Crystallization of calcite spherules around designer nuclei

Article abstract of DOI:10.1002/(SICI)1521-3773(19981116)37:21<3044::AID-ANIE3044>3.0.CO;2-0

Gold colloids modified with p-sulfanylphenol self-assembled monolayers provide novel nanoscopic seeds for the nucleation of calcium and strontium carbonate from aqueous solution. The morphologies of the crystalline precipitates are very different from wha

Full text of DOI:10.1002/(SICI)1521-3773(19981116)37:21<3044::AID-ANIE3044>3.0.CO;2-0

COMMUNICATIONS
and whose solubility in different solvents can be controlled
through the size and functionality of the thiol molecules.[12, 13]
Thiol-coated gold colloids can be purified by chromatography,
evaporated, recondensed, precipitated, and redissolved; this
is behavior that one usually associates with molecules.[12, 14, 15]
Recently, we have been able to functionalize gold colloids
with a Ru-containing polymerization catalyst to obtain a
material that bridges homogeneous and heterogeneous cata-
lytic behavior.[16] When coated with p-sulfanylphenol, the
colloids are soluble in water at a pH value of about 12 or
higher.[13] While solutions of these colloids are homogeneous
in every sense, the colloids possess a surface capable of
heterogeneous nucleation. These thiol-coated gold colloids
can therefore be used to seed the crystallization of materials
from saturated solutions. The different topologies of flat and
spherical templates is to be noted. The use of gold colloids
instead of flat gold surfaces should permit novel crystal or
crystal-aggregate geometries to be obtained through templat-
ed crystallization.
Crystallization of Calcite Spherules around
Designer Nuclei**
Jörg Küther, Ram Seshadri, and Wolfgang Tremel*
Dedicated to Professor Bernt Krebs
on the occasion of his 60th birthday
Living systems exercise considerable control over the
crystallization of certain inorganic minerals.[1] This control
extends to the nature and habit of the mineral polymorph and
is usually guided towards a specific function. Sometimes this is
achieved in seeming contradiction to thermodynamic expect-
ation. An additional level of complexity is reached when
living systems form composite structures that organize
minerals in organic matrices.[2] The utility of organic ±
inorganic interfaces in mimicking these biomineralization
processes have been exploited in the past. Langmuir mono-
layers,[3] protein-coated substrates,[4] polymer dispersions,[5]
micellar media,[6] and self-assembled monolayers (SAMs)[7±9]
have been used as templates or substrates for the crystalliza-
tion of minerals such as calcium carbonate. We demonstrate
here that standard SAM chemistry can be extended to
produce controlled nanoscopic surfaces for the heterogeneous
nucleation of inorganic minerals from aqueous solution. The
seeds are about 5 nm large gold colloids coated with a p-
sulfanylphenol SAM. An interesting aspect here is that while
the nucleation is heterogeneous in the sense that it takes place
at the liquid ± solid interface, it commences from within a
single homogeneous phase.
We have been interested in the crystallization of the
carbonates of calcium and, to a lesser extent, strontium on
self-assembled monolayers of long chain thiols, organized in
two dimensions on gold-coated glass substrates.[8±11] Inspired
in part by the kinds of processes taking place in living systems,
we have examined how different substituents at the positions
of the thiols can influence the nature and morphology of the
calcium carbonate polymorph (calcite, aragonite, or vaterite)
formed. In certain cases, we have also been able to establish a
relationship between the specific 2D structure of the SAM
and that of the growing crystal.[9] These studies complement
existing work and have provided some insight into the
templated crystallization of carbonate biominerals. Further-
more, they suggest mild routes to useful materials.
Fritz et al.[17] have studied the formation of flat pearls on
inorganic substrates inserted between the shell and mantle of
the red abalone. They have determined that oriented calcite
and template protein layers are initially formed. This is
followed by the growth of stacks of aragonite along the
pseudo-sixfold axis of the crystal. The proteins are generated
within epithelial cells in the abalone mantle. Our approach is
in contrast to this. We do not attempt to simulate natural
pearls, but rather seek inspiration from the manner in which
they are formed. The templating organic substance is the p-
sulfanylphenol SAM. This is previously fixed to the substrate
on which the crystallization is carried out rather than being in
solution, which is the case in the formation of natural pearls.
The top right of Figure 1 shows schematically the thiol-
coated particles used to nucleate the precipitation of calcium
1
carbonate. At different seed concentrations (10 ± 100 mgL )
and temperatures (4 ± 458C) the only polymorph we observe
is calcite; a sample X-ray diffraction pattern is displayed in
Figure 1.[18] The effect of the colloid seeds on the crystalliza-
tion is observed in the scanning electron micrographs.
Figure 2a shows rhombohedral crystals of calcite formed
when crystallization was carried out at 228C in the absence of
the colloid seed. This is the expected morphology. The
1
presence of the colloids (here 100 mgL ) initially gives the
solution a deep pink color. By the end of the crystallization,
the solution is colorless and the scanning electron micro-
graphs of the grey-black precipitate display a completely
different morphology (Figure 2b). At higher magnification
(Figure 2c), a typical crystallite assembly shows the individual
spherules of Figure 2b to comprise a complex assembly of the
calcite rhombohedra. The direction radiating out from the
center is the direction connecting opposite edges of the calcite
rhombohedra, that is, the crystallographic [001] direction.
Crystallization experiments carried out under identical
conditions on flat p-sulfanylphenol SAM/gold surfaces (as
opposed to colloidal seeds) do not suggest any sort of epitaxy
between substrate and crystal. The crystallization on the flat
surfaces is inhibited and results in rather ill-defined morpho-
logies, perhaps because the phenolic OH groups are ionized at
the high pH used. This does not seem to affect the efficiency
During the course of these studies, gold colloids attracted
our attention. Their small size (in the range of 2 ± 10 nm for
the methods presented here) ensures that for a given quantity
of gold a relatively large surface area is obtained. The self-
assembly of thiols on gold colloids has in fact provided routes
to preparing particles that are stabilized against aggregation
[*] Prof. Dr. W. Tremel, Dr. J. Küther, Dr. R. Seshadri
Institut für Anorganische Chemie
und Analytische Chemie der Universität
Johann-Joachim-Becher-Weg 24, D-55099 Mainz (Germany)
Fax : (‡ 49)6131-39-3922
E-mail: tremel@indigotrem1.chemie.uni-mainz.de
[**] We thank Degussa (Hanau, Germany) for a generous gift of gold and
Dr. Gabriele Nelles for preliminary AFM measurements. Financial
support from the Fonds der chemischen Industrie is gratefully
acknowledged.
3044
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
1433-7851/98/3721-3044 $ 17.50+.50/0
Angew. Chem. Int. Ed. 1998, 37, No. 21

COMMUNICATIONS
Figure 1. Measured and Rietveld-fitted X-ray diffraction profiles (trans-
mission, q ± 2q) of CaCO3 crystallized from solutions containing thiol-
coated gold colloid seeds. The fit indicates pure calcite; the small hump
marked by an asterisk is due to the (111) reflection of gold. The
1
concentration of seeds used was 100 mgL in solution, and the crystal-
lization was carried out at 228C.
of the seeding, however. The organization of the calcite
rhombohedra therefore arises from crystal ± crystal interac-
tions rather than crystal ± substrate interactions, and from the
fact that the crystals are obliged to pack around a sphere.[19]
The manner in which the solution becomes clear after
precipitation as well as the morphologies obtained in the
presence of the colloid suggest that the colloids act as nuclei or
seeds. We point out that the assemblies display a level of
complexity that cannot be achieved with flat templates. The
assemblies of the calcite rhombohedra are rather fragile. So
far, attempts to fragment the spherules in order to better
understand their structures have always resulted in the
individual crystallites being damaged. It is possible that
well-defined rhombohedra are found only near the outer
surfaces of the spherules, and that within the spherules the
morphology is quite different.
Figure 2. a) Scanning electron micrograph of the crystals formed by
precipitation of CaCO3 from solution at 228C. The rhombohedra are
crystals of calcite displaying {104} faces. b) Scanning electron micrograph of
the crystals formed on crystallization in the presence of the seeds
1
(100 mgL at 228C).
A typical crystallite assembly seen at higher
magnification is shown in c); the rhombohedra arrange spherically around
some central point. The scale bars correspond to 50, 500, and 20 mm,
respectively.
Figure 3a shows the precipitates obtained when the crys-
tallization is carried out at about 48C instead of at 228C and
1
fewer seed particles are used (30 mgL ). The morphology is
still spheroidal, but at higher magnification (Figure 3b) it is
seen to comprise spherulites rather than spherules, that is,
needlelike crystals that radiate from a central point. Despite
the changed morphology, X-ray diffraction experiments
confirm that the samples are pure calcite.
The precipitates formed at 228C, unlike those formed at
about 48C, comprise crystals with well-defined morphologies.
Correspondingly, the X-ray diffraction profiles possess nar-
rower linewidths. This is expected from considerations of the
supersaturation ratios at the two temperatures. However, the
dependence of the size of the precipitates on the amount of
seed colloids used does not follow expectation. This could be
because the distribution of sizes is not monomodal, as
suggested from a closer analysis of the micrographs. Addi-
tionally, the number of seeds used are far in excess of the
spherules that are formed. It is possible that some the seeds
are incorporated into the spherules during the later stages of
the crystallization. The spherelike nature of a large proportion
Angew. Chem. Int. Ed. 1998, 37, No. 21
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
1433-7851/98/3721-3045 $ 17.50+.50/0
3045

COMMUNICATIONS
Figure 4. Scanning electron micrograph of SrCO3 crystals precipitated
1
from solutions containing the colloid seeds (30 mgL at 48C). The scale
bar corresponds to 10 mm. The inset shows a needle (ca. 15 mm long) of
SrCO3 in the strontianite modification obtained from precipitation in the
absence of colloids. The needle grows along a pseudohexagonal [001]
trilling axis.
hence, perfectly wetted. Classical theory predicts that they
should act very much like homogeneous nuclei.[20] This is not
observed in the micrographs. The examples outlined here
testify to the ease with which nanoscopic seeds can be
designed and used for controlled crystallizations.
Experimental Section
Figure 3. Scanning electron micrographs of precipitates obtained from
crystallizations (30 mgL 1) at about 48C. The spherelike morphology is
retained (a), while at higher magnifications (b) differences become visible.
The individual crystallites making up the assemblies are no longer
rhombohedral, but fluffy and branched. Some bilobed structures are also
observed. The scale bars correspond to 200 and 10 mm, respectively.
The gold colloids (diameters near 5 nm) were prepared in two-phase
toluene ± water systems by reported procedures.[11, 12] The organic phase
was isolated and then mixed with a solution of p-sulfanylphenol in toluene,
causing the colloids to precipitate. Surface plasmon spectroscopic measure-
ments confirm that p-sulfanylphenol forms densely packed monolayers on
gold surfaces. Attempts to obtain the 2D structure of the p-sulfanylphenol
SAMs (on gold/mica) through atomic force microscopy failed due to the
high acidity of phenolic OH group. Also the small size of the p-
sulfanylphenol molecule might prevent its forming crystalline monolayers.
The precipitate was washed and dried before being dissolved in 10 mm
solutions of calcium or strontium chloride, whose initial pH value was
raised to 12 ± 13 with aqueous NaOH. The rather alkaline pH are necessary
in order to take the colloids into homogeneous solution.[13] Precipitation of
the carbonate was induced in closed desiccators through the slow diffusion
of CO2 from solid (NH4)2CO3 placed within the desiccator. After 48 h the
precipitates, which were collected on glass slides, were removed and fixed
to aluminum stubs using conducting carbon glue. Before transferring to the
microscope, gold was sputtered on to the glass slides.
of the products also indicates that the spherules fall out of
solution onto the witness glass slides only after nearly all the
calcium in solution is depleted, as there is little or no growth
of the assemblies after they settle.
The high initial pH value at which the crystallizations are
carried out seems to favor the formation of calcite at the
expense of aragonite. To simulate the formation of aragonite
spherules, we crystallized SrCO3, whose stable modification
strontianite has the same structure. The different crystal
structure (orthorhombic Pmcn for strontianite/aragonite,
Å
rhombohedral R3c for calcite) usually results in the phases
Received: May 7, 1998 [Z11828IE]
German version: Angew. Chem. 1998, 110, 3196 ± 3199
precipitating from solution with completely different mor-
phologies.[9] An attempt to make strontianite spherules is
shown in Figure 4. The usual strontianite (or aragonite)
morphology is needlelike. The use of the colloid seeds seems
to induce the needles to cluster around a central point. This
additionally demonstrates the utility of our designer nuclei.
Self-assembled monolayers of thiols on gold (or other)
substrates can induce the controlled crystallization of inor-
ganic materials, whereas gold colloids provide a novel way of
looking at gold surfaces. The marriage of these concepts yields
a method for controlled heterogeneous nucleation with the
result that the crystallite assemblies so formed are very
different from what one obtains (and what one can possibly
obtain) with flat templates. The seeds are in solution and,
Keywords: biomineralization ´ calcium ´ colloids ´ crystal
growth ´ monolayers
[1] On Biomineralisation (Eds.: H. A. Lowenstam, S. Weiner), Oxford
University Press, New York, 1989.
[2] S. Mann, J. Mater. Chem. 1995, 5, 935 ± 946.
[3] B. R. Heywood, S. Mann, Adv. Mater. 1994, 6, 9 ± 20.
[4] L. Addadi, S. Weiner, Angew. Chem. 1992, 104, 159 ± 176; Angew.
Chem. Int. Ed. Engl. 1992, 31, 153 ± 170.
[5] P. A. Bianconi, J. Lin, A. R. Strzelecki, Nature 1991, 349, 315 ± 317.
[6] D. Walsh, S. Mann, Adv. Mater. 1997, 9, 658 ± 662.
[7] D. D. Archibald, S. B. Quadri, B. P. Gaber, Langmuir 1996, 12, 538 ±
546.
3046
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
1433-7851/98/3721-3046 $ 17.50+.50/0
Angew. Chem. Int. Ed. 1998, 37, No. 21

COMMUNICATIONS
[8] J. Küther, R. Seshadri, W. Knoll, W. Tremel, J. Mater. Chem. 1998, 8,
641 ± 650; J. Küther, W. Tremel, Thin Solid Films, in press.
[9] J. Küther, G. Nelles, R. Seshadri, M. Schaub, H.-J. Butt, W. Tremel,
Chem. Eur. J. 1998, 4, 1834 ± 1842.
[10] J. Küther, W. Tremel, Chem. Commun. 1997, 2029 ± 2030.
[11] J. Küther, R. Seshadri, G. Nelles, H.-J. Butt, W. Knoll, W. Tremel, Adv.
Mater. 1998, 10, 401 ± 404.
[12] M. Brust, M. Walker, D. Bethell, D. J. Schiffrin, R. Whymann, J.
Chem. Soc. Chem. Commun. 1994, 801 ± 802.
[13] M. Brust, J. Fink, D. Bethell, D. J. Schiffrin, C. Kiely, J. Chem. Soc.
Chem. Commun. 1995, 1655 ± 1656.
Scheme 1. Palladium complex catalyzed allylic substitution of cyclic allyl
acetates.
[14] R. L. Whetten, J. T. Khoury, M. M. Alvarez, S. Murthy, I. Vezmar,
Z. L. Wang, P. W. Stephens, C. L. Cleveland, W. D. Luedtke, U.
Landmann, Adv. Mater. 1996, 8, 428 ± 433.
[15] A. L. Badia, S. Singh, L. Demers, L. Cuccia, G. R. Brown, R. B.
Lennox, Chem. Eur. J. 1996, 2, 359 ± 363.
low enantioselectivities with these ligands. Because of the
favorable properties of phosphanyldihydrooxazoles, such as
ease of synthesis and high turnover numbers in catalysis, it was
of interest to develop ligands of this type that are suitable for
cyclic substrates. We found such compounds on the basis of
mechanistic considerations.
Key concepts of the reactions with dihydrooxazoles and
other P,N ligands[5] are a) the assumption of preferred attack
of the nucleophile at the carbon atom of the allyl group in the
position trans to the phosphorus atom and b) the postulate
that of the diastereomeric exo- and endo-p-allyl complexes A
(Scheme 2), the exo isomer reacts faster. As the relative
[16] M. Bartz, J. Küther, R. Seshadri, W. Tremel, Angew. Chem. 1998, 110,
2658 ± 2661; Angew. Chem. Int. Ed. 1998, 37, 2466 ± 2468.
[17] M. Fritz, A. M. Belcher, M. Radmacher, D. A. Walters, P. K. Hansma,
G. D. Stucky, D. E. Morse, S. Mann, Nature 1994, 371, 49 ± 51.
[18] When the crystallizations are performed on p-sulfanylphenol SAMs
on gold/glass substrates rather than on colloids, no modification other
than calcite is observed. However, when the pH value is lowered to
about 10.5, the aragonite:calcite ratio of crystals deposited on p-
sulfanylphenol SAMs on gold/glass is about 40:60 at 228C and 85:15 at
458C, as determined from the Rietveld refinement of powder X-ray
profiles. These are the highest aragonite weight fractions that we have
been able to obtain at 228C on SAMs of simple thiols. We are now
attempting to prepare colloids coated with thiols that permit
dissolution of the colloids in water at neutral pH values in the hope
of inducing aragonite.
[19] On the other hand, the use of (flat) SAM/gold templates with a well-
defined hexagonal 2D crystal structure, such as SAMs of hexadeca-
nethiol, result in calcite crystallizing with the [001] direction nearly
always perpendicular to the substrate, suggesting the role of epitaxy.[9]
[20] B. Mutaftschiev, Handbook of Crystal Growth, Vol. 1a (Ed.: D. T. J.
Hurle), North-Holland, Amsterdam, 1993, pp. 187 ± 247.
Enantioselective Allylic Substitution of
Cyclic Substrates by Catalysis with Palladium
Complexes of P,N-Chelate Ligands with a
Cymantrene Unit**
Scheme 2. The exo ± endo isomerism of p-allylpalladium complexes with
phosphanyldihydrooxazole ligands (R' ˆ iPr, tBu).
Steffen Kudis and Günter Helmchen*
reactivities of these diastereomers are nearly equal, the ratio
of their concentrations mainly determines the enantioselec-
tivity of the substitution. The reason for the low selectivity
with cyclic substrates is insufficient differentiation, that is,
equal populations of exo and endo isomers.
With second-generation ligands, the enantioselectivity for
cyclic substrates was increased significantly by substituting the
pseudoaxial phenyl group by a 2-biphenylyl group;[6] ee values
of 50 ± 80% were obtained. The higher enantioselectivity had
been anticipated, since the crystal structure of complex B with
R ˆ iPr and n ˆ 6 displayed only the conformer in which the
terminal phenyl group is directly above the allyl system. This
conformer is the most favorable for achieving high selectivity.
According to 1H NMR NOE measurements,[7] conformer B is
in equilibrium with conformer C, in which the 2-biphenylyl
group is rotated away from the palladium center and does not
interact with the allyl group. To restrict rotational freedom of
Palladium complex catalyzed asymmetric C C bond form-
ing reactions with allylic compounds are being investigated
with great intensity.[1] As ligands, modular C2 diphosphanes[2]
and phosphanylcarboxylic acids[3] gave good results, partic-
ularly with small cyclic substrates. For acyclic substrates, chiral
phosphanyldihydrooxazoles were successfully used as li-
gands,[4] and enantioselectivities of up to more than 99:1 were
achieved. However, cyclic substrates 1 (Scheme 1) furnished
[*] Prof. Dr. G. Helmchen, Dipl.-Chem. S. Kudis
Organisch-chemisches Institut der Universität
Im Neuenheimer Feld 270, D-69120 Heidelberg (Germany)
Fax: (‡49)6221-54-4205
E-mail: en4@ix.urz.uni-heidelberg.de
[**] This work was supported by the Deutsche Forschungsgemeinschaft
(SFB 247) and the Fonds der Chemischen Industrie. We thank Dr. F.
Rominger and Dr. B. Nuber for crystal structure analyses.
Angew. Chem. Int. Ed. 1998, 37, No. 21
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
1433-7851/98/3721-3047 $ 17.50+.50/0
3047