A structural investigation of anion-triazole interactions: Observation of "π-pockets" and "jr-Sandwiches"

Article abstract of DOI:10.1002/ejic.200801260

Eight mononuclear nickel(II) complexes of the ligands 4-amino-3,5-di(2- pyridyl)-1,2,4-triazole (adpt) and 4-pyrrolyl-3,5-di(2-pyridyl)-l,2,4-triazole ipldpti with the anions ClO₄, BF₄, PF₆ and SbF₆ have been prepared. In all cases the metal/ligand ratio is 1:3, and the complexes are of the form [NiL₃](A)₂-soiven(s where L = adpt or pldpt and A = one of the aforementioned anions. Five of these complexes have been structurally characterized by X-ray crystallography: four of these contain pldpt and strong anion-π interactions are observed, with two motifs present in all four structures. One of the anions occupies a "π-pocket" formed by two coordinated triazole rings and one coordinated pyridine ring. The other anion only interacts with one triazole ring, which is involved in the pocket around the first anion, such that the triazole ring is "sandwiched" by two anions. Surprisingly, in all four of these complexes, the two triazole centroid...anion distances in the anion-triazole-anion interactions [2.917(7)-3.005(10) A] are significantly shorter than in any of the other types of triazole-anion interactions [3.164(5)-3.456(9) A].

Full text of DOI:10.1002/ejic.200801260

FULL PAPER
DOI: 10.1002/ejic.200801260
A Structural Investigation of Anion–Triazole Interactions: Observation of
“π-Pockets” and “π-Sandwiches”
Nicholas G. White,[a] Jonathan A. Kitchen,[a] and Sally Brooker*[a]
Keywords: Coordination compounds / Pi interactions / Supramolecular chemistry / Triazoles / N ligands
Eight mononuclear nickel(II) complexes of the ligands 4-
amino-3,5-di(2-pyridyl)-1,2,4-triazole (adpt) and 4-pyrrolyl-
3,5-di(2-pyridyl)-1,2,4-triazole (pldpt) with the anions ClO4,
BF4, PF6 and SbF6 have been prepared. In all cases the
metal/ligand ratio is 1:3, and the complexes are of the form
[NiL3](A)2·solvents where L = adpt or pldpt and A = one of
the aforementioned anions. Five of these complexes have
been structurally characterized by X-ray crystallography:
four of these contain pldpt and strong anion–π interactions
are observed, with two motifs present in all four structures.
One of the anions occupies a “π-pocket” formed by two coor-
dinated triazole rings and one coordinated pyridine ring. The
other anion only interacts with one triazole ring, which is in-
volved in the pocket around the first anion, such that the
triazole ring is “sandwiched” by two anions. Surprisingly, in
all four of these complexes, the two triazole centroid···anion
distances in the anion–triazole–anion interactions [2.917(7)–
3.005(10) Å] are significantly shorter than in any of the other
types of triazole–anion interactions [3.164(5)–3.456(9) Å].
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2009)
interactions will surely be realized in the near future,
as anions play a crucial role in nearly all biological sys-
tems.[13]
Introduction
Since the theoretical description of anion–π (strictly
anion–π*) interactions in 2002[1] numerous theoretical stud-
ies have shown the energetic favourability of contacts be-
tween anions and electron-poor aromatic rings.[2–4] Unam-
biguous crystallographic evidence of anion–π interactions
was provided in 2004,[5,6] and a number of such interactions
have since been observed in the solid state, these interac-
tions often leading to the formation of unusual and interest-
ing supramolecular architectures. Indeed, a recent review of
the Cambridge Structural Database (CSD)[7] shows that
there are approximately 3500 structures that contain close
contacts between six-membered aromatic rings and anions
Ϫ the vast majority of which had not been noted by the
original authors.[8]
Our interest in anion–π systems stems from our research
into spin-crossover (SCO) systems. It is well-established that
enhancing the cooperativity between metal centres in SCO
complexes increases the likelihood of a hysteretic crossover
event,[14,15] a desirable property for technological applica-
tions of SCO.[16] This metal–metal cooperativity may be es-
tablished through covalent bonds i.e. by synthesising oli-
gonuclear or polymeric systems,[15,17] or through non-coval-
ent interactions. Previous work has attempted to use hydro-
gen-bonding[18] or π–π[19] stacking in this manner, but to
the best of our knowledge, no work has investigated the
use of anion–π interactions to induce cooperativity in SCO
systems.[20]
Dunbar et al. have shown that changing anion size and
shape varies the nuclearity of metallacyclophanes, with
these structures being templated by anion–tetrazine interac-
tions.[9] Other authors have used anion–π interactions to
generate structures as diverse as molecular baskets,[5] carou-
sels,[6] and 3D nets.[10] Research has also investigated poten-
tial applications of these interactions, for example, for anion
recognition and binding,[11] and as synthetic trans-
membrane anion channels.[12] Further applications of these
While much research has looked into pyridazine–, tri-
azine– and tetrazine–anion interactions, we have been un-
able to find any reports of triazole–anion interactions, or
indeed of any five-membered heterocycle–anion interac-
tions. Similarly, given how frequently anion–π interactions
occur,[8] it is surprising that, to the best of our knowledge,
no systematic crystallographic investigations of this phe-
nomenon have been conducted; i.e., to date no series of
closely related structures containing anion–π interactions
has been prepared, characterized and analysed. Herein, we
present such a study, of a family of mononuclear nickel(II)
complexes of the ligands L = adpt and pldpt with the anions
[a] Department of Chemistry and MacDiarmid Institute for Ad-
vanced Materials and Nanotechnology, University of Otago,
P. O. Box 56, Dunedin, New Zealand
–
–
–
–
Fax: +64-3-479-7906
E-mail: sbrooker@chemistry.otago.ac.nz
Supporting information for this article is available on the
WWW under http://www.eurjic.org or from the author.
A = BF4 , ClO4 , PF6 , and SbF6 [adpt = 4-amino-3,5-
di(2-pyridyl)-1,2,4-triazole, pldpt = 4-pyrrolyl-3,5-di(2-pyr-
idyl)-1,2,4-triazole, Figure 1]. A detailed investigation of the
1172
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2009, 1172–1180

A Structural Investigation of Anion–Triazole Interactions
interactions between the anions and solvent molecules, and (1BF4)and[Ni(pldpt)3](BF4)2,(2BF4)},843 cm–1{[Ni(adpt)3]-
the triazole and pyridine rings, in the resulting [NiL3]- (PF6)2, (1PF6) and [Ni(pldpt)3](PF6)2, (2PF6)} and 660–
(A)2·solvents complexes is provided.
661 cm–1 {[Ni(adpt)3](SbF6)2, (1SbF6) and [Ni(pldpt)3]-
(SbF6)2, (2SbF6)} were observed, consistent with the respec-
tive non-coordinated anion.
UV/Vis Spectroscopy
Goodwin and colleagues had some success in rationaliz-
ing the magnetic behaviour of [FeIILЈ3]A2 complexes, where
LЈ is a 3-(2-pyridyl) five-membered heterocycle ligand, on
the basis of the UV/Vis spectra of the analogous nickel(II)
complexes.[24,25] Given that we have established a general
synthetic route to 4-substituted 3,5-di(2-pyridyl)-1,2,4-tri-
azole Rdpt ligands,[26] we attempted to distinguish between
the ligand fields imposed by these ligands using UV/Vis
spectroscopy. Hence, we recorded both the solution and so-
lid-state UV/Vis spectra of the nickel(II) complexes 1BF4
and 2BF4Ϫthe ligands adpt and pldpt having triazole N4-
substituents that differ markedly in their electron-donat-
ing/-withdrawing properties (to give a general feel for this
difference the Hammett constants are: NH2, para: –0.66,
ortho: –0.16; pyrrolyl, σ1 = 0.17[27])Ϫin the expectation that
this would inform the subsequent iron(II) chemistry of
these ligands.
Figure 1. Ligands used in this study.
Results and Discussion
Synthesis
The ligand adpt was prepared in two steps following the
procedure of Geldard and Lions:[21] 2-cyanopyridine was
condensed with hydrazine hydrate to give 3,6-di(2-pyridyl)-
1,2,4,5-dihydrotetrazine, which rearranged to give the de-
sired ligand when boiled in aqueous hydrochloric acid
(Scheme S1). A modification of Mandal’s[22] method was
then used to convert adpt into pldpt (Scheme S1); these au-
thors report using 2,5-dimethoxy-2,5-dihydrofuran in a di-
oxane/acetic acid mixed solvent system. We have previously
used Mandal’s procedure to prepare pldpt, except that we
used 2,5-dimethoxytetrahydrofuran in place of the dihy-
drofuran.[23a] While these procedures yield the desired prod-
uct in high yield, the work-up is reasonably time consum-
ing, as a dark by-product is produced, which must be sepa-
rated from pldpt chromatographically. Furthermore, this
method is not particularly robust and reliableϪin some in-
stances, we have found that it yields only the dark by-prod-
uct.
Unfortunately, as shown in Figure S1, the UV/Vis peaks
were too broad to distinguish between the small differences
in 10Dq caused by this remote ligand modification. Both
3
3
complexes show A2g Ǟ T1g transitions at approximately
3
3
19,000 cm–1 and A2g Ǟ T2g transitions at 11,600 cm–1.
X-ray Crystallography
Single crystals of all four pldpt complexes, 2ClO4·
To avoid these problems, we now use ethanol as solvent, CH3CN·0.7(C2H5)2O·0.3H2O, 2BF4·CH3CN·0.5(C2H5)2O,
and add only a small amount of aqueous HCl to initiate 2PF6·CH3CN·(C2H5)2O and 2SbF6·CH3CN·(C2H5)2O, and
the reaction (the reaction does not proceed if no acid is of one adpt complex, 1SbF6·1.5CH3CN·0.58(C2H5)2O, were
added) instead of the 50% by volume acetic acid used in grown by the vapour diffusion of diethyl ether into the ace-
Mandal’s synthesis.[22] This prevents the formation of the tonitrile reaction solution. All complexes contain a central
dark by-product and gives the desired product in slightly nickel(II) ion binding three ligands through just one of the
higher yield than that previously reported (83%, compared two potential bidentate binding pockets per ligand (Fig-
with the 80[22] and 77%[23a] previously achieved). Further- ures 2, 3, 4, and 5; Tables S1–S3 in the Supporting Infor-
more, the synthetic work-up is significantly quicker and eas- mation). The coordination of these bidentate pockets, each
ier, as simply neutralizing the reaction mixture with triethyl- comprising a pyridine nitrogen and an N1 of a triazole ring,
amine and reducing it in volume causes analytically pure gives the nickel(II) ions N6 distorted octahedral geometries
pldpt to crystallize from solution, without the need for (Table S2), with all complexes adopting the meridional con-
chromatography or recrystallization.
figuration. No crystallographically-imposed symmetry is
The family of eight [NiL3](A)2·solvents complexes was present within any of the cations. Two crystallographically
readily formed by stoichiometric reactions carried out in independent non-coordinated counterions per nickel(II) are
CH3CN, and isolated as pink-purple crystalline solids by also present in each structure (Figures 2, 3, 4, 5; S2–S6),
vapour diffusion of diethyl ether into the concentrated reac- along with some solvent molecules. Some structures con-
tion solution.
Elemental analysis, ESI mass spectrometry fragmenta- mation.
tion patterns, and IR spectra of all eight complexes were The sole structure containing adpt contains two Ni-
tained disorder and this is detailed in the Supporting Infor-
2+
consistent with proposed structures. Notably, strong peaks (adpt)3 cations and four SbF6 counter anions, as well as
at 1088–1089 cm–1 {[Ni(adpt)3](ClO4)2, (1ClO4) and some solvent molecules, in the asymmetric unit, and crys-
[Ni(pldpt)3](ClO4)2, (2ClO4)}, 1057 cm–1 {[Ni(adpt)3](BF4)2, tallizes in the triclinic space group P1.
¯
Eur. J. Inorg. Chem. 2009, 1172–1180
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
1173

N. G. White, J. A. Kitchen, S. Brooker
FULL PAPER
Table 1. Comparison of axis cell lengths, unit-cell volume and
anion volume for the structures, 2BF4·CH3CN·0.5(C2H5)2O,
2ClO4·CH3CN·0.7(C2H5)2O·0.3H2O, 2PF6·CH3CN·(C2H5)2O and
2SbF6·CH3CN·(C2H5)2O.
2BF4
16.9235(16) 16.9779(17)
23.175(2) 23.178(2)
15.3132(15) 15.4992(16)
2ClO4
2PF6
2SbF6
a [Å]
b [Å]
c [Å]
V [Å3]
17.4949(11) 17.729(3)
23.3993(14) 23.708(3)
15.8469(9) 16.175(3)
5886.7(6)
54
5534.0(9)
5589.2(9)
47
6131.3(16)
63
Anion volume [Å3][28] 38
As expected, the nickel–nitrogen bond lengths and angles
vary little between the five structures (Table S2). Notably
the nickel–triazole (trz) bond lengths are significantly
shorter than nickel–pyridine (py) bonds in all cases [Ni–trz
2.035(5)–2.078(3) Å, Ni–py 2.100(5)–2.149(5) Å, Table S2].
Unsurprisingly, cis-N–Ni–N bond angles are close to 90°
[largest deviation: 12.70(19)°, mean deviation: 6.6°, Table
S2] and trans-N–Ni–N bond angles are close to 180°
[largest deviation: 18.1(3)°, mean deviation: 12.5°, Table
S2], so that the nickel ions show only a small distortion
from an ideal octahedral geometry.
Figure 2. Perspective view of one of the two crystallographically
independent cations of 1SbF6·1.5CH3CN·0.58(C2H5)2O (hydrogen
atoms omitted for clarity). See Figures S5 and S6 for the structure
of the other cation, and an OFIT diagram of the two cations.
The situation regarding twisting of the individual ligands
away from planarity is more complicated. In the sole adpt-
containing structure (Figure 2, S5 and S6) there are small
twists, of 1–13°, between the coordinated pyridine ring and
triazole ring (Table S3). A somewhat larger twist, of 11–
32°, is seen between the triazole and non-coordinated pyr-
idine ring, but the extent of the twist remains modest as
each of the six non-coordinated pyridine ring nitrogen
atoms hydrogen-bonds to a proton on the amino nitrogen
of the same adpt ligand.
The four pldpt-containing complexes (Figures 3, 4, and
5, S2–4) all show two different types of ligand twist: each
cation contains two ligand strands that have a very small
twist between the coordinated pyridine and triazole rings
(Ͻ 7°), with a much larger twist seen between the non-coor-
dinated pyridine and triazole rings [38.9(5)–50.78(16)°,
Table S3]. However, the third ligand strand behaves dif-
ferently: twists of ca. 14–15° are observed between both the
coordinated pyridine and triazole rings, and between the
non-coordinated pyridine and triazole rings. That is, there
is no significant difference between the triazole–pyridine
mean plane intersects for the coordinated vs. non-coordi-
nated pyridine rings. Both of the twists are in the same di-
rection, so that the two pyridine rings are at angles of ca.
Figure 3. Perspective view of 2BF4·CH3CN·(C2H5)2O (solvent
molecules, hydrogen atoms and non-coordinated pyridine ring twirl
disorder omitted for clarity). The structures of the cations of
2ClO4·CH3CN·0.7(C2H5)2O·0.3H2O, 2PF6·CH3CN·(C2H5)2O and
2SbF6·CH3CN·(C2H5)2O (Figures S2–S4) are very similar.
The four structures containing pldpt all crystallize in the 28–30° to one another. In all four complexes, this ligand
monoclinic space group P21/n and have just one Ni- strand is the only one that contains a coordinated pyridine–
(pldpt)32+ cation and two counter anions, along with solvent anion interaction (see later), so presumably the larger than
molecules, in the asymmetric unit. These pldpt-containing expected coordinated pyridine–triazole twist facilitates a
complexes are structurally very similar to one another in more favourable geometry for this interaction. The non-co-
the solid state. The unit-cell dimensions show only a slight ordinated pyridine on this ligand strand is reasonably close
increase in unit-cell size and axis lengths with increasing to a coordinated pyridine on one side and a region of sol-
anion size (Tables 1, and S1 in the Supporting Infor- vent (partial or full occupancy diethyl ether/water de-
mation). In all four pldpt complexes there is one acetonitrile pending on the structure): it appears that this small tri-
solvate and one further region of solvent (although the ex- azoleϪnon-coordinated pyridine twist arises to prevent un-
act nature and occupancy of the latter site varies between favourable steric repulsions with these groups. In all four
structures) per cation.
pldpt complexes there is a large twist, 80.4(4)–85.3(4)°
1174
www.eurjic.org
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2009, 1172–1180

A Structural Investigation of Anion–Triazole Interactions
(Table S3), around the triazole–pyrrole N–N bond, similar actions with ClO4 anions (trz centroid···X 3.044 Å,[29] trz
to previously observed mean plane intersects for complexes centroid···X 3.138 Å),[30] and we have recently observed ex-
of pldpt [82.4(3)–88.8(1)[22,23a] and for the free ligand in the tremely strong [trz centroid···X 2.790(5) Å] interactions in
solid state [81.9(1) and 82.4(1)].[23a]
the iron(II) analogue of 1BF4, i.e. [FeII(adpt)3](BF4)2·
The only structure to contain significant hydrogen-bond- 0.5(C2H5)2O.[31]
ing was also the only structurally characterized complex of
In all four structures containing pldpt, one of the two
adpt, aSbF6·1.5CH3CN·0.58(C2H5)2O. It contains hydrogen anions in the asymmetric unit interacts with only one aro-
bonds from all twelve amino protons, to SbF6 fluorine matic ring (Figures 4 and 5, Figures S7 and S8). In all cases,
atoms, non-coordinated adpt pyridine nitrogen atoms, a di- this is a triazole ring, and the interaction is strong [trz
ethyl ether oxygen atom and an acetonitrile nitrogen atom centroid···X 2.917(7)–3.005(10) Å, X = O for 2ClO4, X = F
[N–H···A 2.827(8)–3.483(9) Å, angle N–H···A 123.6– for 2BF4, 2PF6 and 2SbF6]. This does not include the minor
170.3°, Table S9]. Interestingly, this structure contained position of the disordered PF6 in 2PF6. For full details of
very little in the way of anion–π interactions, in contrast to all anion–π interactions (see Tables S4–S8). The other anion
all other structures, of pldpt, which showed numerous and sits in a “π-pocket” provided by two triazole rings and one
strong anion–π contacts (see later).
coordinated pyridine ring. One of these triazole rings is the
same one that interacts with the other anion, so it is “sand-
wiched” by these two anions. This second anion–π interac-
tion is also strong [trz centroid···X 2.956(7)–3.005(4) Å], de-
spite the fact that overall this is an anion–π-anion interac-
tion. The interactions with the other triazole and the pyr-
idine of the π-pocket are weaker and vary slightly for the
tetrahedral anions compared to the octahedral anions. The
tetrahedral anions show moderate anion–π interactions
with both the pyridine and triazole rings [py centroid···X
3.202(7), 3.265(5) Å; trz centroid···X 3.164(5), 3.224(4) Å,
for 2ClO4 and 2BF4, respectively]. The octahedral anions
show a moderate interaction with a pyridine ring, coupled
with a much weaker interaction to this same pyridine from
Anion–π Interactions
The two complexes present in the asymmetric unit of the
sole adpt complex only make two anion–π contacts that are
shorter than 3.5 Å. Both are quite long [F···centroid
3.304(7) and 3.492(9) Å], and may well be due to crystal
packing forces rather than genuine anion–π interactions (so
these are not considered further in the anion–π discussion
below). This result quite surprising as other structures con-
taining the ligand adpt have shown moderate anion–π inter-
Figure 4. Diagram showing anion–π and CH3CN–π interactions
present in 2ClO4·CH3CN·0.7(C2H5)2O·0.3H2O. These include both
π-pocket and π-sandwich interactions; the interactions observed in
2BF4·CH3CN·(C2H5)2O (Figure S7) are very similar. Of the three
crystallographically unique triazole rings, the one indicated by an
arrow is not involved in any anion/solvent···π interactions. All num-
bers indicate X···centroid distances. Hollow bonds indicate ligand
strands that have been removed for clarity.
Figure 5. Diagram showing anion–π and CH3CN–π interactions
present in 2SbF6·CH3CN·(C2H5)2O. Of the three crystallographi-
cally unique triazole rings, the one indicated by an arrow is not
involved in any anion/solvent···π interactions. All numbers indicate
X···centroid distances. Hollow bonds indicate ligand strands
that have been removed for clarity. The arrangement in
2PF6·CH3CN·(C2H5)2O is very similar (Figure S8).
Eur. J. Inorg. Chem. 2009, 1172–1180
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
1175

N. G. White, J. A. Kitchen, S. Brooker
FULL PAPER
a different fluorine on the same anion, as well as weak in- rings. Interestingly, the four anions that interact with only
teractions between the anion and the other triazole of the one triazole ring (one anion in each of the four pldpt-con-
π-pocket [2PF6: py centroid···F 3.185(4), 3.433(4) Å; trz taining structures), and as such have more freedom to attain
centroid···F 3.454(13) Å, 2SbF6; py centroid···F 3.218(10), the most favourable orientation to interact with the triazole
3.376(9) Å; trz centroid···F 3.420(10) Å].
ring, have anion central atom-X···centroid angles of
The extent of centering of the anion above the centre of 128.4(3) to 134.8(3)°.
the involved aromatic ring is described by the acute angle
As expected based on the far greater π-acidity[3] of a tri-
of the ring mean plane–ring centroid–X, with an angle of azole ring compared with a pyridine ring, anion–π interac-
90° implying the anion atom is directly above the ring tions involving the triazole rings are generally shorter than
centroid and smaller angles showing correspondingly po- those involving the pyridine rings (Figure S10). The mean
orer centering. In these structures the strong interactions, plane–centroid–X angles are also generally closer to 90° for
where two anions sandwich one triazole ring, have angles the triazole–anion interactions. Only coordinated pyridine
ranging from 70.9 to 83.5°, with the angles for the other, rings are involved in the observed anion–π interactions; this
weaker interactions ranging from 55.1 to 79.3°. The angle is expected, as coordination of a formally positively charged
of the central atom of the anion–X–ring centroid reveals metal ion withdraws electron density from the ring, increas-
how directly X is “pointing” at the ring: these angles range ing its π-acidity, and favouring an interaction with an elec-
between 128.4(3) and 151.7(3) for the strong “sandwiching” tron-rich moiety such as an anion.
interactions, and 100.9(3) and 171.0(3) for the other, weaker
interactions.
Anion–π interactions involving the tetrahedral anions
BF4 and ClO4 vary slightly from those involving the octahe-
For these complexes, shorter anion–π contacts are corre- dral anions PF6 and SbF6. While the anion-sandwiching
lated (P = 0.001) with mean plane–centroid–X angles that interactions are reasonably similar (see next paragraph),
are closer to 90° (Figure 6). While this is the intuitively ex- those involved in the π-pocket are somewhat different (Fig-
pected result: a close (and hence strong[32]) anion–π interac- ure S11). Notably, interactions to the non-sandwiched tri-
tion would be expected to involve an anion positioned close azole are much weaker for the octahedral anions. This is
to the “ideal” location directly above the ring centroid, it presumably because the octahedral anions feature two in-
disagrees with the general trend found by Reedijk et al.[8] teractions with a coordinated pyridine ring (one interaction
in their survey of the CSD.[7] In that survey it seems that comparable with those found in the structures containing
the size of the mean plane–centroid–X angle has very little tetrahedral anion, and one weaker interaction), and so op-
effect on the X···centroid distance, if anything shorter dis- timising the geometries for these interactions comes at a
tances being weakly associated with smaller (further from cost to the triazole–anion interaction.
90°) angles.
The anion sandwiching X···centroid distances increase
slightly ClO4 ≈ BF4 Ͻ PF6 Ͻ SbF6 (mean X···centroid
values: XBF4/ClO4 = 2.956 Å; XPF6 = 2.975 Å; XSbF6
=
2.998 Å). The tetrahedral anions presumably show slightly
shorter interactions as the single negative charge is spread
over four atoms rather than six, making each F/O more δ–
in the tetrahedral anions, than in the octahedral. However,
not too much should be read into any of these differences
as they are very small (≈ 0.04 Å), and only a small number
of data points are being analysed.
The formation of a π-pocket around one anion in each
of the four structures containing pldpt offers an ideal op-
portunity to study anion–π interactions in detail. The anion
is the only moiety present in this pocket, and so should be
able to orient itself purely to optimize anion–π interactions.
This optimized orientation seems to involve close contacts
with two triazole rings and one pyridine ring, as this is ob-
served in all four complexes, whether a tetrahedral or octa-
hedral anion is used. It is surprising, therefore, that by far
the shortest anion–π contact is with the triazole ring that is
already forming a short anion–π contact with the other anion.
Figure 6. Scatter plot and trend line (R2 = 0.48, P = 0.001) showing
relationship between mean plane–centroid–X angle and
X···centroid distance for complexes containing pldpt ligand. The
anion-sandwiching interactions are shown as blue circles, and other
anion–π interactions are shown as red triangles.
A correlation (P = 0.036) is seen between the central Based on the simple electrostatic description of anion–π in-
atom-X···centroid angle and X···centroid distance (Figure teractions to date,[2] this should not occur. Rather, this de-
S9). However, this analysis is complicated as the anions in scription predicts that one anion–π interaction should re-
the π-pocket have more than one oxygen or fluorine atom move much of the aromatic ring’s π-acidity, hence reducing
interacting with aromatic rings and so must compromise its propensity for further anion–π interactions, i.e. in the
and arrange themselves in such a way that all of these atoms systems described here, this would lead to an (incorrect)
are in appropriate locations with respect to the aromatic expectation that either the anion in the π-pocket would
1176
www.eurjic.org
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2009, 1172–1180

A Structural Investigation of Anion–Triazole Interactions
move away from the triazole that is already interacting with 2, n = 2; M = Ni2+, m = 4, n = 0).[23a] However, this is
an anion to form short interactions with the other “free” enforced by the stoichiometry of the system as there are
triazole ring present in the pocket, or the anion that is not four anions and only two triazole rings in each dinuclear
located in the π-pocket would interact with the third tri- complex, and so if each anion is to interact with a π-ring,
azole that is not involved with the π-pocket (and which a sandwiching arrangement must occur.
should therefore be more π-acidic). Such a finding was ob-
What is unusual about the complexes reported herein is
served by Reedijk et al.[3] where a nitrate-triazine-nitrate ar- that we see an anion–π-anion sandwiching arrangement, de-
rangement contained one moderate (O···triazine 3.201 Å) spite the presence of triazole rings that are not involved in
and one weak (O···triazine 3.502 Å) interaction. The au- anion–π interactions. In a search of the CSD,[7] no anion–π-
thors commented that “the longer distance observed, when anion sandwiching interactions (X···centroid Ͻ 3.1 Å) were
two anions are interacting with the triazine, can be ration- found for triazine, tetrazine or triazole rings with halogen
ally explained as follows: the binding of the first nitrate or oxygen containing anions that also had a “spare”
gives the electron density to the electron-poor ring; as a (i.e. not involved in anion–π interactions) ring in the struc-
result, this ring becomes less electron-deficient and the in- ture.
teraction is weaker for the second anion.”[3] However, this
is clearly not the case in our system.
As well as this strong sandwiching interaction, there is
also a weaker sandwiching of a triazole ring, as the triazole
CH3CN–π Interactions
ring that weakly interacts with the anion is also involved in
another interactionϪforming a close contact with the nificant CH3CN–π interactions (N···centroid Ͼ 3.4 Å). All
CH3CN solvate (see next section). four structures containing pldpt show short contacts be-
The structure containing adpt does not contain any sig-
To summarize, there are three triazole rings in each com- tween the nitrogen of an acetonitrile solvate and a triazole
plex: one is involved in two strong contacts to electron-rich ring. These N···centroid distances range from 3.090(9) Å in
moieties [2.917(7)–3.005(4) Å], one is involved in two 2BF4 to 3.212(14) Å in 2SbF6. In all cases, the ring plane-
weaker contacts [3.090(9)–3.454(13) Å], and one is not in- ring centroid-acetonitrile nitrogen is very close to 90° (85.0–
volved in any interactions (closest contact Ͼ 3.6 Å). This 87.3°). Interestingly, all of these angles are much closer to
arrangement occurs even when the size and shape of the anion 90° than any of the anion–π interactions (ring plane–ring
is varied substantially. While we cannot definitively discount centroid–anion atom: 55.1–83.5°).
crystal packing forces as the cause of these interactions,
Heterocycle–CH3CN interactions have previously been
there seems no reason for these to be a factor in our system, calculated to be energetically favourable, and observed crys-
as the anion in the π-pocket is essentially isolated. Instead, tallographically (albeit with a disordered CH3CN molecule)
we suggest that there may be an electronic factor that fav- by Reedijk et al. for complexes involving a 1,3,5-triazine
ours this anion–π-anion sandwiching arrangement. This moiety.[35] These authors calculated that the optimum ge-
cannot be accounted for by the above simple electrostatic ometry for an acetonitrile molecule and a free 1,3,5-triazine
model of anion–π interactionsϪit appears that more com- ring had a centroid–CH3CN nitrogen distance of 3.116 Å,
plex factors are also involved, possibly leading to synergic with a mean plane–centroid–N angle of 90°, and the aceto-
anion–π binding of anions. We hope that this publication nitrile being approximately planar with the triazine ring (i.e.
will help to stimulate further theoretical studies by appro- a CϵN···centroid angle of approximately 90°). Crystallo-
priate experts in the field as it would be valuable to improve graphically, they observed the disordered CH3CN to have
our understanding of the nature of these more complex in- two sites (each at 50% occupancy), with the two centroid–
teractions.
N distances being 3.087(5) and 3.249(5) Å and the mean
There are many examples in the literature of structures plane–centroid–N angles being 75.48(3) and 82.84(3)°,
that show strong anion–π-anion sandwiching interactions. respectively. It is interesting to note that while the mean
For example, a number of metal complexes of 3,6-di(2-pyr- plane–centroid–N angles we report are very close to the 90°
idyl)-1,2,4,5-tetrazine (dptz) or 3,6-di(2-pyrazinyl)-1,2,4,5- calculated as ideal for a 1,3,5-triazine–acetonitrile interac-
tetrazine (dpztz) have been prepared by the groups of Dun- tion (indeed, much closer than those observed by Reedijk
bar[9,32,33] and Champness[34] that contain an anion–tetraz- et al.),[35] the CϵN···centroid angles range from 112.5(11)
ine–anion arrangement with centroid–anion distances as to 119.7(7)°, quite far from 90°. Presumably, this is due to
short as 2.75 Å for both anions. It is worth noting, however, steric repulsion enforced by other parts of the ligand and
that in most structures there is an “excess” of anion, when complex, that would not be involved in Reedijk et al.’s cal-
compared with π-deficient aromatic rings, particularly given culation, which involved only the free 1,3,5-triazine ring.
that one anion may be involved in interactions with a We are aware of only one further report of a crystallograph-
number of π-rings. Hence, in the dptz and dzptz complexes, ically observed CH3CN–π interaction,[36] although that in-
there are no tetrazine rings that are not involved in teraction was reasonably weak [centroid···N 3.348(11) Å,
anion–π interactions. Similarly, the pldpt ligand has pre- mean plane–centroid···N angle 80.16(2)°].
viously shown anion–π-anion sandwiching interactions
(F···centroid 2.89–3.07 Å) in the complexes [M2(pldpt)2- lation was observed between N···centroid distance and
(CH3CN)m(H2O)n](BF4)4 (M = Fe2+ [23b] Co2+, Zn2+, m = mean plane···centroid···N angles, although any possible
Unlike the anion–π interactions, no significant corre-
,
Eur. J. Inorg. Chem. 2009, 1172–1180
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
1177

N. G. White, J. A. Kitchen, S. Brooker
FULL PAPER
of Otago on a Bruker MicrOTOFQ spectrometer in acetonitrile. 1H
NMR spectra were recorded with a Varian INOVA-500 spectrome-
ter at 25 °C.
trend may be hidden by the very small sample size (there are
only four CH3CN interactions across the four complexes).
Despite the small amount of data, a significant correlation
(P = 0.036) is seen between shorter N···centroid distances Single-crystal X-ray diffraction data for all complexes were col-
lected with a Bruker Kappa Apex II area detector diffractometer
(University of Otago) at 83 K. In all cases graphite-monochro-
mated Mo-Kα radiation (λ = 0.71073 Å) was used. All data sets
were corrected for absorption using SCALE.[37] The structures were
solved using SHELXS-97.[38] All structures were refined against F2
using all data by full-matrix least-squares techniques with
SHELXL-97.[38] All non-hydrogen atoms were modelled anisotrop-
ically except for some disordered/partial occupancy solvents or
non-coordinated pyridine rings (see Supporting Information for
full details). Hydrogen atoms were inserted at calculated positions,
and increasing ϽC–N···centroid angles (Figure S12).
This is despite this increase in angle moving the CH3CN
further from the planar (i.e. CϵN–centroid angle of
≈ 90°) arrangement calculated to be favourable by Reedijk
et al.[35]
It is worth noting that while the occupancy and nature
of the second solvent region [0.5 occupancy (C2H5)2O, full
occupancy (C2H5)2O or 0.7(C2H5)2O/0.3H2O] in each of
the four structures varies, there is one full occupancy
CH3CN in each structure, suggesting that this interaction except for the amino protons in 1SbF6·1.5CH3CN·0.58(C2H5)2O,
and rode on the atoms to which they are attached (including iso-
tropic thermal parameters which were equal to 1.2 times the at-
tached non-hydrogen atom). The amino protons in 1SbF6·
1.5CH3CN·0.58(C2H5)2O were found in the difference map and
their coordinates fixed. Crystal structure determination details are
summarized in Table S1.
is favourable enough to prevent disorder of this solvent
within the lattice.
Conclusions
CCDC-696906 [for 2ClO4·CH3CN·0.7(C2H5)2O·0.3H2O], -696907
[for 2BF4·CH3CN·0.5(C2H5)2O], -696908 [for 2PF6·CH3CN·
(C2H5)2O], -696909 [for 2SbF6·CH3CN·(C2H5)2O], -696910 [for
1SbF6·1.5CH3CN·0.58(C2H5)2O], contain the supplementary crys-
tallographic data for this paper. These data can be obtained free
of charge from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif.
Ligand adpt was prepared as described previously.[21] Acetonitrile
(MeCN) was HPLC-grade, and distilled from over calcium hydride
before use; triethylamine was also distilled from calcium hydride
prior to use. All other starting materials or solvents were bought
commercially and used as received.
A series of eight mononuclear nickel(II) complexes has
been prepared. Four complexes containing the ligand pldpt
and one containing adpt have been structurally charac-
terized in the solid state by X-ray crystallography. The com-
plex of adpt contains only very weak anion–π interactions;
the four complexes of pldpt contain both strong anion–π in-
teractions and CH3CN–π interactions. In all four com-
plexes, two types of anion–π interactions are observed: one
anion interacts with only one triazole ring, while the other
anion is situated in a π-pocket provided by two triazole
rings and one pyridine ring. The anion is the only moiety
in this pocket, and so should be free to adopt the most
energetically favourable conformation for anion–π bonding:
surprisingly, the anion forms its shortest interaction with
the triazole ring that is already interacting with an anion.
The other triazole ring involved in the π-pocket around an
anion also interacts with a nitrogen donor from an acetoni-
trile solvate so that this triazole is also sandwiched by elec-
tron-rich moieties. Conversely, the third triazole ring does
not interact with any anions or acetonitrile molecules.
While we cannot definitively discount crystal packing forces
as the cause of the observed sandwiching interactions, we
cannot see why this should be the case in this particular
system, particularly given that this arrangement occurs for
all four complexesϪeven when the size and shape of the anion
is varied substantially. Instead, we suggest that the present
model of anion–π interactions as a simple electrostatic at-
traction between an electron-rich moiety and an electron-
poor π-ring may need to be modified to allow for more
complex electronic effects.
Caution: While no problems were encountered in the course of this
work, reactions involving N2H4·H2O may form potentially explos-
ive mixtures so must be carried out with extreme caution. Similarly,
–
ClO4 salts are potentially explosive so should be handled with ap-
propriate care.
4-Pyrrolyl-3,5-di(2-pyridyl)-1,2,4-triazole (pldpt): Ligand adpt
(0.477 g, 2.00 mmol) was refluxed in ethanol (30 mL) containing
concentrated aqueous hydrochloric acid (≈ 37%, 0.3 mL) and 2,5-
dimethoxytetrahydrofuran (0.31 mL, 0.32 g, 2.4 mmol) for two
hours, during which time all material dissolved to give a pink solu-
tion. This was cooled to room temperature and triethylamine
(1 mL) added, causing the pink solution to turn yellow. Reducing
this solution in volume to 8 mL under reduced pressure caused a
white, crystalline solid to precipitate. This was isolated by filtration,
thoroughly washed with distilled water (25 mL) and dried in vacuo.
Yield 0.481 g (83%). C16H12N6 (288.31): calcd. C 66.65, H 4.20, N
29.15; found C 66.42, H 4.25, N 29.44. 1H NMR spectroscopic
data were consistent with those previously reported.[22,23a]
General Procedure for Synthesis of Complexes 1ClO4, 1BF4, 2ClO4
–
–
and 2BF4: One equivalent of [Ni(H2O)6]X2 (X = ClO4 or BF4 )
was stirred with three equivalents of the appropriate ligand in
MeCN at room temperature. After a few minutes, all material dis-
solved to give a pale pink solution. This was subjected to diethyl
ether vapour diffusion, yielding pale pink-purple microcrystals,
which were isolated by filtration, washed with diethyl ether, and
dried thoroughly in vacuo. Yields ranged from 59% to 73%.
Experimental Section
General Remarks: IR spectra were recorded as pressed KBr discs
with a Perkin–Elmer Spectrum BX FT-IR spectrophotometer be-
tween 400 and 4000 cm–1. UV/Vis/NIR spectra were recorded with
a Varian CARY 500 Scan UV/Vis/NIR spectrophotometer between
[Ni(adpt)3](ClO4)2·3H2O (1ClO4·3H2O): The reaction of [Ni-
200 and 2000 nm. ESI mass spectra were recorded at the University (H2O)6](ClO4)2 (37 mg, 0.10 mmol) and adpt (71 mg, 0.30 mmol) in
1178 Eur. J. Inorg. Chem. 2009, 1172–1180
www.eurjic.org
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

A Structural Investigation of Anion–Triazole Interactions
6 mL of MeCN according to the procedure described above gave
scribed above gave pale pink-purple microcrystals. Yield 37 mg
pale pink-purple microcrystals. Yield 61 mg (59%). IR (KBr): ν =
(43%). IR (KBr): ν = (inter alia) 3313 (m), 3114 (m), 1608 (m),
˜
˜
(inter alia) 3279 (m), 3112 (m), 3079 (m), 3021 (m), 1633 (m), 1607
(m), 1591 (m), 1573 (w), 1495 (m), 1456 (s), 1428 (s), 1294 (w),
1259 (m), 1088 (br. vs), 1016 (m), 793 (m), 749 (m), 700 (m), 624
(m), 608 (w) cm–1. ESI-MS (pos.): m/z = 1085 [Na·Ni(adpt)3-
(PF6)2]+, 647 [Na3(adpt)(PF6)2]+, 267. [Ni(adpt)3]2+·1ClO4·3H2O.
C36H36Cl2N18NiO11 (1026.38): calcd. C 42.13, H 3.54, N 24.56;
found C 41.94, H 3.32, N 24.40.
1592 (m), 1497 (m), 1456 (m), 1430 (m), 1295 (w), 1259 (w),
1055 (w), 1017 (m), 843 (br. vs), 792 (m), 749 (m), 702 (m),
640 (w), 609 (w), 558 (s) cm–1. ESI-MS (pos.): m/z
=
1085 [Na·Ni(adpt)3(PF6)2]+, 267 [Ni(adpt)2]2+·1PF6·0.75H2O.
C36H31.5F12N18NiO0.75P2 (1076.87): calcd. C 40.17, H 2.95, N
23.42; found C 40.39, H 2.70, N 23.11.
[Ni(adpt)3](SbF6)2 (1SbF6): The reaction of NiSO4 (12 mg,
0.08 mmol), NaSbF6 (41 mg, 0.16 mmol) and adpt (57 mg,
0.24 mmol) in 15 mL of MeCN according to the procedure de-
scribed above gave pale pink-purple microcrystals. Yield 39 mg
[Ni(adpt)3](BF4)2 (1BF4): The reaction of [Ni(H2O)6](BF4)2 (34 mg,
0.10 mmol) and adpt (71 mg, 0.30 mmol) in 12 mL of MeCN ac-
cording to the procedure described above gave pale pink-purple
microcrystals, which were washed with ice-cold methanol (2 mL)
(39%). IR (KBr): ν = (inter alia) 3309 (m), 1608 (m), 1592 (m),
˜
and thoroughly dried. Yield 69 mg (73%). IR (KBr): ν = (inter
˜
1496 (m), 1455 (m), 1430 (m), 1293 (w), 1259 (w), 1016 (m), 795
alia) 3294 (m), 3099 (m), 3061 (m), 3007 (m), 1635 (w), 1607 (m),
1592 (m), 1571 (w), 1521 (w), 1496 (m), 1456 (s), 1429 (m), 1295
(m), 1260 (m), 1057 (br. vs), 794 (m), 750 (m), 700 (m), 640 (w),
608 (w), 521 (w) cm–1. Solution Vis/NIR (MeCN): λmax
(ε/Lmol–1 cm–1) = 8700 (54), 11600 (37) cm–1. Solid state vis/NIR
(BaSO4): λmax (relative peak height) = 18900 (1), 11400 (0.67). ESI-
MS (pos.): m/z = 859 [Ni(adpt)3(BF4)]+, 621 [Ni(adpt)2(BF4)]+, 386
[Ni(adpt)3]2+, 267 [Ni(adpt)2]2+·1BF4. C36H30B2F8N18Ni (947.05):
calcd. C 45.66, H 3.19, N 26.62; found C 45.66, H 3.25, N 26.37.
(m), 750 (m), 703 (m), 661 (vs) cm–1. ESI-MS (pos.): m/z = 1009
[Ni(adpt)3(SbF6)]+, 771 [Ni(adpt)2(SbF6)]+, 267 [Ni(adpt)3]2+
·
1SbF6. C36H30F12N18NiSb2 (1244.94): calcd. C 34.73, H 2.43, N
20.25; found C 34.84, H 2.41, N 20.22.
[Ni(pldpt)3](PF6)2 (2PF6·1.5H2O): The reaction of NiSO4 (12 mg,
0.08 mmol), NaPF6 (27 mg, 0.16 mmol) and pldpt (69 mg,
0.24 mmol) in 20 mL of MeCN according to the procedure de-
scribed above gave pale pink-purple microcrystals. Yield 43 mg
(43%). IR (KBr): ν = (inter alia) 3153 (m), 3113 (m), 2872 (m),
˜
[Ni(pldpt)3](ClO4)2 (2ClO4): The reaction of [Ni(H2O)6](ClO4)2
(18 mg, 0.05 mmol) and pldpt (43 mg, 0.15 mmol) in 6 mL of
MeCN according to the procedure described above gave pale pink-
1610 (m), 1588 (m), 1573 (w), 1505 (w), 1452 (s), 1430 (m), 1337
(m), 1298 (m), 1268 (w), 1155 (w), 1103 (w), 1017 (m), 913 (m),
843 (br. vs), 792 (m), 741 (m), 730 (m), 706 (m), 643 (m), 615 (w),
557 (s) cm–1. ESI-MS (pos.): m/z = 1067 [Ni(pldpt)3(PF6)]+, 289
[H(pldpt)]+· 2PF6·1.5H2O. C48H39F12N18NiO1.5P2 (1240.54): calcd.
C 46.47, H 3.17, N 20.32; found C 46.68, H 3.47, N 20.38.
purple microcrystals. Yield 41 mg (73%). IR (KBr): ν = (inter alia)
˜
3170 (m), 2975 (m), 1610 (m), 1588 (w), 1505 (m), 1450 (s), 1430
(m), 1337 (m), 1297 (m), 1268 (w), 1089 (br. vs), 1016 (m), 992 (m),
912 (w), 793 (m), 738 (m), 705 (m), 642 (m), 623 (m) cm–1. ESI-
MS (pos.): m/z
= ,
461 [Ni(pldpt)3]2+ 289 [H(pldpt)]+·2ClO4.
[Ni(pldpt)3](SbF6)2·H2O(2SbF6·H2O): The reaction of NiSO4
(12 mg, 0.08 mmol), NaSbF6 (41 mg, 0.16 mmol) and pldpt (69 mg,
0.24 mmol) in 15 mL of MeCN according to the procedure de-
scribed above gave pale pink-purple microcrystals. Yield 54 mg
C48H36Cl2N18NiO8 (1122.51): calcd. C 51.36, H 3.23, N 22.46;
found C 51.07, H 3.54, N 22.24.
[Ni(pldpt)3](BF4)2·2H2O(2BF4·2H2O): The reaction of [Ni(H2O)6]-
(BF4)2 (34 mg, 0.10 mmol) and pldpt (87 mg, 0.30 mmol) in 12 mL
of MeCN according to the procedure described above gave pale
pink-purple microcrystals, which were washed with ice-cold meth-
anol (2 mL) and thoroughly dried. Yield 70 mg (62%). IR (KBr):
(48%). IR (KBr): ν = (inter alia) 3145 (m), 1595 (m), 1572 (w),
˜
1472 (s), 1430 (s), 1341 (m), 1292 (m), 1244 (w), 1156 (w), 1092
(m), 1076 (m), 1010 (m), 998 (w), 914 (m), 795 (s), 746 (s), 705 (m),
660 (vs) cm–1. ESI-MS (pos.): m/z = 1157 [Ni(pldpt)3(SbF6)]+, 869
[Ni(pldpt)2(SbF6)]+, 461 [Ni(pldpt)3]2+, 311 [Na(pldpt)]+·2SbF6·H2O
C48H38F12N18NiOSb2 (1413.13): calcd. C 40.80, H 2.71, N 17.84;
found C 40.84, H 2.74, N 17.80.
ν = (inter alia) 3141 (m), 3106 (m), 1611 (m), 1587 (w), 1572 (w),
˜
1507 (w), 1479 (m), 1451 (s), 1432 (m), 1337 (m), 1297 (m), 1268
(w), 1057 (br. vs), 912 (w), 792 (m), 739 (m), 707 (m), 643 (m), 615
(m), 521 (w) cm–1. Solution Vis/NIR (MeCN): λmax (ε/Lmol–1 cm–1)
= 18800 (20), 11600 (16). Solid-state Vis/NIR (BaSO4): λmax (rela-
tive peak height) = 19100 (1), 11400 (0.68) cm–1. ESI-MS (pos.):
m/z = 599 [Na(pldpt)2]+, 461 [Ni(pldpt)3]2+, 311 [Na(pldpt)]+·
2BF4·2H2O. C48H40B2F8N18NiO2 (1133.25): calcd. C 50.87, H 3.56,
N 22.25; found C 51.09, H 3.52, N 22.35.
Supporting Information (see also the footnote on the first page of
this article): Organic synthesis (Scheme S1) and reflectance and
solution UV/Vis spectra for 2BF4·2H2O and 1BF4 (Figure S1) are
given. Crystallographic information including complex geometries
and packing interaction data (Tables S1–S9) as well as perspective
views of complexes (Figures S2–S6) are also provided.
Anion···π interaction analyses are presented (Figures S7–S12) as
are color versions of Figures 2, 3 and 6.
General Procedure for Synthesis of Complexes 1PF6, 1SbF6, 2PF6
and 2SbF6: One equivalent of anhydrous NiSO4, two equivalents of
sodium hexafluorophosphate or sodium hexafluoroantimonate,
and three equivalents of the appropriate ligand were refluxed for
one hour in MeCN, protected from atmospheric moisture with a
calcium chloride drying tube. The resulting pink solution contain-
ing a fine white precipitate was cooled to room temperature, and
then filtered to give a clear, pink solution. This was reduced in
volume to ca. 8 mL and subjected to diethyl ether vapour diffusion,
yielding pale purple-pink microcrystals, which were isolated by fil-
tration, washed with diethyl ether, and dried thoroughly in vacuo.
Yields ranged from 39 to 48%.
Acknowledgments
This work was supported by the University of Otago, including a
University of Otago Research Grant (to N. G. W. and S. B.). We
thank the Tertiary Education Commission (New Zealand) for the
award of a Bright Futures Top Achiever Doctoral scholarship to
J. A. K. and the MacDiarmid Institute for Advanced Materials and
Nanotechnology (to S. B.). We are grateful to Associate Professor
Lyall Hanton (University of Otago) for helpful discussions and ad-
vice, and to Professor Helen Nicholson (University of Otago) for
her help with the statistical analyses.
[Ni(adpt)3](PF6)2·0.75H2O (1PF6·0.75H2O): The reaction of NiSO4
(12 mg, 0.08 mmol), NaPF6 (27 mg, 0.16 mmol) and adpt (57 mg,
0.24 mmol) in 10 mL of MeCN according to the procedure de-
Eur. J. Inorg. Chem. 2009, 1172–1180
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
1179

N. G. White, J. A. Kitchen, S. Brooker
FULL PAPER
M. Kojima, F. Dahan, J.-P. Tuchagues, Inorg. Chem. 2003, 42,
8406–8416; Y. Ikuta, M. Ooidemizu, Y. Yamahata, M. Yam-
ada, S. Osa, N. Matsumoto, S. Iijima, Y. Sunatsuki, M. Ko-
jima, F. Dahan, J.-P. Tuchagues, Inorg. Chem. 2003, 42, 7001–
7017.
[1]
D. Quiñonero, C. Garau, C. Rotger, A. Frontera, P. Ballester,
A. Costa, P. M. Deya, Angew. Chem. Int. Ed. 2002, 41, 3389–
3392; M. Mascal, A. Armstrong, M. D. Bartberger, J. Am.
Chem. Soc. 2002, 124, 6274–6276; I. Alkorta, I. Rozas, J. El-
guero, J. Am. Chem. Soc. 2002, 124, 8593–8598; I. Alkorta, I.
Rozas, J. Elguero, J. Org. Chem. 1997, 62, 4687–4691.
B. L. Schottel, H. T. Chifotides, K. R. Dunbar, Chem. Soc. Rev.
2008, 37, 68–83.
P. Gamez, T. J. Mooibroek, S. J. Teat, J. Reedijk, Acc. Chem.
Res. 2007, 40, 435–444.
C. Garau, D. Quiñonero, A. Frontera, P. Ballester, A. Costa,
P. M. Deyà, New J. Chem. 2003, 27, 211–214.
P. de Hoog, P. Gamez, I. Mutikainen, U. Turpeinen, J. Reedijk,
Angew. Chem. Int. Ed. 2004, 43, 5815–5817.
S. Demeshko, S. Dechert, F. Meyer, J. Am. Chem. Soc. 2004,
126, 4508–4509.
[19] J.-F. Létard, P. Guionneau, E. Codjovi, O. Lavastre, G. Bravic,
D. Chasseau, O. Kahn, J. Am. Chem. Soc. 1997, 119, 10861–
10862; Z. J. Zhong, J.-Q. Tao, Z. Yu, C.-Y. Dun, Y.-J. Liu, X.-
Z. You, J. Chem. Soc., Dalton Trans. 1998, 327–328; R. Bocˇa,
M. Bocˇa, L. Dlhánˇ, K. Falk, H. Fuess, W. Haase, R. Jarosˇcˇiak,
B. Papánková, F. Renz, M. Vrbová, R. Werner, Inorg. Chem.
2001, 40, 3025–3033.
[20] Anion ordering/disordering involving a perchlorate anion that
interacts with tetrazole ring has recently been implicated in a
hysteretic SCO: M. Quesada, H. Kooijman, P. Gamez, J. S. Co-
sta, P. J. v. Koningsbruggen, P. Weinberger, M. Reissner, A. L.
Spek, J. G. Haasnoot, J. Reedijk, Dalton Trans. 2007, 5434–
5440.
[21] J. F. Geldard, F. Lions, J. Org. Chem. 1965, 30, 318–319.
[22] S. K. Mandal, H. J. Clase, J. N. Bridson, S. Ray, Inorg. Chim.
Acta 1993, 209, 1–4.
[23] a) M. H. Klingele, P. D. W. Boyd, B. Moubaraki, K. S. Murray,
S. Brooker, Eur. J. Inorg. Chem. 2006, 573–589; b) J. A.
Kitchen, A. Noble, C. D. Brandt, B. Moubaraki, K. S. Murray,
S. Brooker, Inorg. Chem. 2008, 47, 9450–9458.
[2]
[3]
[4]
[5]
[6]
[7]
[8]
F. H. Allen, Acta Crystallogr., Sect. B 2002, 58, 380–388.
T. J. Mooibroek, C. A. Black, P. Gamez, J. Reedijk, Cryst.
Growth Des. 2008, 8, 1082–1093.
C. S. Campos-Fernández, B. L. Schottel, H. T. Chifotides, J. K.
Bera, J. Bacsa, J. M. Koomen, D. H. Russell, K. R. Dunbar, J.
Am. Chem. Soc. 2005, 127, 12909–12923.
[9]
[10] C. A. Black, L. R. Hanton, M. D. Spicer, Chem. Commun.
2007, 3171–3173; I. y. A. Gural’skiy, P. V. Solntsev, H. Krauts-
cheid, K. V. Domasevitch, Chem. Commun. 2006, 46, 4808–
4810.
[11] A. Frontera, F. Saczewski, M. Gdaniec, E. Dziemidowicz-
Borys, A. Kurland, P. M. Deyà, D. Quiñonero, C. Garau,
Chem. Eur. J. 2005, 11, 6560–6567; M. Mascal, I. Yakovlev,
E. B. Nikitin, J. C. Fettinger, Angew. Chem. Int. Ed. 2007, 46,
[24] K. H. Sugiyarto, D. C. Craig, A. D. Rae, H. A. Goodwin, Aust.
J. Chem. 1995, 48, 35–54; K. H. Sugiyarto, H. A. Goodwin,
Aust. J. Chem. 1988, 41, 1645–1663.
[25] K. H. Sugiyarto, H. A. Goodwin, Aust. J. Chem. 1987, 40, 775–
783.
[26] M. H. Klingele, S. Brooker, Eur. J. Org. Chem. 2004, 3422–
3434.
[27] C. Hansch, A. Leo, R. W. Taft, Chem. Rev. 1991, 91, 165–195.
8782–8784; Y. S. Rosokha, S. V. Lindeman, S. V. Rosokha, J. K. [28] M. P. Mingos, A. L. Rohl, Inorg. Chem. 1991, 30, 3769–3771.
Kochi, Angew. Chem. Int. Ed. 2004, 43, 4650–4652.
[12] V. Gorteau, G. Bollot, J. Mareda, A. Perez-Velasco, S. Matile,
J. Am. Chem. Soc. 2006, 128, 14788–14789; V. Gorteau, G. Bol-
lot, J. Mareda, S. Matile, Org. Biomol. Chem. 2007, 5, 3000–
3012.
[29] G. S. Matouzenko, A. Boussekou, S. Borshch, M. Perrin, S.
Zein, L. Salmon, G. Molnar, S. Lecocq, Inorg. Chem. 2004, 43,
227–236.
[30] M. Shakir, S. Parveen, N. Begum, Y. Azim, Polyhedron 2003,
22, 3181–3186.
[13] P. D. Beer, P. A. Gale, Angew. Chem. Int. Ed. 2001, 40, 486–
[31] J. A. Kitchen, N. G. White, B. Moubaraki, K. S. Murray,
P. D. W. Boyd, M. Boyd, S. Brooker, manuscript in preparation.
[32] B. L. Schottel, H. T. Chifotides, M. Shatruk, A. Chouai, L. M.
Pérez, J. Bacsa, K. R. Dunbar, J. Am. Chem. Soc. 2006, 128,
5895–5902.
516.
[14] P. Gütlich, Y. Garcia, H. A. Goodwin, Chem. Soc. Rev. 2000,
29, 419–427; O. Sato, J. Tao, Y.-Z. Zhang, Angew. Chem. Int.
Ed. 2007, 46, 2152–2187.
[15] J. A. Real, A. B. Gaspar, V. Niel, M. C. Muñoz, Coord. Chem.
Rev. 2003, 236, 121–141.
[16] J.-F. Létard, P. Guionneau, L. Goux-Capes, Top. Curr. Chem.
2004, 234, 221–250.
[17] C. Rajadurai, O. Fuhr, R. Kruk, M. Ghafari, H. Hahn, M.
Ruben, Chem. Commun. 2007, 2636–2638; S. Cobo, G. Molnar,
J. A. Real, A. Bousseksou, Angew. Chem. Int. Ed. 2006, 45,
[33] B. L. Schottel, J. Bacsa, K. R. Dunbar, Chem. Commun. 2005,
46–47; C. S. Campos-Fernández, R. Clérac, J. M. Koomen,
D. H. Russell, K. R. Dunbar, J. Am. Chem. Soc. 2001, 123,
773–774; C. S. Campos-Fernandez, R. Clerac, K. R. Dunbar,
Angew. Chem. Int. Ed. 1999, 38, 3477–3479.
[34] N. S. Oxtoby, A. J. Blake, N. R. Champness, C. Wilson, Proc.
Natl. Acad. Sci. USA 2002, 99, 4905–4910.
5786–5789; V. Niel, A. Galet, A. B. Gaspar, M. C. Muñoz, J. A. [35] T. J. Mooibroek, S. J. Teat, C. Massera, P. Gamez, J. Reedijk,
Real, Chem. Commun. 2003, 1248–1249; O. Kahn, C. J. Marti-
Cryst. Growth Des. 2006, 6, 1569–1574.
nez, Science 1998, 279, 44–48.
[36] Z. Lu, J. S. Costa, O. Roubeau, I. Mutikainen, U. Turpeinen,
S. J. Teat, P. Gamez, J. Reedijk, Dalton Trans. 2008, 3567–3573.
[37] G. M. Sheldrick, Univ. of Göttingen, Germany, 1996; R. H.
Blessing, Acta Crystallogr., Sect. A 1995, 51, 33–38.
[38] G. M. Sheldrick, Acta Crystallogr., Sect. A 2008, 64, 112–122.
Received: December 29, 2008
[18] Y. Sunatsuki, Y. Ikuta, N. Matsumoto, H. Ohta, M. Kojima,
S. Iijima, S. Hayami, Y. Maeda, S. Kaizaki, F. Dahan, J.-P.
Tuchagues, Angew. Chem. Int. Ed. 2003, 42, 1614–1618; G. S.
Matouzenko, G. Molnar, N. Bréfuel, M. Perrin, A. Boussek-
sou, S. A. Borshch, Chem. Mater. 2003, 15, 550–556; M. Yam-
ada, M. Ooidemizu, Y. Ikuta, S. Osa, N. Matsumoto, S. Iijima,
Published Online: February 26, 2009
1180
www.eurjic.org
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2009, 1172–1180