On the stereochemical course of self-replication in secondary cycle sharpless aminohydroxylation

Article abstract of DOI:10.1002/adsc.200404209

An investigation on a potential asymmetric self-replication process within the secondary cycle of Sharpless asymmetric aminohydroxylation is discussed. The reaction of two model olefins is investigated within this process and two different models for an a

Full text of DOI:10.1002/adsc.200404209

FULL PAPERS
On the Stereochemical Course of Self-Replication in Secondary
Cycle Sharpless Aminohydroxylation
˜
Kilian Muniz
´
Kekule-Institut fꢀr Organische Chemie und Biochemie, Gerhard-Domagk-Str. 1, 53121 Bonn, Germany
Fax: (þ49)-228-73-5813, e-mail: kilian.muniz@uni-bonn.de
Received: July 3, 2004; Accepted: December 13, 2004
Abstract: An investigation on a potential asymmetric meric excesses as a function of olefin conversion
self-replication process within the secondary cycle of lead to the conclusion that for such processes there
Sharpless asymmetric aminohydroxylation is dis- is no possibility for any inhibitory effect and that
cussed. The reaction of two model olefins is investi- asymmetric self-replication in secondary cycle amino-
gated within this process and two different models hydroxylation reactions is not a feasible process.
for an asymmetric process are discussed. Analysis of
the results on the basis of these models and a mathe- Keywords: amino alcohols; aminohydroxylation; os-
matical description for development of the enantio- mium; oxidation; self-replication
Introduction
more reactive than their heterochiral counterparts re-
sulting in an overproportional removal of the minor
Asymmetric autocatalytic reactions are among the most enantiomer of the catalyst. A variety of other examples
fascinating processes in homogeneous catalysis.[1] Apart of non-linear effects have become available and, as a
from the fundamental interest regarding information on general feature, a delicate equilibrium between associ-
the self-replication of chiral enantiopure molecules, ex- ates of enantiomeric or diastereomeric entities is pre-
plicit knowledge on such processes might ultimately re- requisite for phenomena of this type.[15] On the other
veal an origin of homochirality and, thus, of the origin of hand, catalytic chiral self-replication can be character-
life.[2,3] Certainly, the best studied systems are autocata- ised by a scenario in which a chiral metal-ligand assem-
lytic diisopropylzinc additions to certain aldehydes, a re- bly catalyses the formation of a reaction product identi-
action that was pioneered by Soai.[4,5] Due to the fact that cal to the chiral ligand. In typical template catalysis,[16]
autocatalytic reactions produce additional amounts of such a reaction releases an enantiomerically enriched
chiral catalyst continuously, they are characterised by product which is identical to the chiral ligand that had
an increase in the overall reaction rate upon conversion. been responsible for its asymmetric formation. Unlike
In any event, asymmetric autocatalysis that does not in autocatalysis, this product does not complex a metal
give a 100% ee will inevitably lead to racemic product and thus does not influence subsequent catalytic cycles
unless the process is dominated by a chiral amplifica- (Figure 1).
tion.[2a] An initial model to explain the overall reaction
A single example of this type of reaction was reported
course for chiral autocatalytic dialkylzinc additions by Soai in 1997.[17] He described the enantioselective
was described by Soai.[6] Importantly, Soai has also de- transformation of a-amino ketones employing a chiral
vised autocatalytic reactions which occur with increases reducing agent, which was generated in situ from lithium
in enantioselectivity during the course of catalysis.[7] The aluminium hydride and chiral 1,2-amino alcohols. In
exact nature of such systems is not yet fully under-
stood,[8] and work by Blackmond,[9–11] Brown[9,12] and
Gridnev[12] suggested dimeric or tetrameric homochiral
structures to be the kinetically decisive catalysts. In ad-
ditional, recent experiments Blackmond has obtained
evidence for non-continuous rate increases and for the
involvement of heterochiral precipitates.[13] The origin
of chiral amplification and thus the general details on
the working mode of dimeric catalyst structures in cata-
lytic reactions was uncovered by Noyori.[14] He proved
Figure 1. Schematic representation of chiral autocatalysis
that for diethylzinc additions, homochiral dimers are (top) and chiral self-replication (bottom).
Adv. Synth. Catal. 2005, 347, 275–281
DOI: 10.1002/adsc.200404209
ꢁ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
275

˜
Kilian Muniz
FULL PAPERS
these stoichiometric processes, the original ligands were
reproduced with as high as 95:5 selectivity ratios.
We here discuss an investigation on the course of a
rare example of catalytic asymmetric self-reproduction
in oxidation chemistry and prove that the overall stereo-
chemical course of the reaction is influenced by the chi-
ral product.
Figure 3. Chiral ligands for self-replication via second-cycle
AA reaction.
Results and Discussion
Recently, Sharpless reported the use of certain chiral li-
gands[18] for efficient osmium-catalysed dihydroxylation
(AD) and aminohydroxylation (AA) reactions within
the so-called secondary catalytic cycle.[19–21] All these li-
gands share a free carboxylic acid group as a common
prerequisite for product displacement from the osmium
complex after olefin functionalisation. In view of such a
scenario, aminohydroxylation of acrylates should in
principle represent an ideal reaction for chiral, non-rac-
emic self-replication.[22]
Figure 4. Intermediate bis(azaglycolate) osmium esters.
Aminohydroxylation Reactions
which is not feasible in the traditional 1st cycle of amino-
hydroxylation with Cinchona alkaloids.[21]
The expected catalytic cycle for self-replicative amino-
In a first attempt to validate this assumption of asym-
hydroxylation is depicted in Figure 2 for acrylic acid as metric self-reproduction, aminohydroxylation of so-
the simplest case. In situ formation of the azaglycol os- dium acrylate in the presence of 1 after 46 h led only
mate ester 3 from enantiopure amino alcohol 1 and an an incomplete isolated yield of 31%. As a consequence,
osmium(VI) compound such as 2 is followed by oxida- attention was turned towards more elaborate ligands
tion with chloramine-T to furnish the catalyst 4. Oxida- such as 6 and 7.
tion of an acrylate leads to bis(azaglycolate) 5 which is
Subsequent oxidation reactions of fumarate in the
hydrolysed to give free amino alcohol 1 and regenerate presence of 3-amino-2-hydroxysuccinic acid 6 indeed
catalyst 4. Ideally, this sequence leads to a constant re- convinced us of the applicability of the concept. Free
production of the initial ligand 1 in enantiopure succinate 6 was obtained from dimethyl fumarate via
form.[23] Such a procedure would represent a solution standard Sharpless first cycle AA reaction (80% yield,
to the so far unknown AA reaction of free acrylates 75% ee)[24] followed by alkaline ester cleavage. Esterifi-
cation of the free diacid was accomplished cleanly with
the aid of Meerweinꢂs salt regenerating the original first
cycle AA product 8 without any loss in optical purity.
This method of esterification was subsequently em-
ployed in derivatisation of all the reaction products for
their evaluation via HPLC analysis.
Thus, when sodium fumarate was submitted to stoi-
chiometric or catalytic reactions, the original chiral li-
gand 6 was reproduced in good to high yields and with
an identical absolute configuration. However, the enan-
tiomeric excesses of the isolated amino alcohol 6 were
found to be significantly lower than at the beginning
where enantiomerically pure material was employed.
This reaction outcome can be rationalised since chiral
catalysts such as 4 are non-perfect and catalyse forma-
tion of the undesired enantiomer of the vicinal amino al-
cohol as well. In case of the present aminohydroxyla-
tion, formation of up to six different isomers of the inter-
mediate bis(azaglycolate) osmium ester might be in-
volved (Figure 4).
Figure 2. Catalytic cycle for self-replication in asymmetric
aminohydroxylation.
276
ꢁ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2005, 347, 275–281

Self-Replication in Secondary Cycle Sharpless Aminohydroxylation
FULL PAPERS
Table 1. Self-replication of amino alcohol 6.
S/C
Yield [%][a]
ee [%][b]
ee [%][c]
0.8
2
3
4
6
88
79
77
77
81
65
46
32
17
10
–
44.7
29.5
18.8
7.3
[a]
Isolated yield from complete conversion and after reduc-
tive work-up.
[b]
[c]
Combined ee for original and newly formed amino alco-
hol. Determined after conversion into the corresponding
methyl ester. HPLC conditions for 8: Chiralcel-OG, 2-
PrOH/n-Hex¼20/80, 1.0 mL/min, 6.8 min, 9.1 min.
Calculated value at quantitative conversion according to
Eq. (2) (see below).
The subsequent hydrolysis of these intermediates is
crucial for the overall stereochemical course. For hy-
drolysis rate, two extreme cases are possible that consist
of either exclusive hydrolysis of the homochiral species
5a,b or of their corresponding heterochiral isomers 5c,d.
The former case would represent an ideal scenario of
asymmetric amplification leading constantly to an ex-
clusive reproduction of enantiopure material while the
latter would inevitably give rise to racemic material.
Within this context, it should be noted that these sce-
narios are unrealistic for high S/C ratios since the
amount of active catalyst and thus the overall rate will
decrease constantly over time. Still, for the case of pre-
ferred hydrolysis of the homochiral intermediates, this
pathway devises the simplest model for asymmetric
self-reproduction of chiral entities without the previous-
ly postulated inhibition through formation of dimeric
heterochiral complexes.[14,15]
Scheme 1. Synthesis of ligand 7.
Table 2. Self-replication of amino alcohol 7.
S/C
Yield [%][a]
ee [%][b]
ee [%][c]
0.8
2
3
4
6
84
83
80
80
81
88
87
69
51
39
27.5
11
4
–
50
35
23.8
10.5
2
10
12
<1
0.8
[a]
Isolated yield from complete conversion and after reduc-
tive work-up.
[b]
Combined ee for original and newly formed amino alco-
hol. Determined after conversion into the corresponding
methyl ester. HPLC conditions for 11: Chiralcel-OG, 2-
PrOH/n-Hex¼30/70, 1.0 mL/min, 11.7 min, 18.2 min.
Calculated value at quantitative conversion according to
Eq. (2) (see below).
In order to rationalise the reaction course in the pres-
ent AA of fumarate, a stoichiometric reaction employ-
ing 125 mol % of 6 (S/C¼0.8) was carried out (Table 1,
entry 1). The reaction product was treated with sodium
sulphite and the combined enantiomeric excess of orig-
inal and newly formed 6 was determined after conver-
[c]
sion into the respective methyl ester 8. The obtained liquor, and after 3 consecutive recrystallisations enan-
ee of 65% calculates back to a 21.25% ee for the newly tiomerically pure product was obtained. TFA-induced
formed product. Thus, the overall process of asymmetric cleavage of the ester group gave then the desired ligand
self-replication leads to reproduction of an amino alco- 7 as a single enantiomer. Esterification of 7 with Meer-
hol with identical absolute configuration (er ¼0.60625). weinꢂs salt gave the enantiopure methyl ester 11 as a
In experiments 2–4, the enantiomeric excess of 6 un- standard for HPLC.
derwent a significant drop over time, depending on the
average turnover number.
As already observed for self-replications with 6, com-
pound 7 is again reproduced with a significant drop in
A similar trend was observed for self-replication of enantiomeric excess over time. While the enantiomeric
amino alcohol 7 by oxidation of sodium methacrylate excess for the newly formed product in the stoichiomet-
(Table 2). Ligand 7 was obtained from aminohydroxyla- ric control reaction was about 30%, all catalytic reac-
tion of tert-butyl methacrylate (9) under standard ami- tions gave products with a significantly lower ee than
nohydroxylation conditions (Scheme 1).[24] The respec- the original 100% from the enantiopure starting materi-
tive amino alcohol 10 was isolated in 77% chemical yield al.
with the rather low ee of 39%. Recrystallisation from 2-
This investigation on self-replication has so far been
propanol led to an enantiomerically enriched mother restricted to ligands 1, 6 and 7 since analysis of the reac-
Adv. Synth. Catal. 2005, 347, 275–281
asc.wiley-vch.de
ꢁ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
277

˜
Kilian Muniz
FULL PAPERS
tion course for structurally related olefins is complicated
In the case of reductive work-up as applied in the pres-
by the inherent regioselectivity problem of Sharpless ent case, the remaining catalyst is cleaved and additional
AA reactions.[21,25]
amounts of amino alcohol are released. As mentioned
In view of the observed decrease in amino alcohol ee, before, the enantiomeric excess of this additional
one might conclude identical hydrolysis rates for the ho- amount is identical to that of the newly released amino
mochiral intermediates 5a, b and their heterochiral iso- alcohol in the final turn-over prior to reductive work-
mers 5c, d. Kinetic control experiments indicate that up. Upon inclusion of this extra amount of amino alco-
there is indeed no change in reaction rate for reactions hol, the overall ee is given by:[26]
from Os(VI)/7 with methacrylate at S/C¼20 in the
range of 10–40% conversion (based on a pseudo-first
order rate law). This supports the assumption of identi-
cal hydrolysis rates since any difference in hydrolysis re-
ð2Þ
garding the species 5a, b and 5c, d must inevitably result
in a diminished overall reaction rate.
The observed decrease in ee is the obvious consequence
of non-perfect ligand reproduction. Application of
Eq. (2) to the respective reactions of aminohydroxyla-
tions with catalysts derived from 6 and 7 has been carried
out for all entries in Tables 1 and 2. The apparent good
Mathematical Rationalisation
For any general asymmetric reaction, which involves au- agreements between the enantiomeric excesses of the
tocatalytic behaviour without any kinetic differentiation isolated material and the calculated values imply that
between homo- and heterochiral intermediates whatso- there is indeed no rate difference in the respective hy-
ever, the enantiomeric excess een of the product is given drolysis of intermediates 5 and, hence, no inhibitory ef-
by the following expression:
fect can result. This is not a trivial observation and it
should be noted that even without any rate difference
in hydrolysis and thus without inhibitory effect, the pres-
ent reactions demand an insight into the exact stereo-
chemical course of their intermediates.
ð1Þ
in which ee0 denotes the enantiomeric excess of the initial
It is important to note that the decrease in enantio-
ligand, n the number of average turn-overs [at complete meric excess in the present self-replicating system pro-
conversion, n is given by 100C/S] and er the stereochem- ceeds more slowly than for conventional template catal-
ical reproduction fidelity of the ligand [with 100 er :100 ysis models[17,27] since hydrolysis of (S,R)-complexes
(1 – er) representing the enantiomeric ratio for the newly such as 5c and 5d still results in a 50% reproduction of
formed product in the stoichiometric control reaction].
the original catalyst with correct absolute configuration.
For the present case of ligand self-replication, the ami- This represents an over-proportional reproduction of
no alcohol product arising from hydrolysis is not a cata- the original stereochemistry and, for identical reproduc-
lyst itself and it is only the enantiomeric excess of the chi- tion fidelity, the herein discussed asymmetric self-repli-
ral catalyst that is of interest for the further stereochem- cation processes lead to newly generated catalysts with
ical course. Only on the basis of an identical hydrolysis an ee that is even higher than for related autocatalyses.
rate for homo- and heterochiral intermediates 5a, b However, this inherent advantage of the self-replication
and 5c, d, respectively, there will be release of a product procedure over any asymmetric autocatalysis regarding
with an identical ee with regard to the previous cycle newly formed catalyst is lost. This is due to the fact that
from the homochiral intermediates 5a, b, while hydroly- the original catalyst inevitably does not retain its high in-
sis of 5c, d releases both enantiomers on a 50% statisti- itial enantiomeric excess.
cal chance, thereby generating 50% of the original chiral
The present example thus represents a new, unex-
catalyst and 50% of its enantiomer. Thus, the respective plored example for asymmetric catalytic self-replication
enantiomeric excesses of released amino alcohol from proceeding via an L*2M intermediate which, upon re-
the nth turn-over and osmium-bound amino alcohol lease of one L*, regenerates the catalyst L*2M (Fig-
(i.e., the catalyst) after the nth turn-over are identical. ure 5). Since the original and the newly formed L* in
Therefore, any value for een can be correlated to eenꢀ1
.
L*2M are no longer distinguishable, any catalytic self-
Taking into account the different amounts of intermedi-
ates 5a, b and 5c, d, respectively, which directly result
from the stereochemical reproduction fidelity er, the
enantiomeric excess of all freeamino alcohol product af-
ter n turn-overs is given by Eq. (1) and the mathematical
expression of ee as a function of turn-over is formally
identical to the one for simple asymmetric autocatalysis
without chirality amplification.
Figure 5. Schematic representation of the present self-repli-
cation in asymmetric aminohydroxylation.
278
ꢁ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2005, 347, 275–281

Self-Replication in Secondary Cycle Sharpless Aminohydroxylation
FULL PAPERS
with a 2 M sodium sulphite solution. Extraction with dichloro-
methane, drying over MgSO4 and evaporation to dryness gives
the crude product, which is purified by column chromatogra-
phy (silica gel, ethyl acetate/n-hexane, 1/2, v/v).
replication in such a system without a reproduction fi-
delity of er ¼1 will inevitably suffer from a decrease in
enantiomeric excess as a function of turnover.
Conclusion
General Procedure for Secondary Cycle
Aminohydroxylation
We have described the first detailed investigation on the
stereochemical course of self-replication in secondary
cycle aminohydroxylation reactions. Since the isolation
of any intermediate from this catalytic cycle is still lack-
ing, the good agreement between stereochemical out-
come and mathematical description suggests that, on
the basis of equal hydrolysis rates, the overall catalytic
cycle is indeed best described as depicted in Figure 2.
More importantly, the stereochemical analysis of the
self-reproduction process itself implicitly proofs that
asymmetric self-replication in this catalytic reaction is
not a feasible process. Moreover, the absence of any in-
hibitory effect enforces an unavoidable decrease of
enantiomeric excess over time. Therefore, the search
for self-reproduction of other amino alcohols with high-
er enantioselective reproduction fidelity is of no value.
A solution of potassium osmate and the respective ligand 6 or 7
(1.1 equivs.) in a mixture of water and tert-butyl alcohol (1/1,
v/v) is stirred for 2 h at room temperature. Chloramine-T (1.2
equivs. with regard to the substrate plus 1 equiv. with regard
to the OsVI source) is added in one portion and the resulting
yellow solution is stirred for another 30 min before addition
of sodium methycrylate or sodium fumarate, respectively.
The reaction is left stirring for a period of up to 36 h and is
quenched by addition of solid thiosulphate (1.0 equiv. with re-
spect to chloramine-T). The resulting solution is extracted with
dichloromethane, dried over MgSO4 and evaporated to dry-
ness.
All reactions from Tables 1 and 2 were carried out at least
three times. The given ee values represent average values
and may vary within margins of 5%.
General Procedure for Esterification
A solution of the respective acid in freshly distilled dichloro-
methane (5 mL per mmol) under argon is cooled to 08C and
treated slowly with 1.1 equivs. of Meerweinꢂs salt. After stirring
for 6 h at room temperature, the reaction is extracted with bi-
carbonate solution, washed with brine and water and dried
over MgSO4 and evaporated to dryness to give the analytically
pure methyl esters 8 and 11, respectively, which were analysed
by analytical HPLC on a chiral stationary phase.
Experimental Section
General Remarks
Potassium osmate, tert-butyl methacrylate, sodium methacry-
late, sodium fumarate and chloramine-T were purchased
from Aldrich. Dimethyl fumarate and (DHQD)2PHAL were
purchased from Fluka. Dichloromethane was distilled from
CaH2 under argon. All other solvents were reagent grade and
used as received. Column chromatography was performed
with silica gel (Merck, type 60, 0.063–0.2 mm and Machery Na-
gel, type 60, 0.015–0.025 mm). Optical rotations were meas-
ured on a Perkin Elmer 341 polarimeter. Concentrations are
given in g/100 mL as dichloromethane solutions. NMR spectra
were recorded on a Bruker DPX 300 MHz spectrometer. All
chemical shifts in NMR experiments are reported as ppm
downfield from TMS. IR spectra were recorded on a Nicolet
Magna 550 FT-IR spectrometer. MS and HRMS experiments,
and elemental analysis were performed on a Kratos MS 50 and
an Elementar Analysensystem Vario EL, respectively, within
3-Tosylamino-2-hydroxy-2-methylpropionic Acid (7)
A solution of the ester 10 (990 mg, 3 mmol) in dichlorome-
thane at room temperature is treated with trifluoroacetic
acid/dichloromethane (10 mL, 1/1, v/v) and stirred at room
temperature for a period of 3 h. The solvents are removed un-
der reduced pressure to leave the title compound in form of a
colourless solid; yield: 800 mg (98%); mp 478C (decomp.); 1H
NMR (300 MHz, DMSO-d6): d¼1.28 (s, 3H), 2.44 (s, 3H),
2.80 (m, 1H), 3.04 (m, 1H), 3.73 (s, 3H), 7.43 (d, J¼8.3 Hz,
2H), 7.75 (d, J¼8.3 Hz, 2H); 13C NMR (75 MHz, DMSO-d6):
d¼21.23, 23.27, 51.14, 73.37, 126.98, 129.90, 137.93, 142.93,
175.96; IR (KBr): n¼3358, 3260, 3031, 2581, 1718, 1421,
1305, 1161, 1097, 908, 817, 704 cmꢀ1; MS (EI, eV): m/z (%)¼
228 (81), 184 (100), 155 (92), 91 (88), 57 (71); HR-MS: m/z
calcd. for C10H14NO3S (MꢀCO2): 228.0695; found: 228.0699;
anal. calcd. for C11H15NO5S: C 48.34, H 5.53, N 5.12, S 11.73;
found: C 48.02, H 5.31, N 4.76, S 12.29.
´
the service centres at the Kekule-Department, Bonn.
General Procedure for Sharpless
Aminohydroxylation[24]
A solution of potassium osmate, K2[OsO2(OH)4] (74 mg,
0.2 mmol) and (DHQD)2PHAL (0.2 g, 0.25 mmol) in 10 mL
of a water/tert-butyl alcohol solution (1/1, v/v) at room temper-
ature is stirred until both components are completely dis-
solved. Chloramine-T (4.32 g, 15 mmol, 3.0 equivs.) is added
in one portion and the resulting yellow solution is stirred for
20 min before the olefin (5 mmol) is added in one portion.
The resulting solution is stirred overnight and then quenched
Dimethyl 3-Tosylamino-2-hydroxysuccinate (8)
Synthesised by Sharpless aminohydroxylation according to the
general procedure (80% yield, 75% ee). This reaction had been
previously described.[24] Enantiomerically pure product (>
Adv. Synth. Catal. 2005, 347, 275–281
asc.wiley-vch.de
ꢁ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
279

˜
Kilian Muniz
FULL PAPERS
´
99% ee) was obtained through recrystallisation from metha-
nol. HPLC analysis: Chiralcel-OG, 308C, 1.0 mL/min, n-hex-
ane/2-propanol¼70/30, tR ¼11.7 min, 18.2 min.
[3] a) M. Avalos, R. Babiano, P. Cintas, J. L. Jimenez, J. C.
Palacios, Chem. Commun. 2000, 887; b) C. Bolm, F. Bi-
enewald, A. Seger, Angew. Chem. Int. Ed. 1996, 35,
1657; c) J. S. Siegel, Nature 2001, 409, 777.
[4] K. Soai, T. Shibata, H. Morioka, K. Choji, Nature 1995,
378, 767.
tert-Butyl 3-Tosylamino-2-hydroxy-2-
methylpropionate (10)
[5] a] K. Soai, I. Sato, T. Shibata, Chem. Rec. 2001, 1, 321;
b) K. Soai, T. Shibata, I. Sato, Acc. Chem. Res. 2000,
33, 382; c) K. Soai, T. Shibata, I. Sato, Bull. Chem. Soc.
Jpn. 2004, 77, 1063.
[6] I. Sato, D. Omiya, K. Tsukiyama, Y. Ogi, K. Soai, Tetra-
hedron: Asymmetry 2001, 12, 1965.
Synthesised by Sharpless aminohydroxylation according to the
general procedure (77% yield, 39% ee). Recrystallisation of
this sample from 2-propanol gave enantioenriched material
from the mother liquor. After three to four recrystallisations,
enantiomerically pure material (>99% ee) was obtained.
HPLC analysis: Chiralcel-OG, 208C, 1.0 mL/min, n-hexane/
2-propanol¼80/20, tR ¼15.7 min, 19.2 min; mp 688C (de-
[7] K. Soai, I. Sato, Chirality 2002, 14, 548.
[8] a) I. Sato, D. Omiya, H. Igarashi, K. Kato, Y. Ogi, K.
Tsukiyama, K. Soai, Tetrahedron Asymmetry 2003, 14,
975; b) D. G. Blackmond, Adv. Synth. Catal. 2002, 344,
156; c) T. Buhse, Tetrahedron: Asymmetry 2003, 14, 1055.
[9] D. G. Blackmond, C. R. McMillan, S. Ramdeehul, A.
Schorm, J. M. Brown, J. Am. Chem. Soc. 2001, 123,
10103.
[10] D. G. Blackmond, Proc. Natl. Acad. Sci. 2004, 101, 5732.
[11] F. G. Buono, D. G. Blackmond, J. Am. Chem. Soc. 2003,
125, 8978.
1
comp.); H NMR (300 MHz, CD3OD): d¼1.32 (s, 3H), 1.53
(s, 9H), 2.46 (s, 3H), 2.86 (d, J¼12.3 Hz, 1H), 3.25 (d, J¼
12.3 Hz, 1H), 7.42 (d, J¼8.2 Hz, 2H), 7.77 (d, J¼8.2 Hz,
2H); 13C NMR (75 MHz, CD3OD): d¼21.72, 24.16, 28.43,
75.62, 83.63, 128.31, 131.02, 139.04, 144.99, 175.29; IR (KBr):
n¼3477, 3361, 3282, 1732, 1461, 1415, 1334, 1284, 1160,
704 cmꢀ1. MS (EI, eV): m/z (%)¼329 [M]þ (1), 273 (9), 228
(42), 184 (100), 155 (100), 91 (96), 57 (66); HR-MS: m/z calcd.
for C15H23NO5S: 329.1297; found: 329.1288.
[12] a] I. D. Gridnev, J. M. Serafimov, J. M. Brown, Angew.
Chem. Int. Ed. 2004, 43, 4884; b) I. D. Gridnev, J. M.
Brown, Proc. Natl. Acad. Sci. 2004, 101, 5727.
[13] F. G. Buono, H. Iwamura, D. G. Blackmond, Angew.
Chem. Int. Ed. 2004, 43, 2099.
[14] R. Noyori, S. Suga, H. Oka, M. Kitamura, Chem. Rec.
2001, 1, 85.
[15] a] H. B. Kagan, Adv. Synth. Catal. 2001, 343, 227; 15b) C.
Girard, H. B. Kagan, Angew. Chem. Int. Ed. 1998, 37,
Methyl 3-Tosylamino-2-hydroxy-2-methylpropionate
(11)
Synthesised in optically active form by Sharpless aminohydrox-
ylation according to the generalprocedure.[24] Alternatively, rac-
11 was obtained after secondary cycle aminohydroxylation[18a]
and esterification according to the general procedure detailed
above. HPLC analysis: Chiralcel-OG, 208C, 1.0 mL/min, n-hex-
ane/2-propanol¼80/20, tR ¼6.8 min, 9.1 min; mp 928C (de-
comp.); 1H NMR (300 MHz, CD3OD): d¼1.36 (s, 3H), 2.46 (s,
3H), 2.98 (d, J¼13.0 Hz, 1H), 3.19 (d, J¼13.0 Hz, 1H), 3.73 (s,
3H), 7.41 (d, J¼8.2 Hz, 2H), 7.76 (d, J¼8.2 Hz, 2H); 13C
NMR (75 MHz, CD3OD): d¼21.71, 23.89, 52.31, 53.24, 75.77,
128.31, 131.01, 139.14, 145.01, 176.43. IR (KBr): n¼3520,
´
4000; c) M. Avalos, R. Babiano, P. Ciutas, J. L. Jimenez,
J. C. Palacios, Tetrahedron Asymmetry 1997, 8, 2997;
d) M. Reggelin, Nachr. Chem. Techn. Lab. 1997, 45, 392;
e) C. Bolm, in: Advanced Asymmetric Synthesis, (Ed.: R.
Stephenson), Chapman and Hall, London, 1996, p. 9.
[16] R. Noyori, Catalytic Asymmetric Synthesis, Wiley, New
York, 1994.
[17] T. Shibata, T. Takahashi, T. Soai, Angew. Chem. Int. Ed.
1997, 36, 2560.
[18] a) V. V. Fokin, K. B. Sharpless, Angew. Chem. Int. Ed.
2001, 40, 3455; b) H. C. Kolb, M. G. Finn, K. B. Sharpless,
Angew. Chem. Int. Ed. 2001, 40, 2004.
[19] a) M. A. Andersson, R. Epple, V. V. Fokin, K. B. Sharp-
less, Angew. Chem. Int. Ed. 2002, 41, 472; b) P. Dupau, R.
Epple, A. A. Thomas, V. V. Fokin, K. B. Sharpless, Adv.
Synth. Catal. 2001, 344, 421.
[20] Reviews on AD reactions: a) H. C. Kolb, M. S. VanNieu-
wenhze, K. B. Sharpless, Chem. Rev. 1994, 94, 2483; b) C.
3251, 2983, 1726, 1329, 1286, 1240, 1169, 1092, 839, 704 cmꢀ1
;
MS (EI, eV): m/z (%)¼287 [M]þ (2), 228 (12), 184 (93), 155
(100), 104 (47), 91 (98), 65 (37); HR-MS: m/z calcd. for
C12H17NO5S: 287.0827; found: 287.0832.
Acknowledgements
This work was supported by the Deutsche Forschungsgemein-
schaft (MU-1805/1–1) and the Fonds der Chemischen Industrie
(Liebig fellowship to K. M.). The author acknowledges S. Mu-
´
n˜iz-Fernandez for extensive discussions on mathematical descrip-
tions and Prof. K. H. Dçtz for his ongoing support and interest.
˜
Bolm, J. P. Hildebrand, K. Muniz, in: Catalytic Asymmet-
ric Synthesis, (Ed.: I. Ojima), Wiley-VCH, Weinheim,
2000; c) H. C. Kolb, K. B. Sharpless, in: Transition Metals
for Organic Chemistry: Building Blocks and Fine Chem-
icals, 2nd edn., (Eds.: M. Beller, C. Bolm), Wiley-VCH,
References
˜
Weinheim, 2004, Vol. 2, p. 275; d) K. Muniz, in: Transi-
[1] H. Wynberg, Chimia 1989, 43, 150.
[2] a) F. C. Frank, Biochim. Biophys. Acta 1953, 11, 32; b) M.
Calvin, Chemical Evolution, Oxford University Press,
Oxford, 1969.
tion Metals for Organic Chemistry: Building Blocks and
Fine Chemicals, 2nd edn., (Eds.: M. Beller, C. Bolm), Wi-
ley-VCH, Weinheim, 2004, Vol. 2, p. 298.
280
ꢁ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2005, 347, 275–281

Self-Replication in Secondary Cycle Sharpless Aminohydroxylation
FULL PAPERS
˜
[21] Reviews on AA reactions: a) K. Muniz, Chem. Soc. Rev.
[24] G. Li, H.-T. Chang, K. B. Sharpless, Angew. Chem. Int.
Ed. Engl. 1996, 35, 451.
2004, 33, 166; b) J. A. Bodkin, M. D. McLeod, J. Chem.
Soc. Perkin Trans. 1 2002, 2733; c) H. C. Kolb, K. B.
Sharpless, in: Transition Metals for Organic Chemistry:
Building Blocks and Fine Chemicals, 2nd edn., (Eds.:
M. Beller, C. Bolm), Wiley-VCH, Weinheim, 2004, Vol.
[25] Attempts to investigate related self-replication processes
in dihydroxylation reactions met with complete failure.
For example, when tartaric acid was employed as ligand
in the dihydroxylation of fumaric acid with N-methyl-
morpholine N-oxide as terminal oxidant, only low con-
version and unsatisfying enantiomeric excess was ob-
served. Even for reaction in the presence of 25 mol %
of Os and employing 75% of tartaric acid, only a very
low enantiomeric excess of 5% was determined for the
newly formed tartaric acid.
[26] This equation is derived from the observation that the
enantiomeric excess for the fraction of all amino alcohol
from the (n – 1) turn-overs prior to work-up is calculated
to (n – 1) ee0 ern–1 and the fraction corresponding to the
final amount of amino alcohol is 2 ee0 ern. The sum of
both is then the following: (n – 1) ee0 ern–1 þ2 ee0 ern.
To arrive at the normalised een, this sum is divided by
the total of fractions of amino alcohol which is thus
(nþ1). Eq. (2) is obtained upon simple transformation.
[27] P. D. Bailey, J. Chem. Soc. Chem. Commun. 1995, 1797.
˜
2, p. 309; d) K. Muniz, in: Transition Metals for Organic
Chemistry: Building Blocks and Fine Chemicals, 2nd
edn., (Eds.: M. Beller, C. Bolm), Wiley-VCH, Weinheim,
2004, Vol. 2, p. 326; see also ref.[20b]
[22] A recent review on asymmetric aminohydroxylation has
posed the question on whether the AA could be carried
out as an autocatalytic process: D. Nilov, O. Reiser, Adv.
Synth. Catal. 2002, 344, 1169.
[23] A genuine autocatalytic course would require continuous
complexation of released 1 to additional amounts of os-
mium(VI) precursor 2 in order to re-enter the catalytic
cycle. However, such an additional amount of Os(VI)
species to be added in stoichiometric amounts from the
beginning would not be kinetically inert under the reac-
tion conditions. Instead, it would react with the present
nitrene precursor to form free OsO3NTos or OsO4 and,
thus, would inevitably give rise to a racemic competition
reaction.
Adv. Synth. Catal. 2005, 347, 275–281
asc.wiley-vch.de
ꢁ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
281