Aqueous cholesteric liquid crystals using uncharged rodlike polypeptides

Article abstract of DOI:10.1021/ja047932d

The aqueous, lyotropic liquid-crystalline phase behavior of the α-helical polypeptide, poly(Nε-2-[2-(2-methoxyethoxy)ethoxy]acetyl- lysine) (1), has been studied using optical microscopy and X-ray scattering. Solutions of optically pure 1 were found to fo

Full text of DOI:10.1021/ja047932d

Published on Web 07/02/2004
Aqueous Cholesteric Liquid Crystals Using Uncharged
Rodlike Polypeptides
Enrico G. Bellomo,† Patrick Davidson,‡ Marianne Impe´ror-Clerc,‡ and
Timothy J. Deming*,†
Contribution from the Departments of Materials and Chemistry, Materials Research Laboratory,
UniVersity of California, Santa Barbara, California 93106, and Laboratoire de Physique des
Solides, UMR 8502 CNRS, Baˆt. 510, UniVersite´ d’Orsay, F-91405 Orsay, France
Received April 9, 2004; E-mail: tdeming@mrl.ucsb.edu
Abstract: The aqueous, lyotropic liquid-crystalline phase behavior of the R-helical polypeptide, poly(Nꢀ-
2-[2-(2-methoxyethoxy)ethoxy]acetyl-lysine) (1), has been studied using optical microscopy and X-ray
scattering. Solutions of optically pure 1 were found to form cholesteric liquid crystals at volume fractions
that decreased with increasing average chain length. At very high volume fractions, the formation of a
hexagonal mesophase was observed. The pitch of the cholesteric phase could be varied by a mixture of
enantiomeric samples L-1 and D-1, where the pitch increased as the mixture approached equimolar. The
cholesteric phases could be untwisted, using either magnetic field or shear flow, into nematic phases,
which relaxed into cholesterics upon removal of field or shear. We have found that the phase diagram of
1 in aqueous solution parallels that of poly(γ-benzyl glutamate) in organic solvents, thus providing a useful
system for liquid-crystal applications requiring water as solvent.
Introduction
Aqueous, lyotropic liquid crystals have great potential for a
wide range of applications. These include use as templates for
biomimetic inorganic materials synthesis1 and as electric-field
responsive materials in actuators and devices.2 Furthermore,
cholesteric aqueous liquid crystals would be valuable as chiral
NMR solvents for enantiomer resolution,3 and as a chiral
medium for asymmetric transformations.4 In the area of
polypeptide liquid crystals, poly(γ-benzyl-L-glutamate) (PBLG)
is perhaps the best studied example.5 This rodlike, R-helical
polymer forms cholesteric liquid crystals in a number of organic
solvents at concentrations above ca. 20 wt %, depending on
chain length and temperature.6 Unfortunately, there is no
aqueous phase polypeptide analogue of PBLG, even though the
organic system has been known for over 40 years. A good reason
for this has been the lack of a water-soluble, R-helical
homopolypeptide.
Figure 1. Structure and schematic representation of polymer L-1.
nonionic polypeptides, poly(ω-hydroxyalkyl)glutamines (e.g.,
poly(3-hydroxypropyl)glutamine),8 are only weakly R-helical,
which is likely a consequence of the low hydrophobicity of the
glutamine side-chains. Our solution to this problem was
preparation of polypeptides using ethyleneglycol-modified lysine
residues.9 Short, uniform diethyleneglycol segments provided
excellent water solubility, and the increased hydrophobicity of
a lysine relative to glutamine backbone resulted in increased
helix stability in water. These homopolymers, poly(Nꢀ-2-[2-(2-
methoxyethoxy)ethoxy]acetyl-lysine), 1 (Figure 1), were found
to be miscible with water in all proportions at ambient
temperature, completely R-helical in aqueous solution, and able
to form birefringent liquid-crystalline solutions at high weight
fractions.9 Here, we describe the nature of these aqueous
polypeptide liquid crystals in detail.
The repulsion of abundant-like charges on most water-soluble
homopolypeptides (e.g., polylysine‚HBr or polyglutamate‚Na
salt) is sufficient to prevent R-helix formation in aqueous
solutions of these samples.7 The most studied water-soluble,
† University of California, Santa Barbara.
Results
‡ Universite´ d’Orsay.
Optical Observations. Figure 2 shows test tubes filled with
increasing volume fractions of polymer L-1 in water, viewed
(1) Dujardin, E.; Blaseby, M.; Mann, S. J. Mater. Chem. 2003, 13, 696-699.
(2) Thomsen, D. L., III; Keller, P.; Naciri, J.; Pink, R.; Jeon, H.; Shenoy, D.;
Ratna, B. R. Macromolecules 2001, 34, 5868-5875.
(3) (a) Canet, I.; Meddour, A.; Courtieu, J.; Canet, J. L.; Salau¨n, J. J. Am.
Chem. Soc. 1994, 116, 2155-2156. (b) Merlet, D.; Ancian, B.; Courtieu,
J.; Lesot, P. J. Am. Chem. Soc. 1999, 121, 5249-5258.
(4) Hatano, M.; Nozawa, N. Prog. Polym. Sci. Jpn. 1972, 4, 223-271.
(5) Block, H. Poly(γ-benzyl-L-glutamate) and other Glutamic Acid Containing
Polymers; Gordon and Breach: New York, 1983.
(7) Katchalski, E.; Sela, M. AdV. Protein Chem. 1958, 13, 243-492.
(8) (a) Lupu-Lotan, N.; Yaron, A.; Berger, A.; Sela, M. Biopolymers 1965, 3,
625-655. (b) Lupu-Lotan, N.; Yaron, A.; Berger, A. Biopolymers 1966,
4, 365-368. (c) Okita, K.; Teramoto, A.; Fujita, H. Biopolymers 1970, 9,
717-738.
(9) Yu, M.; Nowak, A. P.; Pochan, D. J.; Deming, T. J. J. Am. Chem. Soc.
1999, 121, 12210-12211.
(6) (a) Robinson, C.; Ward, J. C. Nature 1957, 180, 1183-1184. (b) Robinson,
C. Tetrahedron 1961, 13, 219-234.
9
10.1021/ja047932d CCC: $27.50 © 2004 American Chemical Society
J. AM. CHEM. SOC. 2004, 126, 9101-9105
9101

A R T I C L E S
Bellomo et al.
Figure 2. Test tubes filled with increasing weight fractions of L-1 in
deionized water. Samples were imaged between crossed polarizers. (A) (Mw
) 62 kDa): i ) 32.7%; ii ) 39.3%; iii ) 41.9%; iv ) 43.5%; v ) 45.4%;
vi ) 47.9%. (B) (Mw ) 120 kDa): i ) 15.4%; ii ) 18.7%; iii ) 23.8%;
iv ) 25.2%; v ) 38.5%; vi ) 39.8%. All sample compositions are in weight
percent.
between crossed polarizers. The upper series contains a low
molecular weight sample (Mw ) 62 kDa), and the lower series
contains a high molecular weight sample (Mw ) 120 kDa); note
that the sample concentration ranges differ between the two
series. The most dilute samples appear dark because they are
isotropic liquids, whereas the most concentrated samples appear
bright due to the presence of a liquid-crystalline mesophase.
Between these two regimes, a biphasic domain is found where
samples display macroscopic phase separation, with the more
dense birefringent phase at the bottom of the test tubes. These
observations show the existence of a first-order phase transition
between the isotropic and liquid-crystalline phases, in addition
to confirming that the samples have reached thermodynamic
equilibrium. Similar results were also observed for other L-1
samples with different molecular weights. It should be noted
that the liquid-crystalline phase appears at lower volume
fractions as the molecular weight of L-1 increases. The
properties of bulk L-1 samples at elevated temperatures were
also investigated; however, no thermotropic behavior was
observed. Furthermore, upon heating aqueous L-1 solutions to
ca. 75 °C, the polymers were found to precipitate. This
insolubility was presumably due to the loss of H-bonding
interactions between the ethylene glycol side-chains and water
at elevated temperature.10
Figure 3. Polarized optical micrographs showing the dependence of the
cholesteric pitch on the enantiomeric composition of 1. All samples were
prepared at 60 wt % in deionized water and are defined as the mol % of
L-1 in a L-1 + D-1 mixture. (A) ) 0%; (B) ) 15%; (C) ) 30%; (D) )
40%; (E) ) 60%; (F) ) 70%; (G) ) 85%; (H) ) 100%.
these features are typical of the cholesteric (i.e., chiral nematic)
phase, and the cholesteric pitch could be directly measured from
the fingerprint patterns of these micrographs.11
From a practical standpoint, the two major challenges in
cholesterics are adjusting and controlling the pitch. As well as
being sensitive to solvent and temperature, the pitch also varies
with the optical purity of the sample. If a cholesteric liquid-
crystalline compound is mixed with its opposite enantiomer (i.e.,
a mixture of L-1 and D-1), the pitch should increase and
approach infinity as the mixture becomes racemic, and then the
liquid-crystalline phase will become nematic. To study this point
in detail, two polymers of opposite handedness and almost equal
molecular weights, L-1 (Mw ) 59.5 kDa) and D-1 (Mw ) 64.9
kDa), were separately dissolved in deionized water to form 10
wt % solutions. These solutions were then mixed together in
When examined by polarized optical microscopy, the textures
of the liquid-crystalline samples held in flat glass capillaries
were clear enough to allow mesophase identification. Biphasic
samples displayed banded birefringent liquid spherulites floating
in a dark isotropic liquid. These spherulites, also called tactoids,
are droplets of the anisotropic phase. Fully liquid-crystalline
samples displayed large dark homeotropic regions separated by
bright regions containing fingerprint patterns (Figure 3). All of
(10) Zalipsky, S.; Lee, C. In Poly(EthyleneGlycol) Chemistry: Biotechnical and
Biomedical Applications; Harris, J. M., Ed; Plenum: New York, 1992.
(11) de Gennes, P. G.; Prost, J. The Physics of Liquid Crystals, 2nd ed.;
Clarendon: Oxford, 1993.
9
9102 J. AM. CHEM. SOC. VOL. 126, NO. 29, 2004

Aqueous Cholesteric Liquid Crystals
A R T I C L E S
Figure 4. Cholesteric pitch (µm), measured from optical micrographs, as
a function of the enantiomeric composition of 1 (mol % of L-1 in a L-1 +
D-1 mixture). The solid line is a fit of the data using a hyperbolic divergence
at 50 mol %.
Figure 5. X-ray scattering patterns of liquid-crystalline samples of 1 (60
wt % in deionized water). (A) 100 mol % L-1 with no applied shear (0
s-1); (B) 100 mol % L-1 under high shear (400 s-1); (C) 55 mol % L-1
sample immediately after cessation of applied shear (200 s-1 for 10 min);
(D) 55 mol % L-1 sample aligned by magnetic field (8 T) immediately
after removal from the field. The arrows indicate the direction of the applied
field or shear.
the appropriate volumes to give samples with mole fractions of
L-1 from 0% to 100% in 10-15% increments. Circular
dichroism spectroscopy was used to verify the degree of
enantiomeric mixing in these samples.12 The sensitivity of the
cholesteric pitch to the enantiomeric excess in the samples was
observed in optical micrographs (Figure 3). In our samples, the
dependence of pitch on mixing ratio was symmetrical around,
and increased toward, the 50 mol % mixture (Figure 4). As
optical purity increased toward either pure L-1 or D-1, the
cholesteric pitch decreased to its minimum value. These data
clearly illustrate the fine control of pitch and the cholesteric to
nematic transition that can be obtained with the well-defined,
rodlike polymer samples utilized here.
Sample Alignment. To gain more information on the
structure and stability of these cholesteric phases, we sought to
untwist the pitch by aligning the rodlike polypeptide chains.
For this purpose, it is well-known that magnetic fields can align
R-helical polypeptides parallel to the direction of the applied
field.13 In preliminary studies, we found that a moderate
magnetic field (1.7 T) was not sufficient to align optically pure
L-1 samples, or even the weakest cholesteric samples (i.e., near-
equimolar L-1 + D-1 mixtures). However, when these weak
cholesterics in optical capillaries were placed in a stronger
magnetic field (ca. 8 T magnet of an NMR spectrometer), the
fingerprint textures were found to disappear as the chains were
aligned into a nematic phase. This result showed that strong
enough magnetic fields can unwind the cholesteric helical
organization in L-1 + D-1 mixtures (up to ca. 80 mol % of
either L-1 or D-1). Unfortunately, once the samples were
removed from the field, the fingerprint textures reappeared in a
few minutes as the cholesteric pitch reformed from the untwisted
state. Mesophase alignment and relaxation in these samples was
readily observed by optical microscopy.12 In the case of pure
L-1 or D-1 samples, there appeared to be only partial alignment,
which might have been due to the very fast relaxation of these
samples during transfer from the magnet to the microscope.
The use of a fast CCD camera and the high beam intensity
of a rotating anode X-ray source allowed us to quickly record
the scattering pattern of a mixed L-1 + D-1 sample (55 mol %
L-1) that had been aligned in the 8 T magnetic field and then
removed for analysis. No sharp diffraction spots or lines were
detected, which showed the absence of any positional long-
range order (Figure 5). At rest, the scattering patterns were
isotropic due to a random orientation distribution of liquid-
crystalline domains (Figure 5A). The anisotropic pattern from
field alignment showed that the rodlike polypeptides were
Figure 6. Azimuthal (I versus θ) and radial (I versus q) scans of the X-ray
intensity of a 55 mol % L-1 sample (60 wt % in deionized water) that was
either shear aligned (A + B) or magnetic field aligned (C + D). The radial
scans each show a broad peak with a maximum at 2.6 nm-1, corresponding
to the average distance, 2.4 nm, between the helical rods. The peak width
gives an estimate of the correlation length of the liquidlike positional order
(ca. 9 nm).
aligned parallel to the magnetic field (Figure 5D), as expected
for R-helices.13 This is typical of a lyotropic nematic phase
comprised of rodlike particles.14-17 Thus, the unwinding of the
cholesteric twist by the magnetic field, observed by polarized
light microscopy, was confirmed to arise from the alignment
of the individual helices.
Radial scans (I versus q) of the X-ray intensity scattered in
the diffuse spots for the magnetically aligned sample showed a
broad peak that had a maximum corresponding to the average
distance, 2.4 nm, between the molecular helices. The peak width
gave an estimate, 9 nm, of the correlation length of the liquidlike
positional order (Figure 6D). Azimuthal scans (I versus θ) of
the scattered intensity along a circle going through the maxima
of the diffuse spots were used to derive the nematic order
parameter S, according to classical procedures (Figure 6C).18,19
The value obtained, S ) 0.85 ( 0.05, was very large, but is
(14) Purdy, K. R.; Dogic, Z.; Fraden, S.; Ruhm, A.; Lurio, L.; Mochrie, S. G.
J. Phys. ReV. E 2003, 67, 031708-1-031708-12.
(15) Oldenbourg, R.; Wen, X.; Meyer, R. B.; Caspar, D. L. D. Phys. ReV. Lett.
1988, 61, 1851-1854.
(16) Ao, X.; Wen, X.; Meyer, R. B. Physica A 1991, 176, 63-71.
(17) Keates, P.; Mitchell, G. R.; Peuvrel, E. Polymer 1992, 33, 3298-3301.
(18) Leadbetter, A. J.; Norris, E. K. Mol. Phys. 1979, 38, 669-686.
(19) Davidson, P.; Petermann, D.; Levelut, A. M. J. Phys. II France 1995, 5,
113-131.
(12) See Supporting Information.
(13) (a) Panar, M.; Phillips, W. D. J. Am. Chem. Soc. 1968, 90, 3880-3882.
(b) Orwoll, R. D.; Vold, R. L. J. Am. Chem. Soc. 1971, 93, 5335-5338.
9
J. AM. CHEM. SOC. VOL. 126, NO. 29, 2004 9103

A R T I C L E S
Bellomo et al.
typical of lyotropic nematic phases. The X-ray pattern became
isotropic in less than an hour as the sample relaxed in the
absence of the magnetic field, similar to our optical microscopy
observations.
Due to our inability to record structural information in situ
during magnetic field alignment, a Couette cell setup was used
to shear the liquid-crystalline samples and allow simultaneous
recording of their X-ray scattering patterns. Application of a
shear flow is a very powerful way of aligning viscous meso-
phases and has been successfully applied to liquid-crystal phases
of PBLG in m-cresol.20 For similar alignment of our samples,
a small Couette shear cell was built that can be placed in the
X-ray apparatus and that requires sample volumes of only ca.
80 µL.12 The outer cylinder rotates with a controlled angular
speed with respect to the fixed inner cylinder. Two classical
scattering geometries were used.21,22 In the radial configuration,
the beam goes through the center of the cell and the velocity/
vorticity plane is explored. In the tangential configuration, the
X-ray beam is sent through the gap and the velocity gradient/
vorticity plane is explored. Comparison of data from these two
planes usually allows a full understanding of reciprocal space.
Figure 7. Radial (I versus q) X-ray scattering pattern of a 55 mol % L-1
sample (>70 wt % in deionized water) showing hexagonal symmetry of
mesophase of 1 with diffraction lines in the ratio 1, 31/2, 2, 71/2 (labeled as
the 10, 11, 20, and 21 reflections, respectively). DH ) diffuse halo of
scattering from ethylene glycol side-chains.
found to orient along the velocity direction like the hexagonal
mesophases of other compounds.23-26
Discussion
The phase behavior of concentrated aqueous solutions of L-1,
D-1, and L-1 + D-1 mixtures was found to be that predicted
by both the Onsager model of the isotropic/nematic phase
transition and the numerical simulations on rodlike particles.27,28
Qualitatively, the Onsager model and derived theories predict
a strongly first-order isotropic/nematic phase transition, with
phase coexistence and a large jump of nematic order parameter.
These predictions are valid for very long rodlike particles that
only interact through purely steric excluded-volume effects.
Cholesteric ordering is usually considered as only a small
perturbation to the nematic order, so this model should describe
the isotropic/cholesteric transition as well. The series of test
tubes showing phase coexistence (Figure 2), together with the
large value of the nematic order parameter derived from the
SAXS patterns, are evidence that the isotropic/cholesteric
transition of L-1, D-1, and L-1 + D-1 mixtures can be explained
by the Onsager model. Moreover, we believe that electrostatic
interactions are negligible in this system because similar
properties were observed when samples were prepared in NaCl
solutions (100 mM) where any electrostatic interactions were
screened. An additional prediction of the Onsager model is that
temperature should have no influence on the phase transition.
However, this prediction could not be tested here because all
liquid-crystalline samples became turbid upon heating. This
observation is likely related to the complex miscibility of PEG
with water as a function of temperature.10
In contrast to the magnetic field studies, all samples that were
sheared in the Couette cell did show some degree of alignment.
However, the cholesteric phase of the pure enantiomers was
found to be too viscoelastic to be completely aligned even at a
shear rate of 400 s-1 (Figure 5B). Weak cholesterics (L-1 +
D-1 mixtures) showed the best shear orientation, yet their
alignment became markedly more difficult as optical purity was
increased. The best alignment was seen immediately after the
cessation of shear using a weak cholesteric sample (55 mol %
L-1) (Figure 5C). The scattering pattern in the radial geometry
displayed two symmetrical diffuse spots, whereas the pattern
in the tangential geometry showed no anisotropy. The nematic
order parameter measured from this scattering pattern (S ) 0.88
( 0.05, Figure 6A,B) is comparable to that measured above
using magnetic field alignment. The diffuse spots in Figure 5C
represent the intersection with the Ewald sphere of a diffuse
torus, perpendicular to the nematic director, due to interparticle
interferences. This proves that the rodlike molecules align
parallel to the shear flow, as was intuitively expected.
At very high volume fractions of polypeptide (ca. >70 wt
%), reached by water evaporation from samples in the unsealed
Couette cell, a sharp diffraction line appeared, superimposed
over the diffuse spots. This diffraction line was due to the onset
of some positional long-range order typical of another meso-
phase. Its symmetry was determined by recording additional
diffraction lines on a different X-ray diffraction setup.12 Dif-
fraction lines in a ratio 1, 31/2, 2, 71/2 were observed (Figure 7),
which identified the symmetry of this mesophase as hexagonal.
This is not surprising because suspensions of rodlike particles
often form lyotropic hexagonal mesophases. Upon shearing in
the Couette cell, the hexagonal phase became partially aligned
as shown by the rather small mosaic spread of the reflection
(∼20° fwhm). The R-helices, and therefore the C6 axis, were
To analyze this system more quantitatively, the Onsager
model predicts the volume fractions Φn and Φi of the nematic
and isotropic phases at the transition: Φn ) 4.2 D/L and Φi )
3.3 D/L. For a short L-1 (Mw ) 62 kDa; L ) 32.3 nm and D
) 2.2 nm), we obtained Φn ) 29% and Φi ) 23%. These results
are in fair agreement with the experimental values of 47% and
38%, respectively, all the more because the Onsager model only
gives quantitative predictions for very long rods (L/D > 100).
It is also likely that the experimental values are higher than
predicted because the R-helix does have some finite flexibility,
(23) Impe´ror-Clerc, M.; Davidson, P. Eur. Phys. J. B 1999, 9, 93-104.
(24) Ramos, L.; Molino, F.; Porte, G. Langmuir 2000, 16, 5846-5848.
(25) Hamley, I. W. The Physics of Block-Copolymers; Oxford University
Press: New York, 1998.
(26) Camerel, F.; Gabriel, J. C. G.; Batail, P.; Davidson, P.; Lemaire, B. J.;
Schmutz, M.; Gulik-Krzywicki, T.; Bourgaux, C. Nano Lett. 2002, 2, 403-
407.
(27) Onsager, L. Ann. N.Y. Acad. Sci. 1949, 51, 627-659.
(28) Vroege, G. J.; Lekkerkerker, H. N. W. Rep. Prog. Phys. 1992, 55, 1241-
1309.
(20) Ugaz, V. M.; Cinader, D. K., Jr.; Burghardt, W. R. J. Rheol. 1997, 42,
379-394.
(21) Molino, F. R.; Berret, J. F.; Porte, G.; Diat, O.; Lindner, P. Eur. Phys. J.
B 1998, 3, 59-72.
(22) Hamley, I. W.; Pople, J. A.; Fairclough, J. P. A.; Terrilli, N. J.; Ryan, A.
J.; Booth, C.; Yu, G. E.; Diat, O.; Almdal, K.; Mortensen, K.; Vilgid, M.
J. Chem. Phys. 1998, 108, 6929-6936.
9
9104 J. AM. CHEM. SOC. VOL. 126, NO. 29, 2004

Aqueous Cholesteric Liquid Crystals
A R T I C L E S
especially with the poly(lysine) backbone, which may serve to
effectively decrease L.29 For a longer L-1 (Mw ) 120 kDa; L
) 62.6 nm and D ) 2.2 nm), the predicted volume fractions
were Φn ) 15% and Φi ) 12%, as compared to the
experimental values of 39% and 17%, respectively. Here, the
experimental biphasic gap (Φn/Φi) was much larger than the
prediction, which could be due to sample polydispersity.28
Altogether, these theoretical considerations suggest that the
R-helical polypeptides D-1 and L-1 are essentially rodlike and
that their interactions in water are purely steric.
The formation of a chiral nematic phase in concentrated
solutions is a predictable outcome for these polypeptides. A
typical average distance d between rods was obtained from the
positions of the maxima of the diffuse scattering spots. Assum-
ing a local hexagonal packing of rods, this distance is related
to the sample volume fraction φ by the formula d ) 1.65φ-1/2R,
where R is the apparent rod radius. This gives us R ) 1.1 nm
in good agreement with estimations from molecular modeling
(1.21 nm). The formation of a hexagonal columnar mesophase
at volume fractions above that of the cholesteric phase is also
expected for rods that are monodisperse in diameter. From our
results, it appears that the phase diagram of D-1 and L-1 in
aqueous solution is similar to that of PBLG in organic solvents.20
However, the aqueous system of D-1 and L-1 does show
some notable differences from organic systems employing
PBLG. At comparable chain lengths, PBLG forms the liquid-
crystalline phase at significantly lower volume fractions. This
phenomenon might be due to the increased flexibility of the
R-helix with a lysine backbone as compared to glutamate, as
mentioned above. It might also be related to differences in the
polymer-solvent interactions, which are likely greater for 1 in
water than those for PBLG in organic solvents. The solvent
interactions for 1 are stronger because the ethylene glycol side-
chains are solubilized by H-bonding to water, leading to weaker
interactions between the polypeptide chains and the need for
higher volume fractions to induce ordering. Aside from these
differences, the cholesteric phase of 1 can be well understood
by analogy to the PBLG system. As such, our results show that
there is now a simple homopolypeptide system that provides
robust, water-based cholesteric liquid crystals with predictable
properties. This system may find use in biomimetic composite
materials synthesis, characterization of optically active com-
pounds as an NMR solvent, and for performing asymmetric
catalysis in an aqueous environment.
Acknowledgment. We are deeply indebted to D. Dalle´ and
B. Kasmi for the design of the Couette shear cell apparatus.
This work was supported by the National Science Foundation
MRSEC Program DMR-0080034. E.G.B. thanks the UCSB
MRL for an Executive Vice Chancellor’s Travel Fellowship.
Supporting Information Available: Details of all materials
and measurements. This material is available free of charge via
the Internet at http://pubs.acs.org.
(29) Daniel, E.; Katchalski, E. In Polyamino Acids, Polypeptides, and Proteins;
Stahmann, M. A., Ed.; University of Wisconsin Press: Madison, WI, 1962;
p 183.
JA047932D
9
J. AM. CHEM. SOC. VOL. 126, NO. 29, 2004 9105