Self-assembly of bolaamphiphiles forming alternating layer arrangements with lead and copper divalent ions

Article abstract of DOI:10.1021/jp0116508

Five bifunctional ?±- amino acid-a??-carboxy bolaamphiphiles [(2-R,S, or RS)-(HOOC)-(CH₂)_m-CONH-(CH₂)n-CH(COOH)(NH₂ ) m = 20,22, n = 3,4 labeled (l), (d,l)-C₂₂Orn (l), (d), (d,l)-C₂₂Lys, (d,l)-C₂₄Lys respectively] were synthesized. These molecules were deposited on different aqueous subphases, and studied by means of grazing incidence X-ray diffraction (GIXD). On deionized water, the bolaamphiphiles (d,l)-C₂₂Lys yield a mixture of crystallites: both a monolayer, in which the chains are tilted from the normal, and a multilayer in which the molecules lie parallel to the water. On the other hand, when deposited on mono- or bi-metal ionic subphases, they self-assemble into crystalline multilayer films in which the molecules lie parallel to the aqueous solution surface, linked head-to-head and tail-to-tail in the form of extended chains. The latter are juxtaposed such that the metal ions form sheets separated by the organic molecules, aligned perpendicular to the plane of the aqueous solution. Deposition of either the enantiomerically pure or racemic bifunctional bolaamphiphiles on an aqueous solution of mixed Cu(Acetate)₂ and Pb(Acetate)₂, yields self-assembled crystalline films composed of the two different metal cations, arranged in alternating sheets, separated by the organic spacer. Both GIXD and X-ray-photoelectric-spectroscopy (XPS) studies (after deposition on solid support) demonstrate that the structures of these films differ form the ones formed on either pure Pb(Ac)₂ or on Cu(Ac)₂ solutions, thus excluding the simultaneous formation of the two monometallic crystalline phases.

Full text of DOI:10.1021/jp0116508

J. Phys. Chem. B 2001, 105, 11447-11455
11447
Self-Assembly of Bolaamphiphiles Forming Alternating Layer Arrangements with Lead and
Copper Divalent Ions
†
†
‡
‡
,†
,†
R. Buller, Hagai Cohen, T. R. Jensen, K. Kjaer, M. Lahav,* and L. Leiserowitz*
Materials and Interfaces Department, Weizmann Institute of Science, RehoVot 76100, Israel and Condensed
Matter Physics and Chemistry Department, Risø National Laboratory, DK 4000 Roskilde
ReceiVed: May 1, 2001; In Final Form: September 4, 2001
Five bifunctional R-amino acid-ꢀ-carboxy bolaamphiphiles [(2-R,S, or RS)-(HOOC)-(CH
2
)
m
2 n
-CONH-(CH ) -
CH(COOH)(NH ) m)20,22, n)3,4 labeled (l), (d,l)-C22Orn (l), (d), (d,l)-C22Lys, (d,l)-C24Lys respectively]
2
were synthesized. These molecules were deposited on different aqueous subphases, and studied by means of
grazing incidence X-ray diffraction (GIXD). On deionized water, the bolaamphiphiles (d,l)-C22Lys yield a
mixture of crystallites: both a monolayer, in which the chains are tilted from the normal, and a multilayer in
which the molecules lie parallel to the water. On the other hand, when deposited on mono- or bi-metal ionic
subphases, they self-assemble into crystalline multilayer films in which the molecules lie parallel to the aqueous
solution surface, linked head-to-head and tail-to-tail in the form of extended chains. The latter are juxtaposed
such that the metal ions form sheets separated by the organic molecules, aligned perpendicular to the plane
of the aqueous solution. Deposition of either the enantiomerically pure or racemic bifunctional bolaamphiphiles
2 2
on an aqueous solution of mixed Cu(Acetate) and Pb(Acetate) , yields self-assembled crystalline films
composed of the two different metal cations, arranged in alternating sheets, separated by the organic spacer.
Both GIXD and X-ray-photoelectric-spectroscopy (XPS) studies (after deposition on solid support) demonstrate
2 2
that the structures of these films differ form the ones formed on either pure Pb(Ac) or on Cu(Ac) solutions,
thus excluding the simultaneous formation of the two monometallic crystalline phases.
Introduction
film is composed of two different metal cations, arranged in
alternating sheets, separated by an organic spacer, as shown in
Scheme 1a,b. These metal sheets will serve as precursors for
the eventual formation of the nanocrystallites.
The construction of functional materials via a step by step
process from the molecular level, utilizing supramolecular
architectures, is attracting increasing interest. Among these
1
As the organic matrix, we used a nonsymmetric bolaamphi-
phile, composed of two different functional headgroups where
each selectively binds to a different ion, extracted from a mixture
of two ions in the subphase. As a model system, we made use
of the “bolaamphiphilic series” (Scheme 2 Vi, Vii, Viii)
including (2-R), (2-S) or (2-R,S)-2-amino-7-aza-8-oxo-nona-
cosane-1,29-dicarboxylic acid (HOOC)-C20H40-CONH-C4H8-
CH(COOH)(NH2) [labeled d, l or (d,l)-C22Lys], as well as l or
systems are ordered hybrid organic/inorganic composite materi-
als, in particular quantum size nanoparticles orderly embedded
within rodlike molecular matrixes, which display cooperative
2
,34
optoelectronic properties, and receive special attention.
Recently, we demonstrated that self-assembled organo-
metallic architectures can be generated by depositing the organic
ligand on aqueous subphases containing a metal cation of a
single type. The resulting films formed on the aqueous solution,
contain ions organized in the form of periodic sheets separated
by alkane-dicarboxylic acid molecules, which are aligned
(d,l)-C22Orn and (d,l)-C24Lys. These molecules contain car-
boxylic acid and R-amino acid headgroups at opposite ends of
3
the chain. It was anticipated that when spread on a solution
parallel to the solution surface, or in oriented silver ion grids
4
,5
2+
2+
of (2 × 2) or (3 × 3) in the form of mono- or bi-layers. The
structures of the crystalline films were determined at the sub
nanometer scale. Such architectures are appropriate starting
materials for the preparation, by topotactic reaction, of hybrid
organic/inorganic composites of nanoparticle crystallites sepa-
containing a mixture of Cu and Pb metal ions, the former
will bind exclusively to the amino acid groups, and the latter
to the carboxyl group only, forming a head-to-head and tail-
to-tail arrangement of extended chains. Although there is a
marked difference of 7 orders of magnitude in the dissociation
constants (Kd) of amino acids with each of the two metal ions
6
,7
rated by organic spacers arranged in periodic lattices.
2
+
2+ 8,9
We are currently interested in preparing hybrid systems
comprising two types of nanocrystallites of metal or semicon-
ductor character, arranged in an alternating manner, and
separated by organic spacers. Here, we describe the preparation
and characterization of an ordered three-component film,
prepared by self-assembly, at the air-solution interface. This
(pKd ≈ 15 for Cu and pKd ≈ 8 for Pb ), only a difference
of 0.3 in dissociation constant exists for the complexes of
2+
carboxylic acid with these two ions (pKd ≈ 4 for Pb and pKd
≈ 3.7 for Cu ). In light of these differences, it was logical
to deduce that Cu will bind exclusively to the amino acid
2
+ 10,11
2
+
2+
headgroup, but there was an ambiguity whether Pb would
bind selectively to the carboxylic acid.
*
To whom correspondence should be addressed.
Materials and Interfaces Department, Weizmann Institute of Science.
Condensed Matter Physics and Chemistry Department, Risø National
An indication of selective binding resulted from an analysis
of the metal content of a three-component crystal containing
Cu , Pb and (d,l)-glutamic acid. Atomic absorption experi-
†
‡
2
+
2+
Laboratory.
1
0.1021/jp0116508 CCC: $20.00 © 2001 American Chemical Society
Published on Web 10/31/2001

11448 J. Phys. Chem. B, Vol. 105, No. 46, 2001
Buller et al.
SCHEME 1: Schematic Representation of Two Possible
Packing Arrangements of Bolaamphiphile Molecules
Connecting Two Different Metal Ion Types Arranged in
Alternating Sheets: (a) View Perpendicular to the
Solution Surface, Where the Arrow in the Hydrocarbon
Chain Represents an Amide Moiety, (b) View of the
Arrangement Edge on the Air-solution Interface. (c)
Undesired structure where neither of the metal ionic
sheets is of a single type
SCHEME 2: Synthetic Route for the Preparation of the
Bolaamphiphilic Series
ments show that the crystal contains a 1:1 ratio between copper
and lead. The details are described in the Experimental Section.
The role of the amide group located adjacent to one of the
polar headgroups in the hydrocarbon chain is to promote, by
virtue of the N-H‚‚‚O hydrogen bond, translation symmetry
between molecules along the a axis (Scheme 1a), as opposed
to the undesired structural motif in Scheme 1c, and thus yield
formation of alternating sheets of Pb2+ and Cu ions.
2+
Results
a. Synthesis. The bolaamphiphiles designed (Scheme 2Vi-
iii), are produced by regioselectively condensing the ꢀ amine
of lysine or ornithine methyl esters, with R,ω-dicarboxylic acid
monomethyl ester I, similar to the use of isobutyl chloroformate
as an activating group.1
2,13
The ester protection groups are
removed by hydrolysis, yielding compound V. The overall
process is depicted in Scheme 2.
b. Isotherms. Preliminary evidence for binding of the divalent
copper and lead ions from the aqueous subphase to the
bolaamphiphile at the solution surface was obtained from the
surface pressure-molecular area (π-A) isotherms (Figure 1).
The isotherm for the film on the mixed ionic solution is very
different from that of the film on pure water. It also differs from
that of the film formed when spreading the bolaamphiphile on
pure Cu(Ac)2 and Pb(Ac)2 subphases. Furthermore, the relatively
low limiting areas per molecule of the films on the different
subphases are consistent with multilayer formation.

Self-Assembly of Bolaamphiphiles
J. Phys. Chem. B, Vol. 105, No. 46, 2001 11449
TABLE 1: Analysis of the Separated Peaks
a
b
c
d
literature
(eV)
deionized
water (eV)
Pb(Ac)
mM (eV)
2
Cu(Ac)
1 mM (eV)
2
Cu(Ac)
2
& Pb(Ac)
2
subphase
N(1s)
1 mM (eV)
BE
BE
399.7
%
BE
%
BE
%
BE
%
amide
amine
ammonium
399.6
400.0
401.3
50
399.6
400.2
401.4
53
15
32
399.6
400.2
50
50
399.4
400.0
50
50
401.5(0.1)
50
a
Relative errors in ratios are 5%, errors in peak positions are 0.03 eV unless otherwise stated. Bottom: Experimental spectra and curve fitting
to the N(1s) lines.
Figure 1. Surface pressure-area (π-A) isotherms of bolaamphiphile
d-C22Lys, on: (1) deionized water, (2) Pb(Ac)
2
, (3) Cu(Ac)
2
, and (4)
mixed Cu(Ac) and Pb(Ac) solution as subphase.
2 2
c. X-ray Photoelectron Spectroscopy. Supported multilayers
of d or (d,l)-C22Lys with or without metals were prepared by
spreading the bolaamphiphile, on each of the four room-
temperature subphases: deionized water, solution containing
Pb(Ac)2, Cu(Ac)2, and a 1:1 mixture of Pb(Ac)2 and Cu(Ac)2.
22
Figure 2. Carboxyl binding energy of d-C Lys on different subphases.
(a) pure Cu(Ac)
2
, (b) deionized water, (c) pure Pb(Ac)
2
, (d) mixed
Cu(Ac) and Pb(Ac)
2
2
. Arrows indicate the carboxyl and carboxylate
binding energies. The main aliphatic C(1s) peak is at 284.8 eV.
C22Lys spread on the different subphases (Table 1, Figure 3).
These films were then transferred onto a clean glass support by
20,21
slow draining deposition.14 The glass with the deposited film
Two signals are resolved in the N(1s) line,
the first at 399.6
eV is assigned to amide -CONH-, whereas the second appears
prepared on the metal containing subphase, was then washed
with deionized water to remove excess acetate and metal ions,
and then dried.
-
at 400.0 eV for the amine -CH(COO )(NH2) or at 401.3 eV
-
+
for the ammonium -CH(COO )(NH3 ).
The nitrogen atom serves as a clear indicator of metal binding
because its signal originates only from the bolaamphiphile. Pure
amino acids have a tendency to crystallize as zwitterionic pairs.
According to the XPS spectra, this motif is evident both for
the pure bolaamphiphiles deposited on deionized water (see peak
position at 288.5 eV in Figure 2b, and Figure 3a), and on the
solution containing pure Pb(Ac)2 (see peak position at 288.5
eV in Figure 2c, and Figure 3b). The presence of the weak amine
X-ray photoelectron spectroscopy (XPS) measurements on
the mixed film demonstrate the 1:1 ratio of the two metal types,
as well as the 1:2 ratio between each metal ion and the
bolaamphiphile d-C22Lys. The relative atomic concentrations
of d-C22Lys spread on a 1:1 solution of mixed Pb(Ac)2 and
Cu(Ac)2 and transferred onto a glass support are: Cu:Pb:N:C
1:1:4:56 based on calculation, and 1:1:4:59 as found experi-
mentally. Relative errors in ratio of 3% are found for both the
Pb(4f) peak binding energy (BE) ) 139.0(0.15) eV and the
C(1s) peak BE ) 248.8(0.05) eV; and 10% for both the Cu(2p3/
(NH2) peak in Figure 3b serves as an indication of the binding
of lead to the amino acid functional group, most probably due
8
,9
15
to the high value of pK ≈ 8 of lead amino acid complex.
2
) peak BE ) 934.7(0.20) eV and the N(1s). The concentration
d
The difference in binding energy of 1.3 eV, corresponds to the
of oxygen O(1s) could not be unambiguously determined
because of the overlap by the oxygen signal originating from
the glass support. We note that copper undergoes fast reduction
+
change in ionic character of the amino acid nitrogen (NH3
f
NH2), and is a clear indication of copper amino acid complex-
ation. This difference in binding energy was observed both for
the bolaamphiphile on Cu(Ac)2 subphase (Figure 3c), and on
under the experimental conditions.1
6-19
The C(1s) broad peak around 288.3 eV (Figure 2) originates
-
2 2
the mixed Cu(Ac) and Pb(Ac) subphase (Figure 3d). Despite
from the carboxylate anion (-COO ) line at 288.5 eV, the
the large overlap between the two peaks (in Figure 3c, and in
3d) there is no doubt that they correspond both to the amide
and amine, and thus exclude any possible presence of the
ammonium ion. We may conclude from these data that the
divalent copper and lead ions, each bind to its appropriate ligand,
as depicted in Scheme 3, thus indicating a head-to-head and
tail-to-tail arrangement (Scheme 1a,b).
amide group’s ketone (-CONH-) at 288.0 eV and the
20
carboxylic acid (-COOH) at 289.0 eV when present. A lower
intensity at ∼289.4 eV is observed in curves (c) and (d) (Figure
2
), indicating an induction in ionic character of the carboxylate
-
group, -COOH f -COO . This change occurs only in the
presence of Pb2+ ions, which indicates binding thereof to the
carboxylate group. It should be noted that residual carboxylate
from the unwashed acetate may be present in these samples.
Yet, its quantity is estimated to be less than 15% of the
carboxylate signal.
d. Grazing-Incidence X-ray Diffraction. The anticipated
structure of the metal complex (Scheme 1a,b), was verified via
an analysis of the grazing incidence X-ray diffraction (GIXD)
measurements performed on the various films, in situ, at the
Evidence for the preferential binding of Cu2 to the R-amino
+
2
2,23
acid ligands arises from the XPS signals of films of d or (d,l)-
air-solution interface. The GIXD experiments
were per-

11450 J. Phys. Chem. B, Vol. 105, No. 46, 2001
Buller et al.
Figure 4. Surface plot of the scattered intensity I(qxy, q
0,k} reflections of (d,l)-C22Lys spread on different subphases: (a)
mixed Cu(Ac) and Pb(Ac) , (b) pure Pb(Ac) , and (c) pure
Cu(Acetate)
z
) showing the
{
2
2
2
2
.
first the diffraction spectra obtained from the bimetallic line
array, and then the data obtained from the non metallic species.
d.1. Three-Component System. Deposition of the bolaamphi-
philes on pure Pb(Ac)2 and on Cu(Ac)2 solutions, yields
crystalline films with d-spacings of 76 Å and 71 Å respectively24
(see figures 4, 5). Making use of the mixed Cu-Pb subphase
resulted in the self-assembly of a different crystalline film, that
incorporates both metallic ions. This difference is manifest from
a comparison of the GIXD patterns of d or (d, l)-C22Lys on the
mixed solution of Cu(Ac)2 and Pb(Ac)2, with those of the
bolaamphiphile on the pure Cu(Ac)2 and Pb(Ac)2 subphases.
The GIXD pattern of the mixed Cu-Pb system (Figure 5a,b)
cannot be formed by a superposition of the Bragg peaks of the
Cu and the Pb systems, since they lie at different qxy positions.
This analysis thus excludes the simultaneous formation of two
monometallic crystalline phases, in the mixed Cu-Pb case.
Figure 3. XPS N1s energy measurements of bolaamphiphile d-C22
Lys spread on different subphases: (a) Deionized water, (b) Pb(Ac)
-
2
solution, (c) Cu(Ac)
2 2
solution, and (d) mixed Cu(Ac) and Pb(Ac)
2
solution as subphase.
SCHEME 3: Schematic Representation of the Repeating
Unit of the Bolaamphiphiles on Different Subphases
The X-ray scattered intensity, I(qxy,qz), of the bolaamphiphile
d-C22Lys on Cu(Ac)2 and Pb(Ac)2 mixed subphase, is presented
in Figure 6b in the form of a 2D surface contour plot.
The presence of Cu2 and Pb2+ ions with very different X-ray
scattering amplitudes, arranged in successive sheets in the
crystalline film, facilitates the assignment of the chain axis in
the series. The X-ray diffraction peaks observed at low Bragg
angles (Figure 6b) could all be assigned Miller indices {0,k},
yielding a d(01) lattice spacing between 70 and 80 Å, depending
upon the length of the molecules in the series, indicating that
the chain axes lie parallel to the solution surface. The strong
dependence of d(01), on the molecular length, is consistent with
a structure whose chain axes lies parallel to the solution surface
+
(see Table 2). This follows from the differences in the observed
(01) lattice spacing (Åo), which correspond nicely to the
differences between the calculated lengths (Åc) of the molecules
in the series [dl-C22lys-dl-C22Orn) 3.0 Åo, 2.4 Åc, ddl-C22lys-
ddl-C22Orn ) 3.1 Åo, 2.4 Åc, ddl-C24lys-ddl-C22Lys ) 5.2 Åo, 4.8
Åc]. The d(01) spacing corresponds to the length of two
formed on the liquid surface diffractometer, at the synchrotron
beamline BW1, HASYLAB. A brief description of the method
is given in the Experimental Section. Here, we shall describe

Self-Assembly of Bolaamphiphiles
J. Phys. Chem. B, Vol. 105, No. 46, 2001 11451
Figure 5. (a) The positions (along qxy) and intensities of the Bragg
peaks, integrated along qxy and q
spread on different subphase solutions:mixed Cu(Ac)
Cu(Ac) , Pb(Ac) (b) Corresponding results for (d,l)-C22Lys amphiphile
z
measured for d-C22Lys amphiphile
2
and Pb(Ac)
2
,
2
2
spread on the three different subphases.
bolaamphiphilic molecules whose chain axis lies parallel to the
solution surface. These molecules are linked head-to-head and
tail-to-tail, via two different metal ionic moieties with the
appropriate metals, as shown in Scheme 1a,b.
Figure 6. d-C22Lys on a mixed Cu(Ac)
a) Observed (dots) and calculated (line) Bragg rods I(q
composed of six layers. (b) Low angle 2D surface plot of the scattered
intensity I(qxy, q ) showing the {0,k} reflections.
2
Pb(Ac)
2
subphase solution.
(
z
) for a model
z
There are no significant differences in the GIXD patterns of
the enantiomerically pure and racemic molecules, probably
because both chiral and racemic amino acids tend to form planar
copper complexes with similar geometry. Both bolaamphiphiles
TABLE 2: Long Axial Spacing of the Bolaamphiphilic
Series on Mixed Cu(Ac) and Pb(Ac) Subphase
2
2
measured d
calculated axial
length (Å)
deduced γ
angle (°)
compound
spacing a (Å)
(chiral and racemic) form crystallites with similar d(01) spac-
ings, however, the peaks differ in their relative intensities, which
suggest somewhat different structures. This result also precludes
the possibility of spontaneous resolution of the racemate. The
difference in the structure is manifested in the symmetry element
linking two copper-bound amino acid groups, centered about
the metal ion, which is a 2-fold axis for the enantiomer and an
inversion center for the racemate.
l-C22Orn
(d, l)-C22Orn
l-C22Lys
70.0
70.2
73.0
73.4
73.3
78.5
73.6
73.6
76.0
76.0
76.0
80.8
72
73
74
74
75
76
d-C22Lys
(
(
d, l)-C22Lys
d, l)-C24Lys
a
These values lie within an error range of 0.5 Å, which is acceptable
regarding the broad Bragg peaks and experimental error.
The experimental shape of the Bragg rods, that peak at qz )
-
1
0
Å , implies an arrangement where the bound metal ions are
surface. The former possibility is more probable, in view of
the deduction from the Bragg rods, that the metal sheets are
aligned perpendicular to the aqueous surface (vide supra).
Furthermore, the calculated angle of γ ≈ 74° between the
molecular chain axis and the metal sheets precludes a 2-fold
axis parallel to the solution surface. Both structures were subject
to modeling using Bragg rod profiles for a unit cell of plane
symmetry p2.
aligned in sheets perpendicular to the solution surface. The full
width at half-maxima (fwhm) of these Bragg rods along qz,
provide an estimate of the film thickness Lz ≈ 40 ( 10 Å,
according to the formula Lz ≈ 0.88*2π/[fwhm] (see Experi-
mental Section), which would correspond to about seven layers.
It is possible to estimate the angle γ formed between the
molecular chain axis and the plane of the metal sheet by
comparing the observed d spacing (between the metal sheets)
with the estimated length L of the molecular chain axis. The
angle γ is given by sinγ ) d/L, according to which γ is, on
average, 74° (see Table 2).
A more quantitative analysis of the GIXD data was carried
out for homochiral d-C22Lys on mixed Cu-Pb subphase. The
symmetry element linking two chiral molecules via the metal
ions can only be a 2-fold axis. This symmetry axis may in
principle be either perpendicular or parallel to the solution
Making use of the SHELX97 computer program,25,26
reasonable fit to the observed Bragg rod profiles was obtained
for a unit cell on which the 2-fold axis was aligned perpendicular
to the water plane. The Cu and Pb ions were centered on
the 2-fold axis, and the bolaamphiphiles constrained as rigid
rods (Figure 6a). The model structure best fitted to the
experimental data, suggested that the molecules lie in a plane
parallel to the solution surface, and their long axes making an
angle of γ ) 79° from the plane of the metal sheets (see Scheme
a
2
+
2+

11452 J. Phys. Chem. B, Vol. 105, No. 46, 2001
Buller et al.
Figure 8. (d,l)-C22Lys on pure water. GIXD intensities plotted as a
surface plot against q and qxy
z
An interesting feature observed for all the three-component
systems is the presence of a strong Bragg peak with a d spacing
of ∼41.5 Å, depending on the organic spacer and pressure
applied (see Figure 7a). We attribute this peak to a regular
corrugation of the film with a wavelength of ∼41.5 Å, which
is in keeping with the skewing of the Bragg rods. The fwhm of
the Bragg rod I(qz) is markedly different from those of the other
Figure 7. (a) 2D surface plot of the scattered intensity I(qxy, q
reflections of (d,l)-C22Lys spread on mixed Cu(Ac) and Pb(Ac)
subphase. b) The positions (along θxy) and intensities of the Bragg peaks,
z
) showing
2
2
41
reflections. Rippled hexatic phases were observed in lyotropic
2
7
-
1
-1
liquid crystals. Moreover, a regular corrugation was invoked
integrated along q
z
(between 0 Å and 0.15 Å ) measured for
and Pb(Ac) subphase.
for buckled monolayers of cadmium arachidate at the air-
amphiphile series spread on mixed Cu(Ac)
2
2
2
8
aqueous interface, with a periodicity of 200 Å. A compre-
hensive manuscript on this topic is in preparation.
TABLE 3: Reflections Arising from the Interchain
Arrangement
d.2. Bolaamphiphiles on Deionized Water. We now describe
the observed diffraction spectra, obtained for the non metallic
film of (d,l)-C22Lys.
d (Å)
q
xy
q
z
4.6
4.2
3.8
weak, sometimes present
strong, always present
weak, always present
1.40
1.52
1.67
0< q
0.2
0.3
z
<0.4
We interpret the GIXD pattern (Table, Figure 8) in terms of
two types of arrangements of the bolaamphiphiles with respect
to the water surface: in one, the molecules are arranged with
their long axis parallel to the water surface, in a structure akin
to Scheme 1a,b. The five reflections with Bragg rods, corre-
sponding to average film thickness of 79 Å, correspond to
harmonics of an axis 120.8 Å long. Such a spacing implies that
the molecules lie parallel to the water surface, forming a cell
by adopting a chain arrangement akin to that of the cationic
salts (Scheme 1a,b), but which contains four, rather than two
molecules aligned head-to-head and tail-to-tail, along the lattice
repeat.
TABLE 4: (d,l)-C22Lys on Pure Water: Listed Peak
Analysis of the GIXD Spectrum
assignment
d (Å)
q
x,y (Å-1)
q
z
(Å-1
)
fwhm (Å)
82
85
{
{
{
{
0,6}
20
10
8
0.31
0.63
0.77
1.25
1.34
1.4
0
0
0
0
0.66
0.05
0.05
0.05
0,12}
0,15}
0,24}
70
5
too shallow
highly skewed
40
35
35
4
3
3
4
.5
.4
.7
.0
1.55
1.7
In the second type of arrangement, the molecular crystallize
as a monolayer, in which the chains are aligned with one polar
headgroup in contact with the water surface, and the opposite
1a). The model in which the 2-fold axis is parallel to the solution
surface yields a very poor fit to the observed data. The mismatch
between some of the observed and calculated Bragg rods may
be due to the simplistic constraint imposed on the organic moiety
by assuming it as essentially a straight and rigid rod, which
probably does not reflect its structure at the atomic level.
The GIXD patterns of the amphiphilic series on the mixed
Pb-Cu subphase also exhibit reflections other than those arising
from the long spacing, which we interpret as due to the
arrangement between the chains along the x direction (Scheme
2
9
headgroup pointing toward the air. The film thickness of 37
Å, is consistent with the calculated length of the molecule (35
Å). This type of arrangement was observed for several systems
of bolaamphiphiles on water, among which are tetracosenedioic
3
30
acid and various diols.
Discussion
Pure bolaamphiphile (d,l)-C22Lys on deionized water yields
a mixture of crystalline films where the molecules lie parallel
1
a). All these reflections (Figure 7, Table 3) are extensively
skewed along Scherrer lines of constant total scattering vector
forming a multilayer, and a monolayer where the chains are
2
2
z
3,30,31
qt ) qxy +q . This phenomenon was previously observed in
tilted somewhat from the normal.
On the other hand, these
x
3
the diacid system. The three skewed peaks cannot be associated
with one unit cell; the dominant phase is clearly associated with
the strong peak at qxy ) 1.52 Å , qz ) 0.2 Å . This skewing
is interpreted as a bending of the film about an axis parallel to
the long chain axes of the molecule.
bolaamphiphiles, when deposited on mono- or bi-metal ionic
subphases, yield crystalline multilayer films where the molecules
only lie parallel to the aqueous solution surface, with the
appropriate metal cations intercalated within the structure, as
discussed below.
-1
-1

Self-Assembly of Bolaamphiphiles
J. Phys. Chem. B, Vol. 105, No. 46, 2001 11453
Spreading d or (d,l)-C22Lys on pure Cu(Ac)2 solution, yields
a poorly diffracting film (Figure 4c), with a 35.5 Å lattice
spacing according to the grazing incidence X-ray diffraction
(
(
GIXD) pattern. From the X-ray photoelectron spectroscopy
XPS) data, we deduce that the chains are arranged as in Scheme
3
b, so that the long axis spacing must be 2 × 35.5 Å ) 71 Å
in length.
On substituting the solution subphase for Pb(Ac)2, a clearly
diffracting film is observed by GIXD (Figure 4b), with a long
spacing of 76 Å. Taking into account the XPS results, we deduce
that the lead is bound to the carboxylic moiety only. The GIXD
pattern and the XPS measurements are in agreement,3 resulting
in the arrangement shown in Scheme 3c.
2
Figure 9. Powder diffraction of Cu-Pb-(d,l)-Glutamic acid, taken
with a copper KR radiation of λ)1.542 Å.
We have shown that the bolaamphiphilic series l or (d,l)-
2
+
C22Orn, d, l or (d,l)-C22Lys, (d,l)-C24Lys spread on mixed Cu
2
+
value is sufficiently close to the observed d(010) spacing to
strengthen the overall correctness of the model.
b. Atomic Absorption Experiments. The experiments were
done with solutions prepared in polypropylene bottles. Standards
were measured from Aldrich’s standards for atomic absorption
for Cu and Pb.
Solutions were prepared in the appropriate linear range, with
1% glacial acetic acid in deionized water. The experiments
showed that the crystal contains a 1:1 ratio between copper and
lead, with 4% relative error.
and Pb subphase form crystalline films, where the metal ions
are arranged in an array of alternating sheets, through selective
intercalation of a specific metal cation to a specific anionic
functional group. The evidence for such a structure arises from
the XPS and GIXD measurements, serving as complementary
methods. The XPS data, quantitatively confirm the ratio of two
bolaamphiphiles to one Cu2 ion and one Pb ion. These data
+
2+
2
+
also qualitatively suggest the preferential binding of Pb ions
to the carboxylic acid groups, and of Cu2+ ions to the amino
acid groups. The GIXD studies show the self-assembly of a
new crystalline bimetallic film where the long chain molecules
are arranged head-to-head and tail-to-tail, intercalating the two
different ions in an alternating manner (akin to Scheme 1a,b),
in agreement with the interpretation of the XPS results. The
small differences in the long chain spacing match those of the
chain lengths of the various bolaamphiphiles in the series,
reinforcing the correctness of the designed structure.
c. Isotherms. Solution of d or (d,l)-C22Lys (Scheme 2, V-ii)
0
.7 mL of 0.4mM amphiphile dissolved in 2% TFA in CHCl3),
were spread at room temperature (25 °C) over one of the four
appropriate subphases: deionized water, 1mM solution contain-
ing Pb(Ac)2, Cu(Ac)2, or a 1:1 mixture of the two, in a KSV
Teflon trough, and compressed. The pressure was detected by
a Wilhelmy balance.
d. Synthesis. All the solvents were purchased from Merck
in CP grade, unless otherwise stated. Amino Acids were
purchased from Bachem.
The GIXD patterns from the films formed by the enantio-
merically pure and racemic bolaamphiphiles spread on the
various subphases, differ primarily in the relative intensities of
their Bragg peaks, but barely so in their d spacings. This distinct
difference in Bragg peak intensity excludes the possibility for
spontaneous separation of the two enantiomers, when using the
racemate as a building block for the self-assembled film.
This matrix can potentially be transferred to a solid support,
reacted or reduced in order to form a nanowire grid, or a
nanocrystalline array of metallic or semiconducting character.
These types of arrangements may be applicable for the
construction of nanocrystalline arrays for microelectronics and
d.1. Monomethyl-ester of I-i or I-ii (Scheme 2, II-i, or II-ii).
3
5,36
A modification of an esterification procedure.
Finely ground dioic acid (Scheme 2, I-i or I-ii) (Fluka/Aldrich
respectively) (4.12 mmol) was dissolved in ethylene-chloride
(100 mL). Methanol was added to the clear solution (A. R. 4.12
mmol), then H2SO4 (0.4 mL) was cautiously added, and the
reaction was left for reflux overnight. The solvent was concen-
trated and water was added. Three phases were obtained:
aqueous, organic and a slurry. To the organic phase, which
contained mainly di- and mono-esters, carbon tetrachloride was
added, and placed in a separatory funnel. The organic layer was
washed with water, until reaching pH ) 6, then dried over
MgSO4 and evaporated.
6
,33
micro-detectors.
A schematic view of such an array at the
air-solution interface is shown in Scheme 1a,b. This aspect is
subject to further study, as well as other A-B systems, which
may allow energy transfer between their poles.34
The slurry contained mainly nonreacted diacid, and was
placed for soxlet extraction with n-hexane overnight, extracting
di- and monomethyl esters. The hexane was evaporated to
dryness, and both fractions (from the separatory funnel and
soxlet extraction) were subjected to chromatography on silica
Experimental Section
The water used for all experiments were Millipore purified
and deionized (18.2MΩ), unless otherwise mentioned.
a. 3D Crystals of Glutamic acid with Pb(Ac)2 and Cu(Ac)2.
Equal volumes of (d,l)-Glutamic acid, Cu(Ac)2, and Pb(Ac)2,
3
7
gel S-60 on a multibore column. TLC 1:1 ethyl-acetate:
methylene-chloride, detected by KMnO . The diester was eluted
4
0.05M each, were added successively. The crystals grown were
with methylene-chloride, and the monoester II with ethyl-acetate
hedge-hog in appearance (spherulites).
with 25% chemical yield.
H NMR data for II-i (250 MHz, CDCl3) δ[ppm] ) 3.67(s,
3H, CH3), 2.36(t, 2H, CH2), 2.30(t, 2H, CH2), 1.62(m, 4H, CH2),
1.26(s, 32H, CH2).
The NMR data for II-ii is identical except for 1.26(s, 36H,
CH2).
1
A specimen crystal suitable for a single-crystal X-ray dif-
fraction study could not be obtained, thus we resorted to powder
X-ray diffraction measurements to glean information on the
binding of the metals and amino acids. A set of {0,k,0}, k )
1
-5 reflections with a spacing d(010) ) 15.3 Å were observed
(see Figure 9). A model of the chain axis arrangement akin to
d.2. Hetero Methyl-N-hydroxysuccinimide-ester of II-i or II-
3
8
that in Scheme 1a,b, with each of the two metals (Pb and Cu)
occupying a center of inversion yields an axis of 18.7 Å, this
ii (Scheme 2, III-i or III-ii respectiVely) . A modification of
3
8
an activation procedure.

1
1454 J. Phys. Chem. B, Vol. 105, No. 46, 2001
Buller et al.
Dicyclohexylcarbondiimide, (DCC) (Aldrich, 10.8 mmol) and
(s(broad)t, 1H, chiral CH), 3.39(s(broad), 2H, CH2), 2.45-
(t(broad), 2H, CH2), 2.35(s(broad), 2H, CH2), 2.05(m(broad),
2H, CH2), 1.82(m(broad), 2H, CH2), 1.60(m, 4 H, CH2), 1.26-
(s, 32H, CH2).
II-i or II-ii (10.8 mmol) were dissolved in 90 mL dry methylene-
chloride, in a dry atmosphere. N-hydroxysuccinimide (10.8
mmol) was separately dissolved in 25 mL methylene-chloride,
and was added dropwise at room temperature. The reaction was
left to stir overnight, then water was added and stirring continued
for 15 min. The precipitated dicyclohexylurea (DCU) was
filtered out, yielding a clear methylene-chloride solution which
was dried over MgSO4, and evaporated. Further purification
from DCU was achieved by two recrystallizations: one in
methylene-chloride, and the other in ethyl-acetate. Traces of
DCU still remained. TLC 1:1 ethyl-acetate: methylene-chloride,
detected by KMnO4 or Ninhydrin.
1
H NMR data of C Lys V-ii (250 MHz, CDCl with ∼2%
2
2
3
TFA(D)) δ[ppm] ) 7.624 (s(broad), 1H, NH broad), 4.33-
(s(broad)t, 1H, chiral CH), 3.39(s(broad), 2H, CH ), 2.45-
2
(t(broad), 2H, CH ), 2.35(s(broad), 2H, CH ), 2.05(m(broad),
2
2
2H, CH ), 1.82(m(broad), 2H, CH ), 1.60(m, 4H, CH ), 1.40-
2
2
2
(m(broad), 2H, CH ), 1.26(s, 32H, CH ).
2
2
1
H NMR data of C24Lys V-iii (250 MHz, CDCl3 with ∼2%
TFA(D)) δ[ppm] ) 7.624 (s(broad), 1H, NH broad), 4.33-
s(broad)t, 1H, chiral CH), 3.39(s(broad), 2H, CH2), 2.45-
(t(broad), 2H, CH ), 2.35(s(broad), 2H, CH ), 2.05(m(broad),
(
1
H NMR data for III-i (250 MHz, CDCl3) δ[ppm])3.67(s,
2
2
3
2
H, CH3), 2.85(s, 4H, CH2-CH2), 2.61(t, 2H, CH2), 2.31(t,
H, CH2), 1.62(m, 4H, CH2), 1.26(s, 32H, CH2).
2 2 2
2H, CH ), 1.82(m(broad), 2H, CH ), 1.60(m, 4H, CH ), 1.40-
(m(broad), 2H, CH ), 1.26(s, 36H, CH ).
2
2
-1
The NMR data for III-ii is identical except for 1.26(s, 36H,
I.R. [cm ] (KBr) of the dipotassium salt: 3443, 3303, 2919,
CH2).
d.3. Coupling Reaction of Compound III-i or III-ii with
l-Lysine-Methyl ester (Scheme 2, IV-ii or IV-iii, RespectiVely).
2846, 1640, 1580, 1540, 1507, 1467, 1408, 1348, 710.
+
FABMS calc. For (d,l)-C H N O H[M+1] : 485.73;
2
7
51
2
5
found: 485.34; disintegration point d.p. ) 223 °C.
1
2,13
A modification of a regioselective coupling procedure.
+
calc. For l-C27H51N2O5H[M+1] : 485.73; found:485.28; d.p.
Compound III-i or III-ii (0.14 mmol) was dissolved in 4 mL
methylene-chloride. In a separate flask, lysine-methyl-ester (0.19
mmol) was dissolved in 3 mL methylene-chloride, with addition
of 0.75µL triethylamine (distilled over KOH). The lysine
solution was added to III-i or III-ii, and stirred overnight at room
temperature. The powder obtained after evaporation of the
solvent, was washed with hexane and water. Traces of DCU
were removed by column chromatography with ethyl acetate
elution, while the pure product IV was eluted with acetone.
Chemical yield 92%. TLC 1:1 ethyl-acetate: methylene-
chloride, detected by KMnO4 or Ninhydrin.
) 213 °C.
+
calc. For (d,l)-C28H54N2O5H[M+1] : 499.75; found: 499.41;
d.p. ) 220 °C.
calc. For l-C28H54N2O5H[M+1] : 499.75; found: 499.41;
d.p. ) 224 °C.
calc. For (d,l)-C30H58N2O5H[M+1] : 527.81; found: 527.41;
d.p. ) 220 °C.
+
+
Overall yield ≈ 20%
e. X-ray Photoelectron Spectroscopy (XPS). The crystallites
of the d or (d,l)-C22Lys with or without metals were prepared
by spreading a solution of the amphiphile (0.7 mL of 0.4mM)
dissolved in 2% TFA in CHCl3 (AR) at room temperature, over
one of the four appropriate subphases: deionized water, 1mM
solution containing Pb(Ac)2, Cu(Ac)2, or a 1:1 mixture of the
two, in a Teflon trough. These crystallites were then deposited
on glass which was pre-cleaned by boiling it in a solution of
CHCl3:ethanol 1:1 for 2 h. During the experiment the glass was
placed inside the subphase, on a stainless steel table like mesh.
After 1h at room temperature, the system was cooled to 4 °C,
and left as such for another hour. Then the water was slowly
drained, well below the level of the stainless steel mesh. The
glass with the deposited film prepared on a metal containing
subphase, was left for drying, at least for an hour, and then
washed with water by dipping, and leaving it in a covered Petri
dish for drying. Copper was easily removed by washing, whereas
lead was more difficult to remove, probably due to its binding
to the glass.
1
H NMR data for IV-ii (250 MHz, CDCl3 with ∼2% TFA-
(
D)) δ[ppm])5.85 (t, 1H, chiral CH), 4.02 (s, 1H, NH broad),
3.77(s, 3H, CH3), 3.67(s, 3H, CH3), 3.48(m(broad), 2H, CH2),
3.31(m(broad), 2H, CH2), 2.31(t, 2H, CH2), 2.15(t, 2H, CH2),
1.83(t, 2H, CH2), 1.62(m, 4H, CH2), 1.26(s, 32H, CH2).
The NMR data for IV-iii is identical except for 1.26(s, 36H,
CH2).
d.4. Coupling Reaction of Compound III with l-Ornithine-
Methyl ester (Scheme 2, IV-i)12,13. Compound IV-i was prepared
in an identical manner to c.3., replacing lysine-monomethyl-
ester by ornithine-monomethyl-ester.
1
H NMR data for IV-i (250 MHz, CDCl3 with ∼2% TFA-
(
D)) δ[ppm])5.85 (t, 1H, chiral CH), 4.02 (s, 1H, NH broad),
3.77(s, 3H, CH3), 3.67(s, 3H, CH3), 3.48(m(broad), 2H, CH2),
3.31(m(broad), 2H, CH2), 2.31(t, 2H, CH2), 2.15(t, 2H, CH2),
1.62(m, 4H, CH2), 1.26(s, 32H, CH2).
d.5. Hydrolysis of IV-i, IV-ii or IV-iii. A modification of
XPS measurements were performed on a Kratos AXIS-HS
instrument, using monochromatized Al(KR) source at 75W-
(15KV, 5mA), and detection pass energies of 20-40 eV. A
flood gun was used to compensate extra charges, whereas the
signal energy calibrated as centered to the C(1s) line at 284.8
hydrolysis procedure.39
Finely ground IV-i, IV-ii, or IV-iii (0.35 mole) was dissolved
by heating in methanol (AR, 20 mL). After cooling to 40-45
C, 1M KOH/methanol (18 mL) was added, as well as pure
distilled tetrahydrofuran (20 mL) and water (1 mL). The reaction
mixture was stirred overnight at 40-45 °C. The solvents were
evaporated almost to dryness. Neutralization to pH ) 6-7 was
achieved by an addition of HCl 2M. The product precipitated
as a fine powder, and extracted via centrifugation.
°
1
6-18
eV. To eliminate beam-induced damage,
we have used
repetitive measurements, which presented the evaluation at the
damage on time scales up to 6-9 h. These studies showed
different time constants for different mechanisms, the shortest
2
+
regards reduction of the Cu ions on a scale of a few (∼10)
minutes. Eventually, the damage-free spectra were recorded on
fresh areas at relatively fast scans. Data analysis included Shirley
background subtraction and curve fitting using Gaussian-
Lorenzian line shapes.
The product V was purified by boiling the fine powder in
pure tetrahydrofuran, acetone, methanol and ethanol (AR)
successively. TLC 7:3 Propanol:Ammonia, detected by Ninhy-
drin. Chemical yield 95%.
1
H NMR data of C22Orn V-i (250 MHz, CDCl3 with ∼2%
The concentration of oxygen O(1s) could not be unambigu-
ously determined because of the overlap of the glass.
TFA(D)) δ[ppm] ) 7.624 (s(broad), 1H, NH broad), 4.33-

Self-Assembly of Bolaamphiphiles
J. Phys. Chem. B, Vol. 105, No. 46, 2001 11455
f. Grazing Incidence X-ray Diffraction Experiments
GIXD). The crystallites of the l or (d,l)-C22Orn, d, l or (d,l)-
(10) Thomas, H.; Britton, S.; Meek, F. H. J. Chem. Soc. 1913, 40, 2831-
2
840.
(
(
11) Yazdanian, M.; Yu, H.; Zografi, G. Langmuir 1992, 8, 630-636.
12) Shatzmiller, S.; Confalone, P. N.; Abiri, A. Synlett 1999, S1, 963-
965.
C22Lys, (d,l)-C24Lys with or without metals were prepared by
spreading a solution of 0.7 mL of 0.4mM amphiphile dissolved
in 2% TFA in CHCl3), at room temperature over one of the
four appropriate subphases: deionized water, 1mM solution
containing Pb(Ac)2, Cu(Ac)2, or a 1:1 mixture of the two, in a
Teflon trough. The trough was closed and cooled to 4 °C, and
wt h ae sn rce oa mc hp er de .s sed, until a build up in the pressure of ∼7mN/m
(
(13) Bachovchin, W. W.; Plaut, A. G.; Flentke, G. R.; Lynch, M.;
Kettner, C. A. J. Biol. Chem. 1990, 265, 3738-43.
(14) Majewski, J.; Margulis, L.; Weissbuch, I.; Popovitz-Biro, R.; Arad,
T.; Talmon, Y.; Lahav, M.; Leiserowitz, L. AdV. Mater. 1995, 7, 26.
(15) Brend, P.; Freiser, H. Anal. Chem. 1974, 46, 1147.
(
16) Chusuei, C. C.; Brookshier, M. A.; Goodman, D. W. Langmuir
999, 15, 2806-2808.
17) Usubaliev, B. T.; Amirov, A. S.; Amiraslanov, I. R.; Mamedov,
K. S. Zh. Strukt. Khim. 1989, 30, 175-179.
18) Frydman, E.; Cohen, H.; Maoz, R.; Sagiv, J. Langmuir 1997, 13,
089-5106.
19) Further research is under progress utilizing this phenomenon.
(20) Beamson, G.; Briggs, D. High-Resolution XPS of Organic Polymers.
The Scienta ESCA300 Database; John Wiley & Sons., 1992.
1
(
The GIXD experiments were preformed on the liquid surface
diffractometer at beamline BW1 in HASYLAB at DESY,
Hamburg, Germany. The scattered intensity was detected by a
position sensitive detector (PSD) equipped with Soller slits. This
detector resolves the vertical component of the X-ray scattering
vector qz ≈ (2π/λ)sinRf, Rf being the angle between the plane
of the liquid surface and the diffracted beam. The resolution is
limited to that of the Soller slit. Each sample consists of 2D-
crystallites that are randomly oriented around the normal to the
solution surface, and thus is a “2D powder”. Measurements were
performed by scanning the horizontal scattering vector qxy ≈
(
5
(
(
21) Srivastava, S. Appl. Spectrosc. ReV. 1986, 22, 401.
(22) Als-Nielsen, J.; Jaquemain, D.; Kjaer, K.; Leveiller, F.; Lahav, M.;
Leiserowitz, L. Phys. Rep. 1994, 246, 251.
(23) Weissbuch, I.; Popovitz-Biro, R.; Kjaer, K.; Als-Nielsen, J.; Lahav,
M.; Leiserowitz, L. Molecular Self-Assembly into Crystals at Air-Liquid
Interfaces; Prigogine I., Rice S. A., Ed.; Wiley & Sons: 1997; Vol. 102,
pp 39-120.
(24) The q values of the Bragg peaks of the d-or (d, l)C lys on a
xy
22
(
4π/λ)sinθxy, where 2θxy is the angle between the incident and
Cu(Ac)2 solutions may be interpreted in terms of a lattice spacing of 35.5
Å (d ) 2π/qxy, qxy ) 0.18, 0.35), The XPS results appears to support a
spacing of 2 × 35.5 Å, since we clearly detect the carboxylic acid (-COOH)
signal in case d-C22Lys on pure Cu(Ac)2, which indicates that the carboxylic
acid end is a hydrogen bonded dimer, rather than a copper-dicarboxylate.
diffracted beams projected onto the horizontal xy plane. The
diffraction data are represented by a 2D intensity distribution
plot, showing the diffracted intensity I(qxy,qz) as a function of
qxy and qz. The qxy positions of the Bragg peaks yield the lattice
repeat distances d ) 2π/qxy which can be indexed by two indices
h,k to yield the unit cell. The full width at half-maximum (fwhm)
(
25) Sheldrick, G. M. Acta Crystallogr., Sect. A. 1990, A46, 467-473.
(26) Sheldrick, G. M. “SHELXL97 program for the refinment of crystal
structures,” University of Goetingen, Germany, 1997.
(27) Date, R. W.; Luckhurst, G. R.; Shuman, M.; Seddon, J. M. J. Phys.
II France 1995, 5, 587-605, and references therein.
(28) (a) Fradin, C.; Daillant, J.; Braslau, A.; Luzet, D.; Alba, M.;
Goldmann, M. Eur. Phys. J. B 1998, 1, 57-69. (b) Daillant, J. Synchrotron
Radiation News 1999, 12, 17.
of the Bragg rod intensity modulations (integrated along qxy),
yields a first estimate of the crystallite thickness42 0.88‚2π/
fwhm(qz). More quantitatively, the intensity at a particular qz
value in a Bragg (h,k) rod is determined by the square of the
(29) We can assign rectangular cells using the spacing of 4.0 Å and 4.5
2
2
Å to yield a unit cell with a molecular area of 21.8 Å , and the spacing 3.7
molecular structure factor |Fhk(qz)| , thus allowing its evaluation
2
Å and 4.5 Å to yield a unit cell with a molecular area of 18.3 Å . We
4
0
according to an atomic coordinate model of the molecules.
cannot assign a cell with the 3.4 Å spacing that will yield a reasonable
area.
Acknowledgment. We acknowledge financial support form
the AFIRST (Association Franco-Israelienne pur la Racherche
Scientifique at Technologique) grant no. 9814-1-98, the Israeli
Ministry of Science, the Densyne program of the Danish
Foundation for Natural Sciences and E.U.’s TMR program, and
we thank the HASYLAB, Desy, Hamburg, for synchrotron beam
time at BW1 line, and support.
(30) Popovitz-Biro, R.; Edgar, R.; Majewski, J.; Cohen, S.; Margulis,
L.; Kjaer, K.; Als-Nielsen, J.; Leiserowitz, L.; Lahav, M. Croat. Chem.
Acta, 1996, 69, 689-708.
(
31) Weissbuch, I.; Popovitz-Biro, R.; Lahav, M.; Leiserowitz, L.; Als-
Nielsen, J.; Howes, P. B.; Kjaer, K. HASYLAB Annual Report Part I 1997,
82.
4
(
32) The amino acids which are bound to lead (∼33% see Figure 3b,
Table 1b), are most probably non crystalline, and thus are not observed in
the GIXD spectra.
(
33) Xiong, H.; Cheng, M.; Zhou, Z.; Zhang, X.; Shen, J. AdV. Mater.
998, 10, 529.
34) (a) Sirota, M.; Hensel, V.; Minkin, E.; Lahav, M.; Lifshitz, E. J.
1
References and Notes
(
(
(
(
1) Lehn, J.-M. Supramol. Chem.; VCH: Weinheim, 1995.
2) Ozin, G. A. AdV. Mater. 1992, 4, 612.
3) Weissbuch, I.; Gou, S.; Edgar, R.; Cohen, S.; Howes, P.; Kjaer,
Phys. Chem. B 2001, 105, 6792-6797. (b) Hensel, V.; Godt, A.; Popovitz-
Biro, R.; Cohen, H.; Jaensen, T. R.; Kjaer, K.; Weissbuch, I.; Lifshitz, E.;
Lahav, M. Chem.-A Eur. J. 2001, in print.
K.; Als-Nielsen, J.; Lahav, M.; Leiserowitz, L. AdV. Mater. 1998, 10, 117.
(35) Clinton, R. O.; Laskowski, S. C. Am. Soc. 1948, 70, 3135.
(36) Fieser, L. F.; Fieser, M. Reagents for Organic Synthesis; Wiley:
New York, 1967.
(
4) Weissbuch, I.; Baxter, P. N. W.; Kuzmenko, I.; Cohen, H.; Cohen,
S.; Kjaer, K.; Howes, P. B.; Als-Nielsen, J.; Lehn, J. M.; Leiserowitz, L.;
Lahav, M. Chem.-A Eur. J. 2000, 6, 725-734.
(37) Fischer, G. A.; Kabara, J. J. Analytical Biochemistry 1964, 9, 303-
(5) Weissbuch, I.; Baxter, P. N. W.; Cohen, S.; Cohen, H.; Kjaer, K.;
309.
Howes, P. B.; Als-Nielsen, J.; Hanan, G. S.; Schubert, U. S.; Lehn, J.-M.;
(38) Paquet, A. Can. J. Chem. 1976, 54, 733.
Leiserowitz, L.; Lahav, M. J. Am. Chem. Soc. 1998, 120, 4850-4860.
(39) Bodanszky, M. The Practice of Peptide Synthesis; 2nd, rev. ed.
ed.; Springer-Verlag: Berlin, 1994.
(
6) Gou, S.; Popovitz-Biro, R.; Weissbuch, I.; Cohen, H.; Hodes, G.;
Lahav, M. AdV. Mater. 1998, 10, 121.
7) Samokhvalov, A.; Berfeld, M.; Lahav, M.; Naaman, R.; Rabani,
E. J. Phys. Chem. B 2000, 104, 8631-8634.
8) Greenstein, J. P.; Winitz, M. Chemistry of the Amino Acids; John
Wiley & Sons: 1984.
9) Fowler, B. A. EnVironmental Health PrespectiVes 1998, 106, 1585-
587.
(40) Jacquemain, D.; Grayer-Wolf, S.; Leveiller, F.; Deutsch, M.; Kjaer,
K.; Als-Nielsen, J.; Lahav, M.; Leiserowitz, L. Angew. Chem., Int. Ed. Engl.
1992, 31, 130.
(
(
(41) Buller, R.; Kjaer, K.; Lahav, M.; Leiserowitz L. The 66th Meeting
of the Israel Chemical Society 2001, 176.
(
(42) Guinier, A. X.-Ray Diffraction, W. H. Freeman, 1963, Section 5.5
p 142ff.
1