Synthesis of substituted mandelic acid derivatives via enantioselective hydrogenation: Homogeneous versus heterogeneous catalysis

Article abstract of DOI:10.1002/adsc.200404022

An extensive screening of both homogeneous and heterogeneous catalysts was carried out for the enantioselective hydrogenation of p-chlorophenylglyoxylic acid derivatives. For p-chlorophenylglyoxylic amides only homogeneous Rh-diphosphine complexes gave sa

Full text of DOI:10.1002/adsc.200404022

FULL PAPERS
Synthesis of Substituted Mandelic Acid Derivatives via
Enantioselective Hydrogenation: Homogeneous versus
Heterogeneous Catalysis
Fredrik Cederbaum,a Clemens Lamberth,a Christophe Malan,b Fred Naud,b
Felix Spindler,b Martin Studer,b Hans-Ulrich Blaserb,*
a
Syngenta Crop Protection AG, Research Department, Schwarzwaldallee 215, 4002 Basel, Switzerland
Solvias AG, P. O. Box, 4002 Basel, Switzerland
b
Fax: þ41-61-686-6100, e-mail: hans-ulrich.blaser@solvias.com
Received: January 21, 2004; Accepted: May 7, 2004
Dedicated to Joe P. Richmond, the motor behind ASC, on the occasion of his 60th birthday.
SupportingInformation for this article is available on the WWW under http://asc.wiley-vch.de/.
Abstract: An extensive screeningof both homogene-
4000 and TOFs up to 210 hÀ1. For all catalytic systems
ous and heterogeneous catalysts was carried out for the effects of the metal, the nature of the chiral aux-
the enantioselective hydrogenation of p-chlorophe- iliary and the solvent as well as of the reaction condi-
nylglyoxylic acid derivatives. For p-chlorophenyl- tions were investigated. The homogeneous process
glyoxylic amides only homogeneous Rh-diphosphine was scaled up to the kgscale and the enantiomeric pu-
complexes gave satisfactory results, ees up to 87% rity of the product was enhanced to >99% ee by two
were observed for the cy-oxo-pronop ligand. For recrystallizations of the free p-chlorophenylmandelic
methyl p-chlorophenylglyoxylate both a homogene- acid.
ous as well as a heterogeneous catalyst performed
with ees >90%. A Pt catalyst modified with cinchona
derivatives achieved 93% ee for the (R)- and 87% ee Keywords: cinchona alkaloids; enantioselective hy-
for the (S)-methyl p-chloromandelate. A Ru-MeObi- drogenation; a-keto esters; mandelamide derivatives;
phep catalyst also reached 93% ee with TONs up to MeObiphep; platinum; ruthenium
Introduction
Of the several approaches possible, two were explored
in more detail. The first is an enantioselective Passerini
Specifically substituted mandelamides such as SX three-component condensation of a carbonyl com-
623509 (1) (Fig. 1) constitute a new class of agrochemi- pound, an isocyanide and a chiral carboxylic acid, lead-
cal fungicides, acting specifically against the oomycetes ingdirectly to the desired amides. As a model reaction,
family of phytopathogenic fungi like Phytophthora in- aldehyde 2and isocyanide 3weretransformed with good
festans (potato late blight) and Plasmopara viticola stereoselectivity to the chiral mandelamide 5 by using
(grape downy mildew).[1,2] There are indications in the 1,2,3,4-tetra-O-acetyl-a-d-galacturonic acid (4) aschiral
literature that the two enantiomers of 1 do not contrib- auxiliary (see Fig. 2).[3]
ute equally to the biological efficacy of the racemate.[2]
The second approach involves the enantioselective re-
To check this hypothesis novel concise approaches to duction of suitable a-keto acid derivatives as depicted in
enantiopure mandelamides were needed.
Figure 2. Substituted mandelamides via a Passerini three-
component condensation.
Figure 1. SX 623509 (1), a fungicidally active mandelamide.
842
¹ 2004 WILEY-VCH VerlagGmbH & Co. KGaA, Weinheim
DOI: 10.1002/adsc.200404022
Adv. Synth. Catal. 2004, 346, 842 848

Synthesis of Substituted Mandelic Acid Derivatives
FULL PAPERS
Figure 3. Pathways to substituted mandelic amides via enantioselective reduction of the correspondingphenylglyoxylic acid
derivative.
Figure 3. Three catalytic methods described in the liter-
Here we will focus on the Phase 2, demonstratingthe
ature for the enantioselective hydrogenation of a-keto feasibility of the key enantioselective catalytic step for
acid derivatives were investigated: Homogeneous hy- the two routes sketched in Figure 3. In this context it is
drogenation,[4] homogeneous transfer hydrogenation[5] important to realize that four elements are critical for
and hydrogenation with modified heterogeneous cata- a timely result:
lysts.[6]
In this contribution we describe our efforts to find a
technically feasible access to the desired mandelic
amides in >90% ee. We report screeningresults for
the enantioselective reduction of methyl p-chlorophe-
1. To have a suitable library of chiral ligands and/or
auxiliaries as well as of soluble metal precursors
and/or heterogeneous hydrogenation catalysts.
There will be no time to procure or synthesize
complex chiral structures.
nylglyoxylate (9) and the analogous amide derivatives 2. To have access to modern laboratory equipment
6a c as well as the scale-up of the most effective method
to the kgscale. Furthermore, the enrichment of the free
p-chloromandelic acid from approximately 90% ee to
>99% ee is described.
allowingvery rapid screeningof various catalytic
systems preferably in parallel and with very small
substrate samples.
3. To have very good analytical equipment (especial-
ly chiral GLC or HPLC).
4. Last but not least to have specialists with much
hands-on-experience since findingenantioselective
catalysts is still very much an empirical business
where personal experience is the key to success.
Results and Discussion
Some General Remarks
When developinga process for a new biologically active
compound, there is usually a rather tight schedule since
Hydrogenation of Phenylglyoxylic Amides
the time-to-market is a very important aspect. For the The most elegant solution to the problem at hand would
development chemists this means that he or she cannot be the asymmetric reduction of the substituted amides 6,
spend too much time searchingfor the perfect process;
which, at least in principle, could be prepared via Pd-cat-
rather it is important to be able to give ™yes∫ or ™no∫ an- alyzed double carbonylation of the correspondingaryl
swers within a defined time. In a recent concept paper[7] bromide or iodide in presence of amines 8. For this rea-
we described our approach in more detail. It is useful to son, much emphasis was put on investigating the enan-
divide the development of a manufacturingprocess into
four different phases:
Phase 1: Outliningand assessingpossible synthetic
routes on paper.
tioselective hydrogenation of the keto amides 6 with
the aim to find a catalyst with >90% ee followed by fur-
ther optimization for activity. There are only few but
quite encouraging reports on the enantioselective hy-
drogenation of aryl a-keto amides. Chiba et al.[4b] ob-
Phase 2: Demonstratingthe chemical feasibility of the
key step, usually the enantioselective catalytic reaction. tained 96% ee for N-benzylglyoxylic amide, using Ru-
Phase 3: Optimizingthe key catalytic reaction.
Phase 4: Optimizingthe over-all process.
bichep complexes (for ligand structures, see Fig. 4)while
Rh-amidophosphine-phosphinite (ampp) catalysts were
Adv. Synth. Catal. 2004, 346, 842 848
asc.wiley-vch.de
¹ 2004 WILEY-VCH VerlagGmbH & Co. KGaA, Weinheim
843

Fredrik Cederbaum et al.
FULL PAPERS
Figure 4. Structures of chiral ligands and modifiers listed in Tables 1 3.
reported to give 99% ee for the same substrate[4c] and
87% ee for the p-chloro derivative thereof.[4d] In all cas-
es, rather low TONs (20 200) and TOFs of (3 <100 hÀ1
)
were achieved. The hydrogenation of aliphatic a-keto
amides usinga heteroegneous cinchona-modified Pt
catalyst was reported to occur with ees below 60%.[6b]
Based on these literature precedents, we decided to
screen the three keto amides 6a c, applyingboth Rh
Figure 5. Structure of the byproduct formed in methanol.
and Ru catalysts with a broad array of privileged, elec- all three substrates, but especially for 6a incomplete con-
tron-rich ligands (e.g., of the type: MeObiphep, biphep, versions were observed even after a reaction time of
binap, duphos, josiphos, walphos, plus one available
>15 h. We thought that the free phenolic OH, which is
ampp). In total 80 experiments were carried out on a able to coordinate to the metal center, might be respon-
scale of 0.5 mmol substrate in 5 mL solvent (46 for 6a, sible for this fact. This interpretation is at least partly
25 for 6b, and 9 for 6c) and selected positive results are supported by the higher conversions observed for 6b
compiled in Table 1. The highest enantiomeric excess and 6c. The results with the Ru complexes of atropiso-
(87%, just short of the required 90% ee) was obtained meric diphosphines (especially bichep) were rather dis-
for 6a with an Rh catalyst generated in situ in dichloro- appointingand interestingly, Ru-phanephos (with pla-
methane from the neutral precursor [Rh(nbd)Cl]2 and nar chirality) gave the best results here. Of the solvents
the ligand (R)-cy-oxo-pronop. However, the catalyst tested, dichloromethane was optimal for Rh catalysts
had a rather low activity, and at 408C full conversion while methanol (often requiring1 N HCl as additive)
could only be obtained with a substrate to catalyst ratio was found to be the best for Ru catalysts. In the latter
[4d]
¼
(s/c) of 100. In analogy with the results of Mortreux,
case, the formation of a by-product was observed which,
we tentatively assigned the (S) configuration to the ma- based on 1H NMR, is the a-amido-acetal 11 (Fig. 5), re-
jor product.
sultingfrom the acid-catalyzed addition of methanol to
Except for 6b, Rh complexes generally achieved bet- the keto group of the starting material.
ter enantioselectivities than Ru complexes. Rh com-
A few experiments were carried out with cinchona-
plexes of several ligand classes afforded ees >50% for modified heterogeneous Pt catalysts as described by
844
¹ 2004 WILEY-VCH VerlagGmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2004, 346, 842 848

Synthesis of Substituted Mandelic Acid Derivatives
FULL PAPERS
Table 1. Screeningresults for the hydrogenation of 6a c (effect of metal, ligand, solvent, pressure, temperature).
Amide Catalyst[a] Solvent P/T [bar/8C] ee [%] TON TOF[b] [hÀ1] Abs. conf. Comments[c]
6a
6a
6a
6a
6a
6a
6a
6a
6a
6b
6b
6b
6c
6c
6c
6c
Rh-(R)-cy-oxo-pronop
Rh-(2S,4S)-bcpm
Rh-(S)-R-MeObiphep
CH2Cl2 20/40
THF 10/20
CH2Cl2 20/40
87
74
65
61
57
55
47
47
44
67
62
44
64
54
53
40
100
100
100
120[d]
115[d]
55
100
89
37
100
96
93
93
95
5
5
5
8
7
3
5
5
2
6
5
6
2
2
2
1
(R)
(S)
(S)
(R)
(S)
(S)
(R)
(S)
(S)
(R)
(S)
(S)
(S)
(S)
(R)
(R)
BF4 complex 32% ee
in CH2Cl2 67% ee
R¼cyclopentyl
Ru-(S)-phanephos
MeOH 80/60
60 mL HCl, bp 15%
60 mL HCl, bp 10%
400 mL HCl, bp 15%
Ru complex 12% ee
Ru-(S)-Ar-(MeO)3biphep[e] MeOH 80/60
Ru-(S)-i-Pr-MeObiphep
Rh-(R)-bichep
MeOH 80/60
CH2Cl2 20/40
Rh-(2S,4S)-Ar*-bppm[f]
Rh-(R,R)-Ar*-diop[f]
Ru-(S)-Xyl-phanephos
Rh-(S)-i-Pr-MeObiphep
Ru-(R)-bichep
THF
10/20
CH2Cl2 20/40
MeOH 80/60
CH2Cl2 20/40
MeOH 80/60
at 308C ee 53%
Rh-(2S,4S)-bcpm
THF
10/20
Rh-(S)-i-Pr-MeObiphep
Ru-(S)-phanephos
Rh[h]-mandyphos
CH2Cl2 20/40
[g]
-
80/60
60/25
100
25
THF
[a]
Reaction conditions: s/c 100 (Rh catalysts: PPþ[Rh(nbd)Cl]2; Ru catalysts: PPþ[Ru(p-cymene)I2]2), 0.5 mmol 6, 5 mL
solvent
[b]
[c]
[d]
[e]
[f]
TOF¼TON/reaction time.
mL HCl: addition of 1 N HCl, bp: byproduct (see text).
s/c 200.
Ar¼3,4,5-(MeO)3-C6H2.
Ar*¼3,5-t-Bu2-4-MeO-C6H2.
MeOH/CH2Cl2 (3/1).
[g]
[h]
[Rh(nbd)TFA]2.
Baiker[6b] for aliphatic a-keto amides but even with the firmed, the Rh-catalyzed transfer hydrogenation could
best catalytic system, ees never exceeded 34%.
not be reproduced at all.
Homogeneous hydrogenation: The highest enantiose-
lectivities (up to 94% ee, surpassingthe target of >90%
ee) were obtained with an Ru catalyst generated in situ
from [RuI2(p-cymene]2 and MeObiphep. The observed
Hydrogenation of Methyl Phenylglyoxylates
The hydrogenation of aromatic a-keto ester derivatives absolute configuration is in agreement with those re-
has also been described in the literature.[4] For the parent ported by Mashima et al[4e] but opposite to the correla-
compound methyl phenylglyoxylate (9b) enantioselec- tion of Chiba et al.[4b] Other aryl derivatives of the MeO-
tivities of >95% were reported for several catalytic sys- biphep ligand family and atropisomeric ligands such as
tems. Chiba et al.[4b] achieved >99% ee for an Ru-bi- binap or TMBTP gave medium ees while alkyl-substi-
chep complex, Carpentier and Mortreux[4c] reached tuted MeObiphep ligands gave very poor results. For
95% ee usingan Rh-ampp catalyst and Lemaire et al. [5b] the Ru-MeObiphep catalyst, the addition of catalytic
claimed to have achieved 99% ee for an Rh-Me2dpen- amounts 1 N HCl was essential both for good activity
catalyzed transfer hydrogenation. In addition, a Pt cata- as well as for achievingees >90%. Other acids, like
lyst modified with dihydrocinchonidine was reported to HBF4, HBr, H3PO4 or CF3COOH gave lower ees and/
give 98% ee.[6c] For the hydrogenation of substrate 9b, or activities; racemic product was obtained in presence
Mashima et al.[4e] obtained an ee of 93% with Ru-binap. of Cs2CO3.
In all cases, TON and TOF were too low to be of techni-
cal interest.
Homogeneous transfer hydrogenation of 9a and 9b:
The results with the Rh-Me2dpen hydrogen transfer cat-
Based on these findings three catalyst systems were alysts and isopropyl alcohol as reducingagent were a
screened with 9a as substrate: on the one hand homoge- (negative) surprise. Instead of the expected products
neous Ru complexes with various atropisomeric diphos- 10a or 10b (results not shown), only the corresponding
phines and the Rh-Me2dpen catalyst and on the other isopropyl phenylglyoxylate, i.e., the transesterified ke-
hand the heterogeneous Pt-cinchona system. For the ho- tone, was obtained. Since a blank experiment without
mogeneous catalysts, relevant results are listed in Ta- metal catalysts also afforded 90% of the same product,
ble 2, for the Pt-cinchona in Table 3. While the efficacies we suspect that the literature results might be based
of the Ru-PP and the Pt-cinchona catalysts were con- on a wrongassignment of the reaction product. With
Adv. Synth. Catal. 2004, 346, 842 848
asc.wiley-vch.de
¹ 2004 WILEY-VCH VerlagGmbH & Co. KGaA, Weinheim
845

Fredrik Cederbaum et al.
FULL PAPERS
Table 2. Screening results for the homogeneous hydrogenation of 9a (effect of metal, ligand, reducing agent).
Catalyst[a, b]
10a [%]
ee [%] s/c
TOF[c] [hÀ1
]
Conf. Comments
Ru-(R)-MeObiphep[a]
Ru-(S)-Tol-MeObiphep[a]
Ru-(S)-Ar-MeObiphep[a]
[Ru2Cl4(R)-binap)2(NEt3)(xylene)][a]
Ru-(R)-biphemp[a]
99.5
>99.5
94 100 64 89
100
98
100
<1
2 5
3
94[d]
93
200
3
12
10 12
11
11
10
(R)
(S)
(S)
(R)
(R)
(S)
-
(S)
(R)
(R)
without HCl ee 66%
at 308C 65% conv.
various aryl substituents on P
200
200
200
200
200
20
100 100
100
100
88
87
62
Ru-(S)-tmbtp[a]
Rh-Me2dpen[b]
>90% i-Pr keto ester
95% i-Pr hydroxy ester, ee 6%
80% i-Pr hydroxy ester, ee 23%
Ru-Andersson ligand[b]
Ru-NH2 NTs[b]
60
30
32
80
5
Ru-NH2 NTs[e]
>99
[a]
1.5 mmol 9a; catalyst PPþ[Ru(p-cymene)I2]2, 10 mL MeOH, 60 mL 1 N HCl, 80 bar, 608C.
[b]
[c]
[d]
[e]
Transfer hydrogenation: [RuCl2(p-cymene)]2 þligand or isolated complex, i-PrOH/KOH, r.t.
TOF¼TON/reaction time.
At 308C.
With HCOOH/NEt3 5: 2 in CH2Cl2.
Table 3. Screening results for the heterogeneous hydrogenation of 9a (effect of solvent and modifier).
Solvent
Modifier
9a [g]
Solvent [mL]
Cat.[a] [mg]
ee [%]
Abs. conf.
Comments
Toluene[b]
HCd
iso-Cn
HCn
0.1
0.1
0.1
0.1
0.1
3
1
1
1
1
1
27
27
135
10
10
10
10
10
120
60[d]
600
90
85
75
80
73
87
67
93
(R)
(S)
(S)
(R)
(S)
(S)
(S)
(R)
in AcOH ee 76%
in AcOH ee 71%
in AcOH ee 59%
Toluene[b]
Toluene[a]
Toluene/AcOH 1/1[b]
Toluene/AcOH 1/1[b]
Toluene[c]
HCd
iso-Cn
iso-Cn
iso-Cn
HCd
with HCn 62%
reaction time 45 min
at 5 bar ee 36%
Toluene[c]
3
15
Toluene[c]
reaction time 60 min
[a]
5%Pt/Al2O3 JM 94.
Reaction conditions: 2 mgmodifier, 60 bar, r.t., 30 min.
Reaction conditions: modifier/catalyst 1/10 (w/w), 60 bar, 258C.
Reaction time 180 min.
[b]
[c]
[d]
more active catalysts such as Ru-NH2 NHTs or with the
Under these reaction conditions, full conversion was
exceedingly active Anderson ligand[5a] reduction was obtained. As expected, higher ees were obtained for di-
achieved but transesterification could still not be sup- hydrocinchonidine (HCd) formingpreferentially the
pressed, and in addition, only ees of 23 60% (for 10a) (R)-product than for the pseudo-enantiomeric cincho-
and 6 23% (for the isopropyl ester) were observed. nine (Cn) derivatives where an excess of (S)-hydroxy es-
When the reaction was performed with HCOOH/NEt3 ter was produced. In this case, a newly discovered iso-
as reductant, 10a was obtained in good yields but with meric cinchona derivative (iso-Cn)[6d] gave the highest
only 32% ee.
ees. As observed for aromatic ketones,[6a] toluene gave
Heterogeneous hydrogenation: From the literature,[6a] better results than AcOH (the solvent of choice for ali-
and our own experience, it is well known that only a few phatic analogs). To our disappointment, we could not re-
[6c]
¬
catalysts, modifier and solvent combinations are suita- produce the 98% ee claimed by the group of Bartok
ble for the heterogeneous enantioselective hydrogena- for the hydrogenation of 9b in a mixture of AcOH and
tion of a-keto esters. Based on our vast experience toluene as solvent.
with this catalyst system we chose AcOH and toluene
as solvents and screened a limited number of cinchona carried out for iso-Cn in toluene. A maximum of 88%
alkaloids usinga Pt catalyst from Johnson Matthey ee was reached at relatively low substrate concentra-
A brief optimization of the reaction conditions was
(JM 94). The most important results are listed in Table 3. tions (ꢀ10%) and a catalyst/modifier ratio of 1/10 w/w.
Furthermore, 12 synthetic modifiers from the group of Loweringthe temperature to 5 10 8C did not affect
Pfaltz[6d] were screened, but all of them showed lower the enantioselectivity but led to very viscous reaction
ees than HCd and iso-Cn (results not shown).
solutions which were difficult to stir. Loweringthe cata-
846
¹ 2004 WILEY-VCH VerlagGmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2004, 346, 842 848

Synthesis of Substituted Mandelic Acid Derivatives
FULL PAPERS
lyst loadingto 2% relative to substrate had a detrimental
chonidine gave up to 93% ee, albeit at rather high cata-
effect on both reaction time and ee, especially at 5 bar lyst loadings. While the resulting mandelic ester could
hydrogen pressure. Increasing the catalyst loading not be enriched further, the free acid was successfully re-
from 2% to 4% increased the initial rate from 0.8 to crystallized to give material of >99% ee.
6.6 mmol/min¥gcatalyst, a clear indication of catalyst poi-
soningprobably caused by impurities in the substrate.
Reproduction of several literature results proved to
be difficult and, especially, the high enantioselectivity
This observation confirms the experience that the Pt- claimed for an Rh-diamine-catalyzed homogeneous
cinchona system is quite sensitive to the quality of the transfer hydrogenation could not be confirmed. Also,
startingmaterial. [6e] A similar effect was observed with despite some literature precedents it was not possible
HCd where catalyst loadingof 4% was needed to reach to find a suitable catalyst for the hydrogenation of sever-
93% ee of the (R)-enantiomer (i.e., the goal of >90% ee al phenylglyoxylic amides. In all cases, the resulting ees
was also met with the heterogeneous system).
were below the required 90% and in addition, catalyst
activities were much too low.
Both for the homogeneous as well as the heterogene-
ous catalyst system the nature of the chiral auxiliary was
the decisive factor. For the homogeneous system, both
Preparative and Scale-Up Aspects
For the production of kgamounts of highly enriched (S)- the choice of an Ru catalyst and of the use of methanol
p-chloromandelic acid we used the Ru-(R)-MeObiphep with traces of HCl were important factors, while for the
catalyst. Scale-up experiments showed that s/c up to heterogeneous catalyst the choice of toluene as solvent
4,000 are feasible giving methyl 4-chloromandelate and the catalyst loadinghad a major effect on the cata-
with 90 93% ee with TOFs up to 210 hÀ1. Despite the lyst performance.
fact that this hydrogenation was not fully optimized,
the achieved catalyst performance (ee, TON, TOF) indi-
cate that this hydrogenation might be feasible for the
production of an agrochemical intermediate both from
a technical and an economical point of view.
Experimental Section
Materials
For the Pt-cinchona system, a scale-up from 100 mgto
15 g 9a presented no problems. The non-optimized sys-
tem with HCd showed 93% ee with an estimated TON
of about 2000 and TOF of ꢁ2000 hÀ1 (assumingthat
all surface platinum atoms on the heterogeneous cata-
lyst are active), indicatingthat the Pt-cinchona system
might be a viable alternative for the production of the
(R)-enantiomer.
Methyl p-chloromandelate 10a should be crystalline
(mp 368C) but an extensive screeningof more than 10
solvents and combinations thereof coveringa wide
range of solvent properties showed no success: We
therefore hydrolyzed the ester using0.2 M NaOH and
were able to crystallize the free acid from dibutyl ether
without any problem. In two crystallizations, the ee in-
creased from about 90 to >99% and 150 gof the free
p-chloromandelic acid with 99.3% ee were produced.
At this stage the project was terminated.
Substrates: Methyl phenylglyoxylates (9a and 9b) were pre-
pared accordingto standard Friedel Crafts acylation proce-
dures,[8,9] 4-chlorophenylglyoxylic acid amide (6a; mp 119
1208C) was obtained from 4-chlorophenylglyoxylic acid[9]
and 2-(4-hydroxy-3-methoxyphenethyl)amine[10] under stand-
ard peptide couplingconditions with the aid of Castro×s re-
agent and H¸nig×s base. Subsequent acetylation with acetic an-
hydride and potassium carbonate or benzylation with benzyl
chloride led to the amides 6b (mp 131 1338C) and 6c, respec-
tively.
Catalysts and chiral auxiliaries: 5% Pt/Al2O3 JMC 94[6a]
(Johnson Matthey, pretreated for 2 h at4008C under hydrogen,
dispersion ca. 0.25), [Rh(nbd)Cl]2 (Degussa), [Ru(p-cy-
mene)I2]2 (OMG/Umicore), [Ru(p-cymene)Cl2]2 (Strem).
Biphep derivatives and cy-oxo-pronop (Roche), binap,
phanephos, NH2 NHTs (Strem), diop derivatives (Kanto
Chemicals), bcpm, bppm derivatives (Fuji Chemical Indus-
tries), tmbtp (Chemi), Me2dpen (Acros), all other ligands are
available from Solvias. Cinchona alkaloids were either pur-
chased from Fluka or prepared as described.[6d]
Conclusions
Analytics
The enantioselective hydrogenation of phenyl glyoxy-
late esters is a technically feasible method to produce
the correspondingmandelic acid derivatives with enan-
tioselectivities >90%. Two catalytic systems were
found for the specific case of methyl p-chlorophenyl-
glyoxylate 9a: A homogeneous Ru catalyst with the
atropisomeric MeObiphep ligand achieved ees of 90
Hydroxy ester 10a: ee determination: HP 1100, column Chira-
cel ODH, 0.46ꢂ25 cm, solvent 95% n-hexane, 5% isopropyl
alcohol, flow 0.7 mL/min, detection at 220 nm. Retention
times: 9a 9.4 min, 10a (S)-enantiomer 15.1 min, 10a (R)-enan-
tiomer 17.2 min.
Amides 7a c: Conversion was determined by 1H NMR; by-
products formed in MeOH were estimated by preparative thin
layer chromatography (silica gel with dichloromethane/ethyl
93% with TONs up to 4,000 and TOFs up to 210 hÀ1
.
A heterogeneous Pt catalyst modified with dihydrocin- acetate 1/1). Rf values were as follows: Startingmaterial 6a/
Adv. Synth. Catal. 2004, 346, 842 848
asc.wiley-vch.de
¹ 2004 WILEY-VCH VerlagGmbH & Co. KGaA, Weinheim
847

Fredrik Cederbaum et al.
FULL PAPERS
6b): Rf ¼0.9/0.9; product 7a/7ba: Rf ¼0.4/0.3; ee determination:
HP 1100, column Chirapak AD, 0.46ꢂ25 cm, solvent 50% n-
hexane, 50% isopropyl alcohol, flow 0.5 mL/min, detection at
210 nm. Retention times: 6a: 12.0 min; 7a enantiomer
1:8.4 min, 7a enantiomer 2:11.1 min; 6b: 17.5 min; 7b enan-
tiomer 1:14.4 min, 7b enantiomer 2:21.7 min; 6c: 16.6 min;
7c enantiomer 1:10.2 min, 7c enantiomer 2:17.7 min.
References and Notes
[1] R. G. Griffiths, J. Dancer, E. O×Neill, J. L. Harwood,
New Phytologist 2003, 158, 345.
[2] O. Ort, U. Dˆller, W. Reissel, S. D. Lindell, T. L. Hough,
D. J. Simpson, J. P. Chung, Pesticide Sci. 1997, 50, 331.
[3] R. Frey, S. G. Galbraith, S. Guelfi, C. Lamberth, M. Zel-
ler, Synlett 2003, 1536.
[4] a) For an overview, see: T. Ohkuma, R. Noyori, in: Com-
prehensive Asymmetric Catalysis, Vol. I, (Eds.: E. N. Ja-
cobsen, A. Pfaltz, H. Yamamoto), Springer, Berlin,
1999, p. 199; b) T. Chiba, A. Miyashita, H. Nohira, H. Ta-
kaya, Tetrahedron Lett. 1993, 34, 2351; c) C. Pasquier, S.
Procedures
Here we only describe the screeningexperiments, for a descrip-
tion of the preparative experiments (hydrogenation, hydrolysis
and recrystallization) see SupportingInformation.
Homogeneous hydrogenation: Standard procedure. Two sol-
utions, prepared under argon atmosphere (a) catalyst and (b)
substrate, were transferred under argon with a cannula to an
autoclave. After purging, the desired hydrogen pressure was
applied and the autoclave heated to the desired temperature.
After a defined time (usually 15 21 h) the reaction was stop-
ped and conversion and ee of were determined.
Heterogeneous hydrogenation: All screeningexperiments
were carried out in 3-mL glass vials with magnetic stirring,
placed in a steel autoclave (4 per 50 mL autoclave). After purg-
ing, the desired hydrogen pressure was applied and the reac-
tion started. After 30 minutes the reaction was stopped and
conversion and ee were determined.
¬
Naili, A. Mortreux, F. Agbossou, L. Pelinski, J. Brocard,
J. Eilers, I. Reiners, V. Peper, J. Martens, Organometallics
2000, 19, 5723; d) J. F. Carpentier, A. Mortreux, Tetrahe-
dron: Asymmetry 1997, 8, 1083; e) K. Mashima, K. Kusa-
no, N. Sato, Y. Matsumura, K. Nozaki, H. Kumobayashi,
N. Sayo, Y. Hori, T. Ishizaki, S. Akutagawa, H. Takaya, J.
Org. Chem. 1994, 59, 3064.
[5] a) For a recent overview, see: H.-U. Blaser, Ch. Malan, B.
Pugin, F. Spindler, H. Steiner, M. Studer, Adv. Synth.
Catal. 2003, 345, 103; b) P. Gamez, F. Fache, M. Lemaire,
Tetrahedron: Asymmetry 1995, #6#3, 705.
[6] a) For a recent overview, see: M. Studer, H.-U. Blaser,
Adv. Synth. Catal. 2003, 345, 45; b) G. Z. Wang, T. Mal-
lat, A. Baiker, Tetrahedron: Asymmetry 1997, 8, 2133;
Transfer hydrogenation: Reducing agent i-PrOH/i-PrOK.
The catalysts were either prepared in situ from the metal pre-
cursor and the ligand (i-PrOH, reaction time 1 h, room temper-
ature) or an isolated complex was used.[11] The hydrogenation
was carried out using1 mmol of ketone in 10 mL of isopropyl
alcohol in the presence of 0.06 mL of a 1 M solution of i-
PrOK in i-PrOH. After 1 h stirringat room temperature, the
reaction was quenched with 0.05 mL of acetic acid, the product
extracted with 2 mL of ethyl acetate and the organic phase an-
alyzed.
Reducing agent HCOOH/NEt3: RuCl(p-cymene)(R,R-
NH2 NHTs)] and 1 mmol of ketone were dissolved in 2 mL
CH2Cl2, 1 mL of HCOOH/NEt3 (5:2) was added and stirred
for 16 h at room temperature. The reaction was quenched
with saturated NaHCO3 aqueous solution, extracted with
2 mL of ethyl acetate and the organic phase analyzed.
¬
c) M. Sutyinski, K. Szˆri, K. Felfˆldi, M. Bartok, Catal.
Commun. 2002, 2, 125; d) C. Exner, A. Pfaltz, M. Studer,
H.-U. Blaser, Adv. Synth. Catal. 2003, 345, 1253; e) H.-U.
Blaser, H. P. Jalett, F. Spindler, J. Mol. Catal. A: Chemi-
cal 1996, 107, 85.
[7] F. Spindler, H. U. Blaser, Enantiomer 1999, 4, 557.
[8] O. Itoh, T. Nagata, I. Nomura, T. Takanaga, T. Sugita, K.
Ichikawa, Bull. Chem. Soc. Jpn. 1984, 57, 810.
[9] I. T. Barnish, P. E. Cross, J. C. Danilewicz, R. P. Dickin-
son, D. A. Stopher, J. Med. Chem. 1981, 24, 399.
[10] F. Cederbaum, A. De Mesmaeker, A. Jeanguenat, H.-J.
Kempf, C. Lamberth, A. Schnyder, M. Zeller, R. Zeun,
Chimia 2003, 57, 680.
[11] K-J. Haack, S. Hashiguchi, A. Fujii, T. Ikariya, R.
Noyori, Angew. Chem. Int. Ed. Engl. 1997, 36, 285.
848
¹ 2004 WILEY-VCH VerlagGmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2004, 346, 842 848