A novel chemoentrapment approach for supportless recycling of a catalyst: Catalytic asymmetric dihydroxylation

Article abstract of DOI:10.1002/adsc.200505351

A simple method of recycling a metal catalyst through chemoentrapment in an aqueous layer using ethyl vinyl ether has been developed. Using this new methodology, a highly efficient, filtration-free recycling of osmium for catalytic asymmetric dihydroxylation was accomplished. By means of the formation of a water-soluble OsO₄^2- using EVE, AD reactions of mono- and disubstituted olefins with 1 mol % of OsO₄ proceeded for up to 9 cycles without any loss of yields and enantioselectivities.

Full text of DOI:10.1002/adsc.200505351

COMMUNICATIONS
DOI: 10.1002/adsc.200505351
A Novel Chemoentrapment Approach for Supportless Recycling
of a Catalyst: Catalytic Asymmetric Dihydroxylation
Daewon Lee,a Honggeun Lee,a Seyoung Kim,a Chang-Eun Yeom,a
and B. Moon Kima,*
a
School of Chemistry, College of Natural Sciences, Seoul National University, Seoul 151-747, South Korea
Fax : (+82)-2-872-7505; e-mail: kimbm@snu.ac.kr
Received: September 9, 2005; Revised: March 29, 2006; Accepted: April 6, 2006
Abstract: A simple method of recycling a metal cat-
alyst through chemoentrapment in an aqueous layer
using ethyl vinyl etherhas been developed. Using
this new methodology, a highly efficient, filtration-
free recycling of osmium for catalytic asymmetric
dihydroxylation was accomplished. By means of the
formation of a water-soluble OsO42À using EVE,
AD reactions of mono- and disubstituted olefins
with 1 mol% of OsO4 proceeded for up to 9 cycles
without any loss of yields and enantioselectivities.
lization techniques such as microencapsulation of
[7]
OsO4 in a polymermatrix,
fixation of osmium onto
hindered olefins covalently bound on silica or resin
supports,[8] and use of various ion-exchange supports
have been reported.[9] Although successful, these ap-
proaches require the preparation of the support and
feature relatively short repetition cycles. The use of
an ionic liquid[10] as yet anotherimmobilization
medium forthe Os-ligand complex enabled ercycling
of both components forseveral cycles.
In the search for a new approach for recycling the
catalytic system, we focused ourattention on the de-
tailed catalytic cycle of AD. In a typical AD reaction
Keywords: asymmetric dihydroxylation; chemoen-
trapment; ethyl vinyl ether; osmium tetroxide; sup-
portless recycling
employing K3Fe(CN)6 as
a
secondary oxidant,
Os(VIII)O4, upon reaction with an olefin, is convert-
ed to an osmate ester[Os(VI)], which is hydorlyzed
and oxidized to Os(VIII) in the aqueous layerwith a
secondary oxidant. It then migrates back to the organ-
Developments in the preparation of enantiomerically ic phase to repeat the catalytic cycle.[2b] Aftercon-
pure compounds are becoming increasingly impor- sumption of all the olefin, osmium remains as
tant. Among many methods forthe geneartion of
Os(VIII) due to the excess of K3Fe(CN)6. This
chiral non-racemic compounds, much attention has Os(VIII) can stay as a perosmate anion or neutral
been focused on catalytic asymmetric transformations OsO4 and the majority of the latter would stay in or-
and recently remarkable advances have been made in ganic layer since it is more soluble in the organic sol-
this area.[1] Among the many asymmetric reactions, vent[11] and this may be where a major osmium loss
the osmium-catalyzed asymmetric dihydroxylation occurs when one wants to recycle the catalyst.
(AD) of olefins is one of the most efficient reactions
To verify this, osmium contents in the aqueous
for generating enantiomerically enriched vicinal layerweer measuerd with inductively couple plasma
diols.[2] However, as in many catalytic asymmetric (ICP) mass after the AD reaction of styrene under
transformations, the high cost and toxicity of the the standard conditions employing K3Fe(CN)6.[12]
metal-ligand system have been hurdles against large- After a single AD reaction, only 50% of the original
scale industrial applications of the AD reaction.[3] To osmium was found in the aqueous layer, meaning that
address this issue, various heterogeneous support sys- the other 50% was extracted into organic layer and
tems have been employed including insoluble inor- when the AD was repeated using the recovered aque-
ganic, insoluble organic, and soluble organic materi- ous layer, the reaction yields dropped precipitously at
als.[4,5] Since the heterogeneous supports have been the 3rd run, indicating a significant osmium loss
aimed largely at the recycling of the alkaloid ligand, during the extraction process (entry 1, Table 1). To
recovery of the osmium species still remains to be a overcome this intrinsic problem, it would be necessary
challenging problem. This appears to be closely relat- to confine the osmium in the aqueous layerin a form
ed to the low binding affinity of OsO4 to the mono- which is soluble only in the aqueous layer.
dentate alkaloid ligand, a must forthe catalytic turn-
To chemically trap all the osmium as water-soluble
over.[6] Recently, new, successful examples of immobi- Os(VI), conversion of all the remaining Fe(III) to
Adv. Synth. Catal. 2006, 348, 1021 – 1024
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1021

Daewon Lee et al.
COMMUNICATIONS
Table 1. Recycling results from consecutive use of aqueous layer in ADs of styrene and Os contents in the aqueous layer.
Entry
Reducing agent
Yield [%][a]
1st
Os content in H2O [%][b]
2nd
3rd
EA
TBME
1
2
3
none
Na2SO3
EVE
97
97
88
92
55
50
48
-
89
50
-
83
>90% for up to 9 cycles[c]
[a]
Recycling was performed using the recovered aqueous layer with a new batch of olefin, DHQ2PHAL, K3Fe(CN)6 and
K2CO3.
[b] Determined with ICP mass after extraction with indicated organic solvent from the first reaction.
[c]
For more detailed results, see Table 2.
Fe(II) after the AD reaction would be required. For Table 1) and dihydroxylation products from EVE
this purpose, use of an inorganic reductant would be were mostly water-soluble, leaving almost no isolable
the most straightforward approach. However, when material in the organic layer. ICP mass analysis of the
Na2SO3, a typical inorganic reductant, was employed, product diol showed 7.5% of Os in the organic com-
reduction of remaining Fe(III) using an exact amount pound afterthe treatment of 40 mol% of EVE.
of the inorganic reducing agent was tricky and excess
Encouraged by the above preliminary results, we
Na2SO3 would be harmful in the next reaction. Even performed AD reactions of olefins with 1 mol% of
the carefully controlled use of Na2SO3 did not im- OsO4 in presence of DHQ2PHAL (1.5 mol%) and
prove the recycling efficiency (entry 2, Table 1).
K3Fe(CN)6/K2CO3 (each 250 mol%) in t-BuOH and
In this connection, we set the criteria for a success- water and explored the repetition capability using
ful chemical trapping agent: (a) complete reduction EVE. Aftereach eraction cycle, 40 mol% of EVE
of the secondary oxidant [from Fe(III) to Fe(II)] for was added to the reaction mixture to generate
the generation of Os(VI) from Os(VIII), and (b) no OsO42À. The Os-containing aqueous layerwas almost
trace of the reductant in the aqueous layer after reac- quantitatively recovered from simple phase separation
tion. Use of an external olefin after the AD would using ethyl acetate (EA) or tert-butyl methyl ether
satisfy the above criteria, however, it will be accompa- (TBME) as an extracting solvent. Then the recovered
nied by product contamination from the external aqueous layer was reused for the next reaction with
olefin and its diol product. As a solution to this, new batches of an olefin, t-BuOH, DHQ2PHAL,
ethyl vinyl ether(EVE) was selected as an optimal re-
K3Fe(CN)6 and K2CO3. From the organic layer,
ducing olefin due to its low boiling point (338C) and DHQ2PHAL was also recovered by simple acid/base
the generation of a water-soluble dihydroxylated extraction (>95% recovery) and AD products were
product, glycoacetal, upon dihydroxylation (B to C, obtained without further purification. As recycling
Figure 1).[13] If needed, excess EVE can be easily re- proceeded, cumulative deposition of insoluble
covered through distillation. Indeed when 40 K4Fe(CN)6 in the aqueous layerwas observed, which
mol%[14] EVE was added to the mixture after com- eventually interfered with magnetic stirring and de-
pletion of the AD reaction, a majority of the osmium creased reactivity.[15] To alleviate the deposition prob-
(89%) was found in the aqueous layer(enytr3,
lem, filtration or dilution with an equal volume of
2À
Figure 1. Schematic diagram of recycling osmium through OsO4 using EVE.
1022
www.asc.wiley-vch.de
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2006, 348, 1021 – 1024

Chemoentrapment Approach for Supportless Recycling of a Catalyst
COMMUNICATIONS
Table 2. Results of repetitive asymmetric dihydroxylation using chemoentrapment strategy.
Entry Olefin
Extracting Yields and % ees[a] (in parenthesis) with consecutive use of aqueous layer[b]
solvent
1st
2nd
3rd
4th
5th
6th
7th
8th
1
2
3
4
Styrene
Styrene
EA
97 (96.2) 96 (95.8) 91 (95.7)
97 (97.0) 93 (96.8) 96 (96.6)
96 (89.6) 97 (90.7) 97 (89.9)
94 (95.8) 93[c]
(95.6)
91 (96.5) 91[c]
94
(95.4)
91
94
(95.8)
94
(96.9)
97
(89.5)
-
66 (96.6)
TBME
TMBE
TBME
92[e]
(96.9)
96 (89.6)
(96.3)
97 (90.0) 97
(96.9)
a-Methylstyr-
ene
trans-Stilbene
97[d]
(90.0)
96
(89.8)
-
96
97
95[c]
81
-
(>99.5)
(>99.5)
(>99.5)
(>99.5)
(>99.5)
[a]
Determined from HPLC analysis.
[b] Reactions were carried out with a molar ratio of olefin/OsO4/DHQ2PHAL=1/0.01/0.015 understandard K 3Fe(CN)6 con-
ditions at room temperature. Recycling was performed using a new batch of olefin, DHQ2PHAL and K3Fe(CN)6/K2CO3
(each 2.5 mol equivs.).
[c]
After reaction, insoluble materials were filtered off.
[d] After reaction, the aqueous layer was diluted with an equal volume of water.
[e]
Product of 96.9% ee in 94% yield was obtained at the 9th run.
waterof the aqueous layeronce after4 or5 cycles
In summary, we have developed a new chemoen-
was found to be effective. As shown in Table 2, the trapment strategy for recycling osmium in catalytic
AD reaction of styrene with 1 mol% of OsO4 pro- AD reactions using ethyl vinyl ether as a chemical
ceeded forup to seven cycles without any loss of
trapping agent. The new strategy allows for an effi-
yields in a given reaction time using ethyl acetate as cient recycling of osmium so that AD reactions of
an extracting solvent, however, a decreased yield mono- and disubstituted olefins with 1 mol% of
(66%) was observed in the 8th run (entry 1). When OsO4 proceeded for up to 9 cycles without any loss of
the extracting solvent was switched to TBME, the yields and enantioselectivities. This new catalyst recy-
same reaction proceeded for more than 9 cycles cling system is expected to lay the groundwork to-
(entry 2) presumably due to the lower polarity of wards practical applications of AD. The possibility of
TBME. Recycling of a-methylstyrene, a 1,1-disubsti- extending this chemoentrapment strategy to the cata-
tuted olefin, also exhibited almost the same reactivity lyst recovery of other metal catalytic systems is cur-
upon dilution once of the aqueous layerwithout any
filtration (entry 3). ADs of trans-stilbene, which is
structurally more demanding and thus more sensitive
to the osmium content than the other substrates, pro-
ceeded for 5 cycles (entry 4) with uniform enantiose-
lectivities. This contrasts with the AD of stilbene in
ionic liquid using 1 mol% of OsO4, where a yield de-
crease (70%) and lowered enantioselectivity (98 to
94% ee) were observed at the 4th run.[10b] The chemo-
entrapment of osmium using EVE allows for essen-
tially homogeneous reaction conditions in each recy-
cling experiment, therefore the highest enantioselec-
tivities were warranted. As can be seen from Table 2,
rently being explored.
Experimental Section
Representative Recycling Procedure for ADs of
Olefins using EVE
DHQ2PHAL (6 mg, 0.008 mmol), K3Fe(CN)6 (411 mg,
1.25 mmol), and K2CO3 (157 mg, 1.25 mmol) were dissolved
in t-BuOH-H2O (each 2 mL). Aftercooling, an olefin
(0.500 mmol) and an aqueous solution of OsO4 (4 wt%,
0.005 mmol) were added and the mixture was vigorously
the same levels of enantioselecitivty as those of the stirred at room temperature. After completion of the reac-
tion, EVE (20 mL, 0.200 mmol) was added and the mixture
stirred for 1 h to consume the remaining K3Fe(CN)6. The re-
action mixture was diluted with TBME (4 mL), cooled to
08C and then allowed to stand for 1 h. The extracted organic
layerwas washed with 1 N aqueous HCl solution and brine
to recover DHQ2PHAL. Afterermoval of the solvent and
EVE (bp 338C) under reduced pressure, the desired cis-diol
initial runs were maintained throughout the recycling
with all four olefins examined (entries 1–4). More-
over, products were obtained without contamination
since the AD product of EVE was accumulated in the
aqueous layer. Therefore, in regards to recycling effi-
ciency of the osmium, consistent enantioselectivity,
and ease of operation, this chemoentrapment of
osmium using EVE offers unique benefits compared
to existing osmium immobilization methods including
solid supports and ionic liquids.
was obtained without further purification. Then remaining
2À
aqueous layercontaining the OsO
was reused for the
4
next reaction employing new batches of reactants. When too
much solid was deposited in the aqueous layerupon succes-
Adv. Synth. Catal. 2006, 348, 1021 – 1024
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.asc.wiley-vch.de
1023

Daewon Lee et al.
COMMUNICATIONS
sive recycling, two different methods were used to alleviate
the stirring problem: 1) For entries 1, 2, and 4 in Table 2,
the reaction mixture was filtered to remove the insoluble
precipitate and the filtered insoluble deposition was washed
with t-BuOH-H2O (2.0 and 0.7 mL, respectively). 2) For
entry 3, dilution with water (2 mL) after the 5th cycle was
performed without filtration and the diluted aqueous layer
was used for further recycling experiments. The enantiomer-
ic excesses of the diols were determined from HPLC analy-
sis using chiral columns.
[5] Results from this laboratory: H. M. Lee, S. W. Kim, T.
Hyeon, B. M. Kim, Tetrahedron: Asymmetry 2001, 12,
1537.
[6] H. C. Kolb, P. G. Andersson, K. B. Sharpless, J. Am.
Chem. Soc. 1994, 116, 1278.
[7] a) S. Kobayashi, M. Endo, S. Nagayama, J. Am. Chem.
Soc. 1999, 121, 11229; b) S. Kobayashi, T. Ishida, R.
Akiyama, Org. Lett. 2001, 3, 2649; c) S. V. Ley, C. Ram-
arao, A. Lee, N. Østergaard, S. C. Simth, I. M. Shirley,
Org. Lett. 2003, 5, 185; d) T. Ishida, R. Akiyama, S. Ko-
bayashi, Adv. Synth. Catal. 2003, 345, 576; e) T. Ishida,
R. Akiyama, S. Kobayashi, Adv. Synth. Catal. 2005,
347, 1189.
[8] a) A. Sevreyns, D. E. De Vos, L. Fiermans, F. Verpoort,
P. J. Grobet, P. A. Jacobs, Angew. Chem. Int. Ed. 2001,
40, 586; b) J. W. Yang, H. Han, E. J. Roh, S. G. Lee,
C. E. Song, Org. Lett. 2002, 4, 4685.
[9] a) B. M. Choudary, N. S. Chowdari, M. L. Kantam,
K. V. Raghavan, J. Am. Chem. Soc. 2001, 123, 9220;
b) B. M. Choudary, N. S. Chowdari, K. Jyothi, M. L.
Kantam, J. Am. Chem. Soc. 2002, 124, 5341.
Acknowledgements
This work was supported by the Korea Research Foundation
Grant funded by the Korean Government (MOEHRD)
(KRF-2005–041-C00247). D.L., H.L., S.K. and C.-E.Y. thank
the BK21 fellowships from the Ministry of Education, Re-
public of Korea.
[10] a) L. C. Brano, C. A. M. Afonso, Chem. Commun.
2002, 3036; b) C. E. Song, D. Jung, E. J. Roh, S. Lee,
D. Y. Chi, Chem. Commun. 2002, 3038; c) L. C. Branco,
C. A. M. Afonso, J. Org. Chem. 2004, 69, 4381.
References
[11] Forexample, solubility of OsO in CCl4 and wateris
4
[1] a) Comprehensive Asymmetric Catalysis, (Eds.: E. N.
Jacobsen, A. Pfaltz, H. Yamamoto), Springer, Berlin,
1999, Vols. 1–3; b) Catalytic Asymmetric Synthesis,
(Ed.: I. Ojima), VCH, New York, 2000.
[2] Forreviews, see: a) H. C. Kolb, M. S. Van Nieuwenhze,
K. B. Sharpless, Chem. Rev. 1994, 94, 2483; b) R. A.
Johnson, Catalytic asymmetric dihydroxylation discov-
ery and development, in: Catalytic Asymmetric Synthe-
sis, (Eds.: I. Ojima, K. B. Sharpless), VCH, New York,
2000, pp. 357–398.
[3] a) Z.-M. Wang, K. B. Sharpless, J. Org. Chem. 1994, 59,
8302; b) L. Ahrgren, L. Sutin, Org. Process Res. Dev.
1997, 1, 425; c) X. Lu, Z. Xu, G. Yang, Org. Process
Res. Dev. 2000, 4, 575.
3.75 and 0.07 gmLÀ1, respectively: L. H. Anderson,
D. M. Yost, J. Am. Chem. Soc. 1938, 60, 1822.
[12] a) H. L. Kwong, C. Sorato, Y. Ogino, H. Chen, K. B.
Sharpless, Tetrahedron Lett. 1990, 31, 2999; b) M.
Minato, K. Yamamoto, J. Tsuji, J. Org. Chem. 1990, 55,
766.
[13] N. A. Milas, S. Sussman, H. S. Mason, J. Am. Chem.
Soc. 1939, 61, 1844.
[14] In theory, there should be 0.5equivs. K3Fe(CN)6 re-
maining after the AD reaction in the reaction mixture,
since originally 2.5equivs. K3Fe(CN)6 were used to
ensure excess secondary oxidant. This would require
0.25equivs. of EVE. We used 0.40 equivs. EVE to
ensure to counter this 0.25equivs. excess of oxidant.
[15] Efforts to use N-methylmorpholine oxide as a secon-
dary oxidant in the EVE-mediated chemoentrapment
experiment were unsuccessful. Recycling experiments
showed a precipitous drop in yields of AD at the third
repetition.
[4] Forerviews, see: a) C. Bolm, A. Gelrach,
Eur. J. Org.
Chem. 1998, 21; b) Q. H. Fan, Y. M. Li, A. S. C. Chan,
Chem. Rev. 2002, 102, 3385; c) C. E. Song, S. G. Lee,
Chem. Rev. 2002, 102, 3495; d) D. E. D. Vos, M. Dams,
B. F. Sels, P. A. Jacobs, Chem. Rev. 2002, 102, 3615.
1024
www.asc.wiley-vch.de
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2006, 348, 1021 – 1024