Detection of chiral disorder in langmuir monolayers undergoing spontaneous chiral segregation

Article abstract of DOI:10.1021/ja982123d

Acid - base interactions between two different chiral amphiphilic molecules (p-pentadecylmandelic acid and p-tetradecylphenylethylamine) are used in order to achieve spontaneous chiral segregation from a racemic mixture at the air - water interface, as sh

Full text of DOI:10.1021/ja982123d

J. Am. Chem. Soc. 1999, 121, 2657-2661
2657
Detection of Chiral Disorder in Langmuir Monolayers Undergoing
Spontaneous Chiral Segregation
Ivan Kuzmenko,*,† Kristian Kjaer,‡ Jens Als-Nielsen,§ Meir Lahav,† and
Leslie Leiserowitz*,†
Contribution from the Department of Materials and Interfaces, The Weizmann Institute of Science,
76100 RehoVot, Israel, Department of Solid State Physics, Risø National Laboratory DK-4000,
Roskilde, Denmark, and H. C. Ørsted Laboratory, Niels Bohr Institute, DK-2100 Copenhagen, Denmark
ReceiVed June 18, 1998. ReVised Manuscript ReceiVed December 18, 1998
Abstract: Acid-base interactions between two different chiral amphiphilic molecules (p-pentadecylmandelic
acid and p-tetradecylphenylethylamine) are used in order to achieve spontaneous chiral segregation from a
racemic mixture at the air-water interface, as shown by grazing incidence X-ray diffraction. The extent of
possible chiral disorder for both acid and base components is examined. We show that an oblique lattice
symmetry is insufficient to guarantee spontaneous segregation of chiral molecules in two-dimensional crystals
because of possible chiral disorder.
Introduction
chiral separation in two dimensions. Moreover, the use of this
acid-base binary system has allowed us to probe the extent of
possible chiral disorder.
Spontaneous separation of chiral molecules from a racemic
mixture into two-dimensional crystals of opposite handedness
echoes Pasteur’s famous experiment with sodium ammonium
tartrate tetrahydrate crystals performed 150 years ago.1 Defini-
tive evidence on spontaneous separation in bulk crystals can
be easily obtained by common experimental tools such as optical
rotation measurements, chromatography on chiral columns,
phase diagrams, or X-ray structure analysis. Determination of
chiral separation in monolayer films is far less straightforward.2
Early surface pressure-molecular area (π-A) isotherm mea-
surements provided indirect evidence for spontaneous resolution
of some racemates in two dimensions.3-5 Much more direct
observation of spontaneous separation of enantiomers became
possible only with the advent of modern techniques such as
grazing incidence X-ray diffraction (GIXD) and atomic force
(AFM), scanning tunneling (STM), Brewster angle (BAM), and
epifluorescence microscopies. The latter two methods are optical
and provide information on the shape of monolayer domains at
least a few micrometers in size. Observations of asymmetric
shapes of the polycrystalline monolayer domains in racemic
mixtures were ascribed to the separation into crystalline clusters
of opposite handedness.6 A comparative GIXD and epifluores-
cence microscopy study of racemic diol monolayers on water
showed, however, that such a correlation does not necessarily
hold.7
Experimental Section
1
General Procedures. H NMR spectra were recorded on a Bruker
spectrometer (270 MHz) in CDCl3 using OTS as internal standard.
Multiplicities in 1H NMR are reported as broad (br), singlet (s), doublet
(d), triplet (t), double doublet (dd), quartet (q), double quartet (dq),
and multiplet (m). Thin-layer chromatography (TLC) was performed
on Merck silica gel plates. Column chromatography was performed
on silica gel columns.
p-Tetradecylacetophenone (1). A chloroform solution containing
tetradecylbenzene (15.4 g, 56.2 mmol) and acethyl chloride (11.7 g,
0.169 M) was slowly dripped into a chloroform solution of AlCl3. The
mixture was stirred for 1 h while the temperature was kept below 10
°C and then stirred for another hour at 25 °C. The solution was poured
into ice. The chloroform layer was separated from the aqueous solution
and the latter washed several times with ether. The ether extracts were
collected and combined with the chloroform fraction, treated with
sodium bicarbonate, and dried with MgSO4. The organic solvent was
evaporated and the solid residue was recrystallized from ethyl acetate/
hexane/methanol to yield white solid 1 (13.8 g, 80% yield), mp 54-
55 °C. 1H NMR (CDCl3) δ 7.88 and 7.23 (dd, 4H, aromatic A2B2
system), 2.66 (t, 2H, PhCH2), 1.60 (s, 3H, COCH3), 1.28 (m, 24H,
aliphatic CH2), 0.90 (t, 3H, CH2CH3).
(R,S)-p-Tetradecylphenylethylamine (2). Compound 1 (12.5 g, 41.4
mmol), ammonium acetate (31 g, 378 mmol), and NaBH3CN (1.76 g,
0.279 M) were dissolved in absolute methanol (20 mL) in an inert N2
atmosphere. The solution was stirred for 3 days at 60 °C until the
reaction was complete. The mixture was acidified with concentrated
HCl to pH 2.1. Methanol was evaporated, and the remaining aqueous
solution was washed with CH2Cl2. The hydrochloride salt of 2 was
reacted in water with KOH to release the free amine that was extracted
with CH2Cl2 and dried to yield 2 (8.0 g, 60% yield), mp 68-69 °C. 1H
NMR (CDCl4) δ 7.24 (dd, 4H, aromatic A2B2 system), 4.08 (q, 1H,
NCH), 2.58 (q, 2H, PhCH2), 1.37 (d, 3H, NCCH3), 1.25 (m, 24H,
aliphatic H), 0.88 (t, 3H, CH2CH3).
In this work, we have made use of known acid-base
interactions involving two different chiral molecules to induce
† The Weizmann Institute of Science.
‡ Risø National Laboratory.
§ Niels Bohr Institute.
(1) Pasteur, L. Ann. Phys. 1848, 24, 442.
(2) Kuzmenko, I.; Weissbuch, I.; Gurovich, E.; Leiserowitz, L.; Lahav,
M. Chirality 1998, 10, 415.
(3) Lundquist, M. Prog. Chem. Fats Other Lipids 1978, 16, 101.
(4) Arnett, E.; Thompson, O. J. Am. Chem. Soc. 1981, 103, 968.
(5) Stewart, M.; Arnett, E. Top. Stereochem. 1982, 13, 489.
(6) Brezesinski, G.; Rietz, R.; Kjaer, K.; Bouwman, W.; Mo¨hwald, H.
NuoVo Cimento 1994, 16D, 1487.
(R)-(or S-)p-Tetradecylphenylethamine (3). Racemic 2 was opti-
cally resolved via cocrystallization with optically pure (R)- or (S)-
mandelic acid. Racemic 2 (23.5 g, 11 mmol) and R(-)mandelic acid
(1.5 g, 10 mmol) were dissolved in ethyl acetate (250 mL) at 60 °C.
(7) Mo¨hwald, H.; Dietrich, A.; Bo¨hm, C.; Brezesinski, G.; Thoma, M.
Mol. Membr. Biol. 1995, 12, 29.
10.1021/ja982123d CCC: $18.00 © 1999 American Chemical Society
Published on Web 03/12/1999

2658 J. Am. Chem. Soc., Vol. 121, No. 12, 1999
Kuzmenko et al.
Thin crystalline plates precipitated upon cooling. The salt was recrystal-
lized several times. The free amine was released in 10% aqueous
solution of KOH, extracted with pure CH2Cl2, and washed repeatedly
with water, until all the mandelic acid was removed as indicated by
TLC (CH2Cl2/ethanol 2:1 solution on silica plates). The solution was
dried with MgSO4 to yield (R)-3 with an optical roation of [R]D
+12.95° (c ) 2, CH2Cl2). The H NMR spectrum is identical to that
of racemic 2. The S-enantiomer of 2 was obtained analogously using
(S)-(+)mandelic acid, [R]D25 ) -12.77° (c ) 2, CH2Cl2).
)
25
1
r-Keto-â-(p-pentadecylphenyl)ethyl Propionate (4). AlCl3 (6.86
g, 51.5 mmol) dissolved in chloroform (14 mL) was placed in a three-
neck flask in an inert dry N2 atmosphere. A solution of pentadecyl-
benzene (10 g, 34.2 mmol) and ethyl oxalyl chloride (7.03 g, 51.5
mmol) in chloroform (7 mL) was was slowly added from a dropping
funnel at 0-5 °C. During the exothermic reaction, the yellow reaction
mixture turned dark brown. The reaction was then stirred in the ice
bath for 1 h and then for 3 h at ambient temperature. The reaction
mixture was poured into ice, and organic components were extracted
three times with ether. The organic fractions were combined and washed
with an aqueous solution of sodium bicarbonate and dried with MgSO4,
and the solvent was evaporated. The yellow oil was chromatographed
on a silica column with CH2Cl2/hexane (1:1) to yield 4 (12 g, 91%),
mp 129-131 °C. 1H NMR (CDCl4) δ 7.16 and 7.30 (dd, 4H aromatic
A2B2 system), 5.12 (s, 1H, CH), 4.20 (dq, 2H, COOCH2), 2.59 (t, 2H,
ArCH2), 1.73 (b, 1H, OH), 1.59 (t, 3H, OCH2CH3), 1.25 (m, 26H,
aliphatic CH2), 0.87 (t, 3H, CH2CH3).
(R)-p-Pentadecylmandelic Acid (5). An assymetric reduction of 4
was performed in the following steps. NaBH4 (1.47 g, 38.7 mmol) and
(R,R)-tartaric acid (5.55 g, 38.0 mmol) were dissolved in THF (130
mL) in a three-necked flask and heated in an inert dry N2 atmosphere
at 70 °C for 3 h. The resulting white suspension was cooled to -20
°C. Compound 4 (9 g, 0.023 M) in THF (10 mL) was added dropwise
(exothermic reaction) and the reaction mixture was stirred for 15 h at
-20 °C. Ethyl acetate (60 mL) and 1 M HCl (30 mL) were slowly
added to the mixture at 0 °C. After adding water (50 mL), the organic
layer was separated, washed several times with an aqueous solution of
sodium bicarbonate, and dried with MgSO4. The organic solvent was
evaporated to yield an oil that was chromatographed on a silica column
with ethyl acetate/hexane mixture (1:6). The solvent was evaporated
to yield white crystalline ethyl ester of p-pentadecylmandelic acid (7.5
g, 87%).
Figure 1. (Top) Structural formulas of the C15-MA and C14-PEA
molecules. The asterisk marks the chiral carbon center. (Bottom) Surface
pressure-molecular area (π-A) isotherms at 5 °C for R,R′, R,S′, and
R,S,R′,S′ diastereomeric films formed by C15-MA and C14-PEA
molecules in a 1:1 molar ratio. The molecular area A was calculated
per one amphiphilic molecule, disregarding the difference between MA
and PEA.
temperature (17 °C) on a four-circle Rigaku single-crystal diffractometer
with a 18-kW rotating anode generator. The intensity data of 4015
reflections (0 e h e 67, -10 e k e 0, 0 e l e 8) were measured.
After merging equivalent reflections, 1066 observed reflections re-
mained with I e 2σ(I). The structure was solved with SHELXS978
and refined isotropically with use of SHELXTL97 programs.9 Hydrogen
atoms at C and N atoms were positioned in calculated positions and
refined as riding on the corresponding nonhydrogen atoms with fixed
U ) 0.08. The R-hydroxy oxygen atom was found to be disordered
over two sites (xR and yS in the above formula) was close to 1 and
constrained in the further refinement (y ) (1 - x)) while the U factor
variable was common for both the oxygen sites. The final refinement
resulted in ) 0.45: empirical formula C31H47O3N, Mr ) 481.70 g,
orthorhombic, P212121, a ) 52.140(10) Å, b ) 8.450(2) Å, c )
6.8040(10) Å, V ) 2997.7(10) Å3, Z ) 4, Dx ) 1.067 g/cm3, Mo KR
(0.710 73 Å, 2θmax ) 55°, F(000) ) 1056, R ) 0.139, and S ) 0.884.
Grazing Incidence X-ray Diffraction. The GIXD measurements
were conducted on the liquid surface diffractometer at the undulator
beam line BW1 in HASYLAB at Deutsches Elektronen-Synchrotron
(DESY). The synchrotron radiation beam monochromated to wave-
lengths of 1.336 and 1.452 Å that were used in two synchrotron
sessions. The incident angle was adjusted to Rf ) 0.85Rc, where Rc ≈
0.14°. A detailed explanation of the method and experimental setup is
given elsewhere.10,11 The diastereomeric mixtures were prepared by
dropwise spreading of chloroform solutions at ambient temperature in
a Langmuir trough and then cooled to 5 °C.
The ester (6.1 g, 15.3 mmol) was added in 5 M KOH/methanol
solution (55 mL) with a few drops of water and stirred for 2.5 h at
ambient temperature. White crystals of the salt started precipitating
after several minutes. After 2 h, the solution with crystals was acidified
with HCl and the white crystalline substance was filtered and washed
with water and hexane. The solid was recrystallized from THF to yield
5 (3.0 g, 56%), [R]D25 ) -63.21 (c ) 1.01, THF). The acid was further
optically purified via crystallization with (R)-(+)phenylethylamine for
several times until reaching the constant rotation value of [R]2D5
)
Results and Discussion
-83.11 (c ) 0.965, THF), mp 119-121 °C. 1H NMR (CDCl3) δ 7.08
and 7.27 (dd, 4H, aromatic A2B2 system), 5.06 (s, 1H, CH), 2.49 (t,
2H, ArCH2), 1.16 (m, 26H, aliphatic CH2), 0.78 (t, 3H, CH2CH3).
(R,S)-p-Pentadecylmandelic Acid (6). NaBH4 (0.44 g, 11.61 mmol)
was suspended with THF (40 mL) in a three-neck flask in an inert N2
atmosphere. The mixture was cooled to 0 °C, and 4 (2.84 g, 7.73 mmol)
was added in THF (5 mL) dropwise. The reaction mixture was left to
stir in the bath for 1 h at 0 °C and overnight at ambient temperature.
Ethyl acetate (20 mL) and 1 M HCl (10 mL) were slowly added to the
mixture at 0 °C. Water was added and the organic layer separated,
washed with sodium bicarbonate, dried with MgSO4, and recrystallized
in THF to yield the racemic ethyl ester of p-pentadecylmandelic acid
(1.06 g, 90%). Hydrolysis of the ester was done in the same way as
for (R)-5. 1H NMR of the final product demonstrated the same spectrum
as for 5 and zero optical rotation.
Mandelic acid (MA) and phenylethylamine (PEA), each with
one chiral center (Figure 1, top), can be used for mutual chiral
resolution of their racemic R,S mixtures. All the reported (MA,-
PEA) three-dimensional crystal structures with various chiral
composition have a bilayer arrangement with one crystallo-
graphically independent (MA,PEA) unit in each layer.12-15 The
(8) Sheldrick, G. M. Acta Crystallogr. 1990, A46, 467.
(9) Sheldrick, G. M. SHELXL97 Program for the Refinement of Crystal
Structures; University of Go¨ttingen, Germany, 1997.
(10) Als-Nielsen, J.; Jacquemain, D.; Kjaer, K.; Leveiller, F.; Lahav,
M.; Leiserowitz, L. Phys. Rep. 1994, 246, 251.
(11) Majewski, J.; Popovitz-Biro, R.; Bouwman, W.; Kjaer, K.; Als-
Nielsen, J.; Lahav, M.; Leiserowitz, L. Chem. Eur. J. 1995, 1, 304.
(12) Brianso, M.; Leclercq, M.; Jacques, J. Acta Crystallogr. 1979, B35,
2751.
(13) Larsen, S.; Diego, H. Acta Crystallogr. 1993, B49, 303.
(14) Diego, H. Acta. Chem. Scand. 1994, 48, 306.
(15) Diego, H. 1-Phenylethylammonium, mandelates, a structural and
physico-chemical investigation of a complicated system of diastereomers.
Department of Chemistry, H. C. Ørsted Institute, University of Copenhagen,
1995.
X-ray Crystal Structure Determination. A platelike needle of
{(xR,(1 - x)S)-C15-MA,S-PEA} (x was found in the refinement) of
dimensions 0.2 × 0.4 × 0.8 mm3 was grown by slow cooling of
equimolar mixture of (R)-p-pentadecylmadelic acid (95% ee) and
commercial (S)-PEA from Aldrich (99% ee) dissolved in tetrahydro-
furan. Crystallographic measurements were performed at ambient

Chiral Disorder in Langmuir Monolayers
J. Am. Chem. Soc., Vol. 121, No. 12, 1999 2659
molecules are interconnected by multiple hydrogen bonds within
a two-dimensional network. It is important to note that, in all
these crystal structures, the MA (or PEA) molecules within a
layer have the same absolute configuration. The same structural
feature is also preserved in the 3D structures of the MA salts
with substituted PEA all assuming very similar packing ar-
rangements.16 Thermodynamic properties and crystal structures
of various PEA-MA compositions demonstrated a higher
stability of diastereromeric crystals, in which MA and PEA are
of the same absolute configuration. This structural feature
should, in principle, be preserved in analogous long-chain
compounds which can be used for monolayer studies. Thus,
for 1:1 mixtures of racemic long-chain amphiphiles bearing
mandelic acid and phenylethylamine head groups, we may
expect the monolayer to be separated into crystalline islands
composed of (MA,PEA) units of the same absolute configura-
tion, i.e., R,R′ or S,S′ rather than their diastereomeric17 pairs
R,S′ or S,R′, where R and S denote the absolute chiral
configuration of MA and R′ and S′ that of PEA.
Para-substituted analogues of PEA and MA bearing long
hydrocarbon chains CnH2n+1 (labeled Cn) were synthesized to
form racemic and chiral resolved amphiphilic compounds C14-
PEA and C15-MA (Figure 1, top). Surface pressure-molecular
area (π-A) isotherms of three different mixtures ((R)-C15-MA,
(R)-C14-PEA), ((R)-C15-MA, (S)-C14-PEA), and ((R,S)-C15-MA,
(R,S)-C14-PEA) on Millipore water indicate formation of stable
monolayers with a limiting area per molecule of 25-28 Å2
(Figure 1). All three isotherms have similar shapes. The GIXD
spectra for the three compositions were measured at a surface
pressure π ) 1 mN/m. All the three measured GIXD patterns
display seven narrow Bragg peaks (Figure 2a-c, left) as is
evident from their full width at half-maximums (average fwhm
in Table 1). These diffraction patterns were fully indexed
yielding cell dimensions (Table 1). Each unit cell of area ∼51
Å2 contains two symmetry-independent long-chain molecular
units corresponding to one acid and one amine molecule. The
GIXD patterns for the R,R′ and R,S′ mixtures are clearly
different both in terms of positions of the Bragg peaks, their
integrated intensities (Figure 2a,c, left) and, in particular, the
shapes of the corresponding Bragg rods (Figure 2d), whereas
the mixture R,S,R′,S′ (Figure 2b, left) displays a diffraction
pattern almost identical to that of the R,R′ monolayer. Since
the oblique unit cell for R,S,R′,S′ contains only two independent
molecules and the monolayer is highly crystalline in view of
the narrow Bragg peaks, it is highly plausible that the R,S,R′,S′
mixture segregates into R,R′ and S,S′ crystalline islands of
opposite handedness. The crystallization process for the three
systems is schematically shown in Figure 2a-c (right).
Figure 2. GIXD patterns for the three diastereomeric films R,R′ (thin
line), R,S′ (dotted line), and R,S,R′,S′ (thick line) measured at 5 °C
and surface pressure π ) 1 mN/m. (a-c, left) The observed Bragg
peaks represented by lines of corresponding integrated intensities with
assigned (h,k) indices. (d) Comparison of Bragg rods of the three
diastereomeric systems corresponding to the most intense (1, 0), (0,
2), and (1, -2) reflections, where the type of line corresponds to that
as in (a-c, left). (a-c, right) Schematic views of the molecular
assembly into crystals corresponding to the three diffraction patterns.
Table 1. Crystallographic Data for the Monolayers
parameters
R,R′
R,S′
0.5(R,S)-R′ R,S,R′,S′
lattice spacings (Å)
d01
d10
d11h
d02
d11
d12h
d12
9.26
9.25
9.25
5.46
5.25
4.62
4.31
3.99
3.22
0.0142
480
9.28
5.47
5.24
4.63
4.35
3.98
3.22
0.0143
470
5.48
5.19
4.61
4.37
3.94
3.25
5.38
5.28
4.62
4.20
4.05
3.13
av fwhma (Å-1
)
0.0124 0.0197
av coherence lengthb (Å) 560
unit cell dimens
320
Racemic mixtures of amphiphilic molecules bearing a long
hydrocarbon chain and a chiral hydrophilic head group would
seem, when compared with their water-soluble analogues in the
bulk, to have a lower tendency to separate in two dimensions
since hydrocarbon chains readily pack in a herringbone motif
in which R (right-hand) and S (left-hand) molecules are related
by glide symmetry. Most of the examples of chiral separation
in two dimensions, based on GIXD18,19 and in other reports by
AFM20 and STM,21 have rested primarily upon the finding of
an oblique lattice with one amphiphilic molecule per unit cell
a (Å)
b (Å)
γ (deg)
tilt angle (deg)
tilt azimuthc (deg)
unit cell area, Axy (Å2)
5.59
9.40
101.5
40
42
51.5
5.56
9.55
104.8
39
49
51.3
20.0
5.60
9.47
102.8
41
43
51.7
19.5
5.60
9.48
102.4
40
42
51.8
19.9
chain cross-section area 19.7
A (Å2)
a fwhm, full width at half-maximum of the Bragg peak in Qxy units.
b The coherence length L has been calculated using the standard Sherrer
equation:22 L ) (0.9 × 2π)/(fwhm2 - ∆2)1/2, where the resolution of
the Soller colimator, ∆ ) 0.0081 Å-1 (in Qxy units). c The angle between
the projection of the hydrocarbon chain vector onto the xy plane and
the reciprocal vector a* as determined from the positions of the
maximums qz of the three Bragg rods corresponding to (1, 0), (0, 2),
and (1, -2) reflections.10
(16) Kinbara, K.; Sakai, K.; Hashimoto, Y.; Nohiro, H.; Saigo, K. J.
Chem. Soc. Perkin Trans. 2 1996, 12, 2615.
(17) The term “diastereomeric” refers to optical isomers that are not
mirror images of each other.
(18) Nassoy, P.; Goldmann, M.; Bouloussa, O.; Rondelez, F. Phys. ReV.
Lett. 1995, 75, 457.
leading to an assumption that all molecules are related by
translation symmetry and, therefore, of the same handedness.
(19) Weissbuch, I.; Berfeld, M.; Bouwman, W.; Kjaer, K.; Als-Nielsen,
J.; Lahav, M.; Leiserowitz, L. J. Am. Chem. Soc. 1997, 119, 933-942.

2660 J. Am. Chem. Soc., Vol. 121, No. 12, 1999
Kuzmenko et al.
Figure 4. Layer packing of MA and PEA moieties in the observed
3D crystal structure of ((0.5R,0.5S)-C15-MA,(R)-PEA) with the observed
chiral disorder of the OH group. All hydrogen atoms and hydrocarbon
chains are omitted.
Figure 3. GIXD Bragg peaks (lines) for the two diastereomeric
mixtures: (a, left) 0.5(R,S)-R′ and (c, left) R-0.5(R′,S′), where the
diffraction pattern for R-0.5(R′,S′) is a combination of the R,R′ and
R,S′ phases with the h, k indices denoted above for R,R′ and below for
R,S′. (b) Bragg rod profiles of the 0.5(R,S)-R′ mixture (solid bold line)
vs the corresponding Bragg rods of the R,R′ composition (thin solid
line), originally shown in Figure 2d. (a,c, right) Schematic views of
the molecular assembly into crystals corresponding to the two diffraction
patterns.
phous domains of composition ([(1 - x) R, xS], R′) and ([(1 -
x) S, xR], S′),, where x is a fraction of “foreign” C15-MA
molecules within the R,R′ and S,S′ domains. For the three
compositions R,R′, R,S,R′,S′, and 0.5(R,S)-R′ there is no
significant difference in the Qxy positions of the corresponding
Bragg peaks. Nonetheless, comparison of the sensitive (1, -2)
Bragg rod of these three mixtures suggests that the R,S,R′,S′ is
much closer to R,R′ than to 0.5(R,S)-R′ (cf. Figure 2d and
Figure 3b).
Strong support for the possibility of chiral disorder for the
MA moiety may be drawn from the reported 3D structure of
((R)-MA,(S)-C1-PEA) with intralayer cell dimensions of a )
6.0 Å, b ) 9.1 Å, and γ ) 90°,16 which compares with those
of the diastereomeric monolayers (Table 1). The interchange
of the OH group and the hydrogen atom at the chiral carbon
center with the concomitant switch in chirality leads to formation
of an extra hydrogen bond and does not impair van der Waals
contacts. Experimental proof for such a type of chiral disorder
is provided by observed chiral disorder in the 3D crystal
structure ((0.45R,0.55S)-C15-MA,(R)-PEA), grown from tetra-
hydrofuran solution with a large excess of (S)-C15-MA mol-
ecules (R:S ) 5:95) and enantiomerically pure (R)-PEA. The
structure refinement revealed ∼50% chiral disorder for the OH
group (Figure 4). This structure is therefore a 3D analogue of
the Langmuir phase with the 0.5(R,S)-R′ chiral composition.
Such an assumption neglects potentially high molecular disorder
in monolayer films that are often mesophases rather than
crystalline aggregates. Thus we may envisage, in the ultimate,
a racemic 2D crystal (R:S ) 1:1) with the R and S chiral head
groups randomly distributed, although the long hydrocarbon tails
pack in an oblique unit cell by translation symmetry only.
In order to help evaluate the extent of chiral molecular
disorder for both the acid and amine moieties, two chiral
compositions R-0.5(R′,S′) and 0.5(R,S)-R′ with a 1:1 acid to
amine molar ratio were inspected. Both mixtures display π-A
isotherms similar to those previously described. The GIXD
measurements gave two qualitatively different results. The
mixture 0.5(R,S)-R′ composed of the acid molecules of both
handedness and the amine molecules of single handedness gave
rise to a diffraction pattern (Table 1 and Figure 3a, left, and b)
very similar to that of the R,R′ system. Since there is only one
crystalline phase and the corresponding diffraction pattern has
approximately the same intensity distribution as for R,R′, we
deduce that the (0.5(R,S), R′) phase is isomorphous with R,R′,
but where half the R sites are randomly occupied by S molecules,
as schematically shown in Figure 3a (right).
Conclusions
Diastereomeric interactions between two different chiral
molecules were used for the first time to achieve two-
dimensional spontaneous chiral separation into crystalline islands
of opposite handedness at the air-water interface. The extent
of chiral disorder, as assessed from the GIXD pattern analysis,
was shown to be always low for the amine component and in
some cases high (up to 50%) for the acid component. The results
show that an oblique lattice symmetry, even for highly crystal-
line monolayers, is insufficient proof for spontaneous separation
of chiral molecules in two dimensions. One way to ascertain
whether such a separation occurred would require a determi-
nation of the monolayer crystal structure making use of the
Bragg rod diffraction data, complemented by lattice energy
calculations.
The diffraction spectrum for the R-0.5(R′,S′) mixture,
composed of an acid of single chirality and amine molecules
of both handedness, combines two overlapping GIXD patterns,
characteristic of the R,R′ and R,S′ crystalline phases (Figure 3c,
left). The relative amount of crystalline material of these two
phases is approximately 1.5:1 according to the intensity ratio
of the corresponding diffraction patterns.
The diffraction results for the 1:1 mixtures R-0.5(R′,S′) and
0.5(R,S)-R′ suggest that R,S,R′,S′ separates into enantiomor-
(20) Eckhardt, C.; Peachey, N.; Swanson, D.; Takacs, J.; Khan, M.; Gong,
X.; Kim, J.; Wang, J.; Uphaus, R. Nature 1993, 362, 614.
(21) Stevens, F.; Dyer, D.; Walba, D. Angew. Chem. Int. Ed. Engl. 1996,
35, 900.
(22) Guinier, A. X-ray Diffraction; Freeman: San Francisco, 1968.

Chiral Disorder in Langmuir Monolayers
J. Am. Chem. Soc., Vol. 121, No. 12, 1999 2661
Acknowledgment. We thank Ronith Buller and Edna Shavit
for synthesis of the chiral amphiphilic molecules, Paul B. Howes
for assistance in GIXD measurements, and HASYLAB, DESY,
Hamburg, Germany, for synchrotron beamtime. This research
was supported by the Israel Science Foundation (ISF) and G.
M. J. Schmidt Minerva Center for Supramolecular Architectures
and the Danish Foundation for Natural Sciences.
Supporting Information Available: Tables of crystal data,
atomic coordinates, bond lengths and angles, isotropic displace-
ment parameters for ((xR,(1 - x)S)-C15-MA,(S)-PEA) (x ) 0.45)
(PDF). This material is available free of charge via the Internet
at http://pubs.acs.org.
JA982123D