Stereoretentive C-H bond activation in the aqueous phase catalytic hydrogenation of amino acids to amino alcohols

Article abstract of DOI:10.1021/ol0274211

(Matrix presented) At 100°C and 1000 psi of hydrogen, aqueous L-alanine undergoes facile hydrogenation to L-alaninol over a 5% Ru/C catalyst. In the presence of added acid to protonate the carboxylate moiety, the reaction is faster and more selective than analogous reductions of simple alkanoic acids. Stereochemistry at the α-carbon is retained despite complete exchange of hydrogen at this site, as shown by deuterium incorporation. Similar stereoretentive C-H bond activation at C2 is seen in L-alaninol itself, and when acid is omitted, in L-alanine. These processes reveal a class of mild, highly stereoretentive C-H bond activations occurring in water over a heterogeneous catalyst.

Full text of DOI:10.1021/ol0274211

ORGANIC
LETTERS
2003
Vol. 5, No. 4
527-530
Stereoretentive C−H Bond Activation in
the Aqueous Phase Catalytic
Hydrogenation of Amino Acids to Amino
Alcohols
,‡
Frank T. Jere,† Dennis J. Miller,† and James E. Jackson*
Department of Chemical Engineering and Material Science and Department of
Chemistry, Michigan State UniVersity, East Lansing, Michigan 48824
jackson@cem.msu.edu
Received December 6, 2002
ABSTRACT
At 100 °C and 1000 psi of hydrogen, aqueous L-alanine undergoes facile hydrogenation to L-alaninol over a 5% Ru/C catalyst. In the presence
of added acid to protonate the carboxylate moiety, the reaction is faster and more selective than analogous reductions of simple alkanoic
acids. Stereochemistry at the r-carbon is retained despite complete exchange of hydrogen at this site, as shown by deuterium incorporation.
Similar stereoretentive C−H bond activation at C2 is seen in L-alaninol itself, and when acid is omitted, in L-alanine. These processes reveal
a class of mild, highly stereoretentive C−H bond activations occurring in water over a heterogeneous catalyst.
Lactic acid (2-hydroxypropanoic acid) is easily and selec-
tively hydrogenated to propylene glycol (1,2-propanediol)
in aqueous solution over Ru/C catalyst.1 Such direct hydro-
genation of simple n-alkanoic acids is normally difficult.
Surprised by lactic acid’s mild reaction, we analyzed its
kinetics.2 The resulting optimized conditions (130-170 °C;
1000-2000 psi of H2) modestly improved on values (typi-
cally 150 °C, 7000 psi) obtained by Carnahan et al. for
ruthenium-catalyzed hydrogenation of the closely related
glycolic acid (hydroxyacetic acid).3
hydrogenation to the corresponding amino alcohols. Amino
alcohols are important building blocks in agricultural,4
pharmaceutical, and peptide chemistry,5 and are also utilized
as chiral auxiliaries.6 In their extensive review of amino
alcohol synthesis methods, Ager et al.7 provide many
examples of amino acid reduction, typically in batch reactions
with expensive and/or hazardous hydride reagents. Direct
aqueous phase catalytic hydrogenation of free acids offers
an alternative that is atom economical and amenable to
continuous processing, and that obviates the need for
intermediate esterification, use of organic solvents, and
Prompted by the lactic acid results, we extended our
studies to amino acid substrates, again finding mild, selective
(4) Wu, S.; Takeya, R.; Eto, M.; Tomizawa, C. J. Pestic. Sci. 1987, 12,
221.
† Department of Chemical Engineering and Material Science.
‡ Department of Chemistry.
(1) (a) Zhang, Z.; Jackson, J. E.; Miller, D. J. Appl. Catal. A 2001, 219,
89. (b) Zhang, Z., Ph.D. Dissertation, Michigan State University, East
Lansing, Michigan, 2000.
(5) (a) TenBrink, R. E. J. Org. Chem. 1987, 52, 418. (b) 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. (c) Fincham, C. I.; 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. (d) Auvin-Guette, C.; Rebuffat, S.; Prigent,
Y.; Bodo, B. J. Am. Chem. Soc. 1992, 114, 2170.
(2) Zhang, Z.; Jackson, J. E.; Miller, D. J. Ind. Eng. Chem. Res. 2002,
41, 691.
(3) Carnahan, J.; Ford, T.; Gresham, W.; Grigsby, W; Hager, G. J. Am.
Chem. Soc. 1955, 77, 3766.
(6) Coppola, G. M.; Schuster, H. F. Asymmetric Synthesis of Chiral
Molecules using Amino Acids; Wiley-Interscience: New York, 1987.
10.1021/ol0274211 CCC: $25.00 © 2003 American Chemical Society
Published on Web 01/17/2003

Table 1. Summary of Alanine Reactionsa
substrate
T (°C)
t (h)
pressure (psi)
[phosphoric acid] (M)
alanine conversion
L-alaninol yield
ee (%)
alanine
alanine
alanine
alanine
alanine
alanine
alanine
alanine
alanine
alanine
alanine
alanine
alanine
alanine
alanine
alaninol
alaninol
90
100
100
100
100
100
125
125
125
125
125
125
125
150
150
125
150
6
6
6
6
6
6
2
6
2
6
6
2
6
0.75
6
6
6
1000
1000
1000
1000
1000
1000
250
250
500
500
1000
1800
1800
1000
1000
1000
1000
0.29
0
60
13
46
69
93
60
8
99.7
92.7
97.2
98.7
99.2
99.4
96
89
96
83
90.7
98
0.08
0.15
0.29
0.58
0.29
0.29
0.29
0.29
0.29
0.29
0.29
0.29
0.29
0.29
0.29
37
62
91
68
52
76
74
78
89
96
99
78
38
70
60
>99
88
>99
>99
96
>99
>99
>99
92
89
42
89
40
a Reactions performed in a stirred batch reactor; all feeds are 0.22 M; conversion, yield, and ee values were determined from HPLC data.
byproduct waste streams. Two recent patents by Antons et
al., filed after this work began, describe similar reductions
of lactic and amino acids in good yields and with stereo-
retention, but give no mechanistic details.8
acid is completely retained through the reduction process.
Figure 1 displays alanine and alaninol concentration profiles
for reaction at 150 °C. Alanine conversion is complete at
45 min; at this point the yield of L-alaninol peaks, and then
decreases slowly via racemization to D-alaninol along with
C-C hydrogenolysis to form ethylamine. At no point during
the reaction is D-alanine detected. With L-alaninol used as
the feed, the extent of racemization observed at 150 °C at 4
h was the same as that shown in Figure 1. Hydrogenation of
the carboxylic functionality and loss of optical purity are
thus clearly different processes. In fact, at 100 °C alanine
hydrogenation is roughly 1/8th as fast as at 150 °C, but
product racemization is essentially stopped.
We have probed the mechanism and stereochemistry of
hydrogenation of the simplest chiral amino acid, L-alanine
((S)-2-aminopropanoic acid), to L-alaninol (S-(+)-2-amino-
1-propanol) via isotopic labeling together with analyses of
product chirality and reaction rates. Table 1 summarizes a
series of hydrogenations, run in a stirred batch reactor at
90-150 °C. With 0.22 M alanine in water at 100 °C,
conversion ended after 1 h with low yield (8%) and poor
selectivity (<60%) to L-alaninol. In contrast, conversion of
lactic acid to 1,2-propanediol was complete with selectivity
>90% under similar conditions.
Alanine conversion increased markedly when an excess
of phosphoric acid (0.29 M H3PO4) was added to form
protonated alanine 1 (pKa ) 2.34). To be thermochemically
feasible, hydrogenation requires the carboxylic acid func-
tionality to be undissociated. Reduction of carboxylate anion
to alkoxide would be more endothermic by an energy
difference corresponding to the ∆pKa between the starting
acid and the product alcohol, ca. 15-17 kcal/mol at ambient
temperatures. This effect was clearly seen in our work with
lactic acid and its salts. Neutral aqueous alanine exists as
zwitterion 3 (Scheme 1) so the poor yield and selectivity in
its hydrogenation without added acid are unsurprising. As
expected, substoichiometric quantities of acid allowed only
partial alanine conversion, which stopped when the pH rose
due to basic product formation.
Above 1000 psi rate is independent of hydrogen concen-
tration,9 though an experiment at 250 psi reduced the rate
1
by ca. /2. This finding suggests that catalyst binding sites
for H2 are saturated above 1000 psi. Analysis of reactant
transport using the criteria developed by Weisz and Prater10
indicates no diffusion limitations for either hydrogen or
alanine in this three-phase system. The absence of transport
resistances and the observed plateau in rate as a function of
H2 pressure signify that at sufficient pressures neither H2
nor substrate concentrations limit the reaction rate, as seen
earlier for lactic acid hydrogenation.2
Scheme 1. Alanine‚H+, Alaninol‚H+, and Amino Acid Form
vs pH
With sufficient acid, protonated L-alaninol 2 was formed
with selectivity exceeding 95% and enantiomeric excess (ee)
over 99%. Thus, the stereochemistry of the starting amino
(7) Ager, D. J.; Prakash, I.; Schaad, D. R. Chem. ReV. 1996, 96, 835.
(8) (a) Antons, S. U.S. Patent 5731479, 1998. (b) Antons, S.; Tilling,
A.; Wolters, E. WO 9938838, 1999.
528
Org. Lett., Vol. 5, No. 4, 2003

Scheme 2. Deuterium Incorporation in D2/D2O over Ru/C
labeling is also seen at the C2 positions of alaninol product
(90% complete at 6 h) and alanine feed (100% complete at
3 h).
Figure 1. Secondary racemization of L-alaninol. Reaction per-
formed with 100 mL 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.
Scheme 3. Generalized Amino Acid Hydrogenation Sequence
Showing Aldehyde 5 and Potential Enols 6 and 7
At the outset of this study we sought to probe the
hydrogenation’s stereochemistry via H/D exchange. Loss of
C2-H was expected to lead to racemization, so that D
incorporation could stand in for loss of stereochemical purity.
Subsequent experiments in which the enantiomeric excesses
(ee’s) were explicitly measured by HPLC yielded a sur-
prise: at 100 °C, D incorporation at C2 occurs without loss
of optical purity for both alanine and alaninol. At 150 °C,
both compounds undergo complete H/D exchange at C2 in
less than 1 h; D also appears at the C3 positions with
concomitant racemization of alaninol. Scheme 2 and Table
2 summari ze deuterium (D) incorporation results from
alanine hydrogenation in D2/D2O. In product alaninol, D
incorporation at C1 is complete as expected. Partial D
Scheme 3 shows a proposed general pathway for carboxy-
lic acid hydrogenation, without any explicit representation
of the catalyst’s role. Addition of H2 across the carbonyl
group yields the 1,1-diol 4, which loses H2O to form the
aldehyde 5. A second equivalent of H2 is then added to 5 to
Table 2. Summary of Deuterium Incorporation Experimentsa
% D incorporation at each position
T
t
[acid] alanine L-alaninol alaninol
substrate (°C) (h)
(M)
conv.
yield
ee (%)
alanine C2 alanine C3 alaninol C1 alaninol C2 alaninol C3
alanine
alanine
alanine
alanine
alanine
alanine
alanine
alanine
alaninol
alaninol
alaninol
alaninol
alaninol
alaninol
alaninol
alaninol
100
100
100
100
150
150
150
150
100
100
100
100
150
150
150
150
1
6
1
6
1
6
1
6
1
6
2
6
1
6
1
6
0.27
0.27
0
18
84
8
20
98
99.5
34
65
18
78
4
14
84
74
14
21
>99.9
99.7
>99.9
95.5
95.3
84.1
79.3
58.7
99.9
99.7
>99.9
99.8
95.7
80.7
91.5
60.5
62
100
92
6
2
0
0
100
100
100
100
100
100
100
100
75
100
100
100
100
100
100
100
39
89
*
100
100
100
100
100
42
1
0
0
0
96
0
0.26
0.26
0
25
61
23
52
14
15
0
80
100
31
58
0
0.28
0.28
0
87
100
100
100
100
100
100
0
0
0.29
0.29
0
30
65
7
0
32
a Reactions performed in a stirred batch reactor; alanine feeds 0.26 M; alaninol feeds 0.22 M, in D2O under 1000 psi D2; D-incorporation measured by
1H NMR; an asterisk (*) indicates the peak was obscured by impurity.
Org. Lett., Vol. 5, No. 4, 2003
529

form the product alcohol. Both the dehydration and the
aldehyde reduction steps would be expected to be fast
compared to the first hydrogenation.
bond formation at such sites begun to be exploited, enabling
addition of alkyl or acyl groups at the R C-H bond sites to
amines.15 In these reactions, however, the amines require
attached 2-pyridyl groups, which serve as directing ligands
for the transition metals effecting the C-H bond activation.16
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. We are actively pursuing the
intriguing possibility of stereocontrolled replacement of H
with other functionalities under aqueous heterogeneous
catalytic conditions.
We have been unable to detect 5, but it might be postulated
to explain H/D exchange at C2 via equilibration with the
related enol 7. In principle, enol 6 could also enable C2 H/D
exchange, but this catalyst-independent process is not seen
at the experimentally relevant conditions. However, two
findings clearly show that H/D exchange at C2 occurs
separately from hydrogenation. First, there is little or no loss
of chirality despite the H/D exchange at C2; thus, a free enol,
sp2 hybridized at C2, cannot play a role in this reaction.
Second, if hydrogenation is blocked by omission of acid,
stereoretentive H/D exchange at C2 still proceeds readily.
The results for D incorporation at C2 suggest that the catalyst
directly removes H from the amine-bearing carbon to yield
a surface-bound intermediate that retains the original C2
configurational information. At 100 °C, this species uni-
formly undergoes deuteration on the face from which the H
loss occurred, thus retaining optical purity, and somehow
surviving the elevated temperatures and the aggressive
aqueous acid environment.11 Higher temperatures presumably
then disrupt the surface binding, so that D incorporation can
take place on the opposite face. The only functional group
needed is the amine; besides alanine and alaninol, (S)-2-
aminobutane undergoes H/D exchange at C2 without loss
of chirality.
The 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 we have found little discussion of their
stereochemistry.12 Such reactions are thought to form imine
intermediates, enabling alkyl group transfers via transami-
nation pathways.13 Indeed, alkyl group exchange via catalytic
activation of amines has been explored as a process to
convert primary to secondary and tertiary amines.14 Only in
recent years has the potential for functionalization via C-C
As for the hydrogenation, why are protonated amino and
lactic acid so much more reactive than their simple alkanoic
+
acid counterparts? The obvious explanation is that the NH3
and OH electron-withdrawing groups enhance the neighbor-
ing carbonyl group’s preference for sp3 hybridization, as is
seen in the hydration equilibria of substituted ketones and
aldehydes.17 Two items support this notion: the fact that
alanine’s charged ammonium group is more activating than
the hydroxyl in lactic acid, and the finding by Antons et al.9
that 2-chloropropanoic acid is also reactive toward hydro-
genation. Alternatively, intramolecular hydrogen bonding
might favor hydrogenation.
We are actively exploring the behavior of related systems,
such as the alkylated NR2 and OR analogues of alanine and
lactic acids. Exploratory hydrogenations of glycine, serine,
and phenylalanine have already revealed brisk rates and high
selectivities like those found for alanine. A more detailed
account and mechanistic analysis of this ongoing work is in
preparation.
Acknowledgment. This project has been supported by
USDA-NRICGP grant number 98-35504-6356.
Supporting Information Available: Information regard-
ing experimental and analytical techniques. This material is
available free of charge via the Internet at http://pubs.acs.org.
(9) Hydrogen solubility follows Henry’s law, so H2 concentrations vary
linearly in water.
OL0274211
(10) Weisz, P. B.; Prater, C. D. AdV. Catal. 1954, 6, 143.
(11) We cannot absolutely rule out homogeneous ruthenium complexes
formed in situ, but a rigorous search found no catalyst leaching in our closely
related lactic acid studies. Zhang, Z. Unpublished results.
(12) Shvo, Y.; Thomas, D. W.; Laine, R. M. J. Am. Chem. Soc. 1981,
103, 2461.
(13) Murahashi, S.-I.; Yoshimura, N.; Tsumiyama, T.; Kojima, T. J. Am.
Chem. Soc. 1983, 105, 5002 and references therein.
(14) Wilson, R. B., Jr.; Laine, R. M. J. Am. Chem. Soc. 1985, 107, 361
and references therein.
(15) Doye, S. Angew. Chem., Int. Ed. 2001, 40, 3351.
(16) (a) Chatani, N.; Asaumi, T.; Ikeda, T.; Yorimitsu, S.; Ishii, Y.;
Kakiuchi, F.; Murai, S. J. Am. Chem. Soc. 2000, 122, 12882. (b) Sakaguchi,
S.; Kubo, T.; Ishii, Y. Angew. Chem., Int. Ed. 2001, 40, 2534. (c) Chatani,
N.; Asaumi, T.; Yorimitsu, S.; Ikeda, T.; Kakiuchi, F.; Murai, S. J. Am.
Chem. Soc. 2001, 123, 10935.
(17) (a) Bell, R. P. AdV. Phys. Org. Chem. 1966, 4, 1. (b) Guthrie, J. P.
Acc. Chem. Res. 1983, 16, 122.
530
Org. Lett., Vol. 5, No. 4, 2003