A racemic conglomerate nipped in the bud: A molecular view of enantiomer cross-inhibition of conglomerate nucleation at a surface

Article abstract of DOI:10.1021/cg100806w

The inhibition of conglomerate formation in a racemic monolayer is shown to arise from one enantiomer impeding the growth of homochiral nuclei of the opposite enantiomer. Submolecular level resolution scanning tunneling microscopy (STM) data show that nan

Full text of DOI:10.1021/cg100806w

DOI: 10.1021/cg100806w
A Racemic Conglomerate Nipped in the Bud: A Molecular View of
Enantiomer Cross-Inhibition of Conglomerate Nucleation at a Surface
2010, Vol. 10
4516–4525
Abel Robin,† Patrizia Iavicoli,‡ Klaus Wurst,§ Matthew S. Dyer,† Sam Haq,†
David B. Amabilino,*,‡ and Rasmita Raval*,†
†Surface Science Research Centre, Department of Chemistry, University of Liverpool, L69 3BX
Liverpool, U.K., ‡Institut de Ciencia de Materials de Barcelona (CSIC), Campus Universitari,
08193 Bellaterra, Catalonia, Spain, and Institut fu€r Allgemeine Anorganische und Theoretische Chemie,
Universita€t Innsbruck, A-6020, Innrain 52a, Austria
§
Received June 16, 2010; Revised Manuscript Received July 30, 2010
ABSTRACT: The inhibition of conglomerate formation in a racemic monolayer is shown to arise from one enantiomer
impeding the growth of homochiral nuclei of the opposite enantiomer. Submolecular level resolution scanning tunneling
microscopy (STM) data show that nanometer scale homochiral conglomerate nuclei, containing several tens of molecules, are
obstructed from developing further by enantiomeric blockers in their surroundings. We have observed this in the racemic and
enantiopure samples of a bis-lactate derivative of resorcinol which was prepared because its size and symmetry made it
interesting for the study of chirality in chemisorbed monolayer systems. The solid state racemate forms a racemic compound
containing hydrogen bonded chains. The enantiomers and the racemate were deposited onto a Cu(110) surface under ultrahigh
vacuum conditions and the monolayers formed were studied by STM, low energy electron diffraction (LEED), and reflection
absorption infrared spectroscopy (RAIRS). In particular, STM observations of the racemic monolayer reveal that the
individual enantiomers form small organized aggregates which are interrupted by defects of the opposite enantiomer which
acts as a growth inhibitor. This situation contrasts sharply with the monolayers formed by the enantiopure compounds, which
form large homochiral domains. We believe that this unique nanometer scale view of conglomerate crystallization and its growth
inhibition, with each enantiomer acting as a mutual inhibitor of its mirror counterpart in a situation where diffusion is not fast,
provides important insights for understanding the formation of conglomerates and racemic compounds in bulk crystals.
Introduction
the “wrong” enantiomer. However, for a racemic compound a
“wrong” enantiomer occupying a site only has to translate
across to the “correct” neighboring site to enable racemic
crystal growth to continue. In contrast, during the formation
of a conglomerate, the “wrong” enantiomer has no nearby
“correct sites” available to diffuse to and, thus, can act as a
very effectiveinhibitor. On this basis, they suggested that if the
thermodynamic preference between a conglomerate and a
racemic compound is not great, such inhibition effects affect-
ing the kinetics of formation may be the underlying reason
why bulk crystalline conglomerates are relatively rare.
The only way one could know of such a phenomenon is to be
able to observe directly the nucleation and inhibition process.
Given that nucleation of molecules, in general, is considered to be
a process in which the condensates have dimensions in the
nanometer range,6 this would require techniques that could
probe the system at the nanometer level. During our studies on
the spontaneous resolution of compounds at interphases,7 we
have chanced on an example of the inhibition of long-range
conglomerate formation by mirror enantiomers at the growing
surface-confined crystallite. Thus, the conglomerate nuclei are
observed to be confined to very small aggregates of only tens of
molecules, by enantiomeric blockers in their surroundings. The
technique with which we have observed this process with sub-
molecular resolution is scanning tunneling microscopy (STM).
The unique benefits of STM in the study of chirality within
organized molecular monolayers at surfaces has been demon-
strated in a number of instances, including identification of the
absolute chirality of adsorbates,8 or the observed enantiomor-
phous structures of physisorbed9 and chemisorbed10 enantiomeric
Molecular level understanding of chirality transmission,
and in particular the passage and even amplification of
chirality into homochiral systems, is of great importance in
a number of scientific and technological areas.1 The sponta-
neous resolution of enantiomers in a racemate to form a
conglomerate, which arises from the condensation of mirror
image molecules into separate domains,2 is a crucial step
toward the preparation of homochiral compounds by
crystallization.3 One thing which can either prevent or favor
the emergence of chirality is the inhibition of the growth of
homochiral domains, be they in fibers, monolayers, or bulk
crystals. A key illustration of this effect is the use of tailor-
made inhibitors for the growth of chiral crystals.4 However,
despite great efforts the mapping of chiral nucleation and
crystallization at the nanometer scale remains rudimentary
and a molecular level view of inhibition has not been reported
to the best of our knowledge.
It is particularly interesting to consider what happens in a
racemic system in which enantiomers could act as mutual
inhibitors for the growth of each other’s crystals. Such a situa-
tion would surely disfavor conglomerate formation and,
instead, give rise to the growth of the racemic compound. In
fact, this scenario was raised by Pratt-Brock et al.5 who pointed
out that kinetics may well favor the growth of racemic com-
pound crystals over a conglomerate since half the molecules
arriving at the boundaries of a nucleus site are, statistically, of
*To whom correspondence should be addressed. E-mail: amabilino@
icmab.es (D.B.A.) and Raval@liv.ac.uk (R.R.).
r
pubs.acs.org/crystal
Published on Web 09/03/2010
2010 American Chemical Society

Article
Crystal Growth & Design, Vol. 10, No. 10, 2010 4517
Chart 1
Scheme 1. Synthesis of the Chiral Resorcinol Diacid
Derivative (R,R)-1
and racemic materials. Turning specifically to the creation of
conglomerates, we note that the achiral nature of the surfaces
in these examples has no enantioselectivity toward chiral
molecules. Thus, chirality is introduced by molecular adsorp-
tion at the surface and can be understood in terms of a
symmetry reduction of the surface upon adsorption: any mirror
symmetry of the achiral surface is completely overwritten upon
adsorption of enantiomers of chiral molecules, which internally
have no mirror symmetry. The enantiomers maintain equal
adsorption energy by adopting mirror adsorption geometries
leading to local chiral adsorption motifs which spontaneously
resolve into mirror enantiomorphs.
The molecule we chose to study the formation of racemates
is 1 (see Chart 1), a resorcinol (1,3-dihydroxybenzene) deri-
vative with two chiral lactate acid groups. The choice of the
phenyl spacer was to ensure a sufficient size so as to be able to
distinguish the enantiomers molecule-by-molecule with the
STM on the metal surface. The substitution at the 1,3 posi-
tions of the benzene ring was chosen because the molecular
assembly has a lower symmetry than the corresponding 1,4
derivatives, a feature which was expected to lead to easier identifi-
cation of the enantiomers at the molecular level. Specifically,
the C2 axis running through the center of the benzene ring in the
1,4 derivative is not possible in the 1,3 compound. In this way,
the symmetry axis normal to the surface is avoided.11
The racemic compound (R,R)-1/(S,S)-1 crystallizes in the
orthorhombic Pnna (No. 52) space group with four molecules
in the unit cell. The density of the crystals is higher than those
of the enantiopure compound, conforming to Wallach’s rule.5
The crystal structure reveals molecules whose β-methyl group
is almost in the same plane as the phenyl ring, with the carbo-
xyl groups located in a pseudoanti conformation (Figure 1b),
asisthe casefor the enantiopure compound. The torsionangle
between the carbonyl group and the plane of the phenyl ring
for (R,R)-1 in the racemic crystal (C1C2O3C4 and C01C02O03-
C04) is 72.8ꢀ; for (S,S)-1 the corresponding angle is -72.8ꢀ.
Whereas enantiopure (R,R)-1 crystallizes forming a hydrogen
bondedchain, the (R,R)-1/(S,S)-1 crystal contains a hydrogen
bonded carboxylic acid dimer across a center of inversion (for
all crystallographic information and a details of the packing
see Supporting Information).
The solid state IR spectra of the racemic and enantiopure 1
show the single carbonyl adsorption at 1712 cm-1 (clearly
consistent with strong hydrogen bonding), and the racemic
mixture shows overtones and combination bands of ν(CO)
and OH deformation vibrations between 2500 and 2700 cm-1
and the OH out of plane band at 918 cm-1. The relatively
large bandwidth of the 918 cm-1 band is characteristic of
the centrosymmetric carboxyl dimer. The IR spectra of the
enantiopure compound shows two differences from the race-
mate: first, the carbonyl band is split into two bands at the
lower wavenumber and, second, the OH contribution is
shifted to a lower frequency (881 cm-1). This difference is a
result of the different sheet structures of enantiopure and
racemic crystals (as outlined in the Supporting Information,
the less rigid OH bond in the catemeric structure of the
enantiopure derivative one expects a shift to lower frequencies
in the IR).
Synthesis and Characterization
Compounds (R,R)-1 and (S,S)-1 were prepared by the
synthetic route shown in Scheme 1 for the former (the enantio-
mer was prepared in an identical way using the enantiomer of
methyl lactate as the starting material). In the first step,
resorcinol (1,3-dihydroxy benzene) was condensed with (S)-
or (R)-methyl lactate in the presence of triphenylphosphine and
diisopropyl azodicarboxylate by means of a Mitsunobu
protocol12 giving, respectively, the ester (R,R)-2, where the
stereochemistry is the opposite of the original lactate, in 48%
yield after purification by column chromatography. The com-
pounds were characterized by NMR and IR spectroscopies as well
as by their optical rotation. The second step was a saponification
of these ester derivatives using NaOH in MeOH resulting in
(R,R)-1 and (S,S)-1 as white solids after isolation in 94% yield.
Single crystals of the acids were grown from water. The
enantiopure diacid (R,R)-1 crystallizes in the orthorhombic
C2221 (No. 20) space group with four symmetry-related
molecules of (R,R)-1 in the unit cell. The structure reveals
molecules whose phenoxy group is almost in the same plane
as the β-methyl group, with the carboxyl groups located in a
pseudo-anti conformation (Figure 1a). This conformation is
favored energetically as revealed in theoretical investigations
on related compounds.13 The torsion angle between the carbo-
nyl group and the plane of the phenoxy group (C1C2O3C4 and
C01C02O03C04) is þ70.1ꢀ. In the crystalline structure, the mole-
cules arrange in homochiral chains which run along the crystal-
lographic c axis in which the molecules are linked through
hydrogen bonds between the carboxylic groups (see Supporting
Information).
Monolayers on Cu(110)
Racemic 1 on Cu(110). Racemic (R,R),(S,S)-1 was sub-
limed onto a clean Cu(110) surface at room temperature and
the adlayer was imaged using STM. Images from the as-
deposited sample (Figure 2) show distinct dot-like features,
˚
measuring about 7 A across, which is consistent with the size
of a single molecule of 1 (the distance between the two
stereogenic centers in the crystal structures is approximately

4518 Crystal Growth & Design, Vol. 10, No. 10, 2010
Robin et al.
Figure 1. Views of the (R,R)-1 conformation in the bulk crystals of the enantiopure compound (a) and in the crystal structure of the racemic
compound (R,R)-1/(S,S)-1 (b).
this kind, in this case ranging approximately from 5 to 10 nm
in length.
To gain further insight into the nature of these nanoscale
chiral domains, high-resolution images were obtained from
selected areas (Figure 4). From these pictures. it can be seen
that the circular features seen in the large area STM images in
fact possess significant submolecular features, displaying
five individual elements, of which four give rise to a bulky
square part. The fifth feature is found to be attached to either
the left-hand or the right-hand edge of the square (rather like
a thumbs-up sign), giving rise to two distinct and non-
superimposable mirror forms at the surface which we attri-
bute to each enantiomer of 1. Therefore, a five-spot cartoon
can be used to identify the different types of adsorbed
molecules, as shown for a disordered area of the surface
(Figure 4). We find that four distinct types of images are
actually observed, since each enantiomer is found in two
Figure 2. STM image (24 ꢀ 24 nm2, Ubias = 0.63 V, IT = 0.87 nA)
positions, rotated by 180ꢀ with respect to each other, and as
of racemic (R,R),(S,S)-1 sublimed onto Cu(110) at room tempera-
expected from the 2-fold symmetry of the substrate and
ture.
chirality of the compounds. Thus, we are able to identify
both the chirality and orientation of individual molecules at
the surface. This situation now provides a powerful means to
understand the details of the local chiral organization that
are manifest within the STM images (Figure 4).
˚
7.2 A), aggregated at the surface but displaying no long-
range order. However, amidst this general disorder, occa-
sional small ordered areas are observed, which display short
double-molecule chains oriented along specific chiral direc-
tions that are approximately (46ꢀ away from the [110] Cu
axis (for example, in Figure 2 in the middle there is a row of
six pairs of dots). The lengths of these chains generally do not
exceed 5 nm in length, but their defined growth direction
suggests local aggregation of enantiopure regions; that is,
they may represent the very first steps toward the onset of
conglomerate formation.
In order to enhance such chiral segregation processes, the
sample was annealed to 370 K and the resulting large area
STM image (about 2500 nm2) is shown in Figure 3. This
image is striking and shows that the initial general disorder
observed at room temperature has now been superseded by a
local ordering process in which a multitude of small chiral
domains, oriented either þ46ꢀ or -46ꢀ to the [110] direc-
tion, are present. However, although the number of chiral
domains has increased significantly, their individual size still
remains small compared with other monolayer systems of
High resolution images obtained from a local chiral
domain formed at the surface (Figure 4, bottom) show that
it consists of double-molecule chains. When each molecule
is mapped with its five-spot structure, it becomes apparent
that the double chain structures are enantiopure and are
constructed from antiparallel orientations of the paired
molecules. This observation suggests that local chiral segre-
gation toward conglomerate formation has been initiated
and local concentrations of enantiopure domains are nucle-
ated. Furthermore, it appears that the enantiopure domains
adopt a specific chiral organization. Specific support for
this conclusion comes from STM experiments carried with
enantiopure (S,S)-1 and (R,R)-1, as discussed in the next
section.
Enantiopure (S,S)-1 and (R,R)-1 on Cu(110). Typical STM
images obtained for enantiopure (S,S)-1 and (R,R)-1 on
Cu(110) following deposition at room temperature and

Article
Crystal Growth & Design, Vol. 10, No. 10, 2010 4519
Figure 3. Large-scale STM image (50 ꢀ 50 nm2, Ubias = 0.24 V, IT = 0.65 nA) of racemic (R,R),(S,S)-1 sublimed onto Cu(110) at room
temperature and afterward annealed to approximately 370 K.
annealing to between 340 and 370 K are shown in Figure 5.
These images show that each enantiopure system assembles
into large chiral domains, in which paired-molecule chains
grow along nonsymmetry directions aligned at þ46ꢀ or -46ꢀ
with respect to the [110] direction of the surface, that is,
exactly what is seen in the nanometer scale domains observed
for the racemic mixture (Figure 4). It is clear from a
comparison of the images in Figure 5 that (R,R) and (S,S)
form mirror enantiomorphous domains after moderate
annealing.
The unit cell dimensions measured from the STM images
of (S,S)-1 are approximately 1.85 ( 0.1 nm for a0, and
1.05 ( 0.05 nm for b0. The large scale organization is further
echoed in the low energy electron diffraction (LEED) data
obtained for (S,S)-1 (Figure 6), which show clear diffrac-
tion spots possessing the superstructure unit cell described by
the tensor
!
!
!
3 - 2
a0
b0
a
b
Γr ¼
defined as
¼ Γ
1
5
with a, b describing the unit vectors of the underlying
substrate lattice and a0, b0 the unit cell vectors of the mole-
cula0r superlattice. This 0unit mesh results in unit cell lengths
˚
˚
of a = 18.13 A and b = 10.51 A, which are in excellent
agreement with the measured STM distances for the chiral
structures of (S,S)-1.
To get further insight into the nature of the adsorbed
molecule, we also carried out infrared spectroscopy in the
reflection mode (RAIRS). This allows us to investigate the
chemical identity of the monolayer. Figure 7 shows the RAIRS
spectra obtained of an (R,R)-1 adlayer on Cu(110) at room

4520 Crystal Growth & Design, Vol. 10, No. 10, 2010
Robin et al.
2
˚
Figure 4. Top: Close-up STM image (50 ꢀ 50 A , Ubias = 0.29 V,
IT = 0.85 nA) of racemic (R,R),(S,S)-1 deposited on Cu(110) at
room temperature and afterward annealed to 370 K. The image
depicts an area of molecules of both enantiomers which are identi-
fied by different colors, blue and white, respectively, in (b). Bottom:
Close-up STM image of enantiomer separation and local domain
forming (c, 7.7 ꢀ 7.7 nm2, Ubias = 0.28 V, IT = 0.42 nA). Enantio-
pure double chains consist of up and down running rows, that is,
molecules in one row are rotated by 180ꢀ with respect to the mole-
cules in the other row of the double chain. Single molecule identification
(d) reveals that extended chain formation is hindered either by the
opposite enantiomer or by different, “achiral” molecular conforma-
tions. The main directions of the Cu axes are shown in (d) - the vertical
one is the Æ110æ and the horizontal the Æ001æ - and apply to all images.
Figure 5. Top: STM image of an (R,R)-1 layer annealed to 370 K
(10 ꢀ 10 nm2, Ubias = -0.88 V, IT = -0.27 nA). Bottom: the
corresponding sample of the (S,S)-1 compound.
temperature and upon annealing to the temperature at which
the chiral organization occurs. Of particular note is that the
ν(OH) stretching mode expected at 3513-3518 cm-1 for the
carboxylic group is not observed and neither is the carbonyl
ν(CdO) vibration of the same functionality expected to appear
between 1710 and 1790 cm-1. These observations in combina-
tion with the presence of a strong symmetric carboxylate
stretch, νs(COO-) observed at around 1400 cm-1 suggest
that both carboxylic acid groups of the molecules deproto-
nate on adsorption. Such deprotonation of carboxylic acid
groups to form carboxylate species on Cu(110) is a common
behavior that has been observed for a range of species.14
Apart from these differences, there is good agreement with
the rest of the spectrum, suggesting the remaining molecular
identity is retained upon adsorption. In addition, it can be
seen that annealing to 370 K causes all the bands to sharpen,
indicative of an ordering process.15 This temperature regime
corresponds to a change in the LEED pattern to yield an
extended chiral organization and the formation of the ordered
enantiomorphous chiral phase in STM. Detailed assign-
ments of the RAIRS bands from comparison with calculated
gas phase IR spectra for the geometry optimized symmetric
molecule are given in the Supporting Information.
Figure 6. LEED pattern at 35 eV for (S,S)-1 monolayer on Cu(110)
after annealing to 345 K.
Discussion of Enantiomer and Conglomerate Formation on
Cu(110) by 1
STM data of various domains showing submolecular resolu-
tion provide the basis for a local adsorption model of 1 in
which each carboxylate group is bonded to the surface at a
short-bridge site along the close packed [110] rows of the
Combining knowledge of the nature of the adsorbed species
and measuring the diameter and profile of molecules from the

Article
Crystal Growth & Design, Vol. 10, No. 10, 2010 4521
Figure 7. Comparison of experimental RAIRS data at room temperature and 370 K with calculated IR gas phase spectrum for an optimized
molecular structure of (R,R)-1.
Figure 8. A close-up STM image (Ubias = 0.29 V, IT = 0.85 nA) of racemic (R,R),(S,S)-1 deposited on Cu(110) at room temperature and
afterward annealed to 370 K with an overlaid grid showing the unit cell of the surface structure and the proposed adsorption geometry of the
molecule. The inset (bottom right) shows a profile of the lozenge part of the molecule along the Cu(001) direction.
metal, an adsorption geometry that is found to occur widely
for this functionality at a Cu(110) surface.16 Taking the mole-
cular asymmetry exhibited in the STM images and the possi-
ble molecular configurations into account, we propose the
“diagonal” adsorption geometry as indicated by the mole-
cular sketches in Figure 8 to be the most likely one, in which

4522 Crystal Growth & Design, Vol. 10, No. 10, 2010
Robin et al.
Figure 9. Model of molecular double-chain arrangement of (R,R)-1
on Cu(110) after thermal activation of phase change into chiral
organization by annealing to 370 K (above). Major directions ob-
served coincide with those of an STM image of (R,R)-1 (below).
Figure 10. Model of a molecular double-chain arrangement of
(R,R)-1 in the racemic monolayer showing the interruption of homo-
chiral domains by enantiomers (above), and an area of the STM
image (indicated in green) which reveals a similar area (below). See
also the cartoon in Figure 4d for a comparison of the defect structure.
the two carboxylate functionalities are adsorbed diagonally
across to occupy the short bridge-sites on adjacent [110] rows.
This molecular configuration forms the basis for proposing a
structural model describing the chiral assemblies.
surface.17 Within the double-molecule rows, the geometry
shown in Figure 9 allows the closest possible packing of the
molecule with the two benzene rings facing each other. A
closer packing would be possible were the carboxyl groups to
face each other, but this scenario would be highly unfavor-
able, again, because of the electrostatic repulsion between
the charged carboxylic groups. The juxtaposition of surface
bonded carboxylate functionalities at a surface has recently
been shown to be a driving factor in determining amino-acid
arrangements at surfaces18 and, it would appear, is also
important in this system.
The STM and LEED data for the enantiopure (R,R)-1 and
the local adsorption motif constructed above allows one to
construct an adsorption model for the enantiomer structure as
shown in Figure 9. This model is consistent with the formation
of enantiopure double chains aligned at an angle of -46ꢀ for
(R,R)-1 with respect to the Æ110æ direction. Within the double
chain, all molecules maintain the same adsorption sites (Figure 9);
however, by rotation the molecules prevent Coulomb repul-
sion that would arise from close contact of the charged
carboxylate groups. The same force is proposed to drive
the double chain formation by introducing a “spacer” area
between neighboring chains to counteract the Coulomb
repulsion and to accommodate the compressive stress that
is known to accompany carboxylate bonding on the Cu(110)
From the structural model, it can be seen that there is strong
expression of chirality in the supramolecular organization of
the monolayer, where the major growth direction deviates away
from the Cu <001> symmetry axis by a large angle of -46ꢀ.

Article
Crystal Growth & Design, Vol. 10, No. 10, 2010 4523
There is also, clearly, a direct transmission of chiral information
from the nanoscale molecule with its two chiral centers, to the
macroscale organization with the mirror (S,S)-1 enantiomer,
adopting the mirror organization forming double chains
aligned at an angle of þ46ꢀ with respect to the Æ110æ direction
(the mirror image to the structure in Figure 9).
with great inhomogeneities one enantiomer condenses at one
position and the mirror enantiomer at another place,19 as well
as the observation of the formation of homochiral sequences
of oligopeptides with lower than expected optical purity in
conglomerate-type Langmuir layers which might be explained
by oligomerization in disordered areas of the films which have
small domains.20 The occurrence of this kind of complex
behavior may explain large final enantiomeric excesses seen
in three-dimensional crystallization systems.
Turning now to the racemic mixture, where we were able to
distinguish the different enantiomers by submolecular STM
imaging, we can unambiguously observe the onset of con-
glomerate formation upon annealing. However, clearly the
process has been arrested and, within the temperature range in
which the molecule remains intact, it was impossible to drive
macroscopic conglomerate formation. Nevertheless, the pre-
ference for conglomerate nucleation is clearly evident, and
across the surface one can see the nucleation of enantiopure
areas. Remarkably, even at this very earliest onset of aggrega-
tion, the organized nuclei, containing literally only tens of
molecules, adopt a morphology that is remarkably similar to
that of large 2-D crystals of the enantiopure monolayer.
Thus, homochiral molecular chains are observed, adopting
a growth direction of 44ꢀ with respect to the Cu <001> axis.
In many cases, the nuclei are an almost identical mimic of the
enantiopure phase, and take up the double molecule chain
arrangement observed in the enantiopure chiral phase. Our
observations suggest that even at the earliest inception of
nucleation, way before any recognizable crystallite emerges,
themoleculararrangementstakeup the shapesand habitsthat
resemble those of the fully formed crystals.
The second striking observation from our large scale and
small scale STM data is that the conglomerate we image is
essentially “nipped in the bud”. Discrete organized homo-
chiral nuclei are seen, possessing a boundary that separates
them from their environment. Since crystal growth takes place
at the boundaries, any molecular impurity that blocks growth
atthe interface willact asan inhibitor (see model inFigure 10).
From our data, it would seem that the role of the inhibitor is
very well performed by the mirror enantiomer and that the
major obstacle to large scale conglomerate formation is the
mutual inhibition exerted by one enantiomer on the growth of
the mirror enantiomorph. Thus, each enantiomer is a highly
effective inhibitor for the mirror enantiomorph. This result
can be viewed as an extreme example of the “tailor-made”
additives pioneered by Lahav and co-workers,4,11,19 where
structurally similar molecules are found to be highly effective
suppressors of crystallization. Of perhaps even more general
interest is that our data support the hypothesis put forward by
Pratt-Brock et al.5 that cross-enantiomer inhibitions exact a
very high penalty in the kinetics of conglomerate nucleation
and growth, compared tothe racemic crystals, and this may be
an important factor in determining that less than 10% of
molecular crystals display conglomerate behavior.
The somewhat disordered situation (with very local order)
observed for the racemate contrasts dramatically with the
well-defined monolayers formed by the enantiopure com-
pounds for the same compound in which large domains are
formed. Therefore, it seems that, to quote Pratt-Brock,5 “The
presence of molecules of the “wrong” enantiomer will thus
inhibit formation of the nucleus and possibly also of the
growth of the crystal.” We postulate that in the racemic
monolayer the enantiomers act as “tailor-made” impurities
for each other. Thishypothesisgives weight tothe theories laid
downpreviously forchiral resolutionandcrystallization inthe
presence of chiral impurities,21 which in turn support our
experimental observations. The rate of diffusion on the sur-
face is also expected to play an important role in this system
where the adsorbate interacts strongly with the surface:
Indeed, Monte Carlo simulations have shown remarkably
similar organizational patterns in a racemic system22 to those
shown in Figure 3. In our case, conglomerate nanodomains
are observed, but the formation of racemic lattices possibly as
a result of kinetic hindering has also been postulated.23 In a
separate context, the effect of population differences in affect-
ing the rates of crystallization of opposite enantiomers has
recently been graphically illustrated by STM24 and underlines
the power of scanning probe techniques to map the earliest
stages of nucleation and crystallization. The power of the
STM experiments in probing these fundamental nanometer
scale nucleation and growth phenomena has also been shown
here. We believe that these observations and related ones in other
contexts will aid greatly in the understanding of the nucleation
and growth of racemic compounds and conglomerates.
Experimental Section
Synthesis and Characterization of (R,R)-2. Resorcinol (1.30 g,
11.8 mmol), (S)-methyl lactate (2.7 mL, 28.3 mmol), and triphenyl-
phosphine (7.42 g, 28.3 mmol) were dissolved in dry THF (50 mL)
with stirring under an atmosphere of argon, and the mixture was
cooled in an ice bath. To this mixture, a solution of diisopropyl-
azodicarboxylate (DIAD) (28.3 mmol) in dry THF (10 mL) was
added dropwise over a period of 30 min at 0 ꢀC, and the mixture was
allowed to warm up to room temperature with stirring overnight.
After addition of water (20 mL), THF was removed in a vacuum and
the residue was portioned between dichloromethane and water,
and the aqueous phases were extracted once more with dichloro-
methane. The combined organic phases were dried over sodium
sulfate, filtered, and stripped of solvent. The residue was subjected
to column chromatography (SiO2 EtOAc/hexane 1:4) to give
((R,R)-2 as a clear oil (1.62 g, 48%). M.F.: C14H18O6; M.W.: 282.29;
[R]546 = þ65.6 deg cm2 g-1 (c = 0.108 g/3 cm3), CH2Cl2); 1H
Concluding Remarks
There is a dramatic contrast between the racemic bulk
crystallization of racemic 1 and its adsorption onto Cu(110).
While the former is a racemic compound, the latter shows
signs of a conglomerate, but the very small domains are
curtailed in their growth by enantiomer molecules. Over the
whole racemic monolayer (rather than at the local level), one
would say that a quite chaotic behavior is observed; an
extremely complex system exists with minute areas of what
appear to be conglomerates. This reminds us of a description
of the onset of chirality in other contexts, where in samples
3
3
NMR (250 MHz, CDCl3): 7.11 (t, J = 8.2, 1H, ArH), 6.45 (dd, Ja =
8.2, Jb = 2.2, 2H, ArH), 6.38 (t, J = 2.2, 1H, ArH), 4.71 (q, J = 6.9,
1H, -OCHCH3CO-), 3.72 (s, 3H, -COOMe), 1.57 (d, J = 6.9,
-OCHCH3CO-) ppm. FT-IR (KBr): 2991 (m, -OCHCH3), 2955
(m, -OCHCH3), 1757 (s, COOCH3), 1738 (s, COOCH3), 1673 (w),
1604 (s, phenyl), 1687 (s), 1491 (s), 1454 (s), 1376 (s), 1283 (s), 1209
(s), 1181 (s), 1157 (s), 1135 (s), 1097 (s), 1053 (s), 977 (m), 835 (m),
768 (m), 685 (m) cm-1; Elemental Analysis (%) calculated: C 59.57,
H 6.43, found C 59.73, H 6.22.

4524 Crystal Growth & Design, Vol. 10, No. 10, 2010
Robin et al.
Synthesis and Characterization of (R,R)-1. To a solution of (R,R)-
2 (1.3 g, 4.61 mmol) in methanol (30 mL) 2 N NaOH (6.9 mL) was
added and the mixture was stirred at room temperature overnight.
The reaction mixture was concentrated, diluted with water (20 mL),
and acidified with concentrated HCl to pH 2. After three extractions
with EtOAc, the organic phases were dried over sodium sulfate,
filtered, and the solvent removed in a vacuum. 1.30 g of (R,R)-1 was
obtained as a white solid (94% yield).
STM images were recorded, at room temperature, in an ultra high
vacuum (UHV) chamber, with a base pressure of 2 ꢀ 10-10 mbar,
and fitted with a Specs Aarhus 150 STM operated in constant
current mode with an electrochemically etched tungsten tip. The
bias voltage was applied to the sample.
RAIRS and LEED experiments were performed in a multi-
technique UHV chamber interfaced to a Mattson 6020 FTIR
spectrometer via ancillary optics and KBr windows. A nitrogen
cooled HgCdTe detector allowed the IR spectral range of 650-4000
cm-1 to be accessed. The resolution of the spectrometer was set to 4
cm-1 and 250 scans were coadded. The spectrum of the clean sample
was taken as a background reference, R0, at the beginning of the
experiments. Spectra of the adsorbed layer are displayed as the ratio
(R - R0)/R0 with respect to the clean sample spectrum.
M.F.: C12H14O6; M.W.: 254.08; M.P.:130 ꢀC (lit. 132-140 ꢀC)25
[R]546 = þ38.4 deg cm2 g-1 (c = 0.096 g/3 cm3, CH2Cl2); 1H NMR
3
3
(250 MHz, DMSO-d6): 8.30 (s, OH), 7.14 (t, J = 8.2, 1H, ArH), 6.42
(dd, Ja = 8.2, Jb = 2.2, 2H, ArH), 6.33 (t, J = 2.2, 1H, ArH), 4.75
(q, J = 6.9, 1H, -OCHCH3CO-), 1.46 (d, J = 6.7, -OCHCH3CO-)
ppm; UV-vis (CHCl3) λmax/nm (ε/ mol L-1 cm-1): 206 (788), 273
(220); FT-IR (KBr): 3117 (s, OH), 2992 (m, -OCHCH3), 2942 (m,
-OCHCH3), 1715 (s, COOH), 1670 (m, COOH), 1612 (s, phenyl),
1586 (m, phenyl), 1495 (s), 1481 (s), 1453 (m), 1408 (s), 1372 (w),
1332 (m), 1289 (s), 1232 (s), 1169 (s), 1151 (s), 1087 (s), 1042 (s), 988
(m), 884 (m), 857 (m), 752 (m), 683 (m) cm-1; Elemental Analysis
(%) calculated C 56.69, H 5.55, found C 56.91, H 5.67.
Acknowledgment. The research leading to these results has
received funding from the European Community’s Seventh
Framework Programme under Grant Agreement No. NMP4-
SL-2008-214340, project RESOLVE, and the Marie Curie
Research Training Network CHEXTAN (MRTN-CT-2004-
512161), the EU Marie Curie Host Fellowship Program
TANSAS (MTKD-CT-2005-029689). We are also grateful
to the Generalitat de Catalunya (2009 SGR 158) for research
support for D.B.A., and the UK EPSRC for support for R.R..
We also thank the referees for constructive criticisms which
helped improve the manuscript.
Synthesis and Characterization of (S,S)-2. (S,S)-2 was obtained as
a clear oil following the same synthetic procedure used for (R,R)-2
(48% yield). M.F.: C14H18O6; M.W.: 282.11; [R]546= -73.3
1
deg cm2 g-1 (c = 0.129 g/3 cm3, CH2Cl2); H NMR (250 MHz,
3
3
CDCl3): 7.11 (t, J = 8.2, 1H, ArH), 6.45 (dd, Ja = 8.2, Jb = 2.2, 2H,
ArH), 6.38 (t, J = 2.2, 1H, ArH), 4.71 (q, J = 6.9, 2H,
-OCHCH3CO-), 3.72 (s, 6H, -COOMe), 1.57 (d, J = 6.9, 6H,
-OCHCH3CO-) ppm; UV-vis (CHCl3) λmax/nm (ε/mol L-1 cm-1):
215 (762), 273 (185); FT-IR (KBr): 2991 (m, -OCHCH3), 2955 (m,
-OCHCH3), 1757 (s, COOCH3), 1738 (s, COOCH3), 1673 (w), 1604
(s, phenyl), 1687 (s), 1491 (s), 1454 (s), 1376 (s), 1283 (s), 1209 (s),
1181 (s), 1157 (s), 1135 (s), 1097 (s), 1053 (s), 977 (m), 835 (m), 768
(m), 685 (m) cm-1; elemental analysis (%) calculated: C 59.57, H
6.43, found C 59.78, H 6.18.
Supporting Information Available: Crystallographic information
files; tables of crystallographic data for (R,R)-1 and racemate (R,R)-
1/(S,S)-1; figures of (R,R)-1 and (R,R)-1/(S,S)-1 crystal structures;
(R,R)-1 in the optimized gas phase structure; table of assignments
for IR active modes. This material is available free of charge via the
Internet at http://pubs.acs.org.
Synthesis and Characterization of (S,S)-1. (S,S)-1 was obtained as
white solid following the same synthetic procedure used for (R,R)-1
References
(94% yield). M.F.: C12H14O6; M.W.: 254.08; M.P.: 130 ꢀC; [R]546
=
(1) (a) Collins, A. N.; Sheldrake, G. N.; Crosby, J., Eds. Chirality in
Industry Vols I and II; J. Wiley & Sons: Chichester, 1992 and 1997.
(b) Nugent, W. A.; Rajanbabu, T. V.; Burk, M. J. Science 1993, 259,
479–483. (c) Ding, J.; Armstrong, D. W. Chirality 2005, 17, 281–
292. (d) Amabilino, D. B.; Veciana, J. Top. Curr. Chem. 2006, 265,
253–302. (e) Sancho, R.; Minguillon, C. Chem. Soc. Rev. 2009, 38,
797–805.
-36.6 deg cm2 g-1 (c = 0.091 M, CH2Cl2); 1H NMR (250 MHz,
3
3
DMSO-d6): 8.30 (s, 2H, OH), 7.14 (t, J = 8.2, 1H, ArH), 6.42 (dd,
Ja = 8.2, Jb = 2.2, 2H, ArH), 6.33 (t, J = 2.2, 1H, ArH), 4.75 (q,
J = 6.9, 2H, -OCHCH3CO-), 1.46 (d, J = 6.7, 6H, -OCHCH3CO-)
ppm; UV-vis (CHCl3) λmax/nm (ε/mol L-1 cm-1): 206 (778), 273
(235); FT-IR (KBr): 3117 (s, OH), 2992 (m, -OCHCH3), 2942 (m,
-OCHCH3), 1715 (s, COOH), 1670 (m, COOH), 1612 (s, phenyl),
1586 (m, phenyl), 1495 (s), 1481 (s), 1453 (m), 1408 (s), 1372 (w),
1332 (m), 1289 (s), 1232 (s), 1169 (s), 1151 (s), 1087 (s), 1042 (s), 988
(m), 884 (m), 857 (m), 752 (m), 683 (m) cm-1; Elemental Analysis
(%) calculated C 56.69, H 5.55, found C 56.94, H 5.67.
(2) (a) Jacques, J.; Collet, A.; Wilen, S. H. Enantiomers, Racemates and
ꢀ
Resolutions; Krieger Publishing Co.: Malabar, FL, 1994. (b) Perez-
García, L.; Amabilino, D. B. Chem. Soc. Rev. 2002, 31, 342–356.
ꢀ
(c) Perez-García, L.; Amabilino, D. B. Chem. Soc. Rev. 2007, 36,
941–967.
(3) (a) Collet, A.; Brienne, M.-J.; Jacques, J. Chem. Rev. 1980, 80, 215–
230. (b) Coquerel, G.; Amabilino, D. B. In Chirality at the Nanoscale;
Amabilino, D. B., Eds.; Wiley-VCH: New York, 2009.
(4) Addadi, L.; Berkovitch-Yellin, Z.; Weissbuch, I.; Lahav, M.;
Leiserowitz, L. Top. Stereochem. 1986, 16, 1–85.
IR Spectra of the Racemic Mixture (R,R)-1/(S,S)-1 in Their
Crystal Structure. FT-IR (HATR): 2998 (s, OH), 2943 (m,
-OCHCH3), 2644 (w), 2556 (w), 2465 (w), 1712 (s, COOH), 1599
(s, phenyl), 1583 (s, phenyl), 1493 (m), 1477 (m), 1453 (m), 1422(m),
1370 (w), 1326 (m), 1282 (s), 1233 (s), 1181 (s), 1157 s, 1135 (s), 1090
(5) Pratt-Brock, C.; Schweizer, W. B.; Dunitz, J. D. J. Am. Chem. Soc.
1991, 113, 9811–9820.
(s), 1046 (s), 993 (m), 918 (s), 853 (s), 762 (s), 686 (s) cm-1
.
(6) (a) Davey, R. J.; Allen, K.; Blagden, N.; Cross, W. I.; Lieberman,
H. F.; Quayle, M. J.; Righini, S.; Seton, L.; Tiddy, G. J. T.
CrystEngComm 2002, 257–264. (b) Erdemir, D.; Lee, A. Y.; Myerson,
A. S. Acc. Chem. Res. 2009, 42, 621–629. (c) Kellogg, R. M.; Kaptein,
B.; Vries, T. R. Top. Curr. Chem. 2007, 269, 159–198.
(7) (a) Raval, R. Chem. Soc. Rev. 2009, 38, 707–721. (b) Haq, S.;
Liu, N.; Humblot., V.; Jansen, A. P. J.; Raval, R. Nat. Chem. 2009,
1, 409–414.
(8) (a) Lopinski, G. P.; Moffatt, D. J.; Wayner, D. D. M.; Wolkow,
R. A. Nature 1998, 392, 909–911. (b) Bohringer, M.; Morgenstern,
K.; Schneider, W.-D.; Berndt, R. Angew. Chem., Int. Ed. 1999, 38,
821–823.
Calculation Details for IR Spectra of the Enantiopure (R,R)-1 in
Gas Phase. DFT calculations of (R,R)-1 in the gas phase were
3
˚
performed in a 20 ꢀ 15 ꢀ 15 A supercell. Structures were relaxed
until forces were less than 0.005 eV/atom. The plane wave basis has
been expanded up to a cutoff energy of 400 eV. It was found that the
symmetric structure with two internal hydrogen-bonds is by 60 meV
more stable than the asymmetric structure with only one internal
hydrogen-bond between the carbonyl and the ether group.
Details of Monolayer Studies of 1 on Cu(110). The clean Cu(110)
surface was prepared by argon ion sputtering at 500 eV followed by
annealing to 800 K. Low energy electron diffraction (LEED) was
utilized to check the cleanliness of the sample, with a sharp (1 ꢀ 1)
pattern characteristic of clean Cu(110). Enantiopure and racemic 1
were dosed from an electrically heated glass tube, separated from
the main chamber by a gate valve and differentially pumped by a
turbomolecular pump. Each sample of 1 was thoroughly outgassed
to ensure sample purity prior to dosing. In all experiments, the
Cu(110) crystal was held at room temperature during dosing.
(9) Elemans, J. A. A. W.; De Cat, I.; Xu, H.; De Feyter, S. Chem. Soc.
Rev. 2009, 38, 722–736.
(10) (a) Barlow, S. M.; Raval, R. Surf. Sci. Rep. 2003, 50, 201.
(b) Humblot, V.; Barlow, S. M.; Raval, R. Prog. Surf. Sci. 2004, 76,
1–19. (c) Ernst, K. H. Top. Curr. Chem. 2006, 265, 209–252.
(d) Ortega Lorenzo, M.; Baddeley, C. J.; Muryn, C.; Raval, R. Nature
2000, 404, 376–379. (e) Humblot, V.; Ortega Lorenzo, M.; Baddeley,
C. J.; Haq, S.; Raval, R. J. Am. Chem. Soc. 2004, 126, 6460–6469.

Article
Crystal Growth & Design, Vol. 10, No. 10, 2010 4525
(11) (a) Weissbuch, I.; Kuzmenko, I.; Berfeld, M.; Leiserowitz, L.;
Lahav, M. J. Phys. Org. Chem. 2000, 13, 426–434. (b) Weissbuch,
I.; Leiserowitz, L.; Lahav, M. Top. Curr. Chem. 2005, 259, 123–
165.
(12) (a) Mitsunobu, O. Synthesis 1981, 1–28. (b) Swamy, K. C. K.; Kumar,
N. N. B.; Balaraman, E.; Kumar, K. V. P. Chem. Rev. 2009, 109, 2551–
2651.
(13) Linares, M.; Iavicoli, P.; Psychogyiopoulou, K.; Beljonne, D.;
De Feyter, S.; Amabilino, D. B.; Lazzaroni, R. Langmuir 2008,
24, 9566–9574.
(16) (a) Rankin, R. B.; Sholl, D. S. Surf. Sci. 2004, 548, 301–308.
(b) James, J. N.; Sholl, D. S. Curr. Opin. Colloid Interface Sci.
2008, 13, 60–64. (c) Jones, G.; Jones, L. B.; Thibault-Starzyk, F.;
Seddon, E. A.; Raval, R.; Jenkins, S. J.; Held, G. Surf. Sci. 2006, 600,
1924–1935. (d) Sayago, D. I.; Polcik, M.; Nisbet, G.; Lamont, C. L. A.;
Woodruff, D. P. Surf. Sci. 2005, 590, 76–87. (e) Forster, M.; Dyer, M.;
Persson, M.; Raval, R. J. Am. Chem. Soc. 2009, 131, 10173–10181.
(17) Hermse, C. G. M.; Van Bavel, A. P.; Jansen, A. P. J.; Barbosa, L. A.
M. M.; Sautet, P.; Van Santen, R. A. J. Phys. Chem. B 2004, 108,
11035–11043.
(14) (a) Hayden, B. E.; Prince, K.; Woodruff, D. P.; Bradshaw, A. M.
Surf. Sci. 1983, 133, 589–604. (b) Woodruff, D. P.; McConville, C. F.;
Kilcoyne, A. L. D.; Lindner, T.; Somers, J.; Surman, M.; Paolucci, G.;
Bradshaw, A. M. Surf. Sci. 1988, 201, 228–244. (c) Bowker, M.; Haq,
S.; Holroyd, R.; Parlett, P. M.; Poulston, S.; Richardson, N. J. Chem.
Soc.-Faraday Trans. 1996, 92, 4683–4686. (d) Frederick, B. G.;
Leibsle, F. M.; Haq, S.; Richardson, N. V. Surf. Rev. Lett. 1996, 3,
1523–1546. (e) Ortega Lorenzo, M.; Haq, S.; Bertrams, T.; Murray, P.;
Raval, R.; Baddeley, C. J. J. Phys. Chem. B 1999, 103, 10661–10669.
(f) Ortega Lorenzo, M.; Baddeley, C. J.; Muryn, C.; Raval, R. Nature
2000, 404, 376–379. (g) Barlow, S. M.; Raval, R. Surf. Sci. Rep. 2003,
50, 201–341. (h) Humblot, V.; Ortega Lorenzo, M.; Baddeley, C. J.;
Haq, S.; Raval, R. J. Am. Chem. Soc. 2004, 126, 6460–6469. (i) Haq,
S.; Massey, A.; Moslemzadeh, N.; Robin, A.; Barlow, S. M.; Raval, R.
Langmuir 2007, 23, 10694–10700.
(18) Forster, M.; Dyer, M. S.; Persson, M.; Raval, R. Angew. Chem.,
Int. Ed. Engl. 2010, 49, 2344–2348.
(19) Plasson, R.; Kondepudi, D. K.; Asakura, K. J. Phys. Chem. B 2006,
110, 8481–8487.
(20) 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. 2002, 124, 9093–9104.
(21) (a) Weissbuch, I.; Lahav, M.; Leiserowitz, L. Cryst. Growth Des.
2003, 3, 125–150. (b) Kondepudi, D. K.; Crook, K. E. Cryst. Growth
Des. 2005, 5, 2173–2179.
(22) Unac, R. O.; Vidales, A. M.; Zgrablich, G. Ads. Sci. Technol. 2009,
27, 633–641.
(23) Parschau, M.; Fasel, R.; Ernst, K.-H. Cryst. Growth Des. 2008, 8,
1890–1896.
(24) Haq, S.; Liu, N.; Humblot, V.; Jansen, A. P. J.; Raval, R. Nat.
(15) Raval, R.; Parker, S. F.; Chesters, M. A. J. Chem. Soc. Faraday
Transact. 1996, 92, 2611–2614.
Chem. 2009, 1, 409–414.
(25) Burkard, U.; Effenberger, F. Chem. Ber. 1986, 119, 1594–1612.