A new design strategy for molecular recognition in heterogeneous systems: A universal crystal-face growth inhibitor for barium sulfate [26]

Full text of DOI:10.1021/ja990885i

J. Am. Chem. Soc. 2000, 122, 11557-11558
A New Design Strategy for Molecular Recognition in
11557
Heterogeneous Systems: A Universal Crystal-Face
Growth Inhibitor for Barium Sulfate
Peter V. Coveney,§ Roger Davey,‡ Jonathan L. W. Griffin,#
Yan He,† John D. Hamlin,# Stephen Stackhouse,§ and
Andrew Whiting*,#
Department of Chemistry and
Department of Chemical Engineering
UniVersity of Manchester Institute of Science and
Technology, P.O. Box 88, SackVille Street
Manchester, M60 1QD, U.K.
Centre for Computational Science
Queen Mary and Westfield College
London E1 4NS, U.K.
Department of Chemical Engineering
South West Petroleum Institute, Nachong
Sichuan 637001, Peoples Republic of China
ReceiVed March 18, 1999
ReVised Manuscript ReceiVed August 4, 2000
Controlling crystallization in both chemical production and
product formulations is well-known.1-4 As well as composition
and temperature, specific additives can also exert a powerful
influence on both crystal nucleation and growth rates. Current
design strategies of such additives are based on a molecular
recognition5 in which the additive binds selectively to a single
growing crystal surface. Herein, we report on a new conceptual
approach to heterogeneous recognition based on the rational
design of a molecule which can bind all growing faces of barium
sulfate, resulting in a highly active modifier of barium sulfate
crystal growth in practice.
Eight different faces are thought to be important in the growth
of barium sulfate crystals.6 Since the spacing between nearest-
neighbor sulfate-sulfate sites in each of these faces is different
(see Supporting Information), designing an additive which is
capable of inhibiting all crystal growth faces is a challenging
problem.2,7 The solution is to design a molecule which is capable
of recognizing and binding to all possible growing crystal faces.
This could be achieved in theory by either (a) designing a
molecule with a large number of binding motifs, so that at least
one or two of the groups coincide with binding sites on any
particular crystal surface (schematically shown in Figure 1),8 or
Figure 3. 1,7-Dioxo-4,10-diaza-12-crown-4-N,N′-dimethylenephospho-
nate 6 (in vacuo energy -4402.45141 eV) binding to each of the
crystallographic faces11 of barium sulfate indicated, together with relative
energies (∆E) for each conformation. Length scale is shown in parentheses
in each image. Each snapshot is captured after 10 ps of molecular
dynamics performed at 300 K, following energy minimization. Color
key: gray, carbon; white, hydrogen; red, oxygen; blue in modifier,
nitrogen; pink, phophorus; blue in crystal lattice, barium; yellow, sulfur.
Figure 1. Schematic of a single-face binding agent.
Figure 2. Schematic of a universal face-binding agent.
(b) designing a molecule with a small number of binding motifs,
but which can easily adopt several conformations (preferably of
similar energy) to enable binding to growing surfaces (Figure 2).
The disadvantages of (a) is that large molecules need to be made
and many of the binding motifs might be superfluous by not taking
part in binding. Therefore, the design of a flexible, universal face-
binding agent is more attractive, especially if the molecular weight
* Address correspondence to this author. Fax: (0161) 236-7677. Phone:
(0161) 200-4524. E-mail: a.whiting@umist.ac.uk.
§ Queen Mary and Westfield College.
(4) Coveney, P. V.; Davey, R. J.; Griffin, J. L. W.; Whiting, A. J. Chem.
Soc., Chem. Commun. 1998, 1476-1468.
(5) Weissbuch, I.; Popoviz-Biro, R.; Leiserowitz, L.; Lahav, M. The Lock
and Key Principal; Behr, J. P., Ed.; John Wiley and Sons: Chichester, 1994;
Chapter 6.
(6) Hartman, P.; Perdock, W. G. Acta Crystallogr. 1955, 8, 524-529.
(7) Davey, R. J.; Black, S. N.; Bromley, L. A.; Cottier, D.; Dobbs, B.;
Rout, J. E. Nature 1991, 353, 540-541.
(8) (a) Davey, R. J.; Black, S. N.; Bromley, L. A.; Cottier, D.; Dobbs, B.;
Rout, J. E. J. Chem. Soc., Faraday Trans. 1991, 87, 3409-3414. (b) Bromley,
L. A.; Cottier, D.; Davey, R. J.; Dobbs, B.; Smith, S.; Heywood, B. R.
Langmuir 1993, 9, 3594-3599.
‡ Department of Chemical Engineering, University of Manchester Institute
of Science and Technology.
# Department of Chemistry, University of Manchester Institute of Science
and Technology.
† South West Petroleum Institute.
(1) Davey, R. J. Manuf. Chem. 1990, NoV., 25-29.
(2) Davey, R. J.; Black, S. N.; Godwin, A. D.; Mackerron, D.; Maginn, S.
J.; Miller, E. J. J. Mater. Chem. 1997, 7, 237-241.
(3) Coveney, P. V.; Humphries, W. J. Chem. Soc., Faraday Trans. 1996,
92, 831-841.
10.1021/ja990885i CCC: $19.00 © 2000 American Chemical Society
Published on Web 11/05/2000

11558 J. Am. Chem. Soc., Vol. 122, No. 46, 2000
Communications to the Editor
and number of binding functions can be minimized. It is inherently
more difficult to achieve this goal because it necessarily requires
a more detailed knowledge of the conformational flexibility of
the moleculesan exacting test of predictive computational
molecular design.
The design of this new class of flexible crystal growth inhibitors
started with a search for candidate molecular structures. The basic
structural fragment considered was the poly-aminomethylphos-
phonate system, since such molecules (e.g. 1 and 2) are barium
sulfate scale inhibitors due to replacement of some of the sulfate
sites in the growing crystal by their phosphonate groups.7 Taking
up a theme exemplified by earlier work on the inhibition of
crystallization of ettringite,3,4,9 we examined a range of related
macrocyclic aminomethylphosphonates 3-6. This earlier work
Figure 4. SEM images showing the effect of additive 6 on BaSO4
morphology: (a) native BaSO4; (b) BaSO4 crystallization in the presence
of 0.048 mM 6; (c) BaSO4 crystallization in the presence of 0.096 mM
6. All crystallization were carried out under identical conditions;
precipitates were formed at 70 °C by mixing solutions of BaCl2 and
Na2SO4 according to literature methods.7,8
Scheme 1
The consequence of this should be that the well faceted {001}
rhombs found in pure solutions will be replaced by crystals in
which the growth rates in all directions will be inhibited and
anisotropy of the growth disappears. Such crystals are expected
to be increasingly spherical as the concentration of additive 6 is
increased. Figure 4 shows the comparison. In Figure 4a rhombic
plates grown with additive can be seen; Figure 4b corresponds
to 0.048 mM of 6 with the rhombic outline still evident but with
facets largely absent and a partially spherical morphology; Figure
4c at a loading of 0.096 mM shows crystals which as expected
are virtually spherical and no evidence of faceting remains. These
data are totally consistent with our expectations and fulfill our
success criterion. We note that the spheres produced here are quite
clearly the result of a gradual loss of faceting of single crystals,
in marked contrast to the spheres which are formed from high
loadings of 1 and 2 which are known to be ordered aggregates of
nanocrystals.15 The reduction in sizes with additive concentration
suggests that the crystallization process is inhibited with nucleation
becoming the dominant process.
In summary, we have designed macrocycle 6 which, according
to computer simulations, should recognize and bind to all the
important crystal growth faces of barium sulfate. Subsequent
crystallization experiments clearly show that all faces are modified
by the formation of spherical single crystals, thus showing the
viability of the novel design strategy inherent (Figure 2), i.e., the
application of a universal crystal-face blocking agent. Such an
approach should prove invaluable as a general protocol for the
design of not only new crystal growth inhibitors, but of novel
heterogeneous systems and materials.
showed that molecular modeling can be applied to the rational a
priori design of novel compounds, which acted as efficient
inhibitors of crystalline ettringite formation.9,10
Energy minimization and molecular dynamics simulations at
300 K11 of the binding of molecules 1-6 (among others) on the
different surfaces of barium sulfate revealed that most of these
molecules (1-6 and related structures)2 showed good recognition
for at least one or two faces.2 However, no molecules showed
complementary binding (i.e. all phosphonate groups were capable
of binding into vacant sulfate sites in the lattice) to three or more
faces, the outstanding exception being macrocycle 6. Compound
6 was capable of recognizing and binding to all eight crystal faces
of the barium sulfate lattice without attendant formation of
high-energy, unfavorable conformations (i.e. all conformations
accessible at 300 K). This achievement stems from the fact that
molecule 6 is sufficiently flexible to be able to adopt a wide range
of energetically similar conformations, wherein the relative
separation of the phosphonate groups adjusts to match sulfate-
sulfate distances on each different crystal face. The results of these
simulations are displayed in Figure 3.
Following the prediction that macrocycle 6 should be a highly
efficient barium sulfate binding agent, the synthesis and evaluation
of 6 was undertaken. The synthesis was accomplished using
literature-related methods.12,13 (Scheme 1) and N-phosphono-
alkylation.14 The effect of macrocycle 6 on the growth of barium
sulfate crystals was then studied by performing crystallization
experiments in the presence of varying concentrations of 6.8
To assess the success of our design methods, we prepared
crystals of barium sulfate8 both with and without additive 6, to
compare any morphological changes with those reported7,8 for
linear additives (e.g. 1 and 2). Our criterion for success was that
whereas additives of type 1 and 2 act anistropically, yielding
morphologies in which faces in the [001] zone are selectively
inhibited,7,8 additive 6 is expected to act isotropically, simulta-
neously modifying the growth faces in all three orthogonal zones.
Acknowledgment. We are grateful to the Engineering and Physical
Sciences Research Council (U.K.) and Schlumberger Cambridge Research
for a Total Technology studentship award (to J.G.) (GR94007291).
Supporting Information Available: Experimental procedures and
data (PDF). This material is available free of charge via the Internet at
http://pubs.acs.org.
(9) Billingham, J.; Coveney, P. V. J. Chem. Soc., Faraday Trans. 1993,
89, 3021-3028.
JA990885I
(10) Wattis, J. A. D.; Coveney, P. V. J. Chem. Phys. 1997, 106, 9122-9140.
(11) Molecular modeling3 was carried out using CERIUS2.a Geometry
optimizations were carried out via MOPAC,b using the PM3 semiempirical
method as described in the Supporting Information. (a) CERIUS2 molecular
modeling software, Molecular Simulations Inc., V1.6.2 (1995). (b) Stewart,
J. J. P. MOPAC 6.0, QCPE No. 445, 1990. (c) Rappe, A.; Casewit, C. J.;
Colwell, K. S.; Goddard, W. A., III; Skiff, W. M. J. Am. Chem. Soc. 1992,
114, 10024-10035.
(12) Bogatsky, A. V.; Lukyanenko, N. V.; Basok, S. S.; Ostrovskaya, L.
K. Synthesis 1984, 2, 138-138.
(13) Borjesson, L.; Welch, C. J. Acta Chem. Scand. 1991, 45, 621-626.
(14) Moedritzer, K.; Irani, R. R. J. Org. Chem. 1966, 31, 1603-1607.
(15) Benton, W. J.; Collins, I. R.; Grimsey, I. M.; Parkinson, G. M.; Rodger,
S. A. Faraday Discuss. 1993, 95, 281-297.