Poly(styrene) resin-supported cobalt(III) salen cyclic oligomers as active heterogeneous HKR catalysts

Article abstract of DOI:10.1002/adsc.201200528

A cross-linked poly(styrene) support functionalized with cobalt(III) salen cyclic oligomers that can be used as a catalyst for the hydrolytic kinetic resolution (HKR) of terminal epoxides is reported. This catalyst is the most active heterogeneous catalyst to date for the HKR of terminal epoxides and can be recycled more than six times with excellent enantioselectivities for the HKR of epichlorohydrin. A 3-fold rate enhancement was observed when conducting the HKR reaction with 6 equivalents of water compared to 0.6 equivalents. We hypothesize that this rate enhancement is due to water sequestration of the diol product from the organic phase, thereby maintaining a high local concentration of epoxides and catalyst in the organic phase. Copyright

Full text of DOI:10.1002/adsc.201200528

FULL PAPERS
DOI: 10.1002/adsc.201200528
Poly
N
U
Oligomers as Active Heterogeneous HKR Catalysts
Michael G. C. Kahn,a Joakim H. Stenlid,a and Marcus Wecka,*
a
Department of Chemistry and the Molecular Design Institute, New York University, New York, New York 10003-6688,
USA
Fax : (+1)-212-995-4895; phone (+1)-212-992-7968; e-mail: Marcus.Weck@nyu.edu
Received: June 15, 2012; Revised: August 17, 2012; Published online: November 4, 2012
Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/adsc.201200528.
Abstract: A cross-linked poly
ACHTUNGERTN(NUNG styrene) support func- ducting the HKR reaction with 6 equivalents of
tionalized with cobalt(III) salen cyclic oligomers that water compared to 0.6 equivalents. We hypothesize
AHCTUNGTRENNUNG
can be used as a catalyst for the hydrolytic kinetic that this rate enhancement is due to water sequestra-
resolution (HKR) of terminal epoxides is reported. tion of the diol product from the organic phase,
This catalyst is the most active heterogeneous cata- thereby maintaining a high local concentration of ep-
lyst to date for the HKR of terminal epoxides and oxides and catalyst in the organic phase.
can be recycled more than six times with excellent
enantioselectivities for the HKR of epichlorohydrin. Keywords: bimetallic catalysis; heterogeneous cataly-
A 3-fold rate enhancement was observed when con- sis; recycling; supported catalysis
Introduction
meric supports.[8] A drawback of these soluble sup-
ported catalysts is their removal from the reaction
The production of pharmaceuticals, agro-chemicals, mixture.[9] Products must be removed from the cata-
and natural products often requires compounds with lyst by precipitation, vacuum distillation or other
defined stereochemistry. Enantiopure chiral epoxides, energy-intensive processes that often create signifi-
in particular, are attractive for these applications.[1]
A
cant waste.
wide range of reagents can open chiral epoxides with-
Heterogeneous-supported catalysis is focused on
out racemization making these epoxides very versatile ease of recovery of a catalytic species from the reac-
compounds for the organic chemist.[1a,2] The hydrolytic tion medium through filtration.[9] The Jacobsen
kinetic resolution (HKR) of terminal epoxides using group[5c] has synthesized heterogeneous cobalt salen
a cobalt
opening reaction.[3] It serves on an industrial scale to was that higher loadings were required to obtain the
prepare chiral epoxides and diols.[4] Cobalt
(III) salen performance of the molecular cobalt salen catalyst.
ACHTUNGTRENNUNG(III) salen catalyst is one of these epoxide- catalysts grafted to polyACHTUNGTRENN(UGN styrene) resins. A drawback
AHCTUNGTRENNUNG
complexes have been supported on various materials One reason for the reduced performance is catalytic
with the aim of increasing the activity of the salen cat- site isolation on the support, which does not facilitate
alysts and/or to recover the metal species from the re- the bimetallic interactions necessary for HKR. The
action mixture.[5]
Weck and Jones groups produced a more active heter-
Several research groups have designed homogene- ogeneous HKR catalyst from copolymers composed
ous supports that facilitate the bimetallic mechanism of salen-functionalized styrene and styrene monomers
of the HKR.[6] They were able to synthesize highly grafted from silica.[10] The salen units being in close
active catalysts by placing the metal centers in close proximity by virtue of the polymer backbone allowed
proximity to each other.[7] Our group has reported the enhanced bimetallic cooperation. However, despite
most active HKR catalyst composed of cyclic oligo- several attempts cited in the literature, heterogene-
mers via ring-expanding olefin metathesis of cyclooc- ous-supported catalysts are in most cases not as active
tene salen monomers. The resultant structures have and selective as their homogeneous counterparts.
the salen units freely dangling off the macrocyclic Therefore it would be highly advantageous to obtain
backbone.[7b] We suggest that the high activity of this a HKR catalyst that is both highly active and easily
catalyst is due to the flexibility of these cyclic oligo- recoverable. Herein we report such a system by im-
3016
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2012, 354, 3016 – 3024

PolyACHTUNGTRENNUNG(styrene) Resin-Supported CobaltACHTUTGNREN(NUGN III) Salen Cyclic Oligomers
mobilizing the highly active cyclic oligomeric salen
Initially, we investigated a grafting-to strategy for
catalysts onto insoluble poly(styrene) resins. This the functionalization of Merrifield resins with 8. How-
AHCTUNGTRENNUNG
strategy yields a highly active catalyst that can be ever, the grafting density was unacceptably low and
easily recovered and recycled. Under our reaction we investigated the use of Wang resins instead. Fol-
conditions, we are able to separate the catalyst, the lowing a literature procedure, we modified Wang
enantiopure epichlorohydrin, and the diol product resins to install trichloroimine leaving groups using
from each other using filtration and decantation tech- DBU forming resin 9 (Scheme 2).[12] Using catalytic
niques. The resin-supported catalyst itself is easily BF3, a grafting density of 27% (based on 0.9 mmolgÀ1
synthesized and metallated with only simple filtration initial resin functionalization) of 8 to the surface of
and washing procedures.
the resin was achieved according to CHN analysis.
Compound 12, the oligocyclooctene-functionalized
Wang resin (Scheme 3), was then synthesized by swel-
ling 10 in dichloromethane for one hour followed by
the addition of Grubbs 3rd generation initiator and fi-
Results and Discussion
Aldehyde 3 was synthesized by Williamson ether syn- nally cyclooctene-salen 11 forming an oligomer con-
thesis from compounds 2 and 1. Monomer 8 was syn- taining units of 8 and 11.[13]
thesized by modification of a literature procedure
Proton NMR magic angle spinning spectroscopy ex-
(Scheme 1) and is bifunctional with a polymerizable periments were conducted on functionalized resins 10
cyclooctene group on one side and a C10 alkyl chain and 12 to examine for olefin resonances (Figure 1).
terminated with a hydroxy group for attachment to We rationalized that the resins should only show reso-
the resin on the other.[11]
nances for the carbon-carbon double bonds due to
Scheme 1. Synthesis of monomer 8.
Scheme 2. Functionalization of Wang resin (W.R.) with 8.
Adv. Synth. Catal. 2012, 354, 3016 – 3024
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
3017

Michael G. C. Kahn et al.
FULL PAPERS
Scheme 3. Polymerization on the resin surface.
acetate tetrahydrate. For this, the supported ligands
were swelled in dichloromethane for one hour. Cobalt
acetate was dissolved in methanol and added drop-
wise to the swollen resin under an inert atmosphere.
The resin changed from a yellow to a brick-red color
after several minutes and after an hour to a deep
purple. The reaction mixture was stirred for four
hours followed by washing three times with a 5:1 so-
lution of dichloromethane and methanol. The result-
ing material was filtered and washed thoroughly with
dichloromethane and dried under high vacuum to
obtain pre-catalyst 13 (Scheme 5). Cobalt content was
determined by ICP-MS to be 1.61% cobalt for 13a. In
repeating the synthesis of 13 the cobalt content was
1
Figure 1. Partial magic angle spinning H NMR spectrum of
the olefin region of Wang resins 10 (5.7 ppm) and 12
(5.4 ppm) compared to trichloroimine Wang resin 9.
either the cyclooctene monomers on 10 or the oligo- lower at 1.41%, denoted pre-catalyst 13b. There was
meric cycles for 12. Consistent with what we have ob- also a reduction in nitrogen according to CHN analy-
1
served before in solution using H NMR spectroscopy, sis indicating less ligand on the surface of the resin
the olefin peaks after polymerization shift upfield rather than an inefficient metallation. Differences in
from 5.7 ppm for the monomer to 5.4 ppm for the oli- activity were observed between these catalysts where
gomers.[7b] This finding indicates the formation of resins with higher cobalt content showed better activi-
ROMP products but does not give any details about ty. We hypothesize the correlation between activity
the linear or cyclic structure of the polymers/oligo- and cobalt content is dependent on the local concen-
mers that are formed.
tration of cobaltACTHNGUTERNNU(G III) salen units. Resins with higher
MALDI-TOF spectrometry can determine if the cobalt content have more catalytic sites in closer
styrene group from the initiator and the vinyl group proximity thereby enhancing the bimetallic coopera-
from the terminator are present, indicative of a linear tion necessary for HKR.[5g]
polymer. To obtain a MALDI-TOF of the ROMP
Pre-catalysts 13a and 13b were oxidized to
products on 12, we cleaved them from the benzyl CoACHTUNTGR(EUNNNG III)OAc by swelling them in dichloromethane for
ether groups of the resin using a boron trichloride di- one hour followed by the addition of excessive acetic
methyl sulfide complex. acid and stirring for 20 min under air. The liquids
The MALDI-TOF spectrum of the cleaved material were decanted and the solids dried under high
(see Supporting Information) is consistent with the vacuum for one hour.
formation of cyclic oligomers with ring sizes ranging
To investigate the catalytic activity of our heteroge-
from dimers to pentamers. We observe the cleavage neous catalysts, the catalyst-containing resins
of benzyl ethers at different positions from the sup- (0.01 mol% cobalt basis) were swollen in the epoxide
port as shown in Scheme 4.
of interest for 20 min with chlorobenzene as an inter-
After verifying that we obtained cyclic ROMP- nal standard. To initiate the reaction, 0.6 equivalents
based oligomers containing supported salen ligands of water were added. Aliquots were removed at cer-
on 12, the ligands were metallated using cobalt(II) tain time intervals to determine when the reaction
3018
asc.wiley-vch.de
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2012, 354, 3016 – 3024

PolyACHTUNGTRENNUNG(styrene) Resin-Supported CobaltACHTUTGNREN(NUGN III) Salen Cyclic Oligomers
Scheme 4. Oligomer fragments cleaved from surface of the resin for MALDI-TOF analysis.
Scheme 5. Metallation of resin-supported cyclic oligomers.
reached an enantiomeric excess (ee) of 99%. Epi-
We evaluated our supported catalysts under previ-
chlorohydrin reached 99% ee in 4 h with conversions ously established conditions using epichlorohydrin,
of 53%. This is more than an order of magnitude 0.1 mol% catalyst, 0.6 equiv. water, 258C, and acetate
more active than the best heterogeneous HKR cata- as the counterion.[14] Using catalyst 13a 99% ee was
lyst in the literature.[10] Hexene oxide was also tested reached in less than 20 min, which is comparable to
as a substrate using 13a reaching 99% ee in 240 min the activity of the homogeneous cyclic oligomers.
(Table 1).
Adv. Synth. Catal. 2012, 354, 3016 – 3024
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
3019

Michael G. C. Kahn et al.
FULL PAPERS
Table 1. HKR of epichlorohydrin and hexene oxide using
0.6 equiv. of water and catalyst 13a.
R
Time [min][a]
% ee[b]
% Conversion[b,c]
Cl
n-propyl
240
240
99
99
53
52
[a]
Time to reach 99% ee.
Determined by chiral GC.
Calculated relative to chlorobenzene as an internal stan-
dard.
[b]
[c]
Table 2. HKR of epichlorohydrin and hexene oxide using
6.0 equiv. of water and catalyst 13a.
Figure 2. Optical microscopy image of the emulsion formed
under HKR conditions using 6.0 equiv. of water. Scale bar is
200 mm. Water soluble blue dye used for contrast
R
Time [min][a]
% ee[b]
% Conversion[b,c]
mogeneous cyclic oligomers.[7b] This suggests that the
rate enhancement it is not specific to 13.
Cl
n-propyl
60
180
99
99
50.3
51
To probe the origin of this rate enhancement,
a series of experiments was conducted in which we
gradually reduced the amount of water to investigate
what effect this has on reaction rates. The rates
remain generally the same using 6, 4, and 2 equiva-
lents of water. Only when the reactions were run
using 1 equivalent of water or less did we observe
[a]
Time to reach 99% ee.
Determined by chiral GC.
Calculated relative to chlorobenzene as an internal stan-
dard.
[b]
[c]
A significant rate enhancement was observed when a single phase system. This resulted in as much as 3
conducting the HKR of epichlorohydrin and hexene times slower reaction rates (Table 3).
oxide with a 10-fold excess of water (6 equiv.). The
While the initial rates are similar for entries 1–5
HKR was complete in one hour compared to four (see Supporting Information), entries 4 and 5 have
hours with the standard 0.6 equiv. of water. The HKR significantly reduced rates after 200 min. We hypothe-
of hexene oxide reached 99% ee in 180 min using sized that this reduction in rate might be due to
6 equiv. of water (Table 2).
AcOÀ counterion loss.[16] We theorized that reduced
Our hypothesis to explain the significant rate en- exposure of the catalyst to water would reduce coun-
À
hancement with increased water concentration is terion loss, thus maintaining the balance of Co Ac
based on solubility. When the reaction is carried out and Co OH species needed for high reaction rates.
[6]
À
with 0.6 equivalents of water, an aqueous and an or- Figure 3 shows two identical reactions at 0.6 equiv. of
ganic phase are seen initially. After approximately water. When 5 additional equivalents of water were
10 min, an appreciable amount of diol product accu- added at 120 min (Figure 3, squares) we observed
mulates that acts as an intermediary between the
phases. When the reaction is carried out using
6 equivalents of water, an unstable oil in water emul-
Table 3. Water dependence of the reaction rate for the HKR
of epichlorohydrin using 13b.
sion is formed over the entire course of the reaction.
Figure 2 shows catalyst 13a as brown spheres inside
the colorless organic phase. The catalyst beads appear
to accumulate at the interface of the two phases indi-
cative of a Pickering emulsion. After stirring is
stopped the epoxide droplets begin to coalesce. The
diol product is more soluble in the aqueous phase,
leaving the catalyst and epoxides in the organic
phase[15] (vide infra).
Entry
Equiv. H2O
Time [min][a]
% ee[b]
1
2
3
4
5
6
4
2
1
120
150
180
240
300
99
99
99
99
99
0.6
[a]
A similar rate enhancement was observed when we
conducted the HKR of epichlorohydrin using the ho-
Time to reach 99% ee.
Determined by chiral GC.
[b]
3020
asc.wiley-vch.de
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2012, 354, 3016 – 3024

PolyACHTUNGTRENNUNG(styrene) Resin-Supported CobaltACHTUTGNREN(NUGN III) Salen Cyclic Oligomers
Table 4. Recycling studies for the HKR of epichlorohydrin
with 13a using 6.0 equiv. of water.
Cycle
Time [min][a]
% ee[b]
% Conversion[c]
1
2
3
4
5
6
7
75
90
99
99
99
99
99
95
95
50.6
53.8
52.1
55.9
49.8
57.4
55.2
135
150
180
210
240
[a]
Time to reach 95–99% ee.
Determined by chiral GC.
Calculated relative to chlorobenzene as an internal stan-
dard.
[b]
[c]
Table 5. Elemental analysis of fresh and spent catalyst 13a at
6.0 mol% water content.
Figure 3. Kinetic plot of the HKR of epichlorohydrin. Dia-
monds; 0.01 mol% 13b, 0.6 equiv. water. Squares;
0.01 mol% 13b, 0.6 equiv. water, 5 additional equiv. added
after 120 min.
C[a]
H[a]
N[a]
O[b]
Co
Fresh
Spent
81.82
77.41
7.10
8.02
0.95
0.94
8.53
12.63
1.61[c]
1.00[c]
a significant rate increase. This disproves the counter-
ion loss theory since it is an irreversible process.[16]
Evidence of diol product sequestration into the
aqueous phase was obtained by conducting the HKR
[a]
Determined by CHN analysis.
Calculated.
Determined by ICP-MS.
[b]
[c]
1
of epichlorohydrin using D2O. According to H NMR
spectroscopy experiments, we observe diol product
accumulation in the aqueous phase contaminated over numbers (for all 7 cycles) for 13a were calculated
with approximately 17% epichlorohydrin. The epi- to be 37,480 (Table 4).
chlorohydrin phase was contaminated with 3% diol
product.
While the ees for each cycle are outstanding, the
times required to obtain these yields increased for
The fact that the products of the reaction self-sepa- each cycle suggesting that some catalyst deactivates
rate under reaction conditions can be a very attractive during each cycle. While there are many possible de-
feature to the industrial preperation of enantiopure activation pathways, CHN elemental analysis of the
epoxides as the energy intensive azeotropic distilla- spent catalyst shows a reduction in carbon relative to
tion procedures and the use of solvents for extraction fresh catalyst, indicative of ligand hydrolysis
to separate components from the reaction mixure can (Table 5). Surprisingly the nitrogen content remains
be omitted.[4] We suggest that diffusion is the main unchanged, contrary to what is seen in the litera-
reason behind this rate enhancement. Due to diol re- ture,[10] as nitrogen reductions can be as high as 40–
moval from the organic phase, where the supported 60%.
catalyst resides, the concentration of the epoxides is
To investigate whether the macrocycles on the resin
maintained over the entire course of the reaction are a necessary feature of our catalytic system, a con-
rather than being diluted by the diol product and trol experiment was conducted comparing the HKR
water.
While we were able to demonstrate excellent cata- 10 (Scheme 6). The cobalt content of 14 is 1.1% Co.
lytic activity of 13a using our modified procedure, the Under our standard reaction conditions the HKR
activity of 13a to 14 which is the metallated version of
other main goal of this work was being able to easily of epichlorohydrin using 14 reached 97% ee in 4 h
recover and reuse the catalyst. After each HKR cycle using 6 equiv. water. This is an interesting result due
the catalyst was collected by filtration. Before the cat- to the fact that the small molecule analog, and simi-
alyst was used in the next cycle it was reactivated by larly structured supported catalysts reported in the lit-
stirring in a solution of dichloromethane and glacial erature are significantly less active at loadings of
acetic acid for 20 min.[17] We then decanted the liquids 0.01 mol%.[5c,10] This suggests an increase in bimetallic
and any remaining volatiles were removed under cooperation between adjacent catalyst units grafted
vacuum yielding the reactivated catalytic species. We on the resin surface (Figure 4).
recycled 13a 5 times with ee values above 99% while
Perhaps the high grafting density can be attributed
cycles 6 and 7 yielded ee values of 95%. Total turn- to the leaving group ability of trichloroimine com-
Adv. Synth. Catal. 2012, 354, 3016 – 3024
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
3021

Michael G. C. Kahn et al.
FULL PAPERS
Scheme 6. Metallation of resin-supported monomer 10.
corresponding residual nuclei in deuterated solvents. Solid-
state NMR spectra were obtained using a 750 MHz wide
bore Bruker magnet with a four channel Avance II Bruker
console and a 4 mm HCND HRMAS probe. Spinning fre-
quency was 6 kHz and the temperature 300 K. MALDI-
TOF data were collected on a Bruker OmniFLEX spec-
trometer using anthracenediol as a matrix. Chiral gas chro-
matography data used to determine enantiomeric excesses
were obtained on a Shimadzu 10A instrument with a flame
ionization detector using a b-Dex 120 column purchased
from Supelco using helium as the carrier gas. Straight phase
column chromatography was performed with silica gel 60 ꢁ
(230–400 mesh) from Sorbent Technologies. All solvents
were passed over MBraun copper catalyst for drying and de-
gassed by 3 freeze-pump-thaw cycles. Wang resin was pur-
chased from Nova Biochem, 100–200 mesh, 0.9 mmolgÀ1
substitution. Other reagents were used as received from sup-
pliers. Unless otherwise noted, all reactions were conducted
under Schlenk conditions in a nitrogen atmosphere. All
solid phase reactions were conducted in 15 mL poly(propy-
lene) centrifuge tubes and stirred on a vortex mixer at
2000 rpm. Solid phase products were collected on 3 mL Su-
pelco SPC filtration tubes. Grubbsꢂ 3rd generation initiator
was synthesized according to the literature.[13] Optical mi-
croscopy images were obtained on a Leitz Ergolux micro-
scope at 10ꢃ.
Figure 4. Schematic representation of the bimetallic cooper-
ation between salen units on the resin surface of catalyst 14.
pared to the benzyl chloride of the Merrifield resin.
Additionally, the long, flexible C10 alkyl linker can
have a significant effect on the activity as well.[18] The
macrocyclic resin 13 is twice as active as 14 indicating
the benefit of the cyclic structures for enhanced activ-
ities.
Conclusions
In conclusion, we report a highly active and selective
heterogeneous CoACHTNUGRTENUNG(III) salen catalyst for the HKR of
terminal epoxides. The supported catalyst is an order
of magnitude more active than the best heterogene-
ously-supported HKR catalyst reported in the litera-
ture. A significant rate increase was observed when
a 10-fold excess of water was added which we attri-
bute to the diol product partitioning into the aqueous
phase, maintaining a high concentration of the epox-
ides in the organic phase where the catalyst is located
and active. This partitioning facilitates the separation
of the reaction products increasing the utility of this
Compound 3
Compound 2 was synthesized according to a literature pro-
cedure.[19] 1.62 g (9.3 mmol) of 1,10-decanediol were dis-
solved in 20 mL dry THF by gentle heating. 2.27 g
(37.2 mmol) of NaHCO3 were added and the solution was
allowed to stir at room temperature for 1 hour. Compound
2, 528 mg (2.33 mmol), was dissolved in 10 mL of THF and
injected drop-wise through a septum into the reaction mix-
process for potential industrial preparation of enan- ture. The solution was allowed to stir for 48 h at room tem-
perature with an argon balloon to vent CO2 generated
during the reaction. The bulk of the THF was removed by
rotary evaporation and the mixture was diluted in 200 mL
diethyl ether and washed 2ꢃ with water and 1ꢃ with brine,
dried over anhydrous magnesium sulfate, filtered, the sol-
vent removed by rotary evaporation and dried under high
vacuum. The crude mixture was purified by column chroma-
tography using a 6:4 hexanes:ethyl acetate solvent system
giving a viscous yellow oil; yield: 458 mg (1.52 mmol, 54%).
1H NMR (CDCl3): d=1.27–1.39 (m, 13H, alkyl), 1.41 (s,
9H, CMe3), 1.46–1.67 (m, 4H, alkyl), 1.75 (s, 1H, OH), 3.48
tiomerically pure epoxides.
Experimental Section
General Remarks
Solution phase NMR spectra were recorded on a Bruker
AV-400 (1H: 400.1 MHz; 13C: 100.6 MHz) spectrometer.
Chemical shifts are reported in ppm and referenced to the
3022
asc.wiley-vch.de
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2012, 354, 3016 – 3024

PolyACHTUNGTRENNUNG(styrene) Resin-Supported CobaltACHTUTGNREN(NUGN III) Salen Cyclic Oligomers
(t, J=6.7 Hz, 2H, OCH2), 3.61 (t, J=6.1 Hz, 2H, HOCH2),
4.44 (s, 2H, Bz), 7.38 (d, J=2.1 Hz, 2H, Ph), 9.86 (s, 1H,
OH), 11.76 (s, 1H, O=CH); 13C NMR (CDCl3): d=25.5,
26.2, 29.1, 29.3, 29.4, 29.5, 29.6, 29.7, 32.7, 32.9, 35.0, 6 0.4,
62.9, 70.4, 70.9, 120.2, 125.9, 126.7, 129.3, 131.0, 134.0, 138.3,
160.7, 197.1; HR-MS (ESI): m/z=387.2503, calcd. for
C22H36O4Na+: 387.16.
200 mg (0.25 mmol), was dissolved in 1 mL dichloromethane
and added to the swollen resin. The mixture was stirred for
5 min then 5 mL (0.039 mmol) boron trifluoride etherate
were added and the mixture was stirred for 1 hour. The re-
action mixture was washed 4ꢃ with dichloromethane fol-
lowed by decantation and filtered on a peptide resin filter
giving a yellow granular solid; yield: 580 mg. CHN analysis
found: C 83.08, H 7.64, N 0.84; IR (KBr): n=3060, 3025,
2849, 1945, 1874, 1716, 1602, 1505, 1444, 1368 cmÀ1.
Compound 8
Compounds 6 and 7 were synthesized according to a litera-
ture procedure.[7b] Mono hydrochloride diamine 7, 85.5 mg
(0.57 mmol) was added to a round-bottom flask along with
200 mg 4 ꢁ molecular sieves and purged with nitrogen.
Compound 6, 182 mg (0.57 mmol), was dissolved in 1 mL of
methanol and 0.5 mL dichloromethane and injected through
a septum into the reaction mixture and left to stir at room
temperature for 4 h. Triethylamine, 627 mg (0.57 mmol) was
added and stirred for 5 min. Then, compound 3, 200 mg
(0.57 mmol), was dissolved in 1 mL methanol and 0.5 mL di-
chloromethane and injected into the reaction mixture and
the mixture was stirred for 16 h at room temperature. The
reaction mixture was dissolved in 100 mL of methylene chlo-
ride and washed 2ꢃ with water and 1ꢃ brine, dried over an-
hydrous magnesium sulfate, filtered, and dried by rotary
evaporation followed by drying on high vacuum giving a vis-
cous yellow oil. Material was purified by column chromatog-
raphy using 19:1 hexanes:ethyl acetate solvent system with
0.5% triethylamine giving a yellow foam; yield: 304 mg
(0.39 mmol, 65%). 1H NMR (CDCl3): d=1.27–1.34 (m,
14H, alkyl), 1.37 (s, 9H, CMe3), 1.39 (s, 9H, CMe3), 1.42–
2.72 (m, 25H, alkyl), 3.03 (m, 2H, alkyl), 3.38 (t, J=7.3 Hz,
2H, HOCH2), 3.62 (t, J=7.2 Hz, 2H, HOCH2), 4.33 (s, 2H,
Bz), 5.65 (m, 2H, olefin), 6.71 (d, J=2.9 Hz, 1H, Ph) 6.88
(d, J=2.9 Hz, 1H, Ph), 6.96 (d, J=2.6 Hz, 1H, Ph), 7.21 (d,
J=2.6 Hz, 1H, Ph), 8.21 (s, 1H, OH), 8.27 (s, 1H, OH),
13.77 (s, 1H, N=CH), 13.80 (s, 1H, N=CH); 13C NMR
(CDCl3): d=24.1, 24.4, 25.7, 25.9, 26.2, 27.8, 29.1, 29.3, 29.4,
29.5, 29.6, 29.7, 31.4, 31.5, 32.8, 33.1, 43.7, 34.9, 43.2, 43.3,
63.1, 70.2, 72.2, 72.3, 72,7, 118.1, 121.3, 122.7, 127.6, 129.3,
129.4, 129.5, 130.6, 130.7, 132.3, 137.1, 138.5, 141.7, 157.9,
159.8, 164.7, 165.5, 176.4; HRMS (ESI): m/z=773.5472,
calcd. for C48H72O6 +H: 773.10
Resin 12
Compound 10, 109 mg, was swollen in 1 mL dichlorome-
thane for 60 min. Grubbsꢂ 3rd generation initiator 2.98 mg
(0.033 mmol) was dissolved in 100 mL dichloormethane and
added to the reaction mixture. After 10 seconds, compound
11 was added and the mixture stirred for 30 min. The reac-
tion was quenched with ethylvinyl ether and diluted and
decanted with dichloromethane 3ꢃ and filtered using a pep-
tide resin filter and washed with ether, and dried under high
vacuum affording a dark yellow (slightly greenish) granular
solid. CHN analysis found: C 80.92, H 7.88, N 0.92; IR
(KBr): n=3427, 3059, 3025, 2849, 1944, 1874, 1804, 1749,
1716, 1600, 1506, 1443, 1364 cmÀ1.
General Procedure for Precatalysts 13a/b, 14
In a glove box, 100 mg resin 12 was swollen in 2 mL di-
chloromethane for 60 min. Then 30 mg (0.12 mmol) of co-
balt(II) acetate·4H2O was dissolved in 200 mL methanol.
The cobalt solution was added drop-wise into resin 12. The
color changed from yellow to red in 3 min followed by
a deep purple after 20 min. The reaction was stirred for 4 h.
The liquids were decanted and the resin was washed 4ꢃ
with dichloromethane then filtered onto a peptide resin
filter. The precatalyst was dried thoroughly under high
vacuum before removal from the glove box.
13a CHN analysis found: C 81.82, H 7.10, N 0.95; ICP-
MS cobalt analysis 1.61% Co.
13b CHN Analysis found: C 80.84, H7.22 N 0.93; ICP-MS
cobalt analysis 1.41% Co.
An analogous procedure was used to produce pre-catalyst
14. CHN analysis found: C 83.08, H 7.04, N 0.82; ICP-MS
cobalt analysis 1.10% Co.
Resin 9
Cleavage of Cycles for MALDI-TOF Analysis
Commercially available Wang resin (100–200 mesh, substitu-
tion 0.9 mmolgÀ1, 140 mg) was swollen by vortex mixing in
2 mL dry dichloromethane for 60 min in a poly(propylene)
centrifuge tube. 171 mL (1.7 mmol) of trichloroacetonitrile
were added and the reaction mixture was cooled to 08C.
DBU, 11 mL (0.076 mmol), was added and the reaction mix-
ture was stirred for 1 hour maintaining 08C. The mixture
was washed and decanted with dichloromethane 3ꢃ and
then filtered using a peptide resin filter and washed with di-
ethyl ether followed by drying under high vacuum giving an
off white granular solid. IR (KBr): n=3337, 3062, 2847,
1948, 1804, 1662, 1602, 1510, 1444, 1374 cmÀ1.
20 mg of supported ligand 12 were swollen in 0.5 mL di-
chloromethane for 60 min. Then 10 mg (0.056 mmol) of
boron trichloride dimethyl sulfide complex were dissolved
in 0.5 mL dichloromethane and added to swollen resin 10
and allowed to stir for 20 min. The solution turned from
clear to yellow in color. The resin itself changed from
a yellow to an off-white color, indicating the release of the
ligand from the surface of resin. The solution was filtered
and washed with a solution of saturated sodium bicarbonate,
dried over MgSO4, and the volatiles were removed by rotary
evaporation.
Resin 10
Typical Procedure for HKR of Terminal Epoxides
Trichloroimine functionalized Wang resin 9, 500 mg, was
swollen in 3 mL dichloromethane for 60 min. Compound 8,
In a typical experiment, 10 mg (0.00283 mmol Co basis) pre-
catalyst 13a were swollen in 2 mL dichloromethane and
Adv. Synth. Catal. 2012, 354, 3016 – 3024
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
3023

Michael G. C. Kahn et al.
FULL PAPERS
stirred for 60 min. Then 20 mL of glacial acetic acid were
added and stirred for 20 min open to the air. The liquids
were decanted and the remaining volatiles were removed
under high vacuum for 1 hour. 28.29 mmol terminal epoxide
(epichlorohydrin or hexene oxide) were added to the acti-
vated catalyst with 200 mL of chlorobenzene as an internal
standard. The mixture was stirred for 20 min, and then
305 mg (16.9 mmol) for 0.6 equiv. water or 3050 mg
(169.0 mmol) for 6 equiv. water of water, were added to ini-
tiate the reaction. Aliquots of 5 mL were removed at certain
time intervals and diluted in 300 mL of diethyl ether to mon-
itor the progress of the reaction by chiral GC. For recycling
experiments, the resin was collected on a peptide resin filter
and washed with diethyl ether 3ꢃ and dried under high
vacuum for 5 h. To reactivate the catalyst for the next cycle
it was treated with acetic acid as outlined above.
1212; c) E. N. Jacobsen, Acc. Chem. Res. 2000, 33, 421–
431; d) I. M. Pastor, M. Yus, Curr. Org. Chem. 2005, 9,
1–29.
[2] a) R. M. Harris, T. J. Donohoe, J. Burrows, J. Parker, J.
Am. Chem. Soc. 2006, 128, 13704–13705; b) B. B. Liau,
M. D. Shair, J. Am. Chem. Soc. 2010, 132, 9594–9595.
[3] M. Tokunaga, J. F. Larrow, F. Kakiuchi, E. N. Jacobsen,
Science 1997, 275-277, 936–938.
[4] L. Aouni, K. E. Hemberger, S. Jasmin, H. Kabir; J. F.
Larrow, I. Le-Fur, P. Morel, T. Schlama, Industrializa-
tion Studies of the Jacobsen Hydrolytic Kinetic Resolu-
tion of Epichlorohydrin, in: Asymmetric Catalysis on
Industrial Scale, Wiley-VCH, Weinheim, 2004 pp 165–
199.
[5] a) N. Madhavan, C. W. Jones, M. Weck, Acc. Chem.
Res. 2008, 41, 1153–1165; b) Y. Liu, Y. Wang, Y. Wang,
J. Lu, V. Pinon III, M. Weck, J. Am. Chem. Soc. 2011,
133, 14260–14263; c) D. A. Annis, E. N. Jacobsen, J.
Am. Chem. Soc. 1999, 121, 4147–4154; d) L. Canali,
D. C. Sherrington, Chem. Soc. Rev. 1999, 28, 85–93;
e) B. Clapham, T. S. Reger, K. D. Janda, Tetrahedron
2001, 57, 4637–4662; f) A. Corma, H. Garcia, Adv.
Synth. Catal. 2006, 348, 1391–1412; g) H. Yang, L.
Zhang, L. Zhong, Q. Yang, C. Li, Angew. Chem. 2007,
119, 6985–6989; Angew. Chem. Int. Ed. 2007, 46, 6861–
6865.
[6] L. P. C. Nielsen, C. P. Stevenson, D. G. Blackmond,
E. N. Jacobsen, J. Am. Chem. Soc. 2004, 126, 1360–
1362.
[7] a) J. M. Ready, E. N. Jacobsen, Angew. Chem. 2002,
114, 1432–1435; Angew. Chem. Int. Ed. 2002, 41, 1374–
1377; b) X. Zheng; C. W. Jones, M. Weck, J. Am.
Chem. Soc. 2007, 129, 1105–1112.
[8] Y. Liu, J. Rawlston, A. T. Swann; T. Takatani, D. C.
Sherrill, P. J. Ludovice, M. Weck, Chem. Sci. 2011, 2,
429–438.
13a spent catalyst (after 7 cycles) CHN analysis found: C
77.41, H 8.02, N 0.92; ICP-MS cobalt analysis 1.01% Co.
Epoxide-Diol Partitioning NMR Experiment
To determine how much of the diol product was diffusing
into the aqueous phase; the HKR of epichlorohydrin was
conducted using 6 equiv. D2O. After the reaction was com-
plete the catalyst was filtered on a peptide filter. The filtrate
was centrifuged to separate the emulsion. The epichlorohy-
drin layer was separated, diluted in CDCl3 and dried over
Na2SO4. Integration of the spectra indicated that epichloro-
hydrin (peak used for integration=3.2 ppm, 0.93H) was
contaminated with approximately 3% of the diol product
(peak used for integration=3.9 ppm, 0.03H). The aqueous
1
layer was diluted in D2O for H NMR spectroscopy meas-
urements and showed that the diol product (peak used for
integration=3.9 ppm, 1.0H) was contaminated with approx-
imately 17% epichlorohydrin (peak used for integration=
2.9 ppm, 0.17H, see Supporting Information for spectra).
[9] C. Baleizao, H. Garcia, Chem. Rev. 2006, 106, 3987–
4043.
[10] C. S. Gill, K. Venkatasubbaiah, N. T. S. Phan, M. Weck,
C. W. Jones, Chem. Eur. J. 2008, 14, 7306–7313.
[11] M. Holbach, X. Zheng, C. Burd, C. W. Jones, M. Weck,
J. Org. Chem. 2006, 71, 2903–2906.
[12] S. Hanessian, F. Xie, Tetrahedron Lett. 1998, 39, 733–
736.
[13] J. A. Love, M. S. Sanford, M. W. Day, R. H. Grubbs, J.
Am. Chem. Soc. 2003, 125, 10103–10109.
Acknowledgements
We are thankful to the Department of Energy Office of Basic
Energy Sciences through Catalysis Contract No. DEFG02-
03ER15459 for financial support of this work. This work was
supported partially by the MRSEC Program of the National
Science Foundation under Award Number DMR-0820341.
The Bruker Avance-400 NMR Spectrometer was acquired
through the support of the National Science Foundation
under Award Number CHE-01162222. The Bruker Maldi-
TOF UltrafleXtreme MS spectrometer was acquired through
the support of the National Science Foundation under Award
Number CHE-0958457. We thank Boris Itin and the New
York Structural Biology Center for solid state NMR measure-
ments
[14] X. Zhu, K. Venkatasubbaiah, M. Weck, C. W. Jones,
ChemCatChem 2010, 2, 1252–1259.
[15] J. F. Larrow, E. N. Jacobsen, Top. Organomet Chem.
2004, 6, 123–152.
[16] L. P. Nielsen, D. D. Ford, E. N. Jacobsen, J. Org. Chem.
2012, 77, 2486–2495.
[17] S. Jain, K. Venkatasubbaiah, C. W. Jones, R. J. Davis, J.
Mol. Catal. A: Chem. 2010, 316, 8–15.
[18] N. Madhavan, T. Takatani, C. D. Sherrill, M. Weck,
References
Chem. Eur. J. 2009, 15, 1186–1194.
[19] F. Minutolo, D. Pini, A. Petri, P. Salvadori, Tetrahe-
dron: Asymmetry 1996, 7, 2293–2302.
[1] a) R. E. Parker, Chem. Rev. 1959, 59, 737–799; b) N.
Kasai, K. Sakaguchi, Tetrahedron Lett. 1992, 33, 1211–
3024
asc.wiley-vch.de
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2012, 354, 3016 – 3024