The role of short-range diffusion in solvent-assisted mechanochemical synthesis of metal complexes

Article abstract of DOI:10.1039/b810659m

The role of short-range diffusion in solvent-assisted mechanochemical synthesis is demonstrated in studies of a polymorphic transition and a ligand dissociation reaction involving copper(i) thiocyanate complexes. The Royal Society of Chemistry.

Full text of DOI:10.1039/b810659m

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/dalton | Dalton Transactions
The role of short-range diffusion in solvent-assisted mechanochemical
synthesis of metal complexes†
Graham A. Bowmaker,*a John V. Hanna,b Robert D. Hart,c Brian W. Skeltonc and Allan H. Whitec
Received 23rd June 2008, Accepted 4th August 2008
First published as an Advance Article on the web 19th August 2008
DOI: 10.1039/b810659m
The role of short-range diffusion in solvent-assisted
mechanochemical synthesis is demonstrated in studies of a
polymorphic transition and a ligand dissociation reaction
involving copper(I) thiocyanate complexes.
As an example of a polymorphic transition we have used
the complex [CuSCN(PPh3)2]2. The structure of this compound
as crystallized from dichloromethane solution shows that it
is a centrosymmetric doubly thiocyanate-bridged dimer.6 We
have found that crystallization of this compound from pyridine
produces a different polymorph in which the molecular structure
is closely similar to the one previously determined,6 but is not
centrosymmetric. The X-ray crystal structure of this polymorph
is shown in Fig. 1.‡ The IR spectrum of this compound in the
n(CN) region (Fig. 2(a)) shows two bands at 2082, 2104 cm-1,
consistent with the presence of two inequivalent thiocyanate
groups. Grinding of the dry crystals produced no change in
the IR spectrum. However, addition of a very small amount of
acetonitrile, just sufficient to allow the formation of a paste during
grinding, followed by a few minutes standing of the product
in air to allow evaporative removal of the acetonitrile from the
resulting paste of product, resulted in the complete conversion
of the compound to the symmetrical polymorph. The IR spectra
recorded after three such treatments carried out successively on the
same sample show that this did not occur to any perceptible extent
with one such treatment (Fig. 2(b)), but that the conversion to the
centrosymmetric polymorph was complete after three treatments,
resulting in a single n(CN) band at 2095 cm-1 (Fig. 2(d)).§
There is considerable current interest in the subject of solvent-
free synthetic chemistry.1–3 The main basis for this interest is
the potential reduction in environmental contamination and the
increased convenience associated with the elimination of solvent
from synthetic reactions.1,2 One method of achieving this is by
grinding the solid reactants together (mechanochemistry), and
this has shown promise as a solvent-free synthetic method for
the preparation of a variety of metal complexes.1,2 The possible
benefit of including a minor amount of solvent in such reactions
has been reported,4 and we have recently shown that this can
result in a dramatic increase in the rate of formation of product
in the reaction of metal compounds with ligands to form metal
coordination compounds.5
Little is known about the mechanisms of mechanochemical
reactions. In the case of a reaction involving two or more reactants,
it seems obvious that grinding the reactants together produces a
mixture in which the reactants are in more intimate contact, hence
facilitating reaction between them. However, this does not shed
any light on the detailed mechanisms of the reactions involved.
In order to examine this question further, we have examined the
mechanochemistry of two kinds of reaction in which there is
only one reactant, so that improved intimacy of contact between
different reactants by grinding is not a factor. The two kinds
of reaction that we have studied are polymorphic transitions
and ligand dissociation reactions. In the first of these a single
compound is converted to a different polymorph of the same
compound while in the second a single compound dissociates into
two different compounds. The systems chosen for study involve
complexes of copper(I) thiocyanate with triphenylphosphine and
pyridine. The use of thiocyanate as a ligand provides a useful probe
to follow the reactions via infrared spectroscopy, by monitoring
changes in the n(CN) vibrations.
aDepartment of Chemistry, University of Auckland, Private Bag 92019,
Auckland, New Zealand. E-mail: ga.bowmaker@auckland.ac.nz; Fax: +64
9 3737422; Tel: +64 9 3737599 Ext. 88340
Fig. 1 Structure of the unsymmetrical polymorph (monoclinic P21/c)
of [CuSCN(PPh3)2]2. Cu(1)–P(11,12),S(1),N(2,1¢) are 2.2665, 2.2569(5),
2.4389(7), 2.018(2), 1.99(1); Cu(2)–P(21,22),S(2,1¢),N(1) are 2.2714,
bANSTO NMR Facility, Institute of Materials Engineering, Lucas Heights
Research Laboratories, Private Mail Bag 1, Menai 2234, New South Wales,
Australia
˚
2.2636(5), 2.4746(5), 2.561(4), 1.945(2) A.
cSchool of Biomedical, Biomolecular and Chemical Sciences, Chemistry
M313, The University of Western Australia, Crawley 6009, Western
Australia, Australia
† CCDC reference number 642534. For crystallographic data in CIF or
other electronic format see DOI: 10.1039/b810659m
As an example of a ligand dissociation reaction we have used
the complex [CuSCN(PPh3)(py)2] (py = pyridine). This is a yellow
compound formed in the reaction of CuSCN with PPh3 in a 1 : 1
5290 | Dalton Trans., 2008, 5290–5292
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
Fig. 3 IR spectra in the CN stretching region of [CuSCN(PPh3)(py)2]
(a) untreated, and (b), (c) after successive treatments with acetonitrile and
grinding.
Fig. 2 IR spectra in the CN stretching region of the unsymmetrical
polymorph (monoclinic P21/c) of [CuSCN(PPh3)2]2 (a) untreated, and
(b)–(d) after successive treatments with acetonitrile and grinding.
mole ratio in pyridine, and is a four-coordinate mononuclear
complex.7 It can also be prepared from a 1 : 2 vol/vol mixture
of pyridine and acetonitrile, and washing of the product with
acetonitrile causes no change in the composition of the compound.
However, addition of a very small amount of acetonitrile, just
sufficient to allow the formation of a paste during grinding,
followed by a few minutes standing of the product in air to
allow evaporative removal of acetonitrile and pyridine from the
resulting paste of product, results in the complete conversion of the
compound to the colourless complex [CuSCN(PPh3)(py)] with loss
of 1 mole of pyridine.§ This was confirmed by IR spectroscopy, and
by gravimetric and elemental analysis. The IR spectra recorded
after two such treatments carried out successively on the same
sample (Fig. 3) show that most of the conversion occurs after
the first treatment, and that it is complete after the second
treatment. The 31P CP MAS NMR spectrum of the final product
(Fig. 4(b)) shows that it is a pure, homogeneous phase with a single
copper environment. An attempt to prepare the same compound
by recrystallization of [CuSCN(PPh3)(py)2] from hot acetonitrile
◦
was unsuccessful, and heating of [CuSCN(PPh3)(py)2] to 120 C
resulted in the loss of both pyridine molecules and the formation of
a mixture of symmetrical [CuSCN(PPh3)2]2 and CuSCN, identified
by IR spectroscopy. Thus it appears that the solvent-assisted
mechanochemical reaction readily yields a product that is not
easily obtainable by other synthetic methods, a point that we have
already noted in a previous study.5
It is clear from the results in Fig. 2 and 3 that both reactant and
product phases are present in the reaction mixture at intermediate
stages of the reaction, so material has to physically move from
the reactant phase to the product phase during the reaction. We
propose that the reactions involve dissolution of the reactant in
the solvent and subsequent precipitation of the product within
the voids between the particles of reactant. The two possible
mechanisms of mass transport from the reactant to the product
phase are convection and diffusion. However, convection requires
movement of the liquid relative to the solid phase, and this is
Fig. 4 31P CPMAS NMR spectrum of [CuSCN(PPh3)(py)2] (a) after dry
grinding, and (b) after grinding with acetonitrile.
minimized by the very small amount of solvent used in our
experiments. Also, the movement of solvent relative to a solid
is zero at the solid/liquid interface, so that mass transport across
the interface can only occur by diffusion. This clearly focuses
attention on the importance of molecular diffusion in the reaction
mechanism.
Over distances of the order of centimetres, typical of the size
of reaction vessels used in conventional solution-based chemical
synthesis, diffusion is slow and quite ineffective relative to convec-
tion, hence the need for stirring to dissolve substances, for example.
However, in mechanochemical synthesis, the diffusion takes place
within the inter-particle voids, whose size is comparable to the
particle size when the amount of solvent is sufficiently small to
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 5290–5292 | 5291
©

View Article Online
form a paste. Assuming a typical value for the diffusion coefficient
of a molecule in solution (D = 0.5 ¥ 10-9 m2 s-1), and a mean
diffusion path length of the order of the particle size (say d =
0.1 mm), the diffusion time (t = d2/2D) is 10 s. This is clearly
compatible with the short reaction times that we have observed in
our present and previous studies.5
The factors that determine the rate of mass transport by
diffusion, and hence the reaction rate, are as follows:
(1) Diffusion medium. Addition of a small amount of solvent in
which the reactant is soluble provides a clearly identifiable medium
in which diffusion can take place. The transport rate will increase
with increasing solubility of the reactant in the solvent.
solvent, i.e. they are actually solvent-assisted reactions. A case
in point is the reaction between nickel(II) nitrate hexahydrate and
phenanthroline;1,2,11 this reaction produces water as a by-product,
and this can act in the same way as added solvent. It is interesting
to note that the reported very short reaction time (2 min) is similar
to the times that we have observed in our solvent-assisted reactions.
The above arguments clearly indicate the need for a clearer focus
on the role of molecular diffusion in mechanochemical reactions.
This allows an understanding of observations that are otherwise
difficult to comprehend, and suggests ways in which the scope of
mechanochemistry can be considerably extended.
(2) Surface area. Since diffusion takes place from the surface
of the solid, an increase in surface area results in an increase
in transport rate. Reduction of the particle size d (by grinding)
increases the solid surface area, and it can easily be shown that
the surface area of a given amount of a powdered substance is
proportional to 1/d.
Acknowledgements
We thank AINSE for financial support (Grants No. 02190, 03011).
Notes and references
(3) Diffusion path length. The rate of diffusion is inversely
proportional to the square of the diffusion path length, so that
a reduction in the diffusion path length results in a large increase
in diffusion rate. The diffusion path length is reduced by using the
minimum amount of solvent and by reducing the particle size (by
grinding). The kinetics of diffusion are such that the diffusion rate
is proportional to the inverse square of the diffusion path length8
and hence, according to the above, to 1/d2.
‡ Crystal data: C74H60Cu2N2P4S2, M = 1292.3. Monoclinic, P21/◦c,
˚
a = 23.6170(4), b = 13.5021(2), c = 20.1922(3) A, b = 105.453(2) ,
3
˚
V = 6206 A , T ca. 100 K. 26401 unique CCD reflections (Rint = 0.058;
14188 > 2s(I)), R1 = 0.041, wR2 = 0.103. A disordered component was
resolved for SCN(1), major, minor site occupancies refining to 0.828(2)
and complement; no disorder was discernible elsewhere in the structure.
The phase is isomorphous with its azide counterpart.12
§ About 0.1 mmol of compound was ground using an agate mortar and
pestle of the type normally used to prepare samples for IR spectroscopy.
For the solvent-assisted reactions 1 drop of solvent was added using a small
pipette. The solvent was removed after grinding for about two minutes
by allowing the mixture to stand for a few minutes in air at ambient
temperature. No reaction was observed if large crystals of the reactant
were treated with solvent without grinding. IR spectra were recorded on
dry powders using a Perkin Elmer Spectrum 100 FT-IR spectrometer
equipped with a Universal ATR sampling accessory. The 31P CPMAS
NMR data were acquired at 9.4 T on a Bruker MSL-400 spectrometer
operating at a 31P frequency of 161.92 MHz.
Combining factors (2) and (3) leads to a very strong (1/d3)
dependence of transport rate on paticle size, which provides an
explanation for the dramatic increase in conversion with successive
periods of grinding that is suggested by the IR data in Fig. 2(b)–
(d). The dramatic increase in reaction rate that we observe upon
addition of a very small amount of a suitable solvent and reducing
the reactant particle size by grinding is clearly consistent with the
key role played by diffusion in the reaction mechanism.
1 A. L. Garay, A. Pichon and S. L. James, Chem. Soc. Rev., 2007, 36,
846.
Further consideration of factor number (1) above raises the
question of the relevance of the diffusion mechanism in “dry”
(solvent-free) mechanochemical reactions. Diffusion can take
place in solids as well as in fluids, but the diffusion rate under
normal conditions is negligible. The diffusion coefficient can
however be increased by increasing the temperature, and an
increase in temperature can be caused by frictional heating of
the solid during high energy mechanical milling in a ball mill. This
has been suggested before,9 and provides a possible explanation for
the role of frictional heating in dry mechanochemical synthesis,
where a number of mechanisms involving mass transport have
been suggested in cases where the heating is insufficient to cause
melting.1,2,10
2 Chemistry World, 2007, vol. 4, issue 7 (July), p. 35.
3 G. W. V. Cave, C. L. Raston and J. L. Scott, Chem. Commun., 2001,
2159.
4 N. Shan, F. Toda and W. Jones, Chem. Commun., 2002, 2372.
5 G. A. Bowmaker, N. Chaichit, C. Pakawatchai, B. W. Skelton and A. H.
White, Dalton Trans., 2008, 2926.
6 N. K. Homsy, M. Noltemeyer, H. W. Roesky, H.-G. Schmidt and G. M.
Sheldrick, Inorg. Chim. Acta, 1984, 90, L59.
7 C. Pettinari, C. di Nicola, F. Marchetti, R. Pettinari, B. W. Skelton, N.
Somers, A. H. White, W. T. Robinson, M. R. Chierotti, R. Gobetto and
C. Nervi, Eur. J. Inorg. Chem., 2008, 1974.
8 P. W. Atkins and J. de Paula, Atkins’ Physical Chemistry, Oxford
University Press, Oxford, 8th edn, 2006, pp. 772–782.
9 A. Trovarelli, F. Zamar, J. Llorca, C. De Leitenburg, G. Dolcetti and
J. T. Kiss, J. Catal., 1997, 169, 490.
10 V. P. Balema, J. W. Wiench, M. Pruski and V. K. Pecharsky, Chem.
Commun., 2002, 1606.
It is also possible that enhancement of diffusion by frictional
heating is not necessary in some of the previously reported solvent-
free mechanochemical reactions, because the enhancement may
be caused by the presence of small amounts of adventitious
11 P. J. Nichols, C. L. Raston and J. W. Steed, Chem. Commun., 2001,
1062.
12 R. F. Ziolo, A. P. Gaughan, Z. Dori, C. G. Pierpont and R. Eisenberg,
Inorg. Chem., 1971, 10, 1289.
5292 | Dalton Trans., 2008, 5290–5292
This journal is
The Royal Society of Chemistry 2008
©