Structure and reactivity in langmuir films of amphiphilic alkyl and thio-alkyl esters of ?±-amino acids at the air/water interface

Article abstract of DOI:10.1021/jp049381j

The structure and reactivity of alkyl esters of several ?±-amino acids self-assembled at the air/water interface have been investigated as part of our studies on mechanisms that are possibly relevant for the generation of homochiral prebiotic peptides. Grazing incidence X-ray diffraction (GIXD) studies of monolayers of racemic and enantiopure alkyl esters and thio-esters of alanine on the water surface demonstrated that these racemates self-assemble in the form of mixed solid solutions, because of disorder of the headgroups of the two enantiomers (enantiomeric disorder) within the two-dimensional (2D) crystallites. Matrix-assisted laser-desorption ionization time-of-flight Mass Spectrum (MALDI-TOF MS) analysis of the products collected from the air/water interface indicated the formation of low-molecular-weight oligopeptides (primarily dimers) and, in the case of some of the thioesters, small quantities of trimers and tetramers. Mass spectrometric studies on the diastereoisomeric distribution of the oligopeptides, starting from deuterium enantio-labeled monomers, demonstrated binomial statistics, such as that in reactions occurring in an isotropic environment. The alkyl esters of phenylalanine and tyrosine did not form 2D crystallites at the air/water interface, and, upon polycondensation, they yielded only dipeptides. The enantiomeric disorder within the 2D crystallites of the monomers of the alkyl esters and thioesters of racemic serine was absent. Polycondensation of these esters, however, yielded only dipeptides and tripeptides and they were not investigated further. In contrast to previous reports, the present studies demonstrate that this reaction does not proceed beyond the dipeptide stage and, therefore, cannot be regarded as a plausible system for the generation of prebiotic peptides.

Full text of DOI:10.1021/jp049381j

7228
J. Phys. Chem. B 2004, 108, 7228-7240
Structure and Reactivity in Langmuir Films of Amphiphilic Alkyl and Thio-alkyl Esters of
r-Amino Acids at the Air/Water Interface
†
†
‡
‡
Ran Eliash, Isabelle Weissbuch, Markus J. Weygand, Kristian Kjaer,
†
,†
Leslie Leiserowitz, and Meir Lahav*
Department of Materials and Interfaces, The Weizmann Institute of Science, 76100 RehoVot, Israel, and
Materials Research Department, Risø National Laboratory, 4000 Roskilde, Denmark
ReceiVed: February 11, 2004; In Final Form: March 24, 2004
The structure and reactivity of alkyl esters of several R-amino acids self-assembled at the air/water interface
have been investigated as part of our studies on mechanisms that are possibly relevant for the generation of
homochiral prebiotic peptides. Grazing incidence X-ray diffraction (GIXD) studies of monolayers of racemic
and enantiopure alkyl esters and thio-esters of alanine on the water surface demonstrated that these racemates
self-assemble in the form of mixed solid solutions, because of disorder of the headgroups of the two enantiomers
(enantiomeric disorder) within the two-dimensional (2D) crystallites. Matrix-assisted laser-desorption ionization
time-of-flight Mass Spectrum (MALDI-TOF MS) analysis of the products collected from the air/water interface
indicated the formation of low-molecular-weight oligopeptides (primarily dimers) and, in the case of some of
the thioesters, small quantities of trimers and tetramers. Mass spectrometric studies on the diastereoisomeric
distribution of the oligopeptides, starting from deuterium enantio-labeled monomers, demonstrated binomial
statistics, such as that in reactions occurring in an isotropic environment. The alkyl esters of phenylalanine
and tyrosine did not form 2D crystallites at the air/water interface, and, upon polycondensation, they yielded
only dipeptides. The enantiomeric disorder within the 2D crystallites of the monomers of the alkyl esters and
thioesters of racemic serine was absent. Polycondensation of these esters, however, yielded only dipeptides
and tripeptides and they were not investigated further. In contrast to previous reports, the present studies
demonstrate that this reaction does not proceed beyond the dipeptide stage and, therefore, cannot be regarded
as a plausible system for the generation of prebiotic peptides.
1
. Introduction
such monomers can provide ideal intermediates for the genera-
tion of natural oligopeptides by a process of their unzipping
The emergence of peptides of single handedness and homo-
from membranes or from vesicles. The reported formation of
the peptide bond was demonstrated either by infrared (IR)
chiral sequence (isotactic) from racemic R-amino acids in abiotic
times is one of the most interesting riddles in the field of the
9
-13
8
9
spectroscopy,
ninhydrin analysis, or surface viscosity.
1
,2
origin of life. Recently, we have presented an experimental
model for the formation of enantiopure isotactic oligopeptides
from nonracemic mixtures of activated amino acids bearing long
hydrocarbon chains at their R-carbon.3-7 This methodology is
comprised of two sequential steps: self-assembly of the
amphiphilic amino acids into two-dimensional (2D) crystallites
on the water surface or in a phospholipid environment, followed
by a lattice-controlled polymerization to form the corresponding
oligopeptides. In these systems, however, the oligopeptides
products bear long hydrocarbon chains.
Although these analytical methods reveal the formation of a
peptide bond, they do not provide information regarding the
molecular weight and enantiomeric composition of the peptides.
To probe the potential future application of this reaction as a
useful reaction model system in prebiotic chemistry, we
reinvestigated the reactivity of several molecules previously
9
-13
reported,
as well as some additional alkyl esters and thio-
esters of R-amino acids. We applied the grazing incidence X-ray
diffraction (GIXD) technique to determine the packing arrange-
ment of the monomer molecules within the 2D self-assemblies
and matrix-assisted laser-desorption ionization time-of-flight
mass spectrometry (MALDI-TOF MS) for the determination
of the molecular weight distribution of the peptides. For some
of the systems, the diastereoisomeric composition of the
oligopeptides was deduced by polymerizing monomers enan-
tioselectively labeled with deuterium.
Previously, it has been reported that Langmuir and Lang-
muir-Blodgett films of long-chain alkyl esters of amino acids
both in their enantiopure and racemic forms undergo polycon-
8-11
densation by a simple “unzipping” mechanism,
as illustrated
in Scheme 1. Based on kinetic studies, it was proposed that, in
these systems, the reaction rate was enhanced by the lateral
molecular order within the film, yielding peptides in the form
of R-helixes or â-sheets.1
2,13
The formation of secondary
2. Experimental Section
structure implies that these peptides contained at least eight
2.1. Materials. Materials that were used in this study include
1
4,15
repeating units.
Based on these reports, we considered that
protected amino acids (BACHEM AG and NovaBiochem),
L-alanine(d4) (MSD ISOTOPES), all solvents and reagents
*
Author to whom correspondence should be addressed: Telephone:
(BDH, Fluka, Aldrich, Sigma and Merck).
(
972)-8-934-2111. Fax (972)-8-934-4138. E-mail: meir.lahav@weizmann.ac.il.
†
‡
2.1.1. D-, L-, and DL-Alanine Octadecyl Ester (Ala-C18). The
Weizmann Institute of Science.
Risø National Laboratory.
16
free amine was prepared according to Penney et al. and was
1
0.1021/jp049381j CCC: $27.50 © 2004 American Chemical Society
Published on Web 05/07/2004

Alkyl Esters of R-Amino Acids
J. Phys. Chem. B, Vol. 108, No. 22, 2004 7229
SCHEME 1: Proposed Mechanism of the “Unzipping”
Polymerization, According to Fukuda et al.11
dissolved in 140 mL of dry CH2Cl2. Triethylamine (2.10 g, 20.80
mmol, 1.5 equiv) that was dissolved in 5 mL of CH2Cl2 was
added to the solution, resulting in the precipitation of a white
solid. The reaction was allowed to stir for 2.5 h and then 500
mL of diethyl ether and 100 mL of ethanol were added to form
a clear solution. The organic solution was washed with 0.5 M
HCl (50 mL, three times), water (50 mL, two times) and dried
with magnesium sulfate, and the solvent was evaporated in a
vacuum. The resulting crystalline material was recrystallized
from CH2Cl2 to give 4.01 g (88% yield) of pure octadecanamide
1
N-ethanthiol. H NMR (CDCl3 250 MHz): δ ) 0.88(t, 3H;
CH3-[CH2]14-), 1.25(br m, 28H; CH3-[CH2]14-), 1.64(m, 2H;
-
CH2CH2CONH-), 2.21(t, 2H; -CH2CH2CONH-), 2.67(q,
2
5
3
H; -CONHCH2CH2SH), 3.43(q, 2H; -CONHCH2CH2SH),
+
.86(br s, 1H; -CONH-); ESI-MS: m/z: 328.30[M + H],
50.27[M + Na].
+
2
.1.6. Octadecanamide N-Ethanthiol (C18-ATE). Same pro-
cedure as above, except that 3 equiv of triethylamine and 1.5
equiv of 2-aminoethanthiol chloride were used, and the product
1
was recrystallized from a n-hexane/CH2Cl2 mixture. H NMR
(
CDCl3, 250 MHz): δ ) 0.88(t, 3H; CH3-[CH2]14-), 1.25(br
m, 28H; CH3-[CH2]14-), 1.63(m, 2H; -CH2CH2CONH-),
2
2
5
3
.19(t, 2H; -CH2CH2CONH-), 2.66(br s, 1H; OH), 3.46(q,
H; -CONHCH2CH2OH), 3.75(m, 2H; -CONHCH2CH2OH),
+
.91(br s, 1H; -CONH-); ESI-MS: m/z: 344.52 [M + H],
+
-
66.54 [M + Na], 342.39[M - H]. Silica gel TLC, ethyl
acetate:CH2Cl2 ratio of 1:9, Rf ) 0.5.
2
.1.7. Boc-L-alanine(d4). Boc-L-alanine(d4) was prepared
1
8 1
according to Itoh et al. H NMR (CDCl3, 250 MHz): δ )
1
.45(s, 9H), 6.55(s, 1H; -CONH-), 7.65(br s, 1H; -COOH);
further purified by column chromatography (silica gel, ethyl
-
1
ESI-MS: m/z: 192.08 [M - H].
acetate:n-hexane:acetone ratio of 5:4.5:0.5). H NMR (CDCl3
50 MHz): δ ) 0.88(t, 3H; CH3-[CH2]15-), 1.30(br m, 33H;
2
2.1.8. p-Octadecyl-phenol (C18-PhOH). p-Octadecyl-phenol
(C18-PhOH) was prepared according to Hanabusa et al.1
0 1
H
CH3-[CH2]15-; -CHCH3NH2), 1.64(m, 2H; -CH2CH2OOC-
)
, 3.54(q, 1H; -CHCH3NH2), 4.11(m, 2H; -CH2CH2OOC-);
NMR (CDCl3, 250 MHz): δ ) 0.88(t, 3H; CH3), 1.25(br m,
30H; [CH2]15), 1.56(m, 2H; -CH2CH2Ph), 2.52(t, 2H; -CH2-
CH2Ph), 4.68(s, 1H; -PhOH), 6.75/7.04(double d, 4H; aromatic
+
ESI-MS: m/z: 364.58 [M + Na]. Silica gel TLC, n-hexane:
acetone ratio of 1:1, Rf ) 0.75.
+
+
2
.1.2. L- and DL-Phenylalanine Octadecyl Ester (Phe-C18). The
hydrogen); ESI-MS: m/z: 369.51[M + Na], 385.52[M +
1
6
-
free amine was prepared according to Penney et al. and
purified by column chromatography (silica gel, ethyl acetate:
K], 354.34[M - H]. Silica gel TLC, 10% ethyl acetate in
n-hexane, Rf ) 0.28.
+
n-hexane ratio of 2:3). ESI-MS: m/z: 418.64 [M + H], 440.62
2
.1.9. D-Alanine p-Octadecyl-phenyl Ester Trifluoroacetate
+
[
M + Na]. Silica gel TLC, ethyl acetate:n-hexane ratio of 2:3,
Salt (Ala-C18-Ph). Esterification of the D-alanine p-octadecyl-
Rf ) 0.30.
.1.3. D- and L-Tyrosine Octadecyl Ester (Tyr-C18). The HCl
salt was prepared according to Penney et al. The free amine
was released from the hydrochloride salt by extraction with a
triethylamine/water mixture and recrystallized from n-hexane.
phenyl ester trifluoroacetate salt (Ala-C18-Ph) was accomplished
19
2
by following the method of Neises and Steglich. p-Octadecyl-
phenol (0.080 g, 0.19 mmol, 1 equiv), Boc-D-alanine (0.036 g,
0.19 mmol, 1 equiv), and DCC (0.060 g, 0.28 mmol, 1.5 equiv)
were placed in a 25-mL round-bottom flask that was equipped
with a magnetic stirring bar and a drying tube. The materials
were dissolved in 10 mL of CH Cl through the application of
16
1
H NMR (5% CD3OD in CDCl3, 250 MHz): δ ) 0.88(t, 3H;
CH3-[CH2]15-), 1.25(br m, 30H; CH3-[CH2]15-), 1.60(m, 2H;
2
2
-
CH2CH2OOC-), 2.8-3.1(br m, 2H; -CH2PhOH), 3.73-
gentle heat, and the solution was placed over an ice-water bath.
White material precipitated, and the solution was further cooled
for 10 min. N,N-dimethyl 4-amino pyridine (0.0023 g, 0.019
mmol, 0.1 equiv) dissolved in a small amount of CH Cl was
(
)
[
double d, 1H; -CHNH2COO-), 4.09(t, 2H; -CH2CH2OOC-
, 7.15-7.35(br m; aromatic hydrogens); ESI-MS: m/z: 434.43-
M + H], 456.40[M + Na]. Silica gel TLC, ethyl acetate:
+
+
2
2
n-hexane:acetone ratio of 5:4:0.5, Rf ) 0.50.
added to the stirring mixture. The reaction was continued for 1
h at 0 °C, and then the ice was allowed to melt and the reaction
was continued for another hour. The white precipitate (DCU)
was filtered off and 100 mL of CH Cl were added to the
2
.1.4. L- and DL-Glutamic Dioctadecyl Ester [Glu-(C18)2]. The
1
6
methylsulfonate salt was prepared according to Penney et al.
and recrystallized from CH2Cl2. The free amine was released
from the hydrochloride salt by extraction with a triethylamine/
water mixture and recrystallized from a n-hexane/CH2Cl2
2
2
solution. The solution was washed with 0.5 M HCl (10 mL, 2
times) and sodium bicarbonate (20 mL, 2 times), dried with
magnesium sulfate, and the solvent evaporated. The resulting
material was transferred into a 10-mL round-bottom flask
equipped with a drying tube and a magnetic stirring bar. The
flask was placed over an ice-water bath, and 3 mL of CH2Cl2
and 5 mL of trifluoroacetic acid were added. The solution was
stirred for 1 h and then the solvent was evaporated by bubbling
+
+
mixture. ESI-MS: m/z: 652.96[M + H], 674.92[M + Na].
Silica gel TLC, ethyl acetate:n-hexane:acetone ratio of 2:7:1,
Rf ) 0.6.
2
.1.5. Octadecanamide N-Ethanol (C18-AE). N-Hydroxysuc-
17
cinimide ester of stearic acid (5.29 g, 13.86 mmol, 1 equiv)
and 2-aminoethanol (2.70 g, 44.2 mmol, 3.2 equiv) were

7230 J. Phys. Chem. B, Vol. 108, No. 22, 2004
Eliash et al.
air through the solution. Another portion of CH2Cl2 was added
and evaporated in the same manner. The remaining oil was
transferred to high vacuum for 1 h, and then diethyl ether was
added (7 mL); the oil was dissolved and, after a short time, a
white material precipitated. The mixture was kept at a temper-
ature of 4 °C overnight and the precipitate was filtered, washed
with 10 mL of diethyl ether, recrystallized from a n-hexane/
CH2Cl2 mixture, and dried in a vacuum to give 0.080 g (80%
(qxy)) associated with each h,k reflection. Bragg rod intensity
profiles are the intensity distribution along qz, I(qz), derived by
integrating across the qxy range for each of the Bragg peaks.
The full width at half-maximum of the Bragg rod intensity
profiles, fwhm(qz), gives a first estimate of the thickness of the
2D crystallites: Lz ≈ 0.9(2π/fwhm(qz)). For chainlike molecules,
such as aliphatic hydrocarbons, precise information on the
molecular chain orientation in 2D crystals may be obtained from
the positions of the maxima of Bragg rods, if the chains are
uniformly tilted. The tilt angle t between the molecular axis
1
yield) of pure Ala-C18-Ph. H NMR (5%CD3OD in CDCl3, 250
MHz): δ ) 0.88(t, 3H; CH3-[CH2]15-), 1.25(br m, 30H; CH3-
2
1,22
[
CH2]15-), 1.60(m, 2H; -CH2CH2Ph-), 1.70(d, 3H; -CHCH3-
and the surface normal is given by the equation
+
NH3 ), 2.60(t, 2H; -CH2CH2Ph-), 4.22(q, 1H; -CHCH3-
NH3 ), 6.99/7.19(double d, 4H; aromatic H); ESI-MS: m/z:
4
+
0
qz
+
+
18.52[M + H], 440.53[M + Na].]. Silica gel TLC, ethanol/
cos ψ tan t )
hk
|
q |
hk
CH2Cl2 ratio of 1:9, Rf ) 0.4.
Note: All of the following materials were synthesized in the
same manner as Ala-C18-Ph from the corresponding alcohols
or thio-alcohols and N-Boc protected amino acids.
0
z
where q is the position of the maximum along the Bragg rod
and ψhk is the azimuthal angle between the chain tilt direction
projected onto the xy-plane and the reciprocal vector qhk. More
accurately, the intensity at a particular value of qz in a Bragg
2
.1.10. D- and DL-Alanine Octadecanamid-N-Ethyl Ester
1
Trifluoroacetate Salt (Ala-C18-AE). H NMR (5% CD3OD in
CDCl3, 250 MHz): δ ) 0.88(t, 3H; CH3-[CH2]14-), 1.25(br
m, 28H; CH3-[CH2]14-), 1.57(br d, 5H; -CHCH3NH2; -CH2-
CH2CONH-), 2.16(t, 2H; -CH2CH2CONH-) 3.50(m, 2H;
CONHCH2CH2O-), 4.04(q, 1H; -CHCH3NH2), 4.1-4.4(br
m, 2H; -CONHCH2CH2O-), 6.77(t, 1H; -CONH-); ESI-
rod is determined by the square of the molecular structure factor,
2
|
Fh,k(qz)| . The 2D packing arrangements were determined by
performing X-ray structure factor calculations, using atomic
2
coordinate molecular models constructed using the CERIUS
molecular package, and rigid-body structure refinement, using
the SHELX-97 program adapted for 2D structures.
.3. Polymerization Experiments. The materials were dis-
solved in chloroform, benzene, or chloroform with 4% methanol
to give a solution concentration of 0.5 mM. The solution was
spread on the subphase to the desired area per molecule and
allowed to react for different periods of time, from a few minutes
up to 20 h (see Table 1). The subphase for the different
compounds were as follows: pure water; 0.01 M phosphate
buffer, pH 8.0 (3.4 mL of 1 M NaH2PO4 + 183 mL of 0.2 M
Na2HPO4 into 4 L of water); 1 mM or 0.1 mM copper chloride
-
23
2
4
+
+
MS: m/z: 399.55[M + H], 421.55[M + Na].
.1.11. L-Alanine(d4) Octadecanamid-N-ethyl Ester Trifluoro-
2
2
1
acetate Salt (Ala-C18-AE). H NMR (CDCl3, 400 MHz): δ )
.88(t, 3H; CH3-[CH2]14-), 1.26(br m, 28H; CH3-[CH2]14-
, 1.58(m, 2H; -CH2CH2CONH-), 2.17(t, 2H; -CH2CH2-
CONH-) 3.51(m, 2H; -CONHCH2CH2O-), 4.15-4.35(br m,
H; -CONHCH2CH2O-), 6.68(m, 1H; -CONH-); ESI-MS:
0
)
2
+
+
m/z: 403.82[M + H], 425.86[M + Na].
.1.12. D- and DL-Alanine Octadecanamid-N-ethyl Thioester
2
1
Trifluoroacetate Salt (Ala-C18-ATE). H NMR (5% CD3OD in
CDCl3, 250 MHz): δ ) 0.88(t, 3H; CH3-[CH2]14-), 1.25(br
m, 28H; CH3-[CH2]14-), 1.61(m, 5H; -CHCH3NH2; -CH2-
CH2CONH-), 2.16(t, 2H; -CH2CH2CONH-), 2.9-3.2(br m,
(
CuCl2); 0.01 M carbonate/bicarbonate buffer, pH 10.0 (88 mL
of 0.2 M Na2CO3 + 112 mL of 0.2 M NaHCO3 into 4 L of
water). In some cases, the material was spread on pure water
and then copper chloride was injected from a concentrated
solution to give the desired final concentration; for example, 1
mM (50 mL of 28 mM injected into 1350 mL (the trough
volume)). In these cases, the chloroform was allowed to
evaporate for 10-15 min prior to the injection of the catalyst.
2
4
H; -CONHCH2CH2S-), 3.39(m, 2H; -CONHCH2CH2S-),
.13(q, 1H; -CHCH3NH2), 6.68(t, 1H; -CONHCH2CH2S-);
+
+
ESI-MS: m/z: 415.61 [M + H], 437.59 [M + Na]. Silica
gel TLC, ethanol/CH2Cl2 ratio of 1:9, Rf ) 0.5.
2
.1.13. L-Alanine(d4) Octadecanamid-N-ethyl Thioester Tri-
2
.4. MALDI-TOF MS. After the reaction, the monolayer
+
fluoroacetate Salt (Ala-C18-ATE). ESI-MS: m/z: 419.74[M
films were compressed with the barrier and the material was
collected from the liquid surface, transferred to a glass vial,
and dried by a stream of dry nitrogen. Samples for the MALDI-
TOF MS analysis were then prepared by dissolving the dry
material in chloroform containing 5% trifluoroacetic acid. A
small amount (0.5 µL) of this solution was deposited on top of
a matrix deposit (1:1 v:v of dithranol solution in chloroform
and NaI saturated solution in tetrahydrofuran (THF)) on the
instrument holder. The MALDI-TOF positive-ion mass spectra
were obtained in reflector mode from two different instruments
at the Weizmann Institute (Bruker Reflex III) that were equipped
with a N2 laser. External calibration of the mass spectra was
achieved using a calibrating peptide (Substance P, ACTH 8-39)
+
+
H], 441.71[M + Na].
2
.1.14. D- and L-Serine Octadecanamid-N-ethyl Thioester
1
Trifluoroacetate Salt (Ser-C18-ATE). H NMR (CDCl3, 400
MHz): δ ) 0.82(t, 3H; CH3-[CH2]14-), 1.25(br m, 28H; CH3-
[
CH2]14-), 1.71(m, 2H; -CH2CH2CONH-), 2.60(t, 2H; -CH2-
CH2CONH-), 3.33(m, 2H; -CONHCH2CH2S-), 3.76(t, 2H;
-
-
CONHCH2CH2S-), 4.42(br m, 2H; -CH2OH), 4.67(m, 1H;
+
+
+
CHNH3 -); m/z: 431.79[M + H], 453.76[M + Na].
.2. GIXD Measurements. GIXD measurements were per-
2
formed using the liquid-surface diffractometer at the BW1
synchrotron beamline at HASYLAB, DESY facility in Ham-
burg, Germany. Details about the experimental technique and
the instrument were reported elsewhere.2 The measured GIXD
patterns are represented as 2D contour maps presenting the
scattered intensity (I(qxy, qz)) as a function of the horizontal
0
+
in the studied mass range. Only singly charged ions, (M + H)
and (M + Na) with the expected isotopic pattern, were
+
observed. Mass spectra resulted from a signal average of at least
a few hundred laser shots in different spots of the target to get
reliable statistics about the ion peak. Good statistics are obtained
when the isotopic distribution of an ion species corresponds to
that expected from calculation. Mass assignments were made
using both m/z measurement and isotopic distribution.
(qxy) and vertical (qz) components of the scattering vector. The
unit cell dimensions of the 2D lattice are derived from the qxy
positions of the Bragg peaks. The full width at half-maximum
of the Bragg peaks (corrected for instrument resolution), fwhm-
(qxy), gives the crystalline coherence lengths Lhk ≈ 0.9(2π/fwhm-

Alkyl Esters of R-Amino Acids
J. Phys. Chem. B, Vol. 108, No. 22, 2004 7231
TABLE 1: Experimental Conditions for Polymerization Experiments
mean area,
M (Å )
a
temperature,
materiala
solventb
2
c
duration
T (°C)
catalyst
Ala-C18
1, 3
1
1
1
2
2
1
2
1
25, 28d
35
49
38,35
35
35
35
35
45
45
10 min, 3 h, 5 h
20 h
3 h
10 min, 3 h, 5 h
10 min, 2.5 h
10 min, 3 h
10 min, 3.5 h
10 min, 2.5 h
3 h
20
20
20
20
20
3
20
3
3
water, pH 8, CuCl
water
2
1 mM
Ala-C26-TFAe
Phe-C18
CuCl
CuCl
CuCl
CuCl
2
2
2
2
1 mM
1 mM
1 mM
Tyr-C18
Ala-C18-AE-TFAe
Ala-C18-ATE-TFA
f
e
f
0.1 mM
e
Ala-C18-Ph-TFA
Ser-C18-ATE-TFA
Glu-(C18
Glu-(C18
water
water
CuCl
2
e
)
)
2
1 Mm
-MSAg
1
3 h, 24 h
30
water, pH 10
2
a
See Scheme 1. b The solvents are as follows: 1, chloroform; 2, chloroform/4% methanol; and 3, benzene. The mean area per molecule over
c
d
2
2
e
which the material was spread. The material was spread over an area of 30 Å /molecule and compressed to 28 or 25 Å /molecule. The material
was spread on the surface as a salt with trifluoroacetic acid. A concentrated catalyst solution was injected under the monolayer to achieve the
reported final concentration. The material was spread as a salt of methyl-sulfonic acid.
f
g
SCHEME 2: Molecular Formulas of the Monomers Investigated
We use the following notation code for the oligopeptide
molecules: (h,d) designates a molecule comprising h D-
protonated) repeat units and d L(deuterated) repeat units, with
n ) h + d, being the total number of repeat units. The relative
abundance (r.a.) of each type of oligopeptide (h,d) is obtained
by dividing the intensity of all the ions from a particular
molecule to the total intensity of the ions from all the molecules
of the same length, n. For example, the relative abundance of
a tetrapeptide (4,0) which is composed of four (protonated)
D-alanine molecules, is calculated from the equation
3. Results and Discussion
.1. Ala-C18 and Ala-C26. The first system we investigated
was Ala-C18 (Scheme 2). The GIXD patterns measured from
monolayers of DL- and L-Ala-C18 are shown in Figure 1a and
3
(
1
b, respectively, as 2D contour plots of intensity I(qxy,qz), as a
function of the horizontal qxy and vertical qz components of the
scattering vector. A relatively strong diffraction was obtained
2
at a mean molecular area of 28 and 24 Å (Figure 2). At this
surface coverage, the reported polymerization reaction was
9
optimal. At lower surface coverage, the diffraction intensity
Intensity(4,0)
r.a.(4,0) )
was very weak. Assignment of the (h,k) Miller indices to the
Intensity{(4,0) + (3,1) + (2,2) + (1,3) + (0,4)}
GIXD pattern measured from DL-Ala-C18 yielded a rectangular
2
unit cell (a ) 5.0 Å, b ) 10.5 Å, Am ) 26 Å ; see Table 2) that
The relative abundance of the tetrapeptide (2,2), which is
composed of two (protonated) D-alanine molecules and two
deuterated) L-alanine molecules is calculated from the equation
contained two molecules whose hydrocarbon chains are tilted
along the b-axis by an angle of 40° from the normal to the water
surface (calculated from the qz position of the maximum
(
2
1,22
intensity of the Bragg rods;
see Experimental Section). The
Intensity(2,2)
r.a.(2,2) )
unit cell has typical dimensions for a herringbone packing motif
of the two molecules that are related by glide symmetry along
the b-axis. The GIXD pattern measured from L-Ala-C18 was
assigned according to an oblique subcell (ao ) 5.0 Å, bo ) 5.4
Intensity{(4,0) + (3,1) + (2,2) + (1,3) + (0,4)}
The ionization yield is expected to be very similar for the
h,d) oligopeptides of the same length n ) h + d. In view of
the very similar masses of the compounds and the resulting very
similar ion velocities in the TOF mass spectrometer, we may
assume similar detection efficiencies.25 Thus, the ion intensity
of the different (h,d) oligopeptides are directly and reliably
comparable.
(
2
Å, γo ) 115.5°, Am ) 24 Å ) that can be transformed into a
cell with almost rectangular (pseudo-rectangular) dimensions:
2
ar ) 5.0 Å, br ) 9.8 Å, γr ) 91.7°, Am ) 24 Å . The latter
contains two independent molecules whose hydrocarbon chains
are related by pseudo-glide almost parallel to the b-axis. Note

7232 J. Phys. Chem. B, Vol. 108, No. 22, 2004
Eliash et al.
Figure 1. GIXD patterns I(qxy,q
area of 24 Å and (c) DL-Ala-C26 and (d) D-Ala-C26 at a mean molecular area of 35 Å .
z
) measured from the two-dimensional (2D) crystallites of (a) DL-Ala-C18 and (b) L-Ala-C18 at a mean molecular
2
2
to obtain a reasonable fit between the measured and calculated
Bragg rod intensity profiles (Figure 3a). The refined 2D packing
arrangement is shown in Figure 3b, viewed perpendicular to
the water surface.
The GIXD pattern (Figure 1d) measured from D-Ala-C26 was
analyzed in terms of a pseudo-rectangular unit cell (a ) 5.0 Å,
2
b ) 9.8 Å, γ ) 92.0°, Am ) 24 Å ) that contains two molecules
related by a pseudo-glide plane along the b-axis. The refined
model that best fitted the Bragg rod intensity profiles (Figure
3
c) is shown in Figure 3d, viewed perpendicular to the water
Figure 2. Surface pressure-mean molecular area isotherm of DL-Ala-
18 at 20 °C, showing the points where GIXD and polycondensation
surface.
C
Polymerization experiments performed with D- and DL-Ala-
C26 at 35 and 25 Å /molecule showed that the higher degree of
crystalline order did not contribute to the formation of longer
oligopeptides and, again, only dimers were obtained, according
to the MALDI-TOF MS analysis.
3.2. Ala-C18-AE. According to the GIXD results, DL-Ala-
C18 and DL-Ala-C26 pack in a racemic plane group containing
both enantiomers in the same crystallite. In such systems, the
polymerization can occur between either homochiral or hetero-
chiral molecules, depending on the molecular arrangement,
distance, and orientation of neighboring D-D versus D-L pairs.
A simpler way to obtain homochiral peptides would be when
racemic monomers separate on the water surface into a mixture
of enantiomorphous domains, each containing molecules of a
single-handedness. To achieve such a separation, we introduced
an amide group into the hydrocarbon chain because, as shown
reactions were performed. No diffraction was obtained at the points
2
2
marked with an asterisk, * (at 50 and 36 Å /molecule).
that the GIXD patterns measured from the two-dimensional (2D)
crystallites of DL- and L-Ala-C18 are significantly different from
that measured for the octadecanol precursor,26 indicating that
the two 2D packing arrangements are dependent on the chiral
alanine moiety.
Polymerization experiments performed using DL- and L-Ala-
2
C18 spread over a mean molecular area of 30 Å and compressed
2
to 28 or 25 Å yielded only dimers, as determined by MALDI-
TOF MS analysis of samples taken from the water surface. A
possible reason for this result could be the lower tendency of
these molecules to form ordered crystallites. Therefore, we
investigated Ala-C26 (Scheme 2), which we expected to self-
assemble on the water surface into 2D domains with a higher
degree of order, because of stronger interactions between the
longer hydrocarbon chains.26 Indeed, the GIXD patterns mea-
sured from DL- and D-Ala-C26, spread over a surface coverage
27
previously, this group can promote hydrogen bonding between
molecules related by translation, thus enhancing the possibility
of spontaneous separation of the racemate into a mixture of
enantiomorphous domains.
2
of only 70% (mean molecular area of 35 Å ), showed a strong
diffraction (see Figure 1c, d). The racemic DL-Ala-C26 packs in
a rectangular unit cell (a ) 5.0 Å, b ) 9.6 Å, Am ) 24 Å ) that
contains two molecules tilted along the b-axis by an angle of
Ala-C18-AE (Scheme 2) contains an amide group separated
by a -(CH2)2- “spacer” from the alanine-ester moiety. GIXD
patterns measured from self-assembled 2D crystallites of DL-
and D-Ala-C18-AE are presented in Figure 4a and b, respectively.
For comparison, a significantly different GIXD pattern was
measured from self-assembled 2D crystallites of the C18-AE
precursor alcohol (Scheme 2), as shown in Figure 4c. Initially,
an attempt was made to index the GIXD pattern of DL-Ala-
C18-AE, according to a pseudo-rectangular unit cell (ar ) 4.9
2
3
9° from the normal to the water surface. The molecules are
related by a b-glide symmetry in a herringbone packing motif
as racemic 2D crystallites. To determine the 2D packing
arrangement, X-ray structure factor calculations were performed
24
with the Shelx97 program adapted for 2D crystals. Using an
atomic coordinate molecular model, the structure was refined

Alkyl Esters of R-Amino Acids
J. Phys. Chem. B, Vol. 108, No. 22, 2004 7233
TABLE 2: Unit Cell Dimensions and Molecular Tilt Angles, as Deduced from the GIXD Measurements at the Air/Water
Interface
Coherence Length, Lhk (Å)h
materiala
M
(Å)b
a
(Å)c
b
o
(Å)c
γ
(°)c
a
(Å)d
b
(Å)d
9.8
γ
(°)d
A
(Å2)e t (°)f
ψ
a
(°)g
L
L
L
1h 1
a
o
o
r
r
r
m
01
10
L-Ala-C18
DL-Ala-C18
24
24
5.0
5.4
115.5
5.0
5.0
91.7
90
24
26
35
40
91
90
140
100
170
140
180
i
i
i
i
10.5
D-Ala-C26
DL-Ala-C26
35
35
5.0
5.4
115.4
5.0
5.0
9.8
9.6
92.0
90
24
24
40
39
95
90
100
140
260
240
110
D-Ala-C18-AE
DL-Ala-C18-AE
29
29
4.9
4.9
5.4
5.4
115.8
115.0
4.9
4.9
9.8
9.9
90.9
91.8
24
24
37
37
93
80
250
130
500
340
320
300
L-Ala-C18-ATE
DL-Ala-C18-ATE
29
29
4.9
4.9
5.4
5.4
113.8
113.8
4.9
4.9
9.9
9.9
93.2
93.2
24
24
38
38
79
79
160
170
280
280
320
280
L-Ala-C18-Ph
DL-Ala-C18-Ph
35
35
5.1
5.1
5.3
5.3
113.9
112.6
5.1
5.1
9.8
9.8
94.7
96.3
25
25
36
37
85
88
120
110
250
120
230
90
D-Ser-C18-ATE
DL-Ser-C18-ATE
35
35
4.9
5.4
114.2
4.9
4.9
9.9
9.1
92.9
90
24
22
39
33
79
90
250
400
450
500
310
i
i
j
j
j
j
j
j
j
j
j
i
i
i
i
i
i
L-Glu-(C18
)
2
55
55
45
5.0
5.0
5.0
8.6
8.6
8.3
90
90
90
43
43
42
25
25
18
0
0
90
340
340
80
80
70
110
DL-Glu-(C18
)
)
2
k
DL-Glu-(C18
2
C
C
C
18-AE
29
29
35
35
4.9
4.9
5.1
5.0
9.0
9.0
9.0
7.4
90
90
90
90
22
22
23
18
32
32
32
7
90
90
90
90
470i
410i
510i
420i
560i
510i
560i
360i
d
18-ATE
18-PheOH
hexacosanol
a
Chirality is as indicated. b Mean area per molecule at which the diffraction was measured. Oblique unit cell parameters. Rectangular or
c
e
pseudo-rectangular unit cell parameters calculated from the oblique cell with a
diffraction data. Molecular tilt angle from the normal to the water plane. Azimuthal angle between the projection of the long molecular axes on
the water plane and the a-axis of the unit cell (as a
the full width at half maximum, fwhm(h,k). L02 and L11+ 1h 1, respectively. Rectangular subcell containing one molecule. Subphase is 1 mM CuCl
r
) -a
o
and b
r
) a
o
+ 2b
o
.
Area per molecule calculated from the
f
g
h
r
) -a
o
, this is the same angle for the rectangular and oblique unit cells). As deduced from
i
j
k
2
.
Figure 3. (a) Bragg rod intensity profiles ((×) measured and (s) calculated) and (b) the refined 2D packing arrangements (viewed perpendicular
to the water surface) of DL-Ala-C26; (c) Bragg rod intensity profiles ((×) measured and (s) calculated) and (d) the refined 2D packing arrangements
(viewed perpendicular to the water surface) of D-Ala-C26. Note that, for clarity, only portions of the hydrocarbon chains are shown.
2
Å, br ) 9.9 Å, γr ) 91.8°, Am ) 24 Å ) that contained two
molecules. The calculated molecular tilt angle is 37° from the
normal to the water surface in an azimuth angle of 12° from
the b-axis. However, this unit cell cannot accommodate two
molecules that are related by a pseudo-glide, because the tilt
direction is not along a crystallographic axis. The tilt direction

7234 J. Phys. Chem. B, Vol. 108, No. 22, 2004
Eliash et al.
2
Figure 4. GIXD patterns I(qxy,q
z
) measured at a mean molecular area of 29 Å from the 2D crystallites of (a) DL-Ala-C18-AE, (b) D-Ala-C18-AE,
(c) C18-AE, (d) C18-ATE, (e) DL-Ala-C18-ATE, and (f) L-Ala-C18-ATE.
indicates that Ala-C18-AE molecules can be related by translation
symmetry in the corresponding chiral oblique subcell, of
dimensions ao ) 4.9 Å, bo ) 5.4 Å, γo ) 115.0°. The proposed
molecules, tilted by 37° from the normal to the water surface,
in a direction almost parallel to the b-axis. This tilt direction
indicates that hydrocarbon chains can pack in a pseudo-
herringbone packing motif via a pseudo-glide plane whereas
the Ala-headgroups are related by pseudo-translation. The 2D
molecular packing arrangement, determined by X-ray-structure-
factor-constrained least-squares refinement, is shown in Figure
5c, together with the calculated and measured Bragg rods
profiles (Figure 5d).
2
D packing arrangement is presented in Figure 5b together with
the calculated and measured Bragg rods intensity profiles (Figure
a). In such a molecular arrangement, we can envisage enan-
5
tiomeric disorder within the chiral domains via an interchange
between the CH3 group and the H atom at the stereogenic carbon
to yield a random solid solution of the two enantiomers.
Polymerization experiments of DL-Ala-C18-AE at 20 °C for 2.5
h, initiated by an aqueous solution of CuCl2 (1 mM), yielded
only a mixture of diastereoisomeric dipeptides, as determined
by MALDI-TOF MS analysis of samples prepared using an
L-enantiomer with deuterated L-Ala(d4) (see Experimental
Section). The experimental relative abundance of each of the
dipeptides is given in Figure 6a, in comparison to the theoretical
values expected for a random polymerization process. The
composition of the three diastereoisomeric dipeptides follows
a binomial distribution, as would be expected for self-assembly
of the DL-Ala-C18-AE into a random 2D solid solution. We also
note that the GIXD pattern of D-Ala-C18-AE (Figure 4b) is
different.
3.3. Ala-C -ATE. The polymerization of the alanine long-
18
chain esters described previously yielded only dipeptides. We
expected that longer oligopeptides should be obtained using the
more-reactive corresponding thio-esters. For this purpose, Ala-
C -ATE (Scheme 2, as trifluoroacetate salt) was investigated.
1
8
The GIXD patterns measured from the 2D crystallites self-
assembled from DL- and L-Ala-C -ATE at a mean area of 29
1
8
2
Å /molecule (Figure 4e, f) are very different from the GIXD
pattern (Figure 4d) measured from the corresponding C -ATE
18
thio-alcohol precursor is isostructural to C AE. The diffraction
18
patterns in Figure 4e and f are very similar, indicating that the
Ala-C -ATE racemate might have spontaneously separated into
18
a mixture of enantiomorphous 2D domains. The diffraction
The GIXD diffraction pattern28 of D-Ala-C18-AE (Figure 4b)
patterns were both indexed according to an oblique cell (a )
2
was indexed according to a pseudo-rectangular unit cell (a )
4.9 Å, b ) 5.4 Å, γ ) 113.8°, Am ) 24 Å ). The molecules are
2
4
.9 Å, b ) 9.8 Å, γ ) 90.9°, Am ) 24 Å ) that contained two
tilted by 38° from the normal to the water surface, with an

Alkyl Esters of R-Amino Acids
J. Phys. Chem. B, Vol. 108, No. 22, 2004 7235
Figure 5. (a) Bragg rod intensity profiles of DL-Ala-C18AE ((×) measured and (s) calculated); (b) refined 2D packing arrangements of DL-Ala-
18AE, viewed perpendicular to the water surface; (c) refined 2D packing arrangements of D-Ala-C18AE, viewed perpendicular to the water surface;
and (d) Bragg rod intensity profiles of D-Ala-C18AE ((×) measured and (s) calculated).
C
azimuthal direction of 79° from the a-axis. X-ray-structure-
factor-constrained least-squares refinement of the atomic coor-
dinate model yielded the 2D packing arrangement given in
Figure 7b, together with the Bragg rod intensity profiles (Figure
signals in the MALDI-TOF MS analysis, whereas the intensi-
ties of the trimer, unreacted monomer, and tetramer were in a
descending order (Figure 6c). The experimental results are very
close to the binomial distribution in a random process, indicating
the occurrence of enantiomeric disorder in the 2D crystallites
self-assembled from the racemate. Such disorder can occur via
an interchange of the methyl group with the H atom attached
to the asymmetric C atom. This type of disorder cannot be
determined on the basis of our GIXD data. Another possibility
may be that the polymerization reaction occurred primarily at
the grain boundaries or in the amorphous portion of the
monolayer and not within the 2D crystallites.
7
a).
MALDI-TOF MS of samples collected from the aqueous
surface after 3 h on a 1 mM CuCl2 subphase at 4 °C revealed
the formation of a mixture of diastereoisomeric dipeptides,
tripeptides, and tetrapeptides. The experimental relative abun-
dance of each of the oligopeptides is given in Figure 6b and
compared with the theoretical binomial distribution expected
for a random process. Note that the dimer gave the strongest

7236 J. Phys. Chem. B, Vol. 108, No. 22, 2004
Eliash et al.
Figure 6. Relative abundance of oligopeptides obtained at the air/water interface, as determined by MALDI-TOF MS using deuterium-labeled
2
2 2
L-monomer: (a) Ala-C18-AE (35 Å /molecule, 10 mM CuCl , 4 °C, 3 h) and (b) Ala-C18-ATE (1 mM CuCl , 4 °C, 3 h), average of four independent
samples. (c) MALDI-TOF MS peak intensity of the different oligopeptides and unreacted monomer as formed from the Ala-C18-ATE normalized
to that of the intensity of the dipeptide.
Figure 7. (a) Bragg rod intensity profiles of D-Ala-C18-ATE 2D crystallites self-assembled from either D-or DL-amphiphiles ((×) measured and (s)
calculated). Also shown are the refined packing arrangements of D-Ala-C18-ATE 2D crystallites self-assembled from either D-or DL-amphiphiles,
viewed (b) perpendicular and (c) parallel to the water surface.
3
.4. Ala-C18-Ph. Both Ala-C18-AE and Ala-C18-ATE were
unsuitable for our purposes, because, apparently, both formed
D crystallites comprising enantiomeric disorder. Nevertheless,
monomers, we used another reactive system exhibiting a phenyl
ester as a good leaving group: Ala-C18-Ph (Scheme 2). The
GIXD patterns measured from the 2D crystallites self-assembled
from DL-and L-Ala-C18-Ph at a mean area of 35 Å /molecule
are shown in Figure 8a and b, respectively. For comparison,
2
2
the introduction of a thio-ester moiety in Ala-C18-ATE increased
the length of the longest oligopeptide products obtained from
two to four repeating units. To enhance the reactivity of the
the GIXD pattern measured from the corresponding C18-PhOH

Alkyl Esters of R-Amino Acids
J. Phys. Chem. B, Vol. 108, No. 22, 2004 7237
2
Figure 8. GIXD patterns I(qxy,q
z
) measured at a mean molecular area of 35 Å from the 2D crystallites self-assembled from (a) DL-Ala-C18Ph, (b)
L-Ala-C18Ph, and (c) C18-PhOH.
2
(
Scheme 2) alcohol precursor is very different (Figure 8c),
) 4.9 Å, b ) 9.1 Å, Am ) 22 Å ; see Table 2) containing two
molecules related by b-glide symmetry in a racemic 2D
crystallite.
demonstrating again that the packing arrangement of the alanine
long-chain esters is dependent on the chiral alanine moiety. The
diffraction pattern in Figure 8a was indexed according to a chiral
oblique unit cell a ) 5.1 Å, b ) 5.3 Å, γ ) 112.6°, Am ) 25
This monomer proved to be reactive and the best polymer-
ization results were obtained by spreading the monomer on pure
water at 4 °C when dimers and trimers were formed after 2.5
h.
2
Å . The molecules are tilted by 37° from the normal to the water
surface in an azimuthal direction of 88° from the a-axis. The
GIXD data cannot give information on the degree of order of
the alanine moieties in the crystallites. The unit cell of the
L-crystallites (Figure 8b) are of similar dimensions to the
racemate: a ) 5.1 Å, b ) 5.3 Å, γ ) 113.9°, Am ) 25 Å . The
2
The addition of a catalyst (AgNO or CuCl ) did not give
3
2
any longer oligopeptides. However, there was a strong peak in
the MALDI-TOF MS that corresponds to the disulfide bridge
formation between two C18-ATE thio-alcohol molecules after
possible monomer hydrolysis.
2
D packing arrangement of the enantiomer L-crystallites is given
in Figure 9, together with the measured and calculated Bragg
rod intensity profiles. Polymerization performed on samples
prepared from DL- and L-Ala-C18-Ph yielded only dimers, as
determined by MALDI-TOF MS analysis.
3.7. Glu-(C ) . The diester L-Glu-(C ) , which represents a
1
8 2
18 2
different class of esters bearing two hydrocarbon chains, was
1
1
reported to yield long oligopeptides; therefore, we reinvesti-
gated the polymerization of films of this monomer. This
compound was prepared both as a methylsulfonate salt and as
a free amine. The GIXD patterns measured from the 2D
crystallites self-assembled from the DL- and L-Glu-(C18)2 methyl-
3
.5. Phe-C18 and Tyr-C18. The monomers derived from the
alanine esters or thio-esters formed 2D crystallites comprising
enantiomeric disorder. Such disorder would not be expected in
systems where the R-amino acid side chain is bulky, e.g., Phe-
C18 and Tyr-C18 (Scheme 2). Our polymerization experiments
with these compounds, performed under the same conditions
2
sulfonate salt at a mean molecular area of 55 Å are very similar
(
Figure 11a and b, respectively). The free amine gave a similar
GIXD pattern (not shown). They yielded a rectangular subcell
a ) 5.0 Å, b ) 8.6 Å) that contained one molecule tilted along
(
1
1,12
as previously reported
to yield oligopeptides in form of
the a-axis by 25° from the normal to the water surface. An
atomic coordinate molecular model containing a pseudo-
herringbone arrangement of the hydrocarbon chains was con-
structed. The 2D packing arrangement obtained by X-ray-
structure-factor-constrained least-squares refinement of the
model is shown in Figure 12b and c, together with the
corresponding measured and calculated Bragg rod intensity
profiles (Figure 12a).
R-helix and/or random coil,1
1,12
led to the formation of only
dimers and unreacted monomer, as determined by MALDI-
TOF MS analysis.
The GIXD measurements performed on monolayers of Phe-
C18 and Tyr-C18 did not produce diffraction at any molecular
area. The lack of crystalline order in such monolayers is
presumably due to the steric hindrance introduced by the bulky
Phe and Tyr headgroups.
We also measured the GIXD patterns from DL-Glu-(C18)2
3
.6. Ser-C18-ATE. Another system where we would expect
spread on an aqueous solution of CuCl (1 mM). No diffraction
2
2
the enantiomeric disorder to be reduced is the long-chain ester
of serine Ser-C18-ATE (see Scheme 2). The GIXD patterns
measured from 2D crystallites self-assembled from DL- and
L-Ser-C18-ATE are given in Figure 10a and b, respectively. The
L-crystallites display a diffraction pattern very similar to that
of the Ala-C18-ATE (see Figure 4e, f) with molecules packed
only by translation. By contrast, the GIXD pattern measured
from the racemic amphiphile yielded a rectangular unit cell (a
was observed at a mean molecular area of 55 Å , unlike that
observed for the water subphase. The film was then compressed
2
to a mean molecular area of 45 Å and gave the GIXD pattern
shown in Figure 11c. The calculated subcell is rectangular (a
) 5.0 Å, b ) 8.3 Å) and contains alkyl chains tilted along the
b-axis by an angle of 18° from the normal to the water surface.
Polymerization experiments performed on the water surface
1
1
under conditions as previously reported yielded only dimers

7238 J. Phys. Chem. B, Vol. 108, No. 22, 2004
Eliash et al.
Figure 9. (a) Bragg rod intensity profiles of D-Ala-C18-Ph ((×) measured and (s) calculated). Also shown is the refined packing arrangement of
the D-Ala-C18-Ph viewed (b) perpendicular and (c) parallel to the water surface.
2
Figure 10. GIXD patterns I(qxy,q
z
) measured at a mean molecular area of 35 Å from the 2D crystallites self-assembled at the air/water interface
from (a) DL-Ser-C18-ATE and (b) L-Ser-C18-ATE. Note that the GIXD patterns of the racemate and the corresponding alcohol (C18-ATE, Figure 4d)
are very similar but not identical.
(
Scheme 3), as demonstrated by MALDI-TOF MS analysis.
The degree of order on the water surface under these conditions
30 °C) is expected to be low. Therefore, we also performed
provide suitable architectures for the generation of peptides by
a condensation reactions within a structured environment.8
Here, we report a reinvestigation of some of these reactions by
applying grazing incidence X-ray diffractometry (GIXD) and
matrix-assisted laser deesorption ionization time-of-flight mass
spectroscopy (MALDI-TOF MS) techniques. The packing
arrangements of the two-dimensional (2D) crystallites formed
by self-assembly at the air/solution interface; both of the
amphiphiles reported previously and the new ones were
determined by GIXD. The molecular weights were determined
-11
(
the polymerization on an aqueous subphase at 4 °C, where,
again, even in the presence of 1 mM CuCl2 as a catalyst, only
dimers were formed.
4
. Conclusions
Previous reports have suggested that amphiphilic alkyl esters
of R-amino acids self-assemble at the air/solution interface and

Alkyl Esters of R-Amino Acids
J. Phys. Chem. B, Vol. 108, No. 22, 2004 7239
Figure 11. GIXD patterns I(qxy,q
z
) measured from (a) DL-Glu-(C18
)
2
and (b) L-Glu-(C18
)
2
, as methyl sulfonate at a mean molecular area of 55 Å2,
2
2
and from (c) DL-Glu-(C18 free amine at a mean molecular area of 45 Å on 1 mM CuCl .
)
2
SCHEME 3: Dimerization of Glu-(C18)2, in Contrast to
1
1
a Previously Proposed Polycondensation
labeled racemic monomers is binomial, suggest that the racemate
self-assembles in the form of enantiomerically disordered 2D
crystallites. It is most plausible that the origin of this disorder
is due to the possibility of a methyl group of an Ala moiety of
single-handedness to occupy a site of a H atom attached to the
stereogenic center of the other enantiomer. It was anticipated
that such disorder would not be present in the alkyl esters of
R-amino acids such as Phe, Tyr, and Ser, where the side chains
are bulkier than those in Ala. However, the esters of racemic
and enantiomerically pure Phe and Tyr did not form 2D
crystallites at the air/water interface. The Ser esters formed true
racemic 2D crystallites; however, they did not react beyond the
formation of the corresponding dipeptides and tripeptides.
To rationalize the formation of low-molecular-weight pep-
tides, several possible reasons may be considered. First, these
monomers can undergo hydrolysis in addition to polyconden-
sation. The formation of the dipeptides and the alcohol byprod-
uct may induce defects in the crystallites during the early stages
of the reaction and change the position of the ester group of
the dipeptidesvis a vis, the neighboring amine groupsand, thus,
hinder further propagation. The product distribution seems to
be independent of the coherent length of the various 2D
crystallites (Table 2).
Figure 12. (a) Bragg rod intensity profiles of L-Glu-(C18
measured and (s) calculated). Also shown are the refined 2D packing
arrangements of L-Glu-(C18 , viewed (b) perpendicular and (c) parallel
to the water surface. Note that, for clarity, only portions of the
hydrocarbon chains are shown.
)
2
((×)
)
2
from the MALDI-TOF MS analysis. In all the studied systems,
it was found that the polycondensation reaction of the alkyl
esters does not proceed beyond the formation of dipeptides.
To generate longer oligopeptides, we investigated also a few
alkyl thio-esters, where the alkanethiol is a much better leaving
group than the corresponding alkoxy group. In these systems,
trimers and tetramers were obtained, in addition to dimers.
To probe the feasibility of generating homochiral oligopep-
tides by taking advantage of differences in the packing arrange-
ment of racemic versus enantiomorphous 2D crystallites, we
have investigated the packing arrangement of several racemic
and enantiopure compounds. In the case of Ala-C18-AE, the
packing arrangements of L- and DL-crystallites were similar and,
in the case of Ala-C18-ATE, almost identical. These results,
together with the observation that the distribution of the short
oligopeptides formed in the polycondensation of the enantio-

7240 J. Phys. Chem. B, Vol. 108, No. 22, 2004
Eliash et al.
Finally, the present results demonstrate that, in contrast to
(6) Weissbuch, I.; Rubinstein, I.; Weygand, M. J.; Kjaer, K.; Leiser-
R
owitz, L.; Lahav, M. HelV. Chim. Acta 2003, 86, 3867.
the polycondensation process of amphiphilic thioesters and N -
(
7) Rubinstein, I.; Bolbach, G.; Weygand, M. J.; Kjaer, K.; Weissbuch,
I.; Lahav, M. HelV. Chim. Acta 2003, 86, 3851.
8) Baniel, A.; Frankel, M.; Friedrich, I.; Katchalsky, A. J. Org. Chem.
1948, 13, 791.
ꢀ
carboxyanhydride of γ-stearyl-glutamic acid, and N -stearoyl-
lysine,4 systems where, on reaction, longer oligopeptides were
formed, the alkyl esters and thio-esters of the amino acids
yielded dipeptides and up to tetrapeptides, respectively. There-
fore, in contrast to previous reports, they cannot be regarded as
useful model systems for the formation of prebiotic peptides,
unless more-efficient catalysts for the polycondensation of these
esters are discovered.
,5
(
(9) Fukuda, K.; Shibasaki, Y.; Nakahara, H. J. Macromol. Sci., Chem.
1
981, A15, 999.
(
10) Hanabusa, K.; Oumi, T.; Koyama, T.; Shirai, H.; Hayakawa, T.;
Kurose, A. J. Macromol. Sci., Chem. 1989, A26, 1397.
(11) Fukuda, K.; Shibasaki, Y.; Nakahara, H.; Liu, M. H. AdV. Colloid
Interface Sci. 2000, 87, 113.
(
12) Liu, M. H.; Nakahara, H.; Shibasaki, Y.; Fukuda, K. Chem. Lett.
1
1
7
2
993, 967.
Acknowledgment. We thank Dr. G e´ rard Bolbach (Labora-
toire de Chimie Structurale Organique et Biologique, Universite´
Pierre et Marie Curie, 75252 Paris Cedex 05, France) for
assistance with the MALDI-TOF MS technique development
and analysis. We thank the Israel-Science Foundation (ISF) and
The Danish Foundation for Natural Sciences for financial
support. This work was supported by the IHP (Contract No.
HPRI-CT-1999-00040/2001-00140) of the European Com-
munity. This work is part of a COST D-27 program. We are
grateful to HASYLAB (DESY, Hamburg, Germany) for beam
time at the BW1 beamline.
(13) Liu, M. H.; Nakahara, H.; Shibasaki, Y.; Fukuda, K. Chem. Lett.
994, 783.
(14) Blout, E. R.; Doty, P.; Yang, J. T. J. Am. Chem. Soc. 1957, 79,
49.
(
25.
15) Blair, N. E.; Bonner, W. A. Origin Life EVol. Biochem. 1980, 10,
(
(
16) Penney, C. L.; Shah, P.; Landi, S. J. Org. Chem. 1985, 50, 1457.
17) Lapidot, Y.; Rappoport, S.; Wolman, Y. J. Lipid Res. 1967, 8, 142.
(18) Itoh, M.; Hagiwara, D.; Kamiya, T. Bull. Chem. Soc. Jpn. 1977,
0, 718.
5
(
19) Neises, B.; Steglich, W. Angew. Chem., Int. Ed. Engl. 1978, 17,
22.
20) Kuzmenko, I.; Rapaport, H.; Kjaer, K.; Als-Nielsen, J.; Weissbuch,
5
(
I.; Lahav, M.; Leiserowitz, L. Chem. ReV. 2001, 101, 1659.
(21) Als-Nielsen, J.; Kjaer, K. In NATO AdVanced Study Institute on
Phase Transitions in Soft Condensed Matter (Geilo, Norway, 1989); Riste,
T., Sherrington, D., Eds.; NATO ASI Series, Series B, Physics, Vol. 211;
Plenum Press: New York, 1989; p 113.
Supporting Information Available: Atomic coordinates of
the molecules in the 2D crystallites (PDF). This material is
available free of charge via the Internet at http://pubs.acs.org.
(
(
(
22) Kjaer, K. Physica B 1994, 198, 100.
23) CERIUS2; Accelrys, Inc: San Diego, CA, 1995.
24) Sheldrick, G. M. “SHELXL97” Program for the Refinement of
References and Notes
(1) Frank, P.; Bonner, W. A.; Zare, R. N. On One Hand But Not the
Crystal Structures; University of G o¨ ttingen, Germany, 1997.
25) Brunelle, A.; Chaurand, P.; DellaNegra, S.; LeBeyec, Y.; Parilis,
E. Rapid. Commun. Mass Spectrom. 1997, 11, 353.
26) Majewski, J.; Popovitz-Biro, R.; Bouwman, W. G.; Kjaer, K.; Als-
Other: The Challenge of the Origin and Survival of Homochirality in
Prebiotic Chemistry. In Chemistry for the 21st Century; Keinan, E.,
Schechter, I., Eds.; Wiley-VCH: Weinheim, 2001; p 175.
(
(
(
2) Hitz, T.; Luisi, P. L. HelV. Chim. Acta 2003, 86, 1423.
Nielsen, J.; Lahav, M.; Leiserowitz, L. Chem.-Eur. J. 1995, 1, 304.
(27) Weissbuch, I.; Berfeld, M.; Bouwman, W.; Kjaer, K.; Als-Nielsen,
J.; Lahav, M.; Leiserowitz, L. J. Am. Chem. Soc. 1997, 119, 933.
(28) The GIXD spectra of the D-, L-, and deuterated L-Ala-C18-AE were
measured and yielded identical results. DL-Ala-C18-AE was measured both
from a 1:1 mixture of the enantiomers and from the racemate synthesized
from racemic alanine and yielded identical patterns.
(3) Weissbuch, I.; Zepik, H.; Bolbach, G.; Shavit, E.; Tang, M.; Jensen,
T. R.; Kjaer, K.; Leiserowitz, L.; Lahav, M. Chem.-Eur. J. 2003, 9, 1782.
4) Zepik, H.; Shavit, E.; Tang, M.; Jensen, T. R.; Kjaer, K.; Bolbach,
G.; Leiserowitz, L.; Weissbuch, I.; Lahav, M. Science 2002, 295, 1266.
5) Weissbuch, I.; Bolbach, G.; Zepik, H.; Shavit, E.; Tang, M.; Frey,
J.; Jensen, T. R.; Kjaer, K.; Leiserowitz, L.; Lahav, M. J. Am. Chem. Soc.
002, 124, 9093.
(
(
2