.5 0.21 0.39 2.1 0.40 0.62
8.3 0.11 0.00 5.5 0.40 0.00 2.7 0.39 0.28 2.2 0.58 0.59
8.5 0.10 0.00 6.1 0.45 0.00 3.0 0.50 0.18 2.3 0.72 0.53
8.6 0.14 0.00 6.1 0.45 0.00 3.6 0.62 0.05 2.4 0.85 0.48
8.7 0.13 0.00 6.2 0.46 0.00 4.8 0.69 0.00 2.5 0.93 0.43
Table 3. Measured pH and calculated fraction of protonated alanine for data points in Figure 4.
Reactions performed with 100 grams 0.22 M alanine feed solution, under 1000 psi H2 with 1 g 5%
Ru/C powder catalyst at 100°C.

 

 

 

 

 

 

 

001-th40

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

As shown in Table 3 and Figure 4, when a moderate excess of phosphoric
acid (0.29 M) was present, the reaction proceeded unconstrained to near
completion in 6 hours. The fraction of alanine protonated estimated by the
Henderson-Hasselbach equation shows that during the entire course of the run
there was in excess of 40% of the total alanine in solution protonated. In
comparison, for cases where a sub-stoichiometric amount of phosphoric acid
was fed reactions did not run to completion. As the amount of acid fed was
reduced, as in the cases with 0.15 M and 0.07 M phosphoric acid, the initial rate

of alanine conversion was neariy identical to the base case (0.22 M alanine, 0.29

22

.008? a .289 8:8 as com a s .2.98
Lesson 03m .xb a _. 53> in 89 Co 2335 comet? Love: 60:28 055?. 2 and Co mEEm 2: £5 ooEBtoa 20:83... .22... otocamoca
.2 2114.8 96 ..2 So Thiv ..2 3.2151 ..2 3.21; E29628 65.6.6 S 28 9658.3 6 6:288 a .6 seem .a. 659”.

 

 

 

 

 

 

 

E:
o m a m .6.
< 4
ll
r to
T T
x
. . ed
5 ad
. 3

 

 

23

M phosphoric acid) until the phosphoric acid was consumed by basic product, the
fraction of alanine protonated approached zero, and the reaction stopped.

In the case where no phosphoric acid was present in solution, the
calculated fraction of protonated alanine in the feed was zero. Strangely,
conversion of 12% and yield of 8% was observed after the first hour.
Considering the protonation requirement for hydrogenation, no yield or
conversion was expected. However, acidic sites on the catalyst may donate a
sufficient number of protons to facilitate this limited conversion.

To support the idea that the catalyst imparts a specific “acidity" or enables
a specific molar conversion, two experiments were performed where no
phosphoric acid was added to the feed. In one case the amount of alanine was
doubled to 0.44 M and the other the catalyst loading was doubled. Both
experiments were run at 100°C under 1000 psi H2. In both cases the absolute
molar amount of alanine converted and alaninol formed was nearly the same as
when 0.22 M alanine was hydrogenated with 1 g of catalyst at the same
temperature and pressure. Considering the assumption that the catalyst
imparted a specific acidity, the same molar conversion in the case of 0.44 M
alanine fed was expected. SimilarIy, with a doubling of catalyst, it was expected
that the conversion of alanine and formation of alaninol would double the molar
conversion and yield values observed in the base case, which was not the case.
Therefore, these experiments together do not prove the idea that the catalyst had
a specific acidity. However, an experiment where 0.28 M sodium hydroxide was

added to the solution completely stopped the formation of alaninol. This

24

suggests that the catalyst did impart some acidity, although it is not clear how
this was specifically related to conversion.
3.2.2.2 Effect of molar excess of phosphoric acid on conversion

In the studies presented in this section, 100 grams of a 0.22 M alanine
solution was hydrogenated under varying phosphoric acid concentrations as
specified in the text. In all cases 1000 psi, 100°C, and 1 g of 5% Ru/C catalyst
were used.

In considering the base case of 0.22 M alanine and 0.29 M phosphoric
acid, where the maximum conversion occurs in this set of studies, the fraction of
alanine protonated in the feed was calculated to be roughly 70% and declined as
alaninol was formed. This begs the question of what the effect would be of
increasing the acid concentration above 0.29 M, thus increasing the fraction of
protonated alanine. As shown in Figure 5, increasing acid concentrations to 0.59
and 1.16 M phosphoric acid decreased the conversion rate compared to the case
with 0.29 M phosphoric acid concentration despite increasing the fraction of
alanine protonated, as shown in Table 4. This effect was due to a combination of
the following factors: hydrogen solubility decreased with increased acid
concentration and phosphoric acid occupied active sites on the catalyst.

While both of these factors contributed to the observed reduction in rate,
occupation of reactive sites by phosphoric acid was probably more significant.
This was clear by comparing the final conversion of the three cases. The
difference between the final conversion in the 0.29 M and 0.59 M cases was

25%, whereas the difference in the final conversion between the 0.59 and 1.16 M

25

.98.. 8 .982
not? 4E com m. E $238 5258 03m $2... a F 5? an 89 Co @5826 c3932 59.5 60:20..." oc_:m_m 2 «Nd ho mEmLu 03 5:5 cantata
22681 .28 2333.3 5. N: Torv new ..2 mod Tolv .s. and Alxlv E22058 2.5% :0 Eco otocamoca 8096 he 69cm .m 2:9“.

 

 

 

E:
o m 4 m a F o
b D b b b 0.0
. Nd
. to
X
.. 9o
. ad
a
Si

 

 

 

26

cases was only 13%. If the reduction in rate were due to a decrease in H2
solubility the observed reduction in conversion should be relatively linear with
respect to acid concentration. This observed reduction in rate appears
mathematically to be better explained by an absorption model where phosphoric
acid occupies active sites but does not bind as strongly as alanine. This was

shown to be the case when the kinetic model was constructed in Section 4.2.

 

 

 

 

 

 

 

 

 

0.29 M {039 M 1.2 M
W") PH X CCOOH/Ct PH X Cc00H/C1 PH X CCOOH/Ct
0 2.0 0.00 0.69 1.5 0.00 0.67 1.0 0.00 0.96
1 2.1 0.31 0.64 1.5 0.20 0.66 1.1 0.21 0.95
2 2.1 0.40 0.62 1.6 0.32 0.86 1.1 0.25 0.95
3 2.2 0.56 0.59 1.6 0.42 0.65 1.1 0.37 0.95
4 2.3 0.72 0.53 1.6 0.61 0.84 1.1 0.45 0.95
5 2.4 0.85 0.48 1.1 0.53 0.95
6 2.5 0.93 0.43 1.7 0.70 0.82 1.1 0.61 0.95

 

 

 

 

 

 

 

 

 

 

 

 

Table 4. Measured pH and calculated fraction of protonated alanine for reactions with 0.15 M,
0.29 M, and 0.58 M phosphoric acid. Reactions performed with 100 grams 0.22 M alanine
solution, under 1000 psi H2 with 1 g 5% RulC powder catalyst.

A further discussion of acid-base equilibria calculations is presented in
Section 4.2 and Appendix B, which compares the use of the Henderson-
Hasselbach equation to a solution charge balance calculation that predicts the
ionic state of alanine in solution.

3.2.3 Other acid solutions

In the experiments presented in this section sulfuric acid and triflic acid
were studied as acidulants. In these experiments, 100 grams of a 0.22 M alanine
solution was hydrogenated under 1000 psi H2, at 100°C, with 1 g of 5% Ru/C.

To complement the experiments performed with phosphoric acid added,
two experiments were performed with 0.29 M sulfuric acid and 0.29 M triflic acid

(triflouromethanesulfonic acid) to determine if the use of strong acids would

27

enhance the conversion rate. The strong acidity of sulfuric acid (pKA1=-2,
pKA2=1.9) and triflic acid (pKAz-13) shifts the solution equilibrium of alanine
further to the protonated state and thus a higher conversion rate was expected.

With 0.29 M triflic acid added, the initial pH of the solution was 1.1 and
increased to 1.4 by the end of the experiment (6 hours).~ At the end of the
experiment the fraction of alanine protonated was greater than 90% and the yield
was 46%. Similarly, with 0.29 M sulfuric acid added, the pH of the solution
remained ~1 during the entire experiment, corresponding to greater than 95% of
alanine in solution protonated. A yield of 45% was observed after 6 hours.
Despite the high fraction of protonated alanine the conversion rate in both cases
was roughly two-thirds of when phosphoric acid was used as the acidulant, as
shown in Figure 6. In the kinetic model, Section, 4.2, phosphoric acid is an
adsorbed species in the rate expression to account for the reduction in rate
observed where an excess of phosphoric acid was present in solution (Section
3.2.2.2). Here a similar phenomena was likely observed, where these acids
occupied active sites. However, these acids apparently bind much more strongly
than phosphoric acid which caused the reduction in reaction rate to be much
larger. This reduction in rate could also be attributed to catalyst poisoning from
these sulfur containing acids.
3.2.4 Effect of pressure

In the studies presented in this section the effect of hydrogen pressure on
reaction rate is presented. In these experiments, 100 grams of a 0.22 M alanine

solution was hydrogenated under 1000 psi H2, at 125°C, with 1 g of 5% Ru/C.

28

.28 82.8 2 mad 3 .28 9:5 .2 mad :3 .28 6.588;: .2 88 3 .2885 885.3 an 82 .5298 03:
a\om ho a P .c:_:m_m-._ .2 and h.o 5:28 wEEo cor 5:5 068.. um occtotoa mcozowom .25 Bow :0 2m: :Qflgcoo ho oocouccaco .m 059“.

 

 

 

E:
m v m N _. o
P L I I D ., o
4
. Nd
O
. v.0
X
9 u 0.0
. wd

 

 

29

Hydrogenation rate was only mildly dependent on hydrogen partial
pressure above 1000 psi H2. Because hydrogen solubility follows Henry’s law,
H2 concentration in water varies neariy lineariy with pressure. Figure 7 shows
that the conversion rate declines when hydrogen pressure was reduced to 250
psi; however, above 1000 psi rate was independent of hydrogen concentration
indicating the surface was saturated with hydrogen.

3.2.5 Temperature effect on amino acid hydrogenation

Temperature directly affects the rate at which alanine hydrogenates and
alaninol degrades and was studied over range of 90 to 150°C. In these
experiments 100 grams of a 0.22 M alanine solution with 0.29 M phosphoric acid
was hydrogenated under 1000 psi H2, with 1 g of 5% RulC.

Figures 8 and 9 detail hydrogenation reactions performed at 150 and
100°C, respectively. The upper set of curves in each figure show the
concentrations of L-alanine, L-alaninol, and D-alaninol versus time. The lower
curve in each figure shows alaninol enantiomeric excess versus time as a
measure of the rate of racemization observed.

As shown in the upper graph of Figure 8, at 150°C conversion was
complete in roughly 45 minutes, which equates to a rate roughly eight times
greater than at 100°C. This point corresponds to a maximum yield in L-alaninol
followed by a decrease in L-alaninol with an increase in D-alaninol.
Mechanistically a very important observation was made; the reduction of the
carboxylic functionality and racemization of alaninol are separate processes.

Simply by performing experiments with temperatures at or below 100°C, as

30

28 SF Txiv .56 8w T415 28 83 T1: .5: 88 Ti: 55585 6E3 55.82 29.98 03: 2.8 5 a F .68 .28 25.385 s. 88
65:29.. .2 NNd 5 5:28 mEEa 2: :53 06mm? 5 850:»: 85:80”. .2885 NI 5 SP. 585250 0555 5 352550 K 952...

E:

 

 

 

 

v m N _. o
D! D b o
X X ..No
\ ..Vd
VA
umd
v0.0
i? [F

 

 

31

 

..2 mad 8; 5:28 5 200 05:88.: 82050 .0032. 05m. 2.: 5 a F :5; Doom? 5 .0: 000.. 5 0.5805 A«I 50:: .:0:28
.. .2 mmd m 5 0:55 cow :53 00:555.: 5.5001 .:0:28 5 555.0 5 8090 0:0Eo::0:0 030:0 :55 5030.. 05:50-0 .741:
50.550050 55:20-2 .Tfiiv 80.550800 05:0.0-...T§l. ”:85 :05: 2055.0-.. 5 5.285000: 500:000m .m 0.52“.

05:20-

 

 

 

 

 

 

 

 

 

     

 

E:
m .5 m N F o
_ n p b n - oqoo
. 8.0
a
9
. 8.0
00.:
0::
m .5 m N F o
b b n 0
lil
. mod
0
O
U
. to m
+ u
n
w.
. 3.0 m
w
. No

 

 

mNd

32

..2 mm... 003 5:28 5 200 05:80:: 52050 .0032. 05m. $0 5 a .. :23 0609 5 .00 009 5 050005 «I 000:: .:0:28
05:0.0-.. .2 «Nd 0 5 0505 03 :53 0050.50 5:000”. .:0:28 5 55:05 5 8005 0:0E0::0:0 030:» :55 530.. 05:05-0 :14...
60550050 5555.: xiii. 80.550050 05:05-...Tol. ”505 .0003 55:05-.. 050 >._00_:0E0:0:0 5 5.550“. .m 059".

 

 

 

 

 

 

 

E:
v m « F 0
. . . 00.0
- 00.0
. a
.000
5 x x x .2:
:5.
v 0 « 3 0
a. i ar 0
- 00.0
o
O
U
- :0 m
w
m
- 3.0 m
w
- «.0

 

 

 

mud

33

shown in Figure 9, the stereocenter of alaninol was maintained. Another
important observation made here was that the main side product formed in
alaninol racemization at 150°C was ethylamine. The co-product formed with
ethylamine has not been positively identified because the HPLC method only
detects amine-bearing compounds. However, the co-product was likely
methanol or carbon dioxide.
3.2.6 Effect of increasing alanine concentration on molar conversion rate

In this section three experiments are presented that show the effect of
alanine feed concentration on hydrogenation rate. In each experiment 100
grams of alanine solution was hydrogenated in a phosphoric acid-water solution.
Feed concentrations of 0.22 M and 0.29 M, 0.44 M and 0.59 M, and 1.2 M and
1.5 M of alanine and phosphoric acid, respectively, were used. In all cases the
reaction conditions were 1000 psi H2, 100°C, and 1 g of 5% RuIC catalyst.

Figure 10 shows alanine conversion versus time for each of the reactions
based on the feed concentration. On a molar basis, the initial rate of conversion
in each case was approximately the same. After 2 hours the rate of conversion
diverged. This result was expected because the rate expression is a function of
alanine concentration. As conversion proceeds, the rate must decrease.
3.3 Alaninol degradation
3.3.1 Control experiment with alaninol

As a control experiment, 100 grams of 0.22 M alaninol and 0.29 M
phosphoric acid were fed to the reactor with 1 g of 5% Ru/C catalyst present

under 1000 psi helium. The reactor operated sequentially at 100, 125, and

34

.500 0.55005 .2 .3 0:0 05:05 5. N; .4. .200 0.55005
2 00.0 08 8.8.0 .2 00.0 .5 .28 2.9.89... .2 8.0 08 028.0 .2 ««.0 .6. 58.820808 08". .2885 «I .8. 000. 23.0.8 0...:
$0 5 m 2 .:0:28 05.5.0-.. 00.0.0 00. :53 0603 5 00:20:00 05.500”. .05.. 5.0.0500 5.0E :0 5.5200050 05:05 5 50.5 .3 0.50.“.

 

 

 

0....
0 m a 0 « 2 0
L D D b r o

. «.0

T
- «.0

X

. 0.0
f 0.0
. .

 

 

35

150°C for two hours at each temperature. In this study, no racemization of
alaninol was observed. However, the last sample taken, at 150°C, showed a 5%
yield of ammonia.

3.3.2 Temperature effect on alaninol degradation

An experiment was performed with 100 grams of 0.22 M alaninol fed to
the reactor with 0.29 M phosphoric acid. The experiment was carried out under
hydrogenation conditions of 1000 psi H2 and 150°C with 1 g of 5% Ru/C catalyst.
At 150°C, the rate of alaninol racemization, 41% ee after 6 hours, was nearly the
same as when formed from alanine, 42% ee of alaninol after 6 hours, as
presented in Section 3.2.4.

In the collection of experiments performed in this work, 100°C was the
lower boundary where racemization of alaninol becomes activated by the
catalyst.

Racemization of amino alcohols induced by catalysts at high temperature
has been also been observed by Antons“7 (Ru/C), Hersey56 (Raney Cobalt), and
by Studer46 (Rh/Pt oxide). Interestingly, despite using different catalysts, all
share 100°C as a starting point for product racemization.

3.3.3 Effect of phosphoric acid on alaninol degradation rate

Two experiments were performed with 100 grams of 0.22 M alaninol at
125°C under 1000 psi H2 with 1 g of 5% RuIC catalyst. One experiment
contained 0.29 M phosphoric acid and no acid was added to the other. At the

end of a six hour run, the measured enantiomeric excess of alaninol remaining

 

5‘ Harsey, s. c., u. s. Patent 4,990,666, 1991.

36

was nearly identical in both cases (89% and 91%, for 0.29 M phosphoric and no
phosphoric acid, respectively). However, without phosphoric acid added roughly
24% of the alaninol fed was lost in six hours versus 11% loss of alaninol over six
hours with 0.29 M phosphoric acid added. This was likely due, in part, to
phosphoric acid occupying sites. In both cases the side products are ammonia,

ethylamine, and isopropylamine.
3.4 Hydrogenation of B-Alanine

As a complement to our studies of alanine, two hydrogenation
experiments with B-alanine (3-aminopropanoic acid) were performed.
Experiments were performed with 100 grams of 0.22 M B-alanine solution fed to
the reactor with 0.29 M phosphoric acid. The experiments were carried out
under hydrogenation conditions of 1000 psi H2 with 1 g of 5% Ru/C catalyst at
100 and 150°C, respectively. At 100°C, a 16% yield of 3-aminopropanol product
was observed at 6 hours. L-alanine under identical conditions reached a yield of
91% in 6 hours. At 150°C, B-alanine was completely consumed in six hours.
However, a maximum 3-aminopropanol yield of 49% was observed after 4 hours
followed by degradation of 50% of the product. L-alanine under these conditions
rapidly hydrogenates with high selectivity in the first hour of the reaction followed
by subsequent degradation of the product.

As noted in the Section 1.2.4, a-substituted carboxylic acids are more
easily hydrogenated than normal alkanoic acids and non-a-substituted carboxylic

acids, and thus the observed results are unsurprising.

37

Chapter 4. Kinetics and mass transfer
4.1 Mass transfer calculations

For each reaction performed in the parametric series, a complete set of
mass transfer calculations was performed. To ensure that reactions were run
optimally, there were many conditions within the reactor that were considered in
regards to the catalyst, as well as mass transfer.
4.1.1 Catalyst suspension

Before considering calculation of various mass transfer coefficients,
complete suspension of the catalyst must be ensured. Any calculation in batch
reactors assumes that the contents of the reactor are well mixed. If the catalyst
is suspended, the entire surface of the catalyst is available for reaction. If some
catalyst remains on the bottom of the reactor, reactants must diffuse into those
masses of catalyst which would slow the reaction. In 1958, Zwietering59 studied
the conditions necessary to completely suspend particles in a baffled vessel. For
particles to be considered suspended, no single particle should remain on the
bottom of the reactor for more than 1 or 2 seconds. He proposed the following
correlation for predicting the minimum stirring speed to ensure complete

suspension.

Nm' =B-d8'2-uE'1-g°'45-(pp—p|_)°'45-w'°'13 (2)
'" “0.5511935

 

where Nmin = minimum stirring speed (rotations/sec)
B = constant determined by the impeller geometry

w‘ = catalyst loading (grams catalyst I100 grams solution)

 

5" Zwietering, T. H. Chem Eng. Sci. 1953, 8,244.

38

dp = particle diameter (m)
m = liquid viscosity (kg/m-s)
g = gravitation constant (mlsz)
pp = particle density (kg/m3)
pL = liquid density (kg/m3)
d. = impeller diameter (m)
In our system, 1 g of catalyst in 100 grams of water required a minimum stirring

speed of approximately 475 rpm. Complete suspension was ensured by stirring

at 1200 rpm.
4.1.2. Mass transfer coefficients
In this three-phase system, the following steps occur in the process of
making the alcohol product:
a. Transport of hydrogen from the bulk gas phase to the gas-liquid interface.
b. Transport of hydrogen from the gas-liquid interface to the bulk liquid.
c. Transport of the alanine and hydrogen from the bulk liquid to the catalyst
surface.
d. lntraparticle diffusion of the reactants in the pores of the catalyst.
e. Adsorption of the reactants on the active sites of the catalyst.

f. Reaction on the surface of hydrogen and alanine to produce alaninol.

g. Desorption of products.

The three phases of this system; gas, liquid, and solid, are shown in

Figure 11 along with the film layers that separate them. This figure describes the

39

concentration of the sparingly soluble gaseous hydrogen and dissolved alanine in

the different regions.

 

Concetration
2’

 

H2.
\ HzL
0» E
E :35 33
3g 3 3E 23;;

 

 

 

 

 

 

 

 

 

R

Figure 11. Three phase mass transfer in catalytic system
4.1.3 Reaction rate and mass transfer

In this three phase system, hydrogen must pass through both the gas-
quuid and liquid-solid films while alanine must pass through the liquid-solid film to
reach the catalyst surface. The rate at which these diffusion processes occur is
equal to the observed reaction rate, or the rate at which they are consumed by
reaction. The observed reaction rate is related to these transfer rates by the
following equations:

Gas-liquid transfer:
- RGsz -|- = kLalCHm - CH2,L ) (3)
liquid-solid transfer

‘ R(3412 'L = kst alCHzL " CH2,S ) (4)

40

-RG.A-|- = ks,Aa(CA,L - 0A6) (5)
Where -R3 = observed reaction rate of the species (kmol/mS-s)

L = ratio of catalyst surface area per unit volume fluid (mzlma)

a = mass transfer area per unit volume of reactor (m2/m3)

kLa = the gas-liquid mass transfer coefficient (1/s)

k. = the liquid solid mass transfer coefficient (m/s)

In these expressions, the mass transfer is driven by the concentration
differences between the two phases. The mass transfer coefficients are also
related to the observed reaction rate by the expressions above. They can be
calculated via correlations in the literature or observed experimentally. The
correlations used in this study will be discussed in Sections 4.1.4 and 4.1.5.

Once the mass transfer coefficients are obtained the maximum mass
transfer rate may be calculated. This is achieved by setting the driving force

equal to the value of the higher concentration.

Maxrmum gas - liqmd mass transfer rate = kLH2 a - CH2, G (6)
Maximum liquid - solid mass transfer rate = ks,H2 a - CHz L (7)
= ks,Aa ' CA L (8)

The maximum mass transfer rates calculated from these expressions are

compared to the observed reaction rate. If the maximum rate of mass transfer is

greater than the observed reaction rate in each case, the reaction is not mass

transport limited.

41

4.1.4 Gas-liquid mass transfer coefficient

The gas phase in these studies was composed primarily of hydrogen,
represented as Hz'. Water vapor was also present in this phase, however, in the
temperature range studied, 90 -150°C, the volume of water present in the gas
phase was small and assumed to be zero for calculation purposes.

To enter the bulk liquid phase, hydrogen must pass through the gas and
liquid films, which are also commonly referred to as the gas-liquid film. As shown
in the Figure 11, as hydrogen passes through this layer, its concentration
decreases. In this set of studies centered around 100°C and 1000 psi H2, the
solubility of hydrogen was on the order of 0.05 M. The rate at which hydrogen
passes through this film is called the gas-liquid mass transfer coefficient and is
usually represented as kLaB. To estimate this rate, two well known correlations
have been used.

In 1975 Yagi and Yoshida60 proposed the following correlation for gas

liquid mass transfer, kLaB.

1.5
2 0-19 05 060 0.32
k a d d Np 2 . u .
Sh: LDB l =o,oe.[_L_'-] [fl] (AL) [FL 9] [Mi] (9)
a

 

 

FL 9 PL Da ST ug

Where Sh = Sherwood number (dimensionless)
kL = liquid film mass transfer coefficient (m/s)
a3 = gas-liquid interfacial area per unit volume of reactor (m2lm3)

d. = impeller diameter (m)

 

6° Yagi, H., Yoshida, F. Ind. Eng. Chem. Proc. Des. Dev. 14, 488, 1975.

42

D. = diffusion rate of species a (mlsz)

N = stirring speed (rotations/s)

p, = liquid density (kg/m3)

in = liquid viscosity (kgIm-s)

g = gravitational constant (m/sz)

u9 = gas velocity (mls)

$1 = fluid surface tension (Nlm)
This equation was shown to adequately predict CO; absorption at 30°C for a
glycerol-water system.

In 1976 Bern61 proposed the following equation for a slurry reactor used to
hydrogenate oil in gas-liquid mass transfer controlled regime.

0.011-N1'16 ell-979 03-32

= (10)
kLaB V0.0521
L

 

Where kL = liquid film mass transfer coefficient (cm/s)
a3 = gas-liquid interfacial area per unit volume of reactor (cmzlcm3)
d. = impeller diameter (cm)
N = stirring speed (rotations/s)
pL = liquid density (g/cm3)
u9 = gas velocity (cm/s)

VL = volume of the solution (cm3)

 

6‘ Bern, L., Lidefelt, J. O., and Schoon, N. H. J. Am. Chem. Soc. 53, 463, 1976.

43

This correlation satisfied their data obtained for both 30 and 500-liter reactors. It
also compared well with experimental data obtained from 11 and 14 m3 reactors.
4.1.5 Liquid-solid mass transfer coefficient

From the bulk liquid phase, alanine and hydrogen must pass through the
liquid-solid film (L-S) to reach the catalyst surface. As shown in Figure 11, the
concentrations of both species decrease as they pass though this film and are
defined as A5 and H25 at the catalyst surface. The rate of transport through this
film is generally referred to as the liquid-solid mass transfer coefficient, ks. In this
study, the L-S mass transfer rate for both alanine and hydrogen was estimated

by the Boon-Long62 equation.

 

 

a HL 2 d3 dp PLDa

0.283 0.173 —0.011

2 2 3 0.019 0.461

k d 211 d d N 9 9d

5 P =0.04:5.[____p T J [ L 9] [LVL] _dT (L'- ] (11)
“L 0

Where ks = liquid-solid mass transfer coefficient (m/s)
dp = particle diameter (m)
0.. = diffusion rate of species a (mzls)
d7 = reactor diameter (m)
N = stirring speed (rotations/s)
m = liquid viscosity (kg/m-s)
pL = liquid density (kg/m3)
g = gravitational constant (m/s")

w’ = catalyst loading (kg/100 kg solution)

 

“2 Boon-Long, 8., Laguerie, C., and Couderc, J. P. Chem. Eng. Sci. 1978, 33, 813,

44

vL = liquid volume (m3)

In liquid-solid mass transfer, the mass transfer area of catalyst per unit
volume is calculated from the catalyst properties and experimental conditions as:

a = 521$ (12)
999

Where a = ratio of catalyst surface area to volume of fluid (mzlma)

w’ = catalyst loading (grams catalyst/100 grams feed)

pL = fluid density (kg/m3)

dp = particle diameter (m)

pp = catalyst density (kg/m3)

These correlations for the gas-liquid and liquid-solid mass transfer have
been used to characterize mass transport for all experiments in this study.
4.1.6 lntraparticle mass transfer

Finally, for reaction to occur the reactants must diffuse through the
interstices of the catalyst to the active sites. The Weiss-Prater63 criterion was
used to determine whether the reactions were mass transport or reaction limited
within the catalyst particles by calculation of the observable modulus, 1102.

—r-L2 (13)

2-
ll¢ CS'De

 

Where 11¢2 = observable modulus
r = reaction rate (molIL-s)

L = characteristic length of the catalyst (m), (d.,/6)

 

“3 Weiss, P. 3.; Prater, c. 0. Adv. Catal. 1954, 6, 143.

45

Cs = species surface concentration (mol/L)

o, = effective diffusivity (m2/s), o, = 32 D.

s = support porosity (0.6 for the catalyst used)
The observable modulus is the ratio of the observed reaction rate to the effective
rate of diffusion in the catalyst particle. If the observable modulus is much less
than 1, mass transport rate within the catalyst is greater than the reaction rate,

and conversion rate is not mass transport limited.

4.1.7 Results from mass transfer calculations

Table 5 summarizes calculated values for the calculated maximum mass
transfer rates, observed reaction rates, and maximum observed modulus for
three different experiments. The mass transfer rate of each experiment was
compared to the observed maximum reaction rate to determine if the rate of
conversion was limited by mass transport. It was found that in no case in the
study was the rate of conversion limited by mass transport through the film
layers.

In all reactions performed using alanine, the observable modulus for both
alanine and hydrogen was less than 1, and thus the conversion rates were not
reaction rate limited. In this study, the observable modulus was usually on the

order of 10‘s, and thus not diffusion limited within the catalyst.

46

 

 

 

 

 

 

 

 

 

 

Equation Rate 90°C 100°C 150°C
(fluid basis)

Bern, maximum g-l kLaBCH2- 1.03x10'2 103x10“2 1.03x10’2

mass transfer rate (kmollmss)

of hydrogen

Yagi and Yoshida, kLchHz. 1.77::10'2 2.01 x10'2 3.5x10‘2

maximum g-l mass (kmollmas)

transfer rate

hydrogen

Boon-Long, kLchH2§ 6.23X10'3 7.05::10‘3 1.36::10'2

maximum s-l mass (kmol/m s)

transfer rate of

hydrogen

Boon-Long, kLchAs 3.32x10‘3 3.63x10‘3 5.22x104

maximum s-l mass (kmollmas)

transfer rate of

alanine

Observed maximum RA max 1.32x10'5 3.01 x10‘5 1.81 x10“‘

reaction rate of (kmol/mas)

alanine

Observed maximum RH max 6.BX1O‘° 1.5x10‘5 9.05x1045

reaction rate of (kmol/mas)

hydrogen

110’. observed 1.32x1043 3.54x10" 9.59x10‘°

modulus for

hydrogen

00’. observed 1.05x10‘5 2.09x10“ 7.30x10‘

modulus for alanine

 

 

 

 

 

4.1.8 Limiting reactant

In any reaction system, either of the reactants may limit the reaction rate,

limitations.

The ratio of the observable moduli for hydrogen and alanine identifies the

47

limiting reactant as shown in equation 14.

Table 5. Summary of mass transfer calculations. Reactions performed with 0.22 M alanine in
100 grams 0.29 M phosphoric acid solution with 1 g 5% Ru/C catalyst under 1000 psi H2.

either by stoichiometry, if no mass transfer limitations occur, or by mass transfer

 

2)

00

'Y _( H2 _ "RGin Lz/CsHZDea _ Cs,ADe.A
(11¢sz -b-R3,A Lz/csAoaa sz.H2 De.l-l2

 

 

(14)

In this case b is the stoichiometric coefficient of hydrogen, two in this case. The
calculated ratio in every experiment was greater than 1, so hydrogen was the
limiting reactant. This finding does not mean the reaction was mass transfer
limited in hydrogen; it only serves to give a relative importance to the rates of
transport within the catalyst particles.

4.2 Langmuir—Hinshelwood Model Development

Catalytic reactions proceed through a series of steps at the catalyst
surface called the reaction mechanism. This series of steps serves as the basis
for development of the mathematical description of the reaction kinetics, called
the rate expression. Development of a kinetic model and rate expression is
important because it sheds insight into the mechanism of the reaction and allows
for the planning and development of industrial reactors.

In this section the proposed reaction mechanism, rate expression, and
kinetic parameters that describe alanine hydrogenation are presented.

A Langmuir-Hinshelwood model was employed which includes protonated
alanine (—COOH) and phosphoric acid (H3PO4 form) competing for one type of
site and hydrogen dissociatively adsorbing on another type of site on the catalyst
surface. This proposed model was supported by the following experimental
evidence and literature references.

Experiments with less than a stoichiometric amount of phosphoric acid

halt before alanine was completely consumed, as shown in Figure 4 and Table 3.

48

This effect was explained by the solution behavior of alanine, as discussed in
Section 3.1. In developing this model, therefore, only protonated alanine was
included in the surface reaction step, which makes the rate expression a function
of protonated alanine.

The proposal that alanine and phosphoric acid compete for sites was
supported by the observation that reaction rate decreased as phosphoric acid
concentration was increased above a stoichiometric amount, as shown in Figure
5. Without phosphoric acid competing with alanine for sites, the reaction rate
should increase due to a higher concentration of protonated alanine by further
shifting the acid-base equilibria toward the protonated carboxylic acid form.
Alternatively, this reduction in rate could be an effect of decreased hydrogen
solubility. However, the observed decrease in conversion rate at higher acid
concentration is only partially accounted for by reduced hydrogen solubility.

Finally, modeling hydrogen as dissociated on metal surfaces is widely
accepted in the literature64 and recently the Neurock group65 published
computational results on the hydrogenation of acetic acid over palladium which
suggests that hydrogen and acetic acid adsorb on different sites. Therefore,
hydrogen was modeled as dissociatively adsorbed and occupying different sites
than alanine.

With this evidence, the following reaction mechanism was proposed.

1. A" + $1 = A*-S1 (fast)
2. P + $1 = P-S1 (fast)

 

°‘ Singh, u. K.; Vannice, M.A. Applied Catalysis A: General 2001. 213, 1.
“5 Pallassana, v., Neurock, M. J. Catal. 2002, 209, 239.

49

3. H2 + 2 $2 = 2H'Sz (fast)
4. 2H'Sz + A+'S1 —+ A1+'S1 '1' 2 $2 (SIOW)
5. 2H'Sz + A1+'S1 -> A|0l+'S1 4' 2 82 (fast)

6. Alol*-Si = Alol+ + 81 (fast)

Where A+ is protonated alanine, P is phosphoric acid in H3PO4 form, H2 is
molecular hydrogen, H is atomic hydrogen, $1 is a site occupied by alanine
(-COOH form) or phosphoric acid (H3PO4 form), S2 is a site that is occupied by
hydrogen, and Alol+ is protonated alaninol.

These fundamental steps yield the following rate expression.
KA+CA+CS1 KHZCHZCS2

1+KA+CA++KPCP)G+JKH2CH2y

 

(15)

 

—rA =k$(

Where rA is the reaction rate of alanine, lg is the surface reaction rate coefficient,
CA. is protonated alanine concentration, Kit. is the protonated alanine adsorption
equilibrium constant (—COOH), Co is non-dissociated phosphoric acid (H3PO4)
concentration, Kp is the phosphoric acid adsorption equilibrium constant (H3PO4),
C51 is the concentration of sites on which both protonated alanine and
phosphoric adsorb competitively, CH2 is hydrogen concentration in solution, KH2
is the hydrogen adsorption equilibrium constant, and CS2 is the site concentration
on which hydrogen dissociatively adsorbs.

Several other variations of this model were tested including alternative

species adsorption, such as alanine and phosphoric acid adsorbed in all forms.

50

In these cases, the kinetic parameters determined using the search method
described in Section 4.4 did not describe the experimental data as well as the
proposed expression.
4.3 Solution charge balance

The proposed rate expression is a function of ionic species
concentrations. However, analytical methods measure the overall concentration
of the material, not individual ionic species. The ionic concentrations are
calculated values derived from the overall concentration of all the species in
solution. These concentrations were determined by way of a solution of charge
balance, where the net charge of the solution is zero. The system of equations
used to determine the ionic species concentrations from overall species
concentrations is described in this section.

To estimate the concentration of ionic species in solution for modeling

purposes, the overall charge balance for this system is expressed as follows:
0=H++A++Alol++0H‘-A‘-P'-2P2'-3P3' (16)

where H” is the hydronium cation, A" is protonated alanine, Alol+ is protonated
alaninol, OH' is the hydroxide ion, A' is deprotonated alanine, P' is H2PO4', P2' is
HPO42', and P3‘ is P0432

In this equation there are eight unknowns, therefore seven additional
equations must be defined for the system. For modeling purposes, side product
species were dropped from the overall charge balance and all alanine converted

was assumed to form alaninol. This assumption was made because the rate of

51

formation of these species cannot be predicted as functions of reactant and
product concentrations or from reaction conditions. This assumption will not lead
to appreciable error because of the high selectivities observed and the similar
pKa values of the side products with alaninol.

First in aqueous solution, the product of [H*][OH‘J = 10‘“. Therefore, OH'

may be written as:

10—14

OH-

 

(17)

Alaninol accepts only one proton and rearrangement of the Henderson-
Hasselbach equation yields the general equation:

c _ CAIOITH+
Alol+ ‘ + 18
KAlol 3‘“ ( )

Where CAlol+ is the ionic concentration of alaninol, CAM is the total
concentration of alaninol, and KN... is the acidity constant of alaninol.
Alanine and phosphoric acid exist in three and four ionic states

respectively. The three states of alanine are represented by the following system

of equations: AT = M + A + A‘ (19)
H+ A

[fill (21)

“A” [Al

52

The first equation is an overall balance including each of the three ionic
forms of alanine in solution where AT is total alanine concentration, A+ is
protonated alanine, A is zwitterionic alanine, and A’ is deprotonated alanine. The
second and third equations are equilibrium equations, where KM and KA2 are the
first and second acidity constants of alanine. Solving for each of the three ionic
species as functions of total concentration yields the following system of

equafions:

 

(22)

 

 

(23)

 

(24)

 

Similarly, the ionic states of phosphoric acid are represented by the

following system of equations:

 

 

P= PT
1+ KP1 + KP1KP2 + KP1KP2KP3 (25)
2 3
“+ (it) (it)
P" = PT

 

26
1+ H+ +KP2 +KP2KP3 ( )

KP1 H+ (H+)2

53

P2— 2 PT

 

 

 

27
K H+ (H+)2 ( )
1+ P3 + +
H+ KP2 KP1KP2
P3‘ = PT (28)

ir W (“’13
1+ + +

KP3 KPzKPa KP1KP2KP3

 

where P1 is the total phosphoric acid concentration, P is H3PO4, P' is H2PO4’, P2'
is HPO42', P3“ is PO43’, and KP1, Kp2, Kp3, are the three acidity constants of
phosphoric acid.

There is now a system of eight equations and eight unknowns. Because
of the highly non-linear nature of the charge balance, an analytical solution for H“
is not possible. Therefore, H+ concentration is calculated using Solver in
Microsoft Excel 7.0 by varying H+ until the charge balance equals zero. Since all
other species concentrations are defined as functions of H+ only, determining the
value of H+ simultaneously specifies all other ionic species concentrations.
4.4 Determination of Kinetic Parameters

The matrix of conditions used to determine kinetic model parameters was
derived from 20 experiments for a total of 109 data points. The operating
conditions of the experiments used in parameter determination are shown in

Table 6.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Experiment Alanine H2 solubilty Phosphoric acid TeTnperature Pressure

(M) (M) (M) (K) (Psi)

1 0.226 0.053 0.289 3'? 1000
2 0.231 0.053 0.591 373 1000
3 0.236 0.053 1 .198 373 1000
4 0.468 0.053 0.595 373 1000
5 0.226 0.062 0.289 398 1000
6 0.226 0.056 0.289 383 1000
7 0.226 0.051 0.288 363 1000
8 0.227 0.056 0.289 383 1000
9 0.226 0.122 0.292 398 1800
10 0.227 0.026 0.288 398 500
11 0.226 0.011 0.288 398 250
12 0.226 0.011 0.288 373 250
13 0.226 0.004 0.291 373 100
14 0.227 0.004 0.287 398 100
15 0.227 0.053 0.197 373 1000
16 0.228 0.053 0.147 373 1000
17 0.226 0.053 0.074 373 1000
18 0.227 0.026 0.288 373 500
19 0.226 0.139 0.289 373 2000
20 0.227 0.091 0.289 373 1500

 

 

 

 

 

 

 

 

Table 6. Conditions matrix used for parameter determination

The proposed kinetic model has eight parameters: four pre-exponential
rate constant factors (ks, KM, KH2, and Kp) and four activation energies (Ea,
AH», AHp, and AHHz). The parameter values are selected to predict the
concentration of alanine versus time such that the sum of the squared
differences between experimental alanine concentration and predicted data was

minimized, as described by equation 29.

2
$8 = 2(CAexp erimental _ CApredicted ) (29)
Where Cmpefimema. are experimental alanine concentration data points and
CWW are predicted alanine concentrations.
The reaction trajectories are predicted by a modification to the 4th order

Runge-Kutta numerical method which estimates conversion over a series of

55

small time steps (100 over 6 hours for these studies) using a macro created in
Microsoft Visual Basic for Excel. A more detailed description of the program
created for parameter determination may be found in Appendix B.

To initialize the calculation, the charge balance is satisfied at the initial
conditions using the Excel Solver add-in. The first time step is taken which
predicts the change in alanine concentration. Because of the change in alanine
concentration and formation of alaninol, the charge balance must again be
satisfied by using Solver before the next time step is taken. This sequence is
followed over all time steps.

A random walk technique was used to determine the set of parameters
that minimize the sum of square differences between experimental and predicted
data. For each step in the minimization, one of the eight parameters was
selected randomly, then randomly increased or decreased by a specified
percent. The reaction trajectory for each of the 20 reactions was then predicted
using this new set of parameters. If this parameter change results in a decrease
in the sum of squares, the new value of the parameter was retained; if not, the
initial value was restored. The random walk was continued until no decrease in
error was observed over 83 cycles. Employing standard statistical methods, 83
cycles was determined to give 95% probability that each parameter has been
both increased and decreased at least once.

The fitted constants are given in Table 7. The estimated heats of
adsorption for protonated alanine, phosphoric acid, and hydrogen are 81, 111,

and 69 kJ/mol, respectively, which are reasonable for adsorbed species. The

56

activation energy for the surface reaction rate constant, ks, is determined from
the composite L-H rate constant k1 = kanCa, and was equal to 82 kJ/mol, also a

reasonable value.

Rate or equilibn‘um Value at Activation energy
constant 373 K (kJ/mol)

 

k1(kmol/kg of cat/s) 3.455x10’5 81.54

 

KA.(m3/kmol) 190.62 80.93

 

Kp(m3/kmol) 40.31 110.76

 

 

K82(m3/kmol) 68.78

 

 

 

 

Table 7. Rate Constants for L—H Rate Expression

Comparisons of predicted and experimental conversion data versus time
for several sets of conditions are found in Figures 12 to 15. Over the wide range
of conditions from the 109 points used in 20 chosen experiments, the average
error between the experimental data points and the model prediction was 8%.

Figure 12 compares experimental and predicted conversion versus time
data for experiments that vary in temperature from 90°C to 125°C at 1000 psi H2
with 0.29 M phosphoric acid. Figure 13 compares experimental versus predicted
conversion versus time data for experiments that vary in pressure from 250 to
1800 psi H2 at 125°C with 0.29 M phosphoric acid. Figures 14 and 15 compare
the reactions performed with excess and deficiency of acid. In Figure 14, the
data show conversion rate decreasing with an increase in phosphoric acid
concentration; this trend was accounted for by including phosphate as an
adsorbed species in the rate expression. Figure 15 shows the effect of acid

deficiency on alanine conversion. The model predicts this near stoppage in

57

.0800 60:060.."— .0022

.2- l - IV ”0.00 .0000 5005080.: ”0000 50000.0 _o8_2lTllv ”0.000 .900 Eeoeeoexmli ”0.0: 80200.0 .0851 ..... 0

”0.0: .300 5005000?! ”0.00. 5028.0 .0022 .2-

- - -0 ”0.000 .900 500808010 0500.0 .2 00.0 0:0 080 0520820

.2 mwd £05 .0: an coo? 000:: 00500000 00200000. .0003000080. 0:00.0> 00 0.0.. 00.000250 0500.0 0000:0000 0:00? _0E0E_..0axm .3 00:9“.

 

 

:5.
o m v m N —. o
D I D D D ‘ 0
ts..\\
t\ t\s.
\t\ \\..
\| \ s‘ Q
\. .\ . .00
\‘
t\tl \\ s‘\..
\
t\t \ .\s s
0‘. \ ‘ s
0‘ I \ Q.
\t\ .\ .\ . .vd
‘\. \\ ‘ QQ
|\ \ \ Q
\
\. \ ..4 .0 X
t\tl \ .
|\ ‘ \ s .
I“ \\ \\ Q Imo
\ \ ..O
\ ss‘ s
o\\ \ t
\\ tt ..
\\ .\ .. .00
\ \ \
\\.o. ..\1 o...
‘0\| II!‘
ll-.lll|‘l‘l IIIIIII‘IHUUI, IIIIIIIII ..II UP

 

 

58

.00 000 005.00.. .0851- l - l.

.00 000 0.00 5008009010. .00 000 028.00.. .08....lTll. ..00 000 0.00 50055201.. .00 000. 005.00.. .0851 ..... .
”.00 000. .0000 3005003014. ..00 000. 0000.00... .0851- - - -. ..00 000. 0.00 .scoeeooxml... 0500.0 .2 00.0 000 0.00 0.8200020
.2 00.0 5.3 .0009 .0 00:02.00 0:0..000m. 0.0000... c0020.... 00.2.9 5.? 00. 00.0.0250 0:...0.0 .0EOE..096 0:0.0> 090.00.... .9. 0.00.”.

 

 

 

05.
m m v m N _. o
0 b n n p i0
I\V.0
\\~
\ .
so. \.. 0.0
s. I
t \ d
\ \ \D
s s.
\I \ .
\. \ o\«.
\ \ s 00 IV.O
\\. \ s 0
.\ \ t x...
\ \ ss . VA
.\ I \\ \ .. .00
s\ \ ss‘.s
.\ \ \ s.
\\ \ .\§ 5.
\ \ s
..\0 I \\ \\ s IQ.O
.\.\ \ a \t ..
I\|\ \\ \\ ‘\
I‘ll‘ I \\\ I|\ \t
I‘ll‘l III-““I‘ I*‘II\ II|‘||
11lllllllllll50‘“ ‘ III I I 1| I IIIIII IF

 

59

0.00 0.620820 .2 0.. 00.8.8...

.0851l l l. 0.00 009.0820 5. 0.. 0.00 50028901.. 0.00 082009... .2 0.0 00.8.8.0 .081 ..... . 0.00 020200020 .2 0.0 0.00
5005000011 0.00 082000.... 5. 00.0 00.8.8.0 .0851- - - -. 0.00 08200020 .2 00.0 .0000 .scoetaxwl... 00.8.0 .2 00.0 80 N... .00
0000 .00.... .0003 .0 00:00.00 00050000. 0.00 000200050 .0 0009.0 :0 5.2. 00. 00.0.0500 0c...0.0 0000.090 000.? .0...0E..00xm .3 0.00.".

 

 

 

.5...
o m 0 m N —. o
P p n L n
o
\33
\\.t
\\.u.
\\“\‘“
\\ \ \\ I.
\ 0‘ s No
\* \\ \\
\\ \\ \\
\ 4. t. o
\ s \
..\...“ \ s
\ \0 0 .
\\ ‘ \\ \‘\ [VO
\0\ \t tt
\ \ \
\ 0 \
\ \ \
t‘ ‘0 \ X
\\ ‘\ \\
\\ t ‘t .
4 t\ tt two
|\‘ \\
fl ‘\‘ \\\
II“ .\\\
|\ ‘0‘
0‘ \00‘ IQ.°
“.“

 

 

6O

0.00 0.5000000

.2 000 00.8.8.0 .0851- l - l. 0.00 0.8200020 .2 00.0 0.00 50088001.. 0.00 082000... .2 2.0 00.8.8.0 .0851lll.

0.00 0.820820 .2 E0 0.00 58580014. 0.00 082000.... .2 0.0 8.8.8.0 .0851 ..... . 0.00 082000.... 5. 0.0 .080
58500001... 0.00 0.8200020 5. 000 00.8.8.0 .0851- - - -. 0.00 0.09.0820 .2 00.0 .080 5850001.. 00.8.0 .2 00.0 80 Nr .00
0000 .000: .0000. .0 00000000 000.500”. 0.00 0.5000000 5 20005.50 0 5.3 0.0. 00.0.0>000 00.005 00550.0 000.? 5.005.503 0.. 0.00.0.

 

 

0....
m m v m N F o
p p 5 p In ‘0
\K“
V. .00
..\.
lllllllllllllll 'IIIIIIIIIIIIIUIIIIIIH‘\ .
0 ts
O £tt.t\ ..vd
s s s ...\l\
‘ \0\\\ VA
tt\.
.0M.\ .00
IIIINIIIII III\I‘\\\;\
$ lllll 0\\\\\\
\t.\\\\
...V....\ .00
t .. tI .. t “It-tilt}
.0

 

 

61

alanine conversion at the point that phosphoric acid was consumed due to
alaninol formation forcing the concentration of protonated alanine (—COOH form)
to zero.

The model does not predict alanine conversion without phosphoric acid
added. In experiments without phosphoric acid added to the solution an 8% yield
of alaninol was observed in 6 hrs. As discussed in 3.2.2.1, this result was
surprising because the pH of the 0.22 M alanine feed was 6.9. At this pH, nealiy
100% of the alanine molecules in solution are in the zwitterionic form. Patents by
Antons"'6 also show limited conversion of alanine to alaninol in the presence of
ruthenium oxides and Ru/C catalysts without the presence of acid.

4.5 Confidence intervals

Determining confidence intervals for highly non-linear models with large
numbers of parameters, such as in this study, is difficult. Attempts were made to
use the likelihood ratio and the approximate t-interval to determine confidence
intervals for kinetic parameters. Preliminary investigation into confidence
intervals for this model has yielded intervals greater than 100% of the optimal
values. For this type of model, this is not to be unexpected as noted by the
Belohlav group of Czechoslovakia,66 who have worked extensively in parameter
estimation for rate expression of catalytic hydrogenation systems. Since the
calculated confidence intervals were so large, they are not meaningful and will

not be presented.

 

6" Belohlav, Z.; Zamostny, P.; Kluson, P.; Volf, J. Can. J. Chem. Eng. 1997, 75, 735.

62

Chapter 5. Deuterium incorporation
5.1 Deuterium incorporation in amino acid hydrogenation—mechanistic analysis
5.1.1 Methods

Deuterium incorporation studies were performed in the Parr batch reactor
according to the procedures outlined in Section 2.1.2, with the exception that
hydrogen and water were replaced with deuterium and deuterated water,
respectively.

Proton (1H) NMR spectra were recorded for each sample in the study.
Since 1H NMR is not active for deuterium, peaks in the spectra which correspond
to specific sites on the molecule decrease as H is replaced with D, until the peak
is completely absent when 100% replacement occurs. Acetic acid was added as
an internal standard from which the concentration of H at each site of the
molecules was calculated by evaluating the ratios of the acetic acid methyl peak
with the other relevent peaks in the spectra. This was compared to the
concentrations measured by HPLC to determine the extent of deuteration.
5.1.2 Results

Incorporation studies were performed with alanine and alaninol as feed
materials in a range of temperatures from 100 to 150°C and 1000 psi D2, with a
general summary of results in Figure 16 and numerical details found in Table 8.

Hydrogenation of alanine at 100°C, as expected, added D at the C1
position of alaninol by hydrogenation. Complete incorporation of D was not
observed at the CZ position of product alaninol indicating incorporation at C2

occurs separately from hydrogenation.

63

5500.0 000 0500.0 5.3 00.030 000050500. 5 b0EEam .0 0.00...

 

«:22 I P .3 00.:000E 0000505056 .00 .00 89 .000: ..2 Nu... 0000. .0500? ..2 cud 0000.. 00.00?

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

um 09 000 0.00 o 0 000 .0500?
0. 00.. 000 m. ..0 o F 00.. .0500?
mm 00.. 00.. 0.00 00.0 0 on. .0500?
on 00. 00.. 0.00 00.0 F 00. .0500?
0 000 000 0.00 o 0 000 .0500?
0 000 000 0.0? 0 N 000 .0500?
0.. 00 00 0 0.00 00.0 0 000 .0500?
3 0.. 0.. 0.00 00.0 .. 000 .0500?
mm 000 000 00 00.. 0.00 00 00 0 0 on. 00.00?
mm 000 00 _. 00 00 0.0.. 3 00 0 0 cm. 00.00?
5 000 000 .200 3. 0.00 mm... 0 on. 0500?
mm 00.. 000 0.00 00 00 00.0 F 00. 00.00?
0 00.. 000 o 00 0.00 3 00 0 0 000 00.00?
o 00? 0 mm mdmn v 0 o F 000 0500?
o 00 00.. N 00. 0.00 0.. 00 00.0 0 000 00.00?
0 mm 09 0 No 0.0? or 00 00.0 0 000 00.00?
00 00 .0 00 00 .500 0.0; 580 .5.. 02. .0...

.0500? .0500? .0500? 0500? 00.00? .0500? .0500?... 00.00? 0.0? . ._. 0.000000

005000 0000 .0 000050500. n. 0..

 

 

64

H:\Z—0H 100 c D:\Z—QH__ 150 c DH2C\2—D-oo
.s‘

3N
l 100°C 1 100°C

0 D D
_>
03 +“N“ D 03 N“ =DHZC: + CD
N D3

Figure 16. General incorporation scheme for amino acids

As temperature was increased, the rate of D incorporation into the 02
position of both alanine and alaninol increases. At 150°C, D incorporation also
occurs at the CB position of both alanine and alaninol. Simultaneously,
racemization of product alaninol was observed. An attempt at making a
mathematical or statistical correlation between the extent of racemization and D
content at C3 has shown that racemization does not occur exclusively through
the 02-03 bond.

At 100°C D incorporation at CZ occurs without loss of optical purity for
both alanine and alaninol. This result suggests that the catalyst facilitates
deprotonation of the amino acid or alcohol to yield a surface-bound intermediate
as shown in Figure 17. At 100°C, this species must uniformly undergo
deuteration on the face from which the H loss occurred, thus retaining optical
purity. Higher temperatures would then disrupt the surface binding, so that D
incorporation can take place on the opposite face. At the same time, exchange at

C3 becomes possible, presumably via tautomeric equilibration.

65

  

 

      

 

 

 

H39 0 H39
H3N '5 COO H3N ; COO
e | e
H we 9 D
H H3" a/coo H
<9 I
Ru article .0. -—‘ Ru article . —-—‘ ‘1 Ru particle ' . .

 

Figure 17. Alanine interaction with the catalyst in DleZO

5.2 Method for deuterium labeling

This discovery of stereoretentive C—H bond activation at amine bearing
sp3 C sites was the most unexpected finding in this work. Related catalytic H/D
exchange processes have long been known, but they were typically run in
organic solvents and typically there is little discussion of stereochemistry."1 Such
reactions are thought to form imine intermediates, enabling alkyl group transfers
via transamination pathways."8 Indeed, alkyl group exchanges via catalytic
activation of amines has been explored as a process to convert primary to
secondary and tertiary amines.69 Only in recent years has the potential for
functionalization via C—C bond formation at such sites begun to be exploited,
enabling addition of alkyl or acyl groups at the C—H bond sites to amines.70 In
these reactions, h owever, the a mines require attached 2-pyridyl groups, which
serve as directing ligands for the transition metals effecting the C—H bond

activation."'72'73

 

“7 Shvo, Y.; Thomas, D. W.; Laine, R. M. J. Am. Chem. Soc. 1981, 103, 2461.

°° Marahashi, S. l.; Yoshimura, N.; Tsumiyama, T.; Kojima, T. J. Am. Chem. 8001983, 105,
5002.

”Wilson, R. 8.; Laine, R. M. J. Am. Chem. Soc. 1985, 107, 361.

7° Doye, s. Angew. Chem, Int. Ed. 2001, 40. 3351.

7‘ Chatani, N.; Asaumi, T.; Ikeda, T.; Yoritsu, S.; Ishii, Y.; Kakiuchi, F.; Marai, S. J. Am. Chem.
Soc. 2000, 122, 12882.

’2 Sakaguchi, s.; Kubo, T.; lshii, Y. Angew. Chem, Int. Ed. 2001, 40, 2534.

73 Chatani, N.; Asaumi, T.; Yorimitsu, S.; Ikeda, T.; Kakiuchi, F.; Murai, S. J. Am. Chem. Soc.
2001, 123, 10935.

66

The chemical behavior that has emerged from our own studies of amino
acid hydrogenation mirrors the above C—H bond reactivity, as evidenced by
isotopic exchange, and with the added bonuses of complete retention of
stereochemistry, simple aqueous-phase heterogeneous catalysis, and relatively
mild, green conditions. This discovery has led us to actively pursue the
possibility of stereo-controlled replacement of H with other functionalities under
aqueous heterogeneous catalytic conditions.

5.2.1 Background on amino acid labeling

Deuterium labeled a-amino acids are widely used to probe biological
structure and pathways.”75 Unfortunately, specific labeling by chemical
synthesis typically leads to racemic materials. For biological studies, significant
effort is then required to separate the enantiomeric forms or to differentiate their
biological roles.

In 1977, Maeda, et al.,76 described selective hydrogen-deuterium
exchange of amino acids and aliphatic amines over platinum oxide. After 40
hours at 100°C, incorporation occurs throughout the molecule up to 40% of the
maximum. The Zolorarev group in Russia has published several studies on the
incorporation of deuterium into amino acids. In their US patent,77 they describe a
method in which a biologically active molecule, and a platinum group supported
metal on an inorganic support may be used to deuterate the compound.

Essentially, the compound of interest forms a complex with the metal-inorganic

 

7‘ 75Ramalingam, K.; Najappan, P.; Kalvin, D. M.; Woodard, R. W. Tetrahedron 1992, 44, 5597.

:Pearce, D. A. Hammershoi, A., Harrowfield, J. M., Sargeson, A M. Chem. Com. 2000,2431.
77 °,Maeda M., Ogawa, 0., Kawazoe, Y. Chem. PhamI. Bull. 1977, 25(12), 3329.
Zolotarev, J. A.; Zaitsev, D. A.; Tatur, V. J.; Mysasoedov, N. F. U. S. Patent 5 0,26,909,1991.

67

carrier. The mixture is then exposed to low-pressure deuterium gas between
100—300°C. This process incorporates deuterium into several places in the
molecule with yields of hydrogen replacement up to 90%. Killgore78 studied
labeling of amino acids by the hydrogenation of N=C double bonds. In this
strategy, an a-(N-substituted azomethine)-substituted carboxylic acid equivalent
is dissolved in a deuterated protic solvent and exposed to hydrogen and a metal
catalyst, forming racemic labeled product. There are several examples using
microorganisms79'80'81 to make III-deuterated amino acids. While biological
transformations maintain the stereochemistry, these processes are limited by
which amino acid substrates may be used and their products are difficult to
purify.
5.2.2 Deuteration method

Catalyst and reactor setup are the same as described in Section 2.1.2,
with quantification methods specified in Section 5.1.1. Deuteration experiments
were performed under H2 (30-50 psi) or 02 pressure (1000 psi) in a temperature

range of 30 to 100°C. Deuterations were preferably carried out under 50 psi H2

and 75°C in a 2 wt% amino acid 020 solution with 1 g 5% RuIC catalyst (dry
basis). Deuterations were performed on alanine, cysteine (2-amino-
mercaptopropanoic acid), proline (2-pyrrolindine carboxylic acid), lysine (2,6-
diaminohexanoic acid), and phenylalanine (2-amino-3-phenylpropanoic acid).

Samples were prepared by filtration followed by rotary evaporation.

 

:Kilgore, J. L.; u. s. Patent5,,254 730 1993.
80Lim, Young-Hee, Yoshimura, T., Soda, K., and Esaki, N. J. Pennant. Bioeng. 1998, 86(4), 400.
:Feeney, J., Birdsall, B, Ostler, G" Carr., M. D., and Kairi, M. Febs. Letters 1990, 272(1,,2) 197.
"Mitglovi, G., Lammerhofer, M., Maier, N. M., Lindner, W., J. Labelled de. Radiopharm" 2000,
449.

68

5.2.3 Deuteration results

Alanine undergoes stereorentive D incorporation at C2 at pressures as
low as 30 psi H2 and a temperature range of 50-100°C in deuterated water
without acid added. The combination of low temperature and pressure, and
absence of acid, ensured that no hydrogenation of the carboxylic acid
functionality occured. The D20 served as the deuterium source because H/D
exchange takes place readily between H2 and D20 in the presence of catalyst
and the ratio of H2 to D20 was small (less than 0.5%). Unfortunately this method
was not specific to the 02 site. Deuteration also occurred at sites adjacent to
electron withdrawing groups, as summarized in Table 9. In proline, incorporation
occurred on both carbons adjacent to the nitrogen in the pyrroline ring. In
cysteine, incorporation occurred at the 03 position, adjacent to the mercapto
group. In lysine, deuteration occured at the C2 and 06 position, both adjacent to
amine groups. Phenylalanine has proved to be especially reactive with
ruthenium catalyst, with the ring undergoing rapid reduction and D-incorporation
occuring at all available sites.

The low pressure conditions make this transformation easily performed in

chemical laboratories equipped with standard equipment.

69

Table 9. Deuterium labeling summary

 

Amino acid

Observed structure

Comments

 

L-alanine

D2N/,(D
’ on
if

0

Complete incorporation at 02
in 6 hours at 75°C and 50 psi
H2

 

Cysteine

o N D
ofloo
0

$00

Greater than 80%
incorporation at C2 and 03 in
6 hours at 75°C and 50 psi H2

 

Proline

Complete incorporation at C's
adjacent to N in 6 hours at
75°C and 50 psi H2

 

Lysine

Greater than 75%
incorporation at C2 and 06 at
6 hours at 75°C and 50 psi H2

 

 

Phenylalanine

 

 

 

Incorporation at all positions at

conditions and ring saturation

at conditions as mild as 30°C
and 30 psi H2

 

70

 

Chapter 6. Amino acid hydrogenation economic analysis
6.1 Purpose of economic study

The purpose of this chapter is to compare the raw material costs of
alaninol production from alanine via hydrogenation and borohydride reduction.
This analysis is based on a particular production rate and estimates the cost of
each raw material per pound of alaninol produced. A general process diagram
and process description is included for each process.
6.2 Production rate

In each case, reactions are carried out in a batch reactor and will produce
1000 lbs of alaninol per batch from 1200 lbs of alanine assuming 100%
conversion and selectivity. This process will operate 22 times per month. At this
production rate, 158 tons of alanine will be consumed per year to produce 132
tons of alaninol. As shown in Table 1, the worldwide production alanine is 1,200
tons per year with a price that ranges from 12-29 $/kg (550-1300 $/lb).
Therefore, the cost of alanine per lb of alaninol produced is $5.50-13.00. At the
lower end of the estimate, the purchase price of alanine per year would be ~$1.7
MM/yr. However, if alanine is produced on-site the actual cost would be much
lower.
6.3 Alaninol production via hydrogenation
6.3.1 Hydrogenation process description

A general process diagram is shown in Figure 18. Alanine is produced
on-site via fermentation and is sufficiently purified to be suitable for

hydrogenation. The purified alanine is dissolved in a solution of water and

71

inorganic acid and fed to a reactor containing RulC catalyst. The reactor will

operate under hydrogen pressure at 1000 psi H2 and 100°C. At the conclusion of
the reaction, the reactor is drained and the contents fed to a centrifuge to remove
the catalyst. The inorganic acid is then neutralized using a strong base. Water is

then removed from the product by evaporation.

 

 

 

 

 

 

 

 

 

 

 

 

Water, Acid
7 Fermentation H Purification l——Lpl-—Re:t-o-r—_|' Catalyst
Wafter 03(0H)2
I Evaporator l1———-| Precipitator |.——l Centrifuge l
Alatinol r Gaga 7 7 Caglyst 7

Figure 18. Hydrogenation process diagram

6.3.2 Materials costs for hydrogenation
For this process material costs include alanine, hydrogen, catalyst, sulfuric acid,
and calcium hydroxide. The costs are calculated based on 1200 lbs of alanine
per batch with 22 batches performed per month. This corresponds to a 1000 lbs
and 132 tons of alaninol produced per batch and year respectively. Costs are
presented as per pound of product formed.

Sulfuric acid: The most economical inorganic acid suitable for this
process is sulfuric acid. Because of the strong acidity of both protons in sulfuric

acid only a 1/2 molar equivalent is added to the reactor relative to alanine. This

72

corresponds to 660 lbs/batch of sulfuric acid. At $80/ton, the cost of sulfuric acid
per pound alaninol product 2.6 ¢llb and the plant would require 87 tons/yr of
sulfuric acid. Another advantage is the ease with which sulfuric acid is
neutralized with calcium hydroxide to form calcium sulfate. Calcium hydroxide is
an inexpensive reagent and the cost incurred is less than 1 ¢llb of alaninol
produced.

Hydrogen: Alanine hydrogen requires 2 moles of hydrogen per mole of
alanine reacted. A hydrogen price of $1.00/lb is used as a rule of thumb for
engineering economic estimates. This corresponds to a cost of $27lbatch and 5
¢llb alaninol formed.

Catalyst Ruthenium on carbon catalyst is used for hydrogenation.
Estimating the cost of catalyst is based on the following criteria: the reactant to
catalyst ratio is 10:1, which corresponds to 120 lbs catalyst per batch, purchase
price of fresh catalyst is $50/kg, and ruthenium metal is recovered making the
effective price of the catalyst 1/3ml the fresh cost. The number of times that
catalyst may be reused without significant loss in activity is unknown. However,
estimates for catalyst use based on 1, 4, and 10 times use correspond to a cost
of 91, 31, and 9 ¢llb alaninol formed.

Using this simple economic analysis, the raw material costs are on the
order of $0.17-0.99llb of product produced excluding alanine cost.

6.4 Alaninol production via borohydride reduction
A process for reduction is shown in Figure 19. The process is derived

from a borohydride reduction found in the literature.27 This process was chosen

73

because of the use of relatively inexpensive reagents, no need for intermediate
esterification, and safety considerations.
6.4.1 Reduction process description

Alanine is produced on-site via fermentation and is sufficiently purified to
be suitable for hydrogenation. The purified alanine is dissolved in a solution of
THF and sodium borohydride. Sulfuric acid is slowly added to the reaction
mixture until the reaction reaches completion. The reactor is drained and the
solvent removed by evaporation. Water is then removed from the product by

evaporation.

THF, NaBH4 H2804

Fermentation }—-—O| Purification .-—Lp| Reactor l
Alaninol é-l Evaporator l¢———l Evaporator I

Ir 7 I

Water THF

 

 

 

 

 

 

Figure 19. Reduction process diagram

6.4.2 Materials cost for reduction
Under reduction conditions raw material costs are derived from sodium

borohydride, sulfuric acid, and solvents.

74

Sodium Borohydride: The ratio of sodium borohydride to alanine fed is
251, which corresponds to 1280 lbs/batch. At a cost of $846/ton, the cost of
borohydride is $0.54/lb product.

Sulfuric acid: The ratio sulfuric acid to alanine is 1.25:1. This corresponds
to 1652 lbs/batch. At a price of $80lton, the cost per pound of product is $0.07.

THF: The other cost significant cost is solvent. THF is the preferred
solvent for this reaction and has a selling price of $1 .55/lb. For estimation
purposes, 6000 lbs/batch of THF will be used to make a total batch mass of
10,000 lbs/batch. If an efficient recovery system is made for THF and 95% is
recovered, a solvent cost of 39 ¢llb product is incurred. If the percent of THF
recycled per batch falls to 75 and 50%, the cost of solvent increases to 1.94 and
3.87 $llb of product, respectively.

According to the estimates above the material costs for borohydride
reduction is $1 .00-4.48/Ib alaninol product.
6.5 Economics conclusion

As shown above, on a raw materials cost basis, amino alcohol production
via hydrogenation is competitive with the more often used reduction route.
Hydrogenation offers a “greener” route by the use of heterogeneous catalysts
and water as a solvent. The cost of raw materials for reduction is highly

dependent on solvent recycle and thus the material costs may vary widely.

75

Table 10. Summary of economic evaluation

 

Alanine use and alaninol production

 

Alanine
1200 lbs/batch
316800 lbs/year

Alaninol

998 lbs/batch

 

1r

263407 lbs/year

 

 

 

 

 

 

 

Total Raw Materials
0.99 $llb product

1 time catalyst use

0.31 Mb product 4 time catalyst use

 

 

158.4 tons/yr 132 tons/yr
12 Slkg-low end
5.45 Sllb-low end
29 $lkg-high and
13.18 $llb-high and
1728000 slyr-low and
4176000 $lyr-high end
Hydrowtion Raw Materials
Sulfuric Acid Hydrogen Ca(0H)2 Catalyst
80 Slton 1 $llb 10 Slton 7.58 $/lb
660 lbs/batch 27 lbs/batch 498 lbs/batch 120 lb/batch
174240 lbs/year 13052 lbs/yr 131569 lbs/yr 31680 lb/year
6970 $Iyr 13052 SM 658 SM 240000 $lyear
0.03 $/lb product 0.05 $llb product 0.002 $llb product 0.91 $/lb product 1 reaction
0.23 $Ilb product 4 reactions
0.09 Sllb product 10 reactions

 

 

 

 

 

 

 

 

 

 

0.17 $/lb product 10 time catalyst use
Reduction Raw Materials
Sodium Borohydride Sulfuric Acid THF
846 $Iton 80 Slton 1.55 Mb
1281 lbs/batch 1652 lbs/batch 6000 lbs/batch
338157 lbs/yr 436045 lbs/yr 0.39 $llb product 95% recycle
143041 $lyr 17442 Slyr 1.94 Mb product 75% recycle
0.54 $llb product 0.07 $Ilb product 3.88 Sllb product 50% recycle
Total Raw Materials
1.00 $Ilb product 95% recycle THF
2.55 $llb product 75% recycle THF
4.48 $llb product 50% recycle THF

 

76

 

Chapter 7. Amino acid hydrogenation conclusions and recommendations

Catalytic hydrogenation of amino acids to amino alcohols under
hydrogenation conditions was studied with alanine as the primary reactant of the
study. In acidified aqueous solution, alanine readily hydrogenates to alaninol
under mild conditions with high conversion, selectivity and enantiomeric excess.
7.1 Hydrogenation of alanine

Experiments were performed in stirred batch reactor under hydrogen
pressure. Ruthenium was the most effective catalyst for amino acid
hydrogenation.

Alanine readily undergoes hydrogenation to alaninol in acidified aqueous
solution. For hydrogenation of amino acids to occur, the carboxylic acid
functionality must be in the undissociated state (-000H). To shift the acid-base
equilibrium toward this state, inorganic acid was added to the solution. In this
study maximum hydrogenation rates were observed when phosphoric acid was
used as the acidulant. When less than a stoichiometric amount of phosphoric
acid is added to the solution, conversion ceases before conversion reaches
completion. As phosphoric acid addition was increased above a stoichiometric
amount, the rate decreased owing to phosphoric acid occupying active sites on
the catalyst.

Reaction rate is a strong function of temperature. At temperatures at or
below 100°C selectivities and enatiomeric excess near 100% were observed. At
temperatures greater than 100°C racemization of alaninol, along with side

product formation of ammonia, ethylamine, and isopropyl amine was observed.

77

Experiments at 150°C clearly showed that racemization and hydrogenation were
separate steps.

Reaction rate was essentially independent of hydrogen pressure above
1000 psi H2, indicating the surface was saturated.
7.2 Kinetic modeling

A Langmuir-Hinshelwood (L-H) kinetic model was presented in which
protonated alanine and undissociated phosphoric acid compete for one set of
surface sites and H2 dissociatively adsorbs on a second type of site. Conversion
rate was not limited by mass transport resistances over the range of temperature
(353-398 K), hydrogen pressure (250-2000 psi), alanine feed concentration
(0.22-0.46 M), and phosphoric acid concentration (0-1.2M) investigated. The
average error between experimental and predicted conversion over all data
collected was 8%.
7.3 Deuterium incorporation

Deuterium incorporation experiments were performed on alanine and
alaninol under hydrogenation conditions. Alanine hydrogenation experiments
show addition of D at C1 by reduction. Incorporation of D occurs at 02, the chiral
center, in both alanine and alaninol product without racemization. At
temperatures in excess of 100°C, incorporation occurs at C3. This occurs
simultaneously with racemization, although no direct link is known.

Detuerium incorporation results led to the development of a simple D
labeling method that is easily performed using standard laboratory equipment.

Using this procedure, D20 is used as the D source and H2 gas is supplied at

78

pressures as low as 50 psi. At this pressure, less than 1% of the total pool of H
and D is H and therefore, a high replacement is possible. At temperatures below
75°C and without adding phosphoric acid added, no reduction of the carboxylic
acid occurs.

The observed incorporation at 02 without the occurrence of racemization
suggests that C-C bond formation at the chiral center is possible.
7.4 Recommendations

To optimize this reaction further, it is suggested that investigation of
different acidulants proceed. The greatest liability towards green processing is
the use of inorganic acids. These require stoichiometric addition of base relative
to inorganic acid added for post-reaction neutralization. This creates a significant
amount of waste. A more attractive option would be the use of acetic acid which
would be unreactive under the mild hydrogenation conditions used and is easily
separated from solution.

Further study of amino acids with different side groups, and their effect on
hydrogenation would add further insight to the feasibility of for use of

hydrogenation of a wide variety of amino acids.

79

Chapter 8. PHA hydrogenation

As an extension of our work with amino acids and lactic acid, we have
performed some preliminary hydrogenation studies with two types of PHA's,
polylactic acid (PLA) and poly-3-hydroxybutyric acid (PHB).

8.1 Sustainable chemical processes

Currently, the most favorable means of handling waste PLA and PHB is by
composting, where these PHAs degrade in active composts to carbon dioxide,
water, and organic material. At the near-term projected yearly production of
these renewables-based polymers, composting PHAs would require a significant
effort in both collection and compost volume. Because PHAs are renewables-
based, CO2 released does not represent a net carbon increase in the
atmosphere; however this fixed carbon could be recycled into another application
before eventual return to the biosphere.

There will continue to be a large worldwide demand for PG, and the use of
1,3-BDO will grow over time. As the world chemical market increasingly
demands biobased feedstocks, the use of cropland for chemical and food uses
will have to be well planned to ensure that both needs are met. To reduce the
acreage devoted to chemical needs, discarded bio-based products, such as
PHAs, may be well suited to become feedstocks for other processes.

8.2 Background on feedstocks
8.2.1 PLA
PLA is now being produced at the commercial scale by Cargill-Dow LLC, a

joint venture by Cargill and Dow, in Blair, Nebraska. The plant began operation

80

in November 2001, and at capacity will produce 300 MM lbs/yr. In the long term,
the worldwide production for PLA could reach 10 billion Ibs/yr.82
The current world demand for P0 is on the order 1.25 B lbs/yr, with a

selling price of $0.68/lb83 via propylene oxide hydration. In the Miller/Jackson
group, propylene glycol has been produced from fermentation-derived lactic acid
via aqueous phase catalytic hydrogenation.
8.2.2 PHB

Poly((R)—3-hydroxybutyric acid) is currently produced via fermentation. Under
optimum conditions, PHB accumulates inside microorganisms at up to 80% of
their dry weight. The cells are then harvested and the PHB collected. Lemoigne
first observed PHB in 192784 and it has since been the topic of numerous studies.
Imperial Chemical Industries (lCl) began searching for an economically viable
fermentation route to PHB in 1976. In 1993, the annual production was on the
order of 800 tons per year. In 1996, Monsanto purchased Bl0P0L® (the
tradename for PHB) but stopped production in 2000 because of the high selling
price (~$17/kg) of the product. In May 2001, Metabolix (Cambridge, MA)
purchased the rights to Monsanto’s Bl0POL business, and in September of that
year was awarded $7.4 MM from the US. Department of Energy to continue their
work in PHB production in plants. Metabolix plans to produce PHB in plants with

full-scale production beginning in four to ten years.

 

'2 Testimony before the United States Committee on Agriculture, Nutrition, and Forestry, March
29’", 2001 by Pat Gruber, Cargill Dow LLC.

‘3 Chem. Market Reporter, Jan. 28‘“, 2002.

“ Lemoigne, M, Ann. Inst.,1927, 14, 148.

81

Commercial-scale uses of PHB polymers are on the horizon. Although
PHB is expensive as a polymer, as a chiral intermediate feedstock it is relatively
inexpensive. As an alternative to composting, both optically active and racemic
1,3-BDO (1 ,3-butanediol) could be produced from this othenrvise discarded feed.

(R)-1,3-butanediol is used as the starting material to make azetidinone
derivatives, which are important intermediates for penem and carbapenem

”'86 as well as pheromones,87 fragrances)”89 and insecticides.90

antibiotics,
Racemic 1,3-butanediol is 3 FDA approved food additive and is used as solvent
for natural and synthetic flavors. Chiral 1,3-BDO will find more widespread use if
it can be synthesized via a more economic route.
8.3 Polylactic acid results

The purpose of this section is to provide preliminary results of polylactic
acid hydrogenation. Reactions were performed with lactic acid, lactic acid
oligomers, and three PLA samples with molecular weights of 16,000, 140,000
and 300,000. Experiments were performed in a temperature range of 100-
175°C, under 1000 psi H2, with 1 g of 5% Ru/C catalyst.

There are many challenges in conducting a study the hydrogenation of
PLA because of its physical properties, reactivity, and ability to hydrolyze. PLA is

not water soluble unless it is present as a low chain length oligomer, apparently

n<7 from experimental observations. Therefore, when performing experiments

 

‘5 Nakayama, T.; lwata, H.; Tanaka, R.; lmajo, S.; lshiguro, M. J. Chem. Soc., Chem. Com. 1991,
9, 662.

“8 Tanaka, R.; lwata, H.; lshiguro, M. J. Antibiot.1990, 43, 1608.

“7 Mori, K.; Miyake, M. Tetrahedron 1937, 43, 2229.

8° 0hloff, G.; Giersch, W.; Decorzant, R.; Buechi, G. Helv. Chim. Acta 1980, 63, 1589.

‘9 Ferreira, J.; Ferreira, 3.; Simonelli, F.; Tetrahedron, 1990, 46, 6311.

9° Choi, V.; Elliot, J.; Johnson, W.; Tetrahedron Lett., 1984, 25, 591.

82

with PLA, the contents of the reactor make up a 4 or 5 phase system including
insoluble, solid PLA, insoluble amorphous-leathery PLA, and insoluble liquid
oligomeric PLA depending on the temperature the reaction is carried out at and
the time the material has been in the reactor.

The presence of insoluble, solid PLA in solution is particularly problematic.
It was often observed when opening the reactor after the completion of an
experiment that PLA was stuck in the headspace of the reactor above the level of
liquid. Because of this, the actual yield and conversion could not be determined
accurately for the experiment. The presence of solid masses also made
sampling very difficult because chunks became lodged in the outlet tube, which
limited the number of samples for many reaction conditions.

In the temperature region studied, 100—175°C, PLA goes through a
transition from a hard solid to a “leathery” material, and finally to a liquid. At
150°C, PLA is leathery and sticky. Because of this, during the course of some
reactions the PLA stuck together and formed a sphere. This agglomerate would
get caught in the mechanical workings of the impeller and forced the experiment
to be stopped before completion.

Despite the apparent physical barriers, PLA is rapidly converted to PG
under mild conditions.

8.3.1 Initial results

In the first set of studies, two experiments were performed with PLA with

average molecular weight of ~16,000, which corresponds to a chain length of

~220. In the first experiment, 100 grams of a 1 wt% PLA mixture was

83

hydrogenated with 1 g 5% Ru/C catalyst under 1000 psi H2 at 130°C. For
comparison, the same feed mixture was hydrolyzed at 130°C under 1000 psi
helium. This feed concentration of PLA corresponds to a maximum molar
concentration of monomeric lactic acid, or product PG, of 0.14 M.

As shown in Figure 20, formation of PG was rapid under the
hydrogenation conditions described above and achieved greater than 50% yield
in less than 1 hour before approaching the observed maximum of 77% in 5
hours. As shown in Figure 20, under the hydrolysis conditions described above,
the lactic acid concentration versus time data was exponential in behavior with a
2.5% yield of lactic acid observed after 1 hour and reached 67% at 6 hours.

Under hydrogenation conditions the rate of formation of PG was much
more rapid than the formation of lactic acid under hydrolysis conditions. This
observation answers a fundamental question: will the ester linkages of polylactic
acid hydrogenate? If the ester linkages of PLA do not hydrogenate, the
formation of PG from hydrogenation would be at the same rate, or slower, than
the formation of lactic acid from hydrolysis.

The rate of PG formation from PLA under hydrogenation was surprising
when compared to the formation of PG from lactic acid at the same conditions.
Zhang studied lactic acid hydrogenation extensively and developed a kinetic
model that predicts lactic acid conversion at 130 and 150°C91 using the same 5%
Ru/C catalyst as in this study. Using this model, the formation of PG from lactic
acid was predicted for 100 grams of 0.14 M lactic acid solution at 130°C, under

1000 psi H2, with 1 g of catalyst as shown in Figure 21 (assuming 100%

 

9‘ Zhang, 2.; Jackson, J. E.; Miller, D. J. Ind. Eng. Chem. Res. 2002, 41, 691.

84

.02. 2o...

o_ao_-..Te_=0o. 0.0292: .28 2.81.. .00 00.41.38. 5.0580»: $2.9. an 82 .85 So 02:8 mazes? .5298 030
$0 a P 5.; f E. 82 595 so 8.08 5000823.. 0.09 .0 30 .00; F 0500 8. 0.00223 0:29, 8.008230 Se. .00 2:0".

0 n v

I Lounflll

 

 

 

00:
0

#

 

 

 

.. 00.0

9

8.
0
(w) uonenueouoo

r 00.0

r _..0

 

85

..- - - -. 000 980.

050 2o; 00 8.985 .29» 000-158. 8.08092: .N... 00 88 0:0 0.8. 0 .0038 030 03 5.; 5.5.8 *5 So 0590 8. 5.;
50 00000 0000000000»... 0.00 0000. 0.00 020.005 000 50 E00 00.5000E .0020 00030000 00 0000000.— “.0 0000 05 5 000000E00 .50 0000.”.

050
0 0 e 0 0 F o
b h P h I o

 

 

 

 

 

 

 

86

selectivity). As shown in the figure, formation of PG from PLA was much faster
than predicted from lactic acid. This difference in the initial rate was especially
striking where at 1 hour the yields of PG from PLA and lactic acid are 53% and
26%, respectively.

Another hydrogenation experiment was performed with PLA of average
molecular weight of ~ 300,000, which corresponds to a chain length of ~4,100.
Hydrogenation was carried out with 100 grams of 1 wt% PLA mixture with 1 g 5%
Ru/C catalyst under 1000 psi H2 at 130°C. As shown in Figure 22, the formation
of PG was very slow in the beginning of the reaction when compared to the
previous case of average molecular weight 16,000 PLA. This was not surprising
because hydrolysis rate is related to chain length. However, once hydrolysis has
proceeded for ~6 hours, the formation of PG accelerated rapidly.

8.3.2 Use of tetrahydrofuran as hydrogenation solvent

The observations made in the initial studies of PLA hydrogenation raised
several fundamental questions. Does PLA hydrogenation proceed though one or
a combination of: end group hydrogenation and subsequent hydrolysis or direct
ester linkage hydrogenation? Secondly, if hydrogenation of ester linkages
occurs, what is the maximum chain length hydrogenated, and is the
hydrogenation random? With hydrolysis and hydrogenation occurring
simultaneously, these questions are difficult to answer. Therefore, experiments
were planned where PLA, PLA oligomers, and lactic acid would be hydrogenated
in THF (tetrahydrofuran). Performing the reaction in organic solvent eliminated

hydrolysis and would thus simplify the study. However, it was found that in THF,

87

0..

 

02.00-3130 .0003 058.2: 00.0 0.9. 8.00.8. 0... 0.2.0..
-...l<._0 .000; .0580... .so_ 5.. 5.00.3. 00 e: .8 82 9.0 0.02 .0 .0098 0.30 03 5.; 8:28 Sn. 00; P .o 00.0.0 8. 5.;
50 00.000 0000000020.”: .50 5000. 0.000 .50. 000 00.0 0.00 00.:000E .0000 0003020 00 0000000.. 00 0.0. 000 5 000000.000 .00 0.00.“.

..,...
E 0. 9 0 0 v 0 o

p b 0 p _ o
‘l I

r 0.0

r v.0

_. 0.0

.. 0.0

 

 

88

PLA, PLA oligomers, and LA are not reactive at 130 or 150°C. Therefore, more
modest goals of studying the effect of temperature and addition of wetting agent
were set.

8.3.3 PLA oligomer hydrogenation and hydrolysis

Two experiments were planned using PLA oligomers that were
synthesized in-house by neat condensation. The material chosen had an
average chain length of 3.4 and oligomers of up to chain length 11 were
observed. In the first experiment, 200 grams of a 2.5 wt% OLA mixture was
hydrogenated with 1 g 5% Ru/C catalyst under 1000 psi H2 at 130°C. For
comparison, the same feed mixture was hydrolyzed at 130°C under 1000 psi
helium. This feed concentration of PLA corresponded to a maximum molar
concentration of monomeric lactic acid, or product PG, of 0.34 M.

After six hours under hydrogenation conditions the PG yield from PLA was
62% as shown in Figure 23. In the figure, the predicted rate of formation of PG
from lactic acid was also presented and shows that in this case, the rate of PG
formation from PLA and lactic acid are nearly the same.

That hydrogenation of this OLA was only marginally more rapid than lactic
acid was not surprising after results from the hydrolysis were examined. Under
hydrolysis conditions, the lactic acid concentration was 66% of the maximum
achievable value after 1 hour. The brisk hydrolysis of the oligomeric feed rapidly
shifted the solution mixture toward a high fraction of lactic acid monomer and
enhanced hydrogenation rates as observed previously were not realized. The

rapid hydrolysis rate of short chain oligomers made studying PLA oligomers

89

..- - - -. 28 200. 59. 22> 00

8028.0 .3000-..T0_=8. 0208823: .f 00 82 2.0 .009 .0 .398 030 .03 £0, 8:28 $0; 0.0 0 2:00 80 .03 .8 0208
0000000902.. 0.00 0:00. .000 008.0000 000 (..n. 0000000.? E00 00.50000. .0000 00030000 00 0000000.. “.0 0.0.. 05 0. 000009000 .00 050.“.

050
o m e m N F o
I I I b I o

 

 

. 0.0

 

 

 

90

difficult because below 130°C the hydrogenation rate is relatively slow. Another
difficulty in using PLA oligomers was that they are not water soluble and form
droplets in solution. Samples taken from the reactor would appear clear with
droplets of PLA and catalyst present in the sample vial. The catalyst was
therefore not well dispersed in the reactor and the formation of PG was inhibited
initially.
8.3.4 Temperature effect on PLA hydrogenation

The rate of PG formation from PLA is a function of temperature via four
separate processes. The hydrogenation rate of ester linkages and lactic acid are
both functions of temperatures as is hydrolysis rate. The physical state of PLA
changes over the range of temperature studied from a hard solid at 130°C, to a
leathery material at 150°C, and finally a liquid at 175°C.

The next three sets of studies were all conducted with PLA pellets with a
average molecular weight of ~140,000 which were 2-3 mm in diameter.
8.3.4.1 Hydrogenation of PLA pellets at 130°C

Hydrolysis and hydrogenation experiments were performed at 130°C
under 1000 psi helium and hydrogen, respectively. A feed solution of 200 grams
of a 2.5% PLA solution were fed to the reactor and 1 g of 5% Ru/C catalyst was
used for hydrogenation. At 130°C, PLA remains a hard solid and it appears that
the breakdown of the pellets was due in large part to mechanical impact with the
impeller. The presence of solid material made drawing samples from the reactor

very difficult because of the small diameter of sampling tube used. Therefore, for

91

both hydrogenation and hydrolysis experiments usually only 3 hour and 6 hour
samples were taken, thus limiting the amount of information gained.

At the end of the hydrolysis experiment (6 hours), the measured
concentration of lactic acid in solution was 31% of the maximum value (if all PLA
was hydrolyzed). The hydrogenation experiment at completion (6 hours) yielded
10% lactic acid and 34% PG, of the maximum values. In both cases, there was a
significant amount of solid PLA left in solution. At this temperature, Zhang’s
model would predict a PG yield of roughly 58% if lactic acid was used as a feed
material.

There are two reasons for the relatively slow rates observed in the
hydrolysis and hydrogenation experiments compared to the results presented in
Section 8.3.1. The relatively slow rates are explained by a combination of the
relatively high molecular weight of the material and that the PLA is in pellet form.
As pellets, the amount of material exposed to water is relatively small, and thus
only a small amount was hydrolyzed into solution and available for
hydrogenation.
8.3.4.2 Hydrogenation of PLA pellets at 150°C

Hydrolysis and hydrogenation experiments were performed at 150°C
under 1000 psi helium and hydrogen, respectively. A feed solution of 200 grams
of a 2.5% PLA solution were fed to the reactor and 1 g of 5% Ru/C catalyst was
used for hydrogenation.

Hydrolysis experiments at this temperature were run five times, and were

aborted four times due to agglomeration of pellets which disrupted the stirrer.

92

However, on one of these occasions this did not occur and the PLA hydrolysis
was essentially complete in 4 hours.

Under hydrogenation conditions at this temperature, agglomeration of PLA
did not force the reaction to be aborted. It cannot be said with any degree of
certainty why this did not occur. However, hydrogenation of sticky, short chain
oligomers likely reduced agglomeration.

As shown in Figure 24, formation of PG from PLA under these
hydrogenation conditions was roughly at the same rate as Zhang’s model would
predict from lactic acid. Again in this case the enhanced rates observed in
Section 8.3.1, were not realized because PLA is present in a pellet form.
However, at this temperature the hydrolysis and hydrogenation rates were not as
inhibited as at 130°C.
8.3.4.3 Hydrogenation of PLA pellets at 175°C

Hydrogenation and hydrolysis experiments were performed at 175°C and
a pressure of 1000 psi. The feed solutions were 200 grams of 2.5% PLA and 1 g
of catalyst was used for hydrogenation.

Under these conditions, both hydrolysis and hydrogenation were very
rapid. Under hydrolysis conditions the yield of LA was ~90% at 2 hours and
increased to 95% by the end of the run (6 hours). Hydrogenation also occurs
very rapidly with a maximum yield of 57% of PG observed at two hours with only
a very small amount of lactic acid observed. At 175°C, a significant amount of
gas was formed with the carbon balance only 58% complete at the end of the

reaction (6 hours).

93

..- - - -0 28 209
.80 20$ 00 $0020 ..2000-..T0_=me 5008023: .f 00 82 0:0 0.03 0 0258 030 .03 5E 025.8 $5 0.0 .6 2.00 80 5?
50 00.000 0000000003... 0.00 0000. E00 090.0000 000 (..n. 800 0050000. .0020 00050000 00 00000000 00 0000 05 0. 000009000 .00 0000.".

050
o m 0. m N v o
n n P b h 0.0

 

 

.. wd

 

 

 

0..

94

8.3.5 Effect of PLA surface area on reaction rate

One of the primary issues when considering the combined rate of
hydrolysis and hydrogenation is the PLA surface area accessible to water. Much
of the work in this study has been done with PLA that was in pellet form. This
was done because experiments with powderized PLA would have a significant
amount of PLA stuck to metal in the headspace, particularly at 130°C.

Hydrogenation of powderized PLA was performed at 150°C under 1000
psi H2 with 1 g 5% Ru/C. The feed solution was 200 grams of a 2.5% PLA
solution. Powderized PLA was prepared from the pellets with average molecular
weight of ~140,000.

The yield of PG versus time is shown in Figure 25, with an additional set
of yield versus time data for the rate of PLA pellets under the same conditions.
As shown in the figure, the initial rate of PG formation was much more rapid with
PLA powder than with PLA in pellet form. This was expected because the
powder will have much more surface than the pellet and thus more opportunity
for water to attack the PLA for hydrolysis and subsequent hydrogenation.
However, as shown in the Figure 24 the yield only reached roughly 65%. This
was due in some part to the formation of gaseous products but primarily because
some PLA remained physically trapped in the headspace.

8.3.6 Addition of surface wetting agent

Two hydrogenation experiments were performed, one with the addition of

glycerol and the other propylene glycol. Due to the hydrophobic nature of PLA, it

was thought that addition of a material to improve the wettability of the PLA

95

00-3
[.838 Sn. .5.. 20... 0... .00-...T0__8 50 0.9. 22.. on. .f .3 82 9.0 0.02 0 00.0.8 030 $0 a F 5.; 5.028 00; 0.0 .0 2.0.0
000 5.; 50 00.000 000000090»... .50 03000300 000 00:00 Sn. 0.00 .0000 0005080 00 00000000 00 000.. 0.0 0. 000000000 .00 0.50.“.

 

 

 

 

050
0 m .V m N v o
. . . . l 0.0
. «.0
r «.0
IA
. 0.0
. 0.0

 

 

 

oé

96

would speed up the overall reaction rate. One experiment was run with PLA at
130°C, with 200 grams of 2.5% PLA fed under 1000 psi H2 with 1 g of 5% Ru/C
catalyst. The feed solution contained 5% glycerol by weight. In this case, at the
conclusion of the run (6 hours) the yield to PG was 25%. For comparison in the
case presented in Section 6.3.4.1, where no glycerol was added, a yield of 34%
at 6 hours was achieved. Similarly, a hydrogenation experiment was performed
at 150°C with 200 grams of a 2.5 wt% PLA solution with 2.5 wt% PG added to
solution. In this case, the yield of PG by reaction was 68% at 6 hours
(subtracting the PG fed). Under identical conditions presented in Section 8.3.4.2,
with no PLA added to the reactor, a yield of 72% was observed.

In the two cases studied no improvement in conversion rate was made by
adding a wetting agent. Possible improvement could be made by using different
materials.

8.3.7 Conclusion for PLA

Hydrogenation of PLA may be carried under mild conditions, similar to
those used for lactic acid. Several factors determine the rate of PG formation
from PLA, however the most important is temperature which controls the rate of
hydrogenation, and hydrolysis, and physical properties. With a Ru/C carbon
catalyst, 150°C was the preferred temperature because reaction rates are high,
without significant product degradation, and the physical state of PLA was

favorable for hydrolysis.

97

8.4 Results for PHB hydrogenation

Before direct hydrogenation of PHB was carried out, two hydrogenation
experiments with monomeric 3-hydroxybutyric acid were carried out at 100 and
150°C. A 100 9 feed solution of 1 wt%, racemic 3-HBA was hydrogenated in
water under 1000 psi H2 with 1 g 5% Ru/C catalyst. Phosphoric acid (0.08 M)
was added to the feed to ensure the carboxylic moiety was not deprotonated. At
both 100 and 150°C, 5% yield of 1,3—butanediol was observed at six hours.

Lactic acid under similar conditions, reacts much more readily than 3-
HBA. This is attributable to 3-HBA being a B-hydroxyacid, which are less
reactive than a-hydroxy acids. Similar differences in hydrogenation behavior
were noted between alpha and beta alanine as noted in Section 3.4.

An attempt at direct PHB hydrogenation was performed with 1 wt% PHB in
water at 150°C with and without phosphoric acid present with 5% RulC catalyst
under 1000 psi H2. This material had an average molecular weight of 437,000.
With 0.22 M phosphoric acid, yields of 15% 3-HBA and 13% 1,3-BDO were
observed after 20 hours. Without acid present, these yields fell to 2 and 4%,
respectively.

The slow rate observed for the PHB hydrogenation was probably due in
part to the high molecular weight of PHB used. As noted in Section 8.3.1, as
chain length increases, formation of alcohol product is delayed until hydrolysis
proceeds to a sufficient degree. Under the conditions studied, PHB was

apparently less susceptible to hydrolysis and hydrogenation than PLA. Because

98

of the low reactively of PHB, no further studies were performed and PLA became
the focus of the PHA hydrogenation study.
8.5 Recommendation for PHA hydrogenation studies

The results of this study showed the viability of polylactic acid
hydrogenation. The most important fundamental observation was that
hydrogenation of ester linkages occurs.

This work clearly showed that hydrogenation of polylactic acid and lactic
acid does not occur in organic solvents. This begs the question of whether these
organic solvents occupy active sites or if the presence of water is important to the
formation of reaction intermediates. Understanding the role of water in

hydrogenation would add greatly to the fundamental understanding of this class

of reactions.

99

Appendix A Alanine hydrogenation results
Table 11. Alanine hydrogenation: 100 grams solution, 0.22 M alanine, 0.29 M
phosphoric acid, 100°C, 1000 psi H2, 1 g 5% Ru/C catalyst

Table 12 Alanine hydrogenation: 100 grams solution, 0.22 M alanine, 0.59 M
phosphoric acid, 100°C, 1000 psi H2, 1 g 5% Ru/C catalyst

Table 13 Alanine hydrogenation: 100 grams solution, 0.22 M alanine, 1.2 M
phosphoric acid, 100°C, 1000 psi H2, 1 g 5% Ru/C catalyst

Table 14. Alanine hydrogenation: 100 grams solution, 0.46 M alanine, 0.59 M
phosphoric acid, 100°C, 1000 psi H2, 1 g 5% Ru/C catalyst

Table 15. Alanine hydrogenation: 100 grams solution, 0.22 M alanine, 0.29 M
phosphoric acid, 100°C, 2000 psi H2, 1 g 5% Ru/C catalyst

Table 16. Alanine hydrogenation: 100 grams solution, 0.22 M alanine, 0.29 M
phosphoric acid, 100°C, 1500 psi H2, 1 g 5% Ru/C catalyst

Table 17. Alanine hydrogenation: 100 grams solution, 0.22 M alanine, 0.29 M
phosphoric acid, 100°C, 500 psi H2, 1 g 5% Ru/C catalyst

Table 18. Alanine hydrogenation: 100 grams solution, 0.22 M alanine, 0.29 M
phosphoric acid, 100°C, 250 psi H2, 1 g 5% Ru/C catalyst

Table 19. Alanine hydrogenation: 100 grams solution, 0.22 M alanine, 0.19 M
phosphoric acid, 100°C, 1000 psi H2, 1 g 5% Ru/C catalyst

Table 20. Alanine hydrogenation: 100 grams solution, 0.22 M alanine, 0.15 M
phosphoric acid, 100°C, 1000 psi H2, 1 g 5% Ru/C catalyst

Table 21. Alanine hydrogenation: 100 grams solution, 0.22 M alanine, 0.08 M
phosphoric acid, 100°C, 1000 psi H2, 1 g 5% RulC catalyst

Table 22. Alanine hydrogenation: 100 grams solution, 0.22 M alanine, 0.29 M
phosphoric acid, 125°C, 1000 psi H2, 1 g 5% Ru/C catalyst

Table 23. Alanine hydrogenation: 100 grams solution, 0.22 M alanine, 0.29 M
phosphoric acid, 90°C, 1000 psi H2, 1 g 5% RulC catalyst

Table 24. Alanine hydrogenation: 100 grams solution, 0.22 M alanine, 0.29 M
phosphoric acid, 110°C, 1000 psi H2, 1 g 5% Ru/C catalyst

100

Table 25. Alanine hydrogenation: 100 grams solution, 0.22 M alanine, 0.29 M
phosphoric acid, 125°C, 1800 psi H2, 1 g 5% Ru/C catalyst

Table 26. Alanine hydrogenation: 100 grams solution, 0.22 M alanine, 0.29 M
phosphoric acid, 125°C, 500 psi H2, 1 g 5% Ru/C catalyst

Table 27. Alanine hydrogenation: 100 grams solution, 0.22 M alanine, 0.29 M
phosphoric acid, 125°C, 250 psi H2, 1 g 5% Ru/C catalyst

Table 28. Alanine hydrogenation: 100 grams solution, 0.22 M alanine, 0.29 M
phosphoric acid, 125°C, 100 psi H2, 1 g 5% Ru/C catalyst

Table 29. Alanine hydrogenation: 100 grams solution, 0.22 M alanine, 0.29 M
phosphoric acid, 100°C, 100 psi H2, 1 g 5% RulC catalyst

Table 30. Alanine hydrogenation: 100 grams solution, 0.22 M alanine, 0.29 M
phosphoric acid, 100°C, 100 psi H2, 1 g 5% Ru/C catalyst

101

.055... 0.3... $0 a F .f .8 8.: 0.8: .55 05585 s. 0~... .0555 .2 -... 88.8 0.5... 8: 5558.2... 8.52 : 555

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

~80 PB... 2.... ~80 9.... 0.~ 0
.8... 8.... ~20 ~00... 3... v.~ m
08... 2.0... 8.... E... 8... ~.~ 4
08... 03... 03... 03... 8... ~.~ n
08... v8... 8...... 08... ~00 :.~ ~
.8... 2.0... :~... :0... .0... :.~ :
8.... 8.... 8... ...~ ..

0.2850 25> 8.0.850 08. 0000

8 3552.0 0552.0 0552... 5 2 505.". 1.. 255
~80... :8... 88... 88... :8... 88... 3.0.... 0
88... 58... 88... ~02... 0:8... 88... 3.8... m
-8... 58... 38... 5:... :8... 88... 88... v
88... E8... 88... 08:... 88... 88... 88... n
-8... 28... ~08... 05.... 0.8... 88... 82... ~
28... :8... 88... 5...... 88... 88... 82... :
88... 88... o

.58.. .58... .52.... .58.. 358.. .52... :52... 2....-
<0_ 05.5050 0552-0 5552.0 55502 8.52.0 8.52... .

 

 

 

 

 

 

 

102

0020.00 03m ..\..0 0 w .~I .00 000—. 0.00.. .200 000000000 2 00.0 6500.0 2 NNd 0000.00 00.0.0 00.. H0000000000»; 00.00.< NF 0.00...

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

50... 05.. 08... ~....... ~..... 0. 0
.8... 0.0... 20... :0... 2.... 0.. v
.8... 8.... 0...... -v... 2... 0.. 0
08... $0... :~... 0...... 2.... 0.. ~
~00... ~02. 8.... 8~... 2.... 0.. .
8.... 5...... 0...... 8.... E... 0.. 0...
.. .. ..l..... 0.. ..
a... 0.5800 22> 5.0.850 05.1000 1.. 955
0552.0 0552.0 0552-. 5 2 505.".
58... 88... 2.0.... 0.8... 0.8... 0
88... 88... 8:... 88... 88... a...
-8... 88... ~50... 88... 52... 0
28... ~08... 58... 8...... :0... ~
:8... 88... $8... 88... .5.... .
:8... S8... 8...... 88... 5...... 0...
88... 0.20.. ..
.28.. .22... .28.. 385. .22... 0...:
0552.50 3552.... 0552.0 8.52.0 2552-.. .

 

 

 

 

 

 

 

 

103

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

....30 ....20050 0.20050 >330...0 0.20050. 2.20050. 9.2.0350 .0)
.300... .300... .3050 .3050 .3050 A300... .300...

0 0.0000 0.0000
. 0.000 0.0000 0.0000 0.00.0 0.0000 0.00.0 0.00.0
0 0.. .2. 0.0000 0.0000 0.0000 0.0000 0.00.0 0.0000
0 0.300 0.0000 0.0000 0.000. 0.000. 0.0000 0.0000
0 9.000 0.0000 0.0000 0.0000 0.000. 0.000. 0.0000
0 0.. ..0 0.0000 0.0000 0.300 0.0000 0.0000 0.0000 .
0 0.00.. 0.0000 0.00.0 0.03 0.0000 0.0000 0.00.00
. 3.00.50 2 5 ...20050 0.20050. 0.20050.

....30 0: 000... 53,. 00308.0: <5... 00.00.23. 00
0 . .0 0.00 0 0 ,
. ... 0.00 0.0.. 0.. 00 0.000 ..000
0 . .. 0.00 0.000 0.000 0.000 . .000
0 ... 0.00 0.000 0.00. 0.000 0.000
0 ... 0.00 0.000 0.0. 0 0.0. 0 0.000
0 ... 0.00 0.000 0.000 0.000 0.000
0 ... 0.00 0.0.0 0.000 0.000 0.000

 

 

 

 

 

 

 

 

400.0 .0 20050 3.0300035... .00 0330 00.000... 0.00 2. 0.0050. ..0 2. 0000000.... 00.... .006. .000 00. ...0. . 0 0.... .050 00.0.0.0.

104

r->_m=.=m 0-22.50 >339...» m->_m=5o_ 2-22.50. mssmasm

 

33.6: >_ S r->_m:.=a mica—8. m.>_m:_=o_

u... 000: 3:: 00:58.2. <35

 

.306 I 23.3 33.82.38? So cBBm 8.5.0:. 93 2. £3.39 Fun .5 gown—83d moi. 896. 88 B. In. a a mi. 2E0 88:3

105

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.238 723.3 0.23.3 2:335 w.>_ma=o_ m->_ma:o_ $333.8 =u>
38:5 38:5 33:5 35:5 35:5 35:5 36.5
c ...NNNA ll. 2&5 |..
A ...rum.‘ ..E 3F?» 90on .88 9E boom
u. .13; .BB .88 0.8% .88 PE .88
u . .ommn .olobm .88 o.. mg _ .oooo Ebola .88
A .E .88 .88 P. We . .08» 9E .83
m .83 .E .89 0.. gm _ .88 2E .83
m .33 _ .88 .82 0., 8m _ .88 Peace _ .oonm

 

 

 

:1
3
a:

.305 5 23.3 33258283” :5 aBSw 8.5.0:. oh» 3 2359 Fun 2. 383020 35. 396. 38 um. I». A a 9* 3:0 882m.

n:

38.6: >_ 5

0001 8:: 00:38.0:

$23.3

m.>_ma:o_
mamas:

 

106

r->_m:3m 0.23:5 >332? @2332 $22.30. $353.3

 

33:0: >_ S _.->_ma=m m->_ma:o_ m.>_ma=o_

2.. 00013:: 08338: <55

 

Asia 3 23.3 33333383” :5 cB3m 8.58:. oh» z. 2359 Pun z. 38252.0 no.9 48.0. 58 um. I». .. a 3.. min SEE

107

723.3 0.23.3 >33o:.m m-2m_..3o_ 2-23.30. $3335... .3»
.325 305.. 30:... A325 30:... 30.5 .32....

 

.582. 2 ... .5225; w.>_m=...o.<.¢.n $22.50.

u... 000... 33. 00:53.2. 9.8.2.2

 

...mZo 3 22:3 =<a8oo=o=oa .8 o83m 8.5.2.. ohm .<. 9.2.3? 0.8 2. 2.2.2510 no.9 .86. moo no. In. . n m... NEG 88.6.

108

..-200.00 0200.00 >3300.0 $200.00. $200.00. m.0<.03.00 =u>
30.5 30.5 30.5 30.5 30.5 30.5 30.5

 

300000 .0 ...200.00 $200.00. $200.00.

0: 000... 3.3 0853.2. .00... 00.8.2.0.

 

0.00.0 .0 200.00 308000000? .8 9030 00.5.00. PB .s 0.00.00. 93 .<. 000000000 00.0. 30.0. N6 00. I». .. a mi. NEG 00.0.5.

109

r-200.00 0.200.00 >3300.0 $200.00. 3200.00. m.0<_03.00 .0.»
30.5 30.5 30.5 30.5 30.5 30.5 30.5

 

030000 2 .0 ...200.00 $200.00. $200.00.

0... 000... 8:0 00053.00 <.0.0

 

...00.0 .0 200.00 3030000000. .8 0330 00.5.00. 0.00 z. 0.00.00. 9.0 3 000000000 00.0. .86. .80 00. I». . 0 0* NEG 00.006.

110

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

....30 0200.00 $200.00 $200.00. $200.00. @3033
30.5 30.5 30.5 30.5 30.5
0 0.0000 0.0000
0.0 0.. 00. 0.88 0.00.8 0.003 0.00..
. 0.. 0.. 0.0000 0.03. 0.0000 0.0000
0.0 0.03 0.0000 0.300 0.0000 0.0000
0 0.. .00 0.0000 0.00. 0.002 0.0000
0 0.0000 0.0000 0.000 0.0000 0.0000
0 0.3.0 0.0000 0.000 0.00.0 0.000..
....30 3000080 1.00000 2 .0 ...200.00 $200.00. $200.00. 00
0.... 000... 80.0 00053.00 $0.0 $200000.
0 0.0 0.3 0 0
0.0 0.0 0.8 0.30 0.00 0.000 0.000
. 0.0 0.00 0.000 0.. 00 0.0. 0 ..000
0.0 0.0 0.00 0.000 0.000 0.000 ..000
0 0.0 0.0 0.000 0.50 0.000 0.000
0 0.0 0.00 0.00. 0.000% 0.000 0000']
0 0.0 0.00 0.000 0.0.0 0.000 0.000

 

 

 

 

 

 

 

 

400.0 00 200.00 0330000000“ .00 0830 00.5.00. 0.00 .s 0.00.00. 0.... 2. 000000000 00.0. 30.0. .000 00. ...». . 0 00.. .050 00.0.5...

111

0200.00 0200.00 >3300.0 $200.00. $0200.00. 900.0330 ..u>

0... 00000.00 2 .0 0200.00 $200.00. $200.00.
00013.3 00038.00 <.0.0

 

0.00.0 0. 200.00 00080000000. .00 0330 00.5.00. 0.00 .s 0.00.00. 0.00 .5 000000000 00.0. 80.0. .000 00. I». . 0 0* 3E0 0000.00.

112

0200.00 0200.00 >3300.0 $200.00. $200.00. m.0<.03.00 ..u>
30.5

305.00 2 .0 0200.00 $200.00. $200.00.
00010003 000500.00 <.0_0 9.00.23.

0...

 

000.0 00 200.00 00030000000. .00 0330 00.5.00. 0.00 z. 0.00.00. 0.00 2. 000000000 00.0. .005. .000 00. I”. . 0 00.. .050 00.0.00.

113

0200.00 0200.00 $200.00. $200.00. m.0<_03.00 .0)
30.5 30.5 30.5 30.5 30.5 30.5

0.00000 2 .0 0200.00 $200.00. $200.00.

0... 000... 33. 0853.2. <55

 

400.0 00 200.00 00080000003 .00 0.030 00.5.00. 0.00 z. 0.00.00. 0.00 .s 000000000 00.0. 00.0. .000 00. I». . 0 00.. ....50 00.0.00.

114

..200.00 0200.00 20300.0 m200.00_ 2200.00. 93.03.00
3.0.5

 

0... 3.00000 2 .0 ..20050 m200.0o_ $200.00.
0001.020 00053.00 <.0_0 9.00.2.2

 

0.00.0 00 200.00 3030000009 .00 03.00 00.5.00. 0.00 z. 0.00.00. 0.00 2. 000000000 00.0. 30.0. .000 00. In. . 0 00.. 3.00 00.020.

115

..->.00.00 20080.0 $200.00. $200.00. m.0<.03.00
200.5 200.5

....30

 

3.00000 2 .0 7200.00 $200.00. $200.00.

0: 000... 800 085.080 0020 00.8.2.2

 

...020 00 200.00 303000000? .00 0330 00.5.00. 0.00 .s 0.00.00. 0.00 .5 000000000 00.0. .006. .000 00. In. . 0 00.. 250 00.020.

116

...30 ..-200.00 0-200.00 >3300.0 0200.00. 0200.00. 902.0330 .0».

305.00 2 .0 r200.00 0.200.00. 0.200.00.
000... .23 000510.00 <.0.0 00.8.2.2

0...

 

...00.0 no 200.00 02030000000. .00 0330 00.5.00. 0.»... z. 0.00.00. 0.00 .5 000000000 00.0. 30.0. 000 00. In. . 0 0.... 050 00.020.

117

7200.00 0200.00 >3300.0 0200.00. 0200.00. m.0<_03.00 .3»
30.5 30.5 .325 30.5 30.5 30.5 30.5

4.30

300000 2 .0 ...200.00 0200.00. 0200.00.
0001.03 <.0_0

 

...020 nu 200.00 02080000003 .00 0330 00.5.00. 0....» .<. 0.00.00. 0.00 2. 000000000 00.0. $0.0. ~00 00. In. . 0 00.. 050 00.020.

118

...200.00 0.200.00 >3300.0 0200.00. 0200.00. m.0<.03.00 ..u>

 

030000 2 .0 r.200.00 0200.00. 0200.00.

0... 0001 .23 00.0

 

...020 00 200.00 0208000000? .00 0330 00.5.00. 0.00 z. 0.00.00. 000 2. 000000000 00.0. 30.0. .00 00. In. . 0 000 250 00.020.

119

..200.00 0200.00 >3300.0 0200.00. 3200.00. m.0<_03.00 =0)

....30

 

3.00000 2 .0 7200.00 0200.00. 0200.00.

0... 000... 8:: <0...

 

0.00.0 no 200.00 0380000000“ .00 0830 00.5.00. PM» 2. 0.00.00. 0.00 2. 000000000 00.0. 30.0. .8 00. ...». . 0 0* NEG 00.020»

120

7200.00 0200.00 >3300.0 0200.00. $200.00. 90:03.00 .0.»

4.30

3.00000 2 .0 r200.00 0.200.00. 0.20050.

01 000... 8:0 .003

 

...020 00 200.00 00380000000” .8 0330 00.5.00. ohm z. 0.00.00. 0.8 2. 000000000 00.0. 30.0. .00 00. In. . 0 $0 .050 008.5»

121

Appendix B. Acid-Base calculations

As discussed throughout this work, determining the ionic state of alanine
in solution is critical to understanding amino acid hydrogenation and predicting
the kinetics of alanine conversion.

The Henderson-Hasselbach equation is used throughout this work to
estimate the fraction of alanine protonated from measured sample pH. Another
option for estimating the ionic states of all species in solution is to use an overall
charge balance. This method is described in Section 4.3 for use as part of the
kinetic modeling strategy. This calculation may also be used to determine the
ionic state of all species in solution from concentrations determined by HPLC
with some modification to system of equations in section 4.3.

To estimate ionic species the overall charge balance is expanded to

included all measured species in solution.

0=H+ +A+ +Alor +Am+ +EA+ +IPA+ +OH’ —A' -P' -2P2‘ -3P3' (30)

Where the symbols are as previously described in Section 4.3 with the addition
of Am+ as protonated ammonia, EA“ as protonated ethylamine, and IPA” as
protonated isopropyl amine. The charge balance now consists of 11 unknowns.
The species concentrations of ammonia, ethylamine, and isopropylamine are
defined similarly as alaninol is in section 4.3 by rearrangement of the Henderson-

Hasselbach equation.

122

There is now a system of 11 equations and 11 unknowns. This method is

solved in the same manner as discussed in section 4.3. The purpose of this

section is to compare the estimates of the fraction protonated from measured pH

values with the pH and fraction protonated from the overall charge balance. As

shown in Table 31, there is good agreement between the Henderson-Hasselbach

and charge balance method. Therefore, the charge balance could be used for

kinetic model development with confidence.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

10.15 M 0.29 M 0.58 M
Measured Calculated Measured Calculated Measured Calculated
to") pH CCOOH/Ct pH CCOOH/Ct DH Ccoon/Cr PH CCOOHlCtl pH CCOOH/Ct PH CCOOH/Ct
O 2.4 0.46 2.5 0.43 2.0 0.69 2.2 0.60 1.5 0.87 1.8 0.77
1 2.5 0.40 2.5 0.39 2.1 0.64 2.2 0.58 1.5 0.87 1.8 0.77
2 2.5 0.39 2.6 0.35 2.1 0.62 2.3 0.53 1.5 0.86 1.8 0.76
3 2.7 0.28 2.7 0.28 2.2 0.59 2.4 0.47 1.6 0.86 1.8 0.76
4 3.0 0.18 3.0 0.19 2.3 0.53 2.5 0.41 1.6 0.85 1.9 0.75
5 3.6 0.05 3.3 0.10 2.4 0.48 2.6 0.37 1.6 0.84 1.9 0.75
6 4.8 0.00 4.2 0.01 2.5 0.43 2.7 0.29 1.7 0.82 1.9 0.72

 

Table 31. Comparison of the estimated fraction of protonated alanine by use of the Henderson-
Hasselbach equation and the charge balance equations.

123

 

Appendix C. Excel program for kinetic modeling

The kinetics model derived for alanine hydrogenation, as discussed in
Chapter 4, involves a differential equation, the rate expression, which is
dependent on a highly non-linear algebraic equation, the charge balance. This
type of problem is not easily solved using the usual computational tools.
Therefore, a hybrid Runge Kutta method was developed and encoded for use in
the Visual Basic for Application environment of Excel. Since this represents a
novel computational technique and tool, the visual basic code and discussion of
the spreadsheet interface is presented.

C.1 Set Up of the spreadsheet

As shown in Figure 26, an Excel worksheet was used as the interface to
the macro that performs the calculations. This worksheet is broken into two
sections, one for entering independent variable information and the other for
entering dependent variables, constants, initial values, differential equation, and
charge balance.

In the independent variable section, the initial and final values of the
independent variable are entered. In this case the variable name given to time is
tau with intial and final values of 0 to 6. The number of steps is entered below as
100, which has proved adequate for this system. Also located in this box is the
“Solve” button, which initiates the macro, and the “Proof Formulas” button which
copies the formulas entered in the “Input” column into the “Display Formula”

column.

124

Figure 26. Kinetic parameter Set Up page

 

IEnter Independent Variable lnformatlon

 

 

 

 

 

 

 

 

 

 

 

 

I Name Inltlal Value Flnal Value
tau 0 6

INum of Steps 100

Step Size (calc'dl r 0.06

Number of Points in Summary 10

Enter Constants, Inltlal Values, Females, and Dlfierentlal Equations

Name Input D! I Formula

c.aO 0.226

c.ptot 0.291

c.h2 0.004

c.alol 0.000 =c.aO-c.a

KAI1 0.005

KAI2 0.000

KP1 0.007

KP2 0.000

KP3 0.000

KAlol 0.000

hplus 0.005

.p.H 2.276 =-LOG(hplue)

c.p1 0.124 =c.ptotl(1+KP1IhpIus+KP1'KPZ/hplus“2+KP1'KP2'KP3Ihplus“3)
c.p2 0.167 =c.ptot/(1+hplus/KP1+KP2Ihplus+KP3'KP2/hplus“2)
c.p3 0.000 =c.ptot/(1+KPan1plus+hpluleP2+hplusA2l(KPZ*KP1))
c.p4 0.000 =c.ptot/(1+hpluleP3+hplus/(KP2'KP3)+hplus*3/(KP1'KPZ‘KP3))
c.a 0.226

c.a1 0.121 =c.al(1+KAl1/hplus+KAl1'KAlZ/hplus*2)

c.a2 0.105 =c.al(1+hpluleAl1+KAl2/hplus+KAl2/hplus)

c.a3 0.000 =c.al(hplus*2l(KAl1‘KAl2)+hplus/KAI2+1)

temp 373.000

law 0.008

HP 1 1 1.310

H.Aplus 80.929

H.h2 68.777

E.s 81.526

k.s_373 0.124

k.Aplus_373 190.617

k.P_373 40.319

kh2_373 101.801

k.s 0.124 =k.s_373'EXP(E.s/laW‘(-1ltemp+1/373))

k.Aplus 190.617 =k.Aplus_373‘EXP(-H.Aplusllauf(1l373-1hemp»

k.P 40.319 =k.P__373'EXP(-H.Pllavr(1/373-1/temp))

k.h2 101.801 =k.h2_373'EXP(-H .h2/Iaw‘(1/373-1Itemp))

chargebal -0.040 =hplus+c.a1+hplus'c.aloll(KAlol+hplus)—(104-14Ihp|us+c.a3+c.p2+2‘c.p3+3‘c.p4)
d(c.a)/d(tau) -0.015 =-k.e‘k.Aplus'c.a1'k.h2‘c.h2/(1+kAplus'c.a1+k.P'c.p1)l(1+(k.h2‘c.h2)*0.5)"2
oonv 0.000 =1-c.alc.ao

 

 

 

 

125

 

In the section below the remainder of the problem is specified. In the first
column variable names and equation names are specified. In the second
column, the variable values are specified and equations entered. The third
column displays the formulas entered in the “Input” column. It should be noted
thatin addition to the code used in this program, the “accept labels” option under
the Tools—+Options—>Calculations menu. This handy option allows the user to
essentially use variable names without explicitly defining them. For example, the
first variable is the feed concentration of alanine and called c.aO. The numerical
value of c.a0 is entered into the adjacent cell. Now, the variable name “c.a0”
may be used in equations entered in the “Input” column instead of a cell
reference. The “Display Formula" column shows equations as entered into the
“Input” column.

There are several other worksheets in this program. The “Data”
worksheet contains the initial conditions and experimental data for the 20
experiments used for parameter determination. The output of the macro is
written to a different worksheet for each of the twenty experiments.

0.2 Description of calculation procedure

Calculations are performed in the following manner. For each of the 20
experiments used to determine the kinetic parameters, alanine concentration
versus time is predicted. These trajectories are compared to the concentration
data points from experimental works. The sum of the squared differences (SS)
between the predicted and 109 experimental points are calculated at the

conclusion of each run through the entire set experiments. The goal is then to

126

minimize the 88, which is done by random walk where at the beginning of each
set of calculations, one of the 8 kinetic parameters is randomly chosen and then
randomly increased or decreased by a specific percent. If at the conclusion of
the run the sum of squares is reduced, the parameter change is retained, if the
SS increases the original value is restored. If after 83 cycles the SS is not
reduced, the set of parameters have reached a minimum and the calculation is
terminated.

The mathematics involved in predicting these trajectories makes use of
the Runge Kutta method in conjunction with the Excel Solver add-in. Runge
Kutta is a predictor method where small time steps are taken and estimates of
the derivative at these points is used to predict the next value of the dependent
variable. In this particular case, between time steps the Solver add-in is called to
set the solution charge balance to zero.

In the next section, the code is provided as it appears in the program file.
The code contains comment lines which explain the function of various section of

code.

0.3 Visual Basic Code

Public rowdeq(1 To 10)

Public f(10) As Double 'holds derivative values
Public z(1 To 50) As Double 'Variable used to hold onto initial values
Public initial 'holds the initial value of the indep variable

Public columnvar

Public CurrentRow

Public rowvar(1 To 10) As Single
Public expnames(21) As Double
Sub Caller()

127

row1 = ActiveWorkbook.Names("startingpt").RefersToRange.Row + 1
column1 = ActiveWorkbook.Names(“startingpt").RefersToRange.Column

For i = 0 To 50

If (Not lsEmpty(Worksheets("Set Up").Cells(row1 + i, column1))) _
And (Worksheets(“Set Up").Cells(row1 + i, column1)) _

Like “d(*" Then

subcaller = Worksheets(“Set Up“).Cells(row1 + i, column1).Value
End If

Next i

For i = 0 To 50

If (Not lsEmpty(Worksheets("Set Up").Cells(row1 + i, column1))) _
And (Worksheets("Set Up“).Cells(row1 + i, column1)) _

Like “f(*" Then

subcaller = Worksheets(“Set Up“).Cells(row1 + i, column1).Value
End If

Next i

If subcaller Like “d(*" Then
Call RungeKutta4
End If

If subcaller Like "f *“Then

Dim solverpath As String

On Error Resume Next

If Application.lsNA(Workbooks("Solver.xla")).Name Then
solverpath = Application.LibraryPath & "\solver\solver.xla"
ThisWorkbook.VBProject.References.AddFromFile solverpath

End If

On Error GoTo 0
Call SolverSub
End If

End Sub

Sub FuncList(t, y, f, i)

temp = Worksheets("Set Up“).Cells(rowdeq(i), columnvar).Value
If (Not |sNumeric(temp)) Then

msg = "There was a problem with the calculation. "

msg = msg & “At least one derivative“

msg = msg & " has become non-numeric due to an error.“

msg = msg & " You will need to reinitialize all dependent variables after"
msg = msg & " you find the error. Be sure to reinitialize the“

msg = msg & " independent variable also.“

notice = MsgBox(msg, vbOKOnly, "Runtime Error")

128

End

End If

f(i) = Worksheets(“Set Up“).Cells(rowdeq(i), columnvar).Value
End Sub

Sub Addworksheets()

jump = 1

Fori = 1 To 19

expnames(i) = Worksheets(“Experiments").Cells(1, jump).Value
Sheets.Add.Name = expnames(i) 8. “Output“

jump = jump + 5

Next i

End Sub

Sub RungeKutta4()

'Call displayformula
Worksheets("Output").UsedRange.ClearContents
Worksheets("Summary').UsedRange.ClearContents

'Adapted using notation and methods described in

'Referenoe B. Camahan, H. Luther, J. Wilkes, “Applied Numerical

' Methods“, Wiley, New York, 1969.

Dim rowvar(1 To 50) 'Used to keep keep track of rows in loop below
Dim y(1 To 50) As Double 'Array variable used for transfering dependent
variable values

Dim k1(10) As Double 'k values are part of Runge Kutta calculation
Dim k2(10) As Double

Dim k3(10) As Double

Dim k4(10) As Double

Dim expnames(1 To 22) As Variant

Dim cao(22) As Double

Dim cpo(22) As Double

Dim ch2(22) As Double

Dim temperature(22) As Double

Dim parameters(9) As Double

walks = 125

' start number of random steps

initialave = Worksheets("Summary").Cells(27, 2).Value
frankjump = 0

For frank = 0 To walks

Worksheets("Set Up“).Range("h15“).Value = frank

For pmer = 1 To 9

parameters(pmer) = _

Worksheets(”Set Up“).Cells(30 + pmer, 8).Value

Next pmer

129

'Select cell to change randomly
Randomize

random1 = |nt((38 - 31 + 1) * Rnd + 31)
random2 = |nt((2 -1 + 1) * Rnd +1)

If random2 = 1 Then

Worksheets("Set Up“).Cells(random1, 8).Value = _
Worksheets(“Set Up”).Cells(random1, 8).Value * 1.005
Else

Worksheets(“Set Up“).Cells(random1, 8).Value = _
Worksheets(“Set Up“).CeIls(random1, 8).Value * 0.995
End If

For pmer = 1 To 8

Worksheets(“Set Up").Cells(30 + pmer, 2).Value = __
Worksheets("Set Up").Cells(30 + pmer, 8).Value
Next pmer

jump = 1

Fori = 1 To 21

expnames(i) = Worksheets(“Experiments“).Cells(1, jump).Value
cao(i) = Worksheets(”Experiments").Cells(6, jump + 1).Value

cpo(i) = Worksheets("Experiments").Cells(5, jump + 1).Value

ch2(i) = Worksheets("Experiments").Cells(4, jump + 1).Value
temperature(i) = Worksheets("Experiments“).Cells(2, jump + 1).Value

jump = jump + 5
Next i
runs = 20

For reaction = 1 To runs

Worksheets("Set Up“).Range(“h14").Value = reaction

If reaction = 5 Then

Worksheets(”Set Up").Cells(44, 2).Fon'nula = Worksheets(“Set Up").Cells(45,
7).Formula

End If

If reaction = 8 Then

Worksheets("Set Up”).Cells(44, 2).Formula = Worksheets("Set Up“).Cells(46,
7).Formula

End If

i=0

130

i=0

m = 0

Worksheets(“Set Up”).Range("b9").Value = cao(reaction)
Worksheets(“Set Up").Range("b25").Value = cao(reaction)
Worksheets(“Set Up“).Range(“b10").Value = cpo(reaction)
Worksheets("Set Up“).Range("b11").Value = ch2(reaction)
Worksheets(“Set Up“).Range("b29").Value = temperature(reaction)

'Puts the name of the independent variable into the Output Sheet
Worksheets(expnames(reaction) & " Output“).Cells(2, 1) = Worksheets(“Set
Up").Range("indepvar").Value

'Puts the initial value of the independent variable into the Output Sheet
Worksheets(expnames(reaction) 8 " Output“).Cells(3, 1) = Worksheets(“Set
Up”).Range("initialindep“).Value

'These variables used to keep track of the rows and columns of variables in Set
Up

row1 = ActiveWorkbook.Names("startingpt").RefersToRange.Row + 1

column1 = ActiveWorkbook.Names("startingpt").RefersToRange.Column

'columnvar holds the value of the column with variable values
columnvar = column1 + 1

'The following two nested loops find the initial values and names of the
'dependent variables

For i = 0 To 50

If (Not IsEmpty(VVorksheets("Set Up").Cells(row1 + i, column1))) _
And (Worksheets("Set Up").Cells(row1 + i, column1)) _

Like 'd(*" Then

'store the rows of the diffeqs

fl=fl+1
'check number of diffeqs
If (jj > 10) Then
notice = MsgBox("You have more than 10 Diff Eqns. The program can only solve
10.", vbOKOnly, "Error")
End
End If
rowdeq(jj) = row1 + i
temp = Worksheets(“Set Up").Cells(row1 + i, column1).Value
temp = Right(temp, Len(temp) - 2)
temp = Left(temp, InStr(temp, “)“))
temp = Left(temp, Len(temp) - 1)
For k = 0 To 50
If (Not |sEmpty(Worksheets(“Set Up“).Cells(row1 + k, column1))) _
And (Worksheets("Set Up”).Cells(row1 + k, column1)) _

131

Like temp Then
i =1 + 1
‘rowvar holds the rows that correspond to dependent variable derivatives.
'20) holds the initial value
rowvar(j) = row1 + k
20) = Worksheets(”Set Up“).Cells(row1 + k, column1 + 1)
Worksheets(expnames(reaction) & " Output“).Cells(3, 1 + j) = _
Worksheets("Set Up").Cells(row1 + k, column1 + 1)
Worksheets(expnames(reaction) & " Output“).Ce|ls(2, 1 + j) = _
Worksheets(“Set Up").Cells(row1 + k, column1).Value
End If
Next k
End If
Next i ‘
'check that initial values exist for all indep variables
If (jj <> j) Then
notice = MsgBox(“The program cannot continue because initial values were not
found for every dependent variable“, vbOKOnly, ”Error“)
End
End If

'This loop finds the location of the formulas
Dim rowform(1 To 50) As Single
For i = 0 To 50
If (Not Worksheets("Set Up“).Cells(row1 + i, column1).Value Like “d(*)") _
And Worksheets("Set Up“).Cells(row1 + i, column1 + 1).HasFormula Then
m=m+1
rowform(m) = row1 + i
Worksheets(expnames(reaction) & " Output").Cells(3, 1 +j + m) = _
Worksheets("Set Up“).Cells(row1 + i, column1 + 1).Value
Worksheets(expnames(reaction) & " Output").Cells(2, 1 +j + m) = _
Worksheets(“Set Up”).Cells(row1 + i, column1).Value
End If
Next i

'Formats Runge Kutta Sheet

Worksheets(expnames(reaction) & " Output").Range("A1") = expnames(reaction)
& " Output“

Worksheets(expnames(reaction) & “ Output").Range("A1“).HorizontalAlignment =
leeft

Worksheets(expnames(reaction) & " Output").Range("A1").Font.Bold = True
Worksheets(expnames(reaction) & " Output”).Range("A1“).Font.ltalic = True
Worksheets(expnames(reaction) & " Output").Range("A1").Font.Size = 12
Worksheets(expnames(reaction) & " Output“).Range("2:2“).HorizontalAlignment
= xlCenter

132

Worksheets(expnames(reaction) & " Output“).Range("2:2").Font.Bold = True
Dim CurrentRow

row1 = 3

row2 = row1 + 1

column1 = 1

Stoprow = row1 + Worksheets(”Set Up“).Range("numsteps")

' Sets up for main calculation loop

h = Worksheets(“Set Up“).Range("stepsize")

CurrentRow = row1

initial = Worksheets(”Set Up").Range("initialindep").Value
Worksheets("Output").Cells(3, 12).Value = Worksheets(“Set
Up“).Range("b25").Value

'Main calculation loop

'Solves inital charge balance

SolverReset
SolverOptions Scalingz=True

SolverOk SetCell:=Worksheets("Set Up").Range("b43"), MaxMinVal:=3,
ValueOf:=0, _

ByChange:=Worksheets(“Set Up”).Range("b19")

SolverAdd CellRef:=Worksheets("Set Up").Range("b19"), Relation:=3,
FonnulaText:=".000000001 "

SolverAdd CellRef:=Worksheets("Set Up").Range("b19"), Relation:=1,
FormulaText:=".99"

SolverSolve UserFinishz=True

SolverDelete CellRef:=Worksheets(“Set Up“).Range(“b19"), Relation:=3,
ForrnulaText:=".000000001 “

SolverDelete CellRef:=Worksheets("Set Up“).Range("b19"), Relation:=1,
FormulaText:=".99"

Do While CurrentRow < Stoprow

t = Worksheets(expnames(reaction) & " Output").Cells(CurrentRow,
column1).Value

Worksheets(“Set Up").Range("initialindep").Value = t

Fori = 1 Toj

y(i) = Worksheets(expnames(reaction) & " Output").Cells(CurrentRow, column1 +
i).Value

Next i

For i = 1 To j
Worksheets("Set Up").Cells(rowvar(i), columnvar) = y(i)

133

Next i

SolverReset
SolverOptions Scaling:=True

SolverOk SetCell:=Worksheets("Set Up“).Range(“b43"), MaxMinVal:=3,
ValueOf:=0, _

ByChange:=Worksheets("Set Up“).Range("b19")

SolverAdd CellRef:=Worksheets("Set Up“).Range("b19"), Relation:=3,
FormulaText:=“.000000001 "

SolverAdd CellRef:=Worksheets(“Set Up“).Range("b19"), Relation:=1,
FonnulaText:=".99“

SolverSolve UserFinishz=True

SolverDelete CellRef:=Worksheets("Set Up”).Range("b19"), Relation:=3,
FormulaText:=“.000000001 "

SolverDelete CellRef:=Worksheets("Set Up").Range("b19"), Relation:=1,
FonnulaText:=".99"

Fori = 1 To m
Worksheets(expnames(reaction) 8 " Output").CelIs(CurrentRow, column1 +j + i)

Worksheets("Set Up“).Cells(rowform(i), 2).Value
Next i

Fori = 1 To j

FuncList t, y, k1, i

FuncList t + h / 2, y(i) + h * k1 (i) I 2, k2, i
FuncList t + h / 2, y(i) + h * k2(i) I 2, k3, i
FuncList t + h, y(i) + h * k3(i), k4, i

Next i

CurrentRow = CurrentRow + 1

Worksheets(expnames(reaction) & " Output“).Cells(CurrentRow, column1).Value
=t+h

Worksheets(“Set Up“).Range("initialindep").Value = t + h

Fori = 1 To m
Worksheets(expnames(reaction) & " Output").Cells(CurrentRow, column1 +j + i)

Worksheets(“Set Up").Cells(rowform(i), 2).Value
Next i

134

Worksheets(expnames(reaction) 8 " Output“).Cells(CurrentRow, column1 + 1) =
y(1) + h / 6 * (k1(1) + 2 * k2(1) + 2 * k3(1) + k4(1))

Loop

'End of main calculation loop

'Adds final values of dependent variables to Runge Kutta Sheet

For i = 1 Toj

y(i) = Worksheets(expnames(reaction) 8 " Output“).Cells(Stoprow, column1 +
i).Value

Next i

For i = 1 Toj

Worksheets(“Set Up“).Cells(rowvar(i), columnvar) = y(i)

Next i

Fori = 1 To m

Worksheets(expnames(reaction) 8 " Output“).Cells(Stoprow, column1 + j + i) = _
Worksheets(“Set Up“).Cells(rowfonn(i), 2).Value
Next i

'Resets initial values

Fori = 1 Toj

Worksheets(“Set Up”).Cells(rowvar(i), columnvar) = z(i)
Next i

Worksheets(“Set Up").Range("initialindep") = initial
Worksheets(“Set Up“).Activate

Next reaction

'Graphs Values
jump2 = 1

Fori = 1 To 20

expnames(i) = Worksheets(“Experiments”).Cel|s(1,jump2).Value
'cao(i) = Worksheets(“Experiments").Cells(6, jump2 + 1).Value

'cpo(i) = Worksheets("Experiments").Cells(5, jump2 + 1).Value

'ch2(i) = Worksheets("Experiments").Cells(4, jump2 + 1).Value
temperature(i) = Worksheets("Experiments“).Cells(2, jump2 + 1).Value

jump2 = jump2 + 5

Next i
runs2 = 20
jump2 = 1

For reaction2 = 1 To runs2

135

'Clears previous chart values
Charts(expnames(reaction2) 8 " Plot“).ChartArea.ClearContents

'Adds dependent variable series to chart

Worksheets(expnames(reaction2) 8 " Output“).Activate

With Charts(expnames(reaction2) 8 " Plot”).SeriesCollection.NewSeries
.Name = ”Predicted“

.Values = Worksheets(expnames(reaction2) 8 " Output“).Range(Cells(3, 2),
Cells(1003, 2))

.XValues = Worksheets(expnames(reaction2) 8 " Output“).Range(Cells(3, 1),
Cells(1003, 1))

End With

Worksheets(“Experiments”).Activate

With Charts(expnames(reaction2) 8 " Plot").SeriesCollection.NewSeries

.Name = “Data“

.Values = Wod<sheets("Experiments").Range(Cells(9, jump2 + 1), Cells(17,
iumpZ + 1))

.XValues = Worksheets("Experiments“).Range(Cells(9, jump2), Cells(17, jump2))
End With

Wod<sheets("Experiments“).Activate

With Charts(expnames(reaction2) 8 " Plot').SeriesCollection.NewSeries

.Name = ”Fit“

.Values = Worksheets(“Experiments“).Range(Cells(9, jump2 + 2), Cells(17,
iumpZ + 2))

.XValues = Worksheets(”Experiments").Range(Cells(9, jump2), Cells(17, jump2))
End With

'Gives X axis name of independent variable and sets to the range to
'be the initial to final value

If Worksheets(“Set Up").Range("finalindep").Value > _
Worksheets("Set Up“).Range(“initialindep").Value Then

minx = Worksheets("Set Up“).Range("initia|indep“).Value

ma)o< = Worksheets(“Set Up").Range("finalindep").Value

Else

minx = Worksheets("Set Up").Range("finalindep").Value

maxx = Worksheets(“Set Up“).Range("initialindep").Value

End If

With Charts(expnames(reaction2) 8 " Plot“).Axes(xlCategory)
.HasTitle = True
.AxisTitle.Caption = "t(hr)"
.MinimumScale = minx

136

.MaximumScale = maxx
End With

With Charts(expnames(reaction2) 8 " Plot").Axes(lealue)
.HasTitle = True
.AxisTitle.Caption = "C.a(M)"
.MinimumScale = 0

End With

With Charts(expnames(reaction2) 8 " Plo ")
.HasTitle = True -
.ChartTitIe.Text = expnames(reaction2)

End With

jump2 = jump2 + 5
Next reaction2
jump3 = 1

Fori = 1 To 20

expnames(i) = Worksheets(”Experiments“).Cells(1, jump3).Value
cao(i) = Worksheets(“Experiments“).Cells(6, jump3 + 1).Value

cpo(i) = Worksheets(“Experiments").Cells(5, jump3 + 1).Value

ch2(i) = Worksheets("Experiments").Cells(4, jump3 + 1).Value
temperature(i) = Worksheets(“Experiments").Cells(2, jump + 1).Value

jump3 = jump3 + 5

Next i
runs3 = 20
jump3 = 1

For reaction3 = 1 To runs3

For p = 0 To 9

If lsEmpty(VVorksheets("Experiments").Cells(8 + p, jump3).Value) Then
Exit For

End If

For q = 0 To 1000

If Worksheets(expnames(reaction3) 8 " Output").Cells(3 + q, 1).Value >= _
Worksheets(”Experiments').Cells(8 + p, jump3).Value Then

137

Worksheets("Experiments“).Cells(8 + p, jump3 + 3).Value = _
Worksheets(expnames(reaction3) 8 " Output").Cells(3 + q, 2).Value
Exit For

End If

Next q

Next p

jump3 = jump3 + 5

Next reaction3
Worksheets(“Set Up“).Activate
jump4 = 5

Fori = 1 To 20

expnames(i) = Worksheets("Experiments“).Cells(1, jump4 - 4).Value
Worksheets(“Summary").Cells(9 + i, 1 + frankjump).Value = _
Worksheets("Experiments").Cells(1, jump4).Value
Worksheets("Summary').Cells(9 + i, 2 + frankjump).Value = _
Worksheets("Experiments").Cells(18, jump4).Value

jump4 = jump4 + 5

Next i

For k = 0 To 7

Worksheets(“Summary").Cells(2 + k, frankjump + 1).Value = _
Worksheets("SetUp").Cells(31 + k, 1).Value
Worksheets(“Summary").Cells(2 + k, frankjump + 2).Value = _
Worksheets(“Set Up“).Cells(31 + k, 2).Value

Next k

Worksheets(”Summary").Cells(27, frankjump + 1).Value = “SSq”
Worksheets("Summary").Cells(27, frankjump + 2).Value = _
Worksheets("Experiments").Cells(18, 2).Value
Worksheets("Summary").Cells(28, frankjump + 1).Value = "Desired”
Worksheets("Summary").Cells(28, frankjump + 2).Value = _
Worksheets(“Experiments").Cells(20, 2).Value

If parameters(9) < _
Worksheets("Experiments').Cells(18, 2).Value Then
For pmer = 1 To 9

Worksheets("Set Up“).Cel|s(30 + pmer, 8).Value = _

138

parameters(pmer)
Next pmer
End If

If parameters(9) > _
Worksheets(”Experiments“).Cells(1 8, 2).Value Then
Worksheets("Set Up“).Cells(39, 8).Value = _
Worksheets(“Experiments“).Cel|s(1 8, 2).Value

End If

Worksheets(”Set Up”).Cells(40, 8).Value = _
Worksheets(”Experiments“).Cells(1 9, 2).Value
Worksheets("Set Up").Activate

frankjump = frankjump + 2

Next frank

End Sub

Sub Grapher()

Dim expnames(1 To 19) As Variant
Dim cao(19) As Double

Dim cpo(19) As Double

Dim ch2(19) As Double

Dim temperature(19) As Double

'Get names, put in array, add sheets
jump2 = 1

Fori = 1 To 19

expnames(i) = Worksheets("Experiments").Cells(1, jump2).Value
cao(i) = Worksheets("Experiments“).Cells(6, jump2 + 1).Value

cpo(i) = Worksheets("Experiments").Cells(5, jump2 + 1).Value

ch2(i) = Worksheets("Experiments").Cells(4, jump2 + 1).Value
temperature(i) = Worksheets("Experiments“).Cells(2, jump2 + 1).Value

jump2 = jump2 + 5

Next i
mns2 = 19
jump2 = 1

For reaction2 = 1 To mns2

'Clears previous chart values
Charts(expnames(reaction2) 8 " Plot").ChartArea.ClearContents

'Adds dependent variable series to chart

Worksheets(expnames(reaction2) 8 " Output").Activate
With Charts(expnames(reaction2) 8 " Plot").SeriesCol|ection.NewSen‘es

139

.Name = “Predicted“

.Values = Worksheets(expnames(reaction2) 8 " Output“).Range(Cells(3, 2),
Cells(1003, 2))

.XValues = Worksheets(expnames(reaction2) 8 " Output“).Range(Cells(3, 1),
CellS(1003, 1))

End With

Worksheets(”Experiments“).Activate

With Charts(expnames(reaction2) 8 " Plot').SeriesCollection.NewSeries

.Name = “Data”

.Values = Worksheets(“Experiments“).Range(Cells(9, jump2 + 1), Cells(17,
jump2 + 1))

.XValues = Worksheets("Experiments“).Range(Cells(9, jump2), Cells(17, jump2))
End With

Worksheets(“Experiments").Activate

With Charts(expnames(reaction2) 8 " Plot“).SeriesCollection.NewSeries

.Name = "Fit”

.Values = Worksheets("Experiments“).Range(Cells(9, jump2 + 2), Cells(17,
iumpZ + 2))

.XValues = Worksheets("Experiments“).Range(Cells(9, jump2), Cells(17, jump2))
End With

'Gives X axis name of independent variable and sets to the range to
'be the initial to final value

If Worksheets("Set Up").Range("finalindep").Value > _
Worksheets("Set Up“).Range("initialindep").Value Then

minx = Worksheets("Set Up").Range("initialindep“).Value

maxx = Worksheets("Set Up").Range("finalindep").Value

Else

minx = Worksheets("Set Up”).Range("finalindep“).Value

maxx = Worksheets("Set Up").Range("initialindep").Value

End If

With Charts(expnames(reaction2) 8 " Plot“).Axes(xlCategory)
.HasTitle = True
.AxisTitle.Caption = "t(hr)"
.MinimumScale = minx
.MaximumScale = maxx
End With

With Charts(expnames(reaction2) 8 " Plot").Axes(lealue)
.HasTitle = True
.AxisTitle.Caption = “C.a(M)"
.MinimumScale = 0

140

End With

With Charts(expnames(reaction2) 8 " Plot")
.HasTitle = True
.ChartTitIe.Text = expnames(reaction2)
End With

jump2 = jump2 + 5

Next reaction2

End Sub

Sub Fitter()

Dim expnames(1 To 19) As Variant
Dim cao(19) As Double

Dim cpo(19) As Double

Dim ch2(19) As Double

Dim temperature(19) As Double

'Get names, put in array, add sheets
jump3 = 1

For i = 1 To 19

expnames(i) = Worksheets(“Experiments”).Cells(1,jump3).Value
cao(i) = Worksheets("Experiments”).Cells(6, jump3 + 1).Value

cpo(i) = Worksheets(“Experiments“).Cells(5, jump3 + 1).Value

ch2(i) = Worksheets("Experiments").Cells(4, jump3 + 1).Value
temperature(i) = Worksheets(“Experiments“).Cells(2, jump + 1).Value

jump3 = jump3 + 5

Next i
runs3 = 19
jump3 = 1

For reaction3 = 1 To runs3
For p = 0 To 9
If IsEmpty(VVorksheets("Experiments").Cells(8 + p, jump3).Value) Then

Exit For
End If

141

Forq = 0 To 1000

If Worksheets(expnames(reaction3) 8 " Output“).Cells(3 + q, 1).Value >= _
Worksheets(”Experiments").Cells(8 + p, jump3).Value Then

Worksheets(“Experiments“).Cells(8 + p, jump3 + 3).Value = _
Worksheets(expnames(reaction3) 8 " Output").Cells(3 + q, 2).Value
Exit For

End If
Next q
Next p

jump3 = jump3 + 5
frankjump = frankjump + 2
Next reaction3

End Sub

Sub SolverSub()

Call displayformula
Worksheets("Output").UsedRange.ClearContents
Worksheets(“Summary").UsedRange.ClearContents

'These variables used to keep track of the rows and columns of variables in Set
Up

row1 = ActiveWorkbook.Names("startingpt").RefersToRange.Row + 1

column1 = ActiveWorkbook.Names("startingpt").RefersToRange.Column

'columnvar holds the value of the column with variable values
columnvar = column1 + 1

'The following two nested loops find the initial values and names of the
'dependent variables

For i = 0 To 50

If (Not |sEmpty(Worksheets("Set Up").Ce|ls(row1 + i, column1))) _
And (Worksheets(“Set Up").Cells(row1 + i, column1)) _

Like "f(*" Then

'store the rows of the diffeqs

fi=fi+1

'check number of diffeqs

lfjj = 1 Then

142

balance1 = Cells(row1 + i, column1 + 1).Address
End If

If jj = 2 Then
balances = balances 8 Cells(row1 + i, column1 + 1).Address 8 “2“
End If

If (jj > 10) Then

notice = MsgBox("You have more than 10 Diff Eqns. The program can only solve
10.", vbOKOnly, "Error")

End

End If

rowdeq(jj) = row1 + i
temp = Worksheets("Set Up").Cells(row1 + i, column1).Value
temp = Right(temp, Len(temp) - 2)
'temp = Left(temp, InStr(temp, ")"))
temp = Left(temp, Len(temp) - 1)
For k = 0 To 50
If (Not lsEmpty(Worksheets("Set Up“).Cells(row1 + k, column1))) _
And (Worksheets("Set Up").Cells(row1 + k, column1)) _
Like temp Then
i =1 + 1
'rowunk(j) = row1 + k
varied = varied 8 Cells(row1 + k, column1 + 1).Address 8 ","

' rowvar holds the rows that correspond to dependent variable derivatives.
'20) holds the initial value

rowvar(j) = row1 + k

20) = Worksheets(“Set Up").Cells(row1 + k, column1 + 1)

' Worksheets(“Output").Cells(3, 1 +j) =_

' Worksheets("Set Up“).Cells(row1 + k, column1 + 1)
Worksheets("Output").Cells(2, 1 + j) = _
Worksheets(”Set Up”).Cells(row1 + k, column1).Value
End If
Next k

End If
Nexti

'check that initial values exist for all indep variables

If (jj <> j) Then

notice = MsgBox("The program cannot continue because initial values were not
found for every dependent variable“, vbOKOnly, ”Error")

End

143

End If

'This loop finds the location of the formulas
Dim rowform(1 To 50) As Single
For i = 0 To 50
If (Not Worksheets(“Set Up").Cells(row1 + i, column1).Value Like "d(*)") __
And Worksheets(“Set Up“).Cells(row1 + i, column1 + 1).HasFormula Then
m=m+1
rowfomn(m) = row1 + i
' Worksheets(“Output").Cells(3, 1 +j + m) = _
' Worksheets(“Set Up“).Cells(row1 + i, column1 + 1).Value
Worksheets(“Output").Cells(2, 1 +j + m) = _
Worksheets(“Set Up“).Cells(row1 + i, column1).Value
End If
Next i

'Formats Runge Kutta Sheet

Worksheets(“Output").Range(“A1 ") = “Solver Output“
Worksheets(“Output").Range(“A1“).HorizontalAlignment = leeft
Worksheets(“Output”).Range("A1“).FontBold = True
Worksheets(“Output").Range(“A1“).FontJtalic = True
Worksheets("Output").Range(“A1”).Font.Size = 12
Worksheets("Output").Range(“2:2“).HorizontalAlignment = xlCenter
Worksheets("Output").Range("2:2”).Font.Bold = True ‘

Dim CurrentRow

'row1 = ActiveWorkbook.Names("indep0").RefersToRange.Row
'row2 = row1 + 1

'column1 = ActiveWorkbook.Names("indep0").RefersToRange.Column
'stoprow = row1 + Worksheets("Set Up“).Range("numsteps")

' Sets up for main calculation loop

Range(balance1 ).Select
bal1 = Selection.Address
ActiveSheet.Names.Add Name:="bal1", RefersTo:="=“ 8 bal1

balances = balances 8 Cells(rowdeqfij), column1 + 1).Address 8 “2"
balances = Left(balances, Len(balances) - 1)
Range(balances).Select

bals = Selection .Address

ActiveSheet.Names.Add Name:="bals", RefersTo:="=" 8 bals

varied = Left(varied, Len(varied) - 1)

Range(varied).Select
vary = Selection.Address

144

ActiveSheet.Names.Add Name:="vary", RefersTo:="=" 8 vary
final = Worksheets(“Set Up").Range(“finalindep“).Value
'Main calculation loop

CurrentRow = 3

h = Worksheets(”Set Up“).Range("stepsize")

initial = Worksheets(“Set Up“).Range(“initialindep“).Value
initial2 = Worksheets(“Set Up").Range(“initialindep").Value
Stoprow = 4 + Worksheets(“Set Up").Range("numsteps")
initial = Worksheets("Set Up").Range(“initialindep“).Value
initial2 = Worksheets(“Set Up").Range(“initialindep").Value

If Not (lsEmpty(VVorksheets("Set Up“).Range("indepvar"))) Then

Do While CurrentRow < Stoprow

SolverReset

SolverOptions Scaling:=True
SolverOk SetCell:=Range("bal1"), MaxMinVal:=3, ValueOf:=0, _
ByChange:=Range("vary")

'ByChange:="c.h, c.x, c.h"
SolverAdd CellRef:=Range("bals"), Relation:=2, FormulaText:="0"
'SolverAdd CellRef:=Range(“$b$4:$b$6“), Relation:=3, FormulaText:="0"
SolverSolve UserFinishz=True

SolverDelete CellRef:=Range("bals"), Relation:=2, FormulaText:="0"
'SolverDelete CellRef:=Range(“$b$4:$b$6"), Relation:=3, ForrnulaText:="0"
Worksheets(“Output").Cells(CurrentRow, 1).Value = _
Worksheets("Set Up”).Range("initialindep").Value

For i = 1 Toj

Worksheets(”Output").Cells(CurrentRow, 1 + i).Value = _
Worksheets("Set Up“).Cells(rowvar(i), 2).Value

Next i

For i = 1 To m

Worksheets(“Output").Cells(CurrentRow, 1 +j + i).Value = _
Worksheets(”Set Up“).Cells(rowform(i), 2).Value

Next i

CurrentRow = CurrentRow + 1

Worksheets(“Set Up”).Range(“initialindep").Value = _
Worksheets(“Set Up”).Range(“initialindep").Value + h

Loop

145

Worksheets("Set Up").Range(“initialindep").Value = initial2
Fori = 1 Toj

Worksheets(“Set Up“).Cells(rowvar(i), 2).Value = z(i)
Next i

'Summary

Worksheets(“Summary“).Range(“A1 ") = “Summary of Output“
Worksheets("Summary").Range(“A1“).HorizontalAlignment = leeft
Worksheets(“Summary").Range(“A1”).FontBold = True
Worksheets(“Summary').Range(“A1“).Font.ltalic = True
Worksheets(“Summary").Range(“A1“).Font.Size = 12

'Resets initial values

Fori = 1 Toj

Worksheets("Set Up“).Cells(rowvar(i), columnvar) = z(i)
Next i

Worksheets("Set Up“).Range(“initialindep") = initial2

'Creates Summary Table

summarypts = Worksheets(“Set Up“).Range(“summarypts").Value
For i = 1 To summarypts

Worksheets(“Summary").Cells(2, i) = _
Worksheets("Output").Cells(2, i)

Worksheets("Summary").Cells(2, i).Font.Bold = True
Worksheets("Summary").Cells(2, i).HorizontalAlignment = xlCenter
Next i

step = Worksheets("Set Up").Range(“numsteps").Value

For i = 0 To summarypts

Worksheets("Output").Activate
Worksheets(“Output").Range(Cells(3 + step / summarypts * i, 1), Cells(3 + step /
summarypts _

* i, 20)).Copy Worksheets("Summarfl.Cells(3 + i, 1)

Next i

End If

lf (lsEmpty(Worksheets("Set Up“).Range("indepvar"))) Then

SolverReset

SolverOptions Scaling:=True
SolverOk SetCell:=Range("ba|1"), MaxMinVal:=3, ValueOf:=0, _
ByChange:=Range("vary")

'ByChange:="c.h, c.x, c.h"
SolverAdd CellRef:=Range("bals"), Relation:=2, FormulaText:="0"

146

'SolverAdd CellRef:=Range("$b$4:$b$6"), Relation:=3, FormulaText:="0"
SolverSolve UserFinishz=True

SolverDelete CellRef:=Range(“bals"), Relation:=2, FormulaText:="0“
'SolverDelete CellRef:=Range("$b$4:$b$6"), Relation:=3, FormulaText:="0"
End If

'Clears previous chart values
'Charts("RK PIot").ChartArea.ClearContents

'Adds dependent variable series to chart

'For i = 1 Toj

With Charts(“RK Plot“).SeriesCollection.NewSeries

'.Name = Worksheets(”Output").Cells(2, 1 + i)

'.Values = Worksheets(“Output").Range(Cells(3, 1 + i), Cells(3 + stoprow, 1 + i))
'.XValues = Worksheets(“Output").Range(Cells(3, 1), Cells(3 + stoprow, 1))
'End With

'Next i

'Gives X axis name of independent variable and sets to the range to

'be the initial to final value

'If Worksheets("Set Up“).Range("finalindep").Value > _

Worksheets(“Set Up“).Range(“initialindep").Value Then

'minx = Worksheets("Set Up").Range("initialindep“).Value

'maxx = Worksheets(“Set Up").Range(finalindep").Value

'Else

'minx = Worksheets(“Set Up”).Range(“finalindep").Value

'ma>o( = Worksheets("Set Up").Range(“initialindep").Value

'End If

With Charts(“RK Plot").Axes(xlCategory)

' .HasTitle = True
' .AxisTitle.Caption = Won<sheets("Output").Cells(2, 1).Value
' .MinimumScale = minx

' .MaximumScale = maxx

'End With

Worksheets(”Set Up“).Activate

End Sub

'Displays formulas in Set Up Sheet

Sub displayformula()

Worksheets(“Output").UsedRange.ClearContents

row1 = ActiveWorkbook.Names("startingpt").RefersToRange.Row + 1
column1 = ActiveWorkbook.Names("startingpt").RefersToRange.Column
For i = 0 To 50

If Worksheets("Set Up").Cells(row1 + i, column1 + 1).HasFormula _

147

Then

Worksheets(”Set Up”).Cells(row1 + i, column1 + 2) = ""' 8 Cells(row1 + i,
column1 + 1).Formula

Else

Worksheets(”Set Up“).Cells(row1 + i, column1 + 2) = ""

End If

Next i

End Sub

148

REFERENCES

1 Grewal, H.; Kalra, K. Biotechnol. Adv. 1995, 13,209.
2 Pazouki, M.; Panda, T. Bioprocess Eng. 1999, 19,435.

3 Milson, P. E.; Meers, J. L. Gluconic and ltaconic Acids. Comprehensive
Biotechnology, Moo-Young, M.; Blanch, H. W., Drew, H. W., Wang, D. I. C., Eds.;
Pergamon Press: New York, 1985; Vol. 3, Chapter 35.

4 Boyaval, P.; Corre, c. Left 1995, 75,453.
5 Jin, z. w.; Yang, s. T. Biotechnol. Prog. 1993, 14, 457.
6 Rehberger, J. L.; Glatz, B. A. J. Food Prot. 1998, 61 , 211.

7 Vandak, o.; Zigova, .1.; Sturdik, E.; Schlosser, 3. Process Biochem. 1997, 32,
245.

8 Hwang, S.; Hansen, C. L. Biotechnol. Bioeng. 1997, 54,451.

9 Du, J. X.; Cao, N. J.; Gong, C. S.; Tsao, G. T.; Yuan, N. J. Appl. Biochem.
Biotechnol. 1997, 54, 451.

1° Datta, R.; Tsai, S. P. ACS. Symp. Ser. #666, 1997, 224.
‘1 Gorkam, R. R.; Eitemann, M. A., Sridhar, J. ACS Sypm. Ser. #6661997, 237.

‘2 Nghiem, N. P.; Davison, B. H.; Richardson, G. R. Appl. Biochem. Biotechnol.
1997, 63-65, 565.

‘3 Anastas, P. T.; Warner, J. C. Green chemistry: theory and practice; Oxford
University Press: New York, NY, 1998.

1‘ http:I/www.chemicalvision2020.orgltechroadmaps.html

‘5 Amino Acids. Ulmann’s Encylopedia of Industrial Chemistry, Kadheinz, D.;
Berd, H.; Kleeman, A.; Krimmer, H.; Leuchtenberger, W.; Weckbecker, C., 7'".
ed.; Wiley: New York, 2002.

‘6 Furui, M.; Yamashita, K. J. Ferment. Technol. 1983, 61 , 587.

‘7 Hashimoto, S.; Katsumata, R. Proc. Ann. Meet. (Agrfc. Chem. Soc. Jpn.),
1994, 341 .

‘8 TenBrink, R. E. J. Org. Chem. 1987, 52, 418.

149

‘9 Nicollaides, E. D.; Tinney, F. J.; Kaltenbronn, J. S; Repine, J. T.; DeJohn, D.
A.; Lunney, E.A.; Roark, W. H.; Marriot, J. G.; Davis, R. E.; Voigtmen, R. E. J.
Med. Chem. 1986, 29,959.

2° Fincham, C. l.; Higginbottom, M; Hill, D. R.; Horwell, D. C.; O'Toole, J. C.;
Ratcliffe, G. S.; Rees, D. C.; Roberts, E. J. Med. Chem. 1992, 35, 1472.

2‘ Auvin-Guette, C; Rebuffat, s.; Prigent, Y; Bodo, B. J. Am. Chem. Soc. 1992,
114.2170.

22 Wu, 3.; Takeya, R.; Eto, M.; Tomizawa, c. J. Pestic. Sci. 1987, 12,221.

23 Coppola, G. M.; Schuster, H. F. Asymmetric Synthesis of Chiral Molecules
using Amino Acids; Wiley-lnterscience: New York, 1987.

2‘ Ager. o. .1.; Prakash, I; Schaad, o. R. Chem. Rev. 1996, 96,835.
25 Rouhi, A. M. Chemical and Engineering News 2003, 81 (18), 45-55.

26 Karrer, P.; Karrer, W.; Thomann, H.; Harlacher, E.; Mader, W. Helv. Chim.
Acta. 1921, 4, 76.

27 Abiko, A.; Masamune, S. Tetrahedron Lett. 1992, 33, 5517.
2" McKennon, M. J.; Meyers, A. l. J. Org. Chem. 1993, 58, 3568.

29 Meyers, A. 1.; Dickman, o. A.; Bailey, T. R. J. Am. Chem. Soc. 1985, 107,
7974.

3° Dickman, o. A.; Meyers, A. 1.; Smith, G. A.; Gawley, R. E. Org. Synth. 1990,
Coll Vol. 7, 530.

3‘ Evans, D. A.; Takacs, J. M.; McGee, L. R.; Ennis, M. D.; Mathre, D. J.; Bartroli,
J. Pure Appl. Chem. 1981 , 53, 1 109.

32 Evans, 0. A.; Bartroli, .1.; Shih, T. L. J. Am. Chem. Soc. 1981, 103, 2127.
33 Evans, 0. A.; Ennis, M. 0.; Mathre, o. J. J. Am. Chem. Soc. 1982, 104, 1737.
3‘ Evans, D. A.; Nelson, J. V.; Taber, T. R. Top. Stereochem. 1982, 13, 1.

35 Lane, C. F.; Myatt, H. L.; Daniels, J.; Hopps, H. B. J. Org. Chem. 1974, 40,
3527.

3’6 Poindexter, G. S.; Meyers, A. l. Tetrahedron Lett. 1977,3527.
3’ Smith, G. A.; Gawley, R. E. oig. Synth. 1985,63, 136.
38 Brown, H. C.; Choi, Y. M.; Narasimhan, S. J. Org. Chem. 1982, 47, 3153.

150

39 Gage, J. R.; Evans, 0. A. Org. Synth. 1990, 68,77.

4° Nicolas, E.; Russell, K. c.; Hruby, v. J. J. Org. Chem. 1993, 58,766.

4‘ Karrer, P.; Portmann, P.; Suter, M. Helv. Chim.Acta1949, 32, 1156.

‘2 Karrer, P.; Naik, A. R. Helv. Chim. Acta 1948, 31, 1617.

‘3 Seki, H.; Koga, K.; Matsuo, H.; Yamada, S. Chem. Pharrn. Bull, 1965, 13,995.

4‘ Ager, D., ed., Handbook of Chiral Chemicals, Marcel Dekker, Inc., New York,
N.Y.,1999.

‘5 Adkins, H.; Pavlic, A. A. J. Am. Chem. Soc. 1947, 69,3040.

‘6 Studer, M., Burkhardt, S., Blaser, H. U. Adv. Synth. Catal. 2001, 343(8), 802.
‘7 Antons, s.; Beitzke, B. US. Patent 5,536,879, 1996.

‘8 Antons, S., Tilling, A., Wolters, E. US. Patent 6,310,254, 2001.

‘9 Bowden, E., Adkins, H. J. Am. Chem. Soc. 1934, 56, 689.

5° Carhnahan, J. E.; Ford, T. A.; Gresham, W. F.; Grigsby, W. E.; Hager, G. F. J.
Am. Chem. Soc. 1955, 77, 3766.

5‘ Broadbent, H. s.; Campbell, G. c.; Bartley, w. J.; Johnson, J. H. J. Am. Chem.
Soc. 1959, 24, 1847.

52 Zhang, Z. 6., Jackson, J. E., Miller, D. J. Appl. Catal. A-Gen. 2001, 219, 89.
53 Zhang, Z. 6., Ph.D. Dissertation, Michigan State University, 2000.
5‘ Antons, S.; Tilling, A. S.; Wolters, E. US. Patent 6,355,848, 2002.

55 Cortright, R. D.; Sanchez-Castillo, M.; Dumesic J. A. Applied Catal B: Env.
2002, 39, 353.

56 Gal, J. J. Liquid Chromatography 1986 9, (2-3), 673.

5’ Bada, J. L. Annu. Rev. Earth Planet Sci. 1985, 13,241.

58 Harsey, s. c., u. S. Patent 4,990,666, 1991.

59 Zwietering, T. H. Chem Eng. Sci. 1958, 8, 244.

6° Yagi, H., Yoshida, F. Ind. Eng. Chem. Proc. Des. Dev. 14, 488, 1975.

6‘ Bern, L., Lidefelt, J. O., and Schoon, N. H. J. Am. Chem. Soc. 53,463, 1976.

151

62 Boon-Long, S., Laguerie, C., and Couderc, J. P. Chem. Eng. Sci. 1978, 33,
813,

63 Weiss, P. B.; Prater, C. D. Adv. Catal. 1954, 6, 143.
6‘ Singh, U. K.; Vannice, M.A. Applied Catalysis A: General 2001, 213, 1.
65 Pallassana, V., Neurock, M. J. Catal. 2002, 209, 289.

66 Belohlav, Z.; Zamostny, P.; Kluson, P.; Volf, J. Can. J. Chem. Eng. 1997, 75,
735.

6’ Shvo, Y.; Thomas, o. w.; Laine, R. M. J. Am. Chem. Soc. 1981, 103, 2461.

68 Marahashi, S. l.; Yoshimura, N.; Tsumiyama, T.; Kojima, T. J. Am. Chem.
300.1983, 105, 5002.

69 Wilson, R. B.; Laine, R. M. J. Am. Chem. Soc. 1985, 107,361.
7° Doye, S. Angew. Chem, int. Ed. 2001, 40, 3351.

7‘ Chatani, N.; Asaumi, T.; Ikeda, T.; Yoritsu, S.; lshii, Y.; Kakiuchi, F.; Marai, S.
J. Am. Chem. SOC. 2000, 122, 12882.

72 Sakaguchi, S.; Kubo, T.; lshii, Y. Angew. Chem, Int. Ed. 2001, 40, 2534.

73 Chatani, N.; Asaumi, T.; Yorimitsu, S.; Ikeda, T.; Kakiuchi, F.; Murai, S. J. Am.
Chem. Soc. 2001, 123, 10935.

7" Ramalingam, K.; Najappan, P.; Kalvin, D. M.; Woodard, R. W. Tetrahedron
1992, 44, 5597.

75 Pearce, D. A., Hammershoi, A., Harrowfield, J. M., Sargeson, A. M. Chem.
Com. 2000,2431.

76 Maeda, M. Ogawa, O., Kawazoe, Y. Chem. Pharm. Bull. 1977, 25(12), 3329.

7’ Zolotarev J. A.;1Zaitsev, o. A.; Tatur, v. J.; Mysasoedov, N. F. u. s. Patent
5,026, 909,

7° Kilgore, J. l..; u. s. Patent5,254,730, 1993.

79 Lim Youn -Hee Yoshimura, T., Soda, K. and Esaki, N. J. Ferment. Bioeng.
1998,86), 400.

8° FeeneyJ Birdsall B. Ostler, 6., Cam, M. D, and Kairi, M. Febs. Letters
1990, 272(1,12),19.‘7’

81Mitulovi,<3..,Lammerhoier M., Maier, N. M., Lindner, w., J. Labelled de.
RadiOp'har'm” 2000, 43, 449."

152

82 Testimony before the United States Committee on Agriculture, Nutrition, and
Forestry, March 29‘“, 2001 by Pat Gruber, Cargill Dow LLC.

‘33 Chem. Market Reporter, Jan. 28‘“, 2002.
8‘ Lemoigne, M, Ann. Inst., 1927, 14, 148.

85 Nakayama, T.; lwata, H.; Tanaka, R.; lmajo, S.; lshiguro, M. J. Chem. Soc.,
Chem. Com. 1991, 9,662.

86 Tanaka, R.; lwata, H.; lshiguro, M. J. Antibiot.1990, 43, 1608.
‘37 Mori, K.; Miyake, M. Tetrahedron 1987, 43, 2229.

88 Ohloff, G.; Giersch, W.; Decorzant, R.; Buechi, G. Helv. Chim. Acta 1980, 63,
1589.

89 Ferreira, J.; Ferreira, B.; Simoneili, F.; Tetrahedron, 1990, 46, 6311.
9° Choi, V.; Elliot, J.; Johnson, W.; Tetrahedron Lett, 1984, 25, 591.
9‘ Zhang, 2.; Jackson, J. E.; Miller, D. J. Ind. Eng. Chem. Res. 2002, 41,691.

153

ll

881

ll

 

ll