Sorption of small molecule vapours by single crystals of [Pt{4′-(Ph)trpy}(NCS)]SbF6 where trpy = 2,2′:6′, 2′′-terpyridine: A porous material with a structure stabilised by extended π-π Interactions

Article abstract of DOI:10.1039/c2dt12398c

Treatment of [Pt{4′-(Ph)trpy}Cl]SbF₆ with AgSCN in a metathesis reaction in refluxing acetonitrile affords, after work-up, single crystals of [Pt{4′-(Ph)trpy}(NCS)]SbF₆·CH₃CN, where trpy is 2,2′:6′,2′′-terpyridine. These crystals lose solvent to give single crystals of [Pt{4′-(Ph)trpy}(NCS)]SbF ₆ (1). An X-ray crystal structure determination of 1 shows that the SCN^- ion is N-bound and that the cation as a whole is approximately planar. Compound 1 is porous with "empty" channels that corkscrew through the crystal: this crystal structure is stabilised by extended π-π interactions between the planar cations. When a single crystal of 1 is exposed to vapours of acetonitrile the vapours are sorbed without loss of single crystallinity, as confirmed by crystal structure determinations of 1 and 1·CH₃CN using the same single crystal. Similarly, single crystals of 1 sorb vapours of methanol without loss of single crystallinity, as confirmed by a crystal structure determination of 1·CH₃OH. We also report the crystal structure of 1·(CH₃)₂CO; however, in this case the single crystal was grown directly from acetone. Compound 1 and its solvates are all yellow. Nevertheless, there are differences between the emission spectra recorded for 1 and its solvates in the solid state. Thus, whereas 1 exhibits very weak multiple emission from ³MLCT (MLCT = metal-to-ligand charge transfer) and excimeric ³π- π* excited states, 1·CH₃CN and 1·(CH ₃)₂CO both exhibit more intense ³MLCT emission; and the emission by 1·CH₃OH is complicated by the presence of metallophilic interactions in the crystal. We discuss the role of the solvent in causing these differences.

Full text of DOI:10.1039/c2dt12398c

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
Dalton
Transactions
Cite this: Dalton Trans., 2012, 41, 5486
www.rsc.org/dalton
PAPER
Sorption of small molecule vapours by single crystals of [Pt{4′-(Ph)trpy}
(NCS)]SbF6 where trpy = 2,2′:6′,2′′-terpyridine: a porous material with a
structure stabilised by extended π–π interactions†
John S. Field,* Orde Q. Munro and Bradley P. Waldron
Received 12th December 2011, Accepted 5th March 2012
DOI: 10.1039/c2dt12398c
Treatment of [Pt{4′-(Ph)trpy}Cl]SbF6 with AgSCN in a metathesis reaction in refluxing acetonitrile
affords, after work-up, single crystals of [Pt{4′-(Ph)trpy}(NCS)]SbF6·CH3CN, where trpy is 2,2′:6′,2′′-
terpyridine. These crystals lose solvent to give single crystals of [Pt{4′-(Ph)trpy}(NCS)]SbF6 (1). An
X-ray crystal structure determination of 1 shows that the SCN− ion is N-bound and that the cation as a
whole is approximately planar. Compound 1 is porous with “empty” channels that corkscrew through the
crystal: this crystal structure is stabilised by extended π–π interactions between the planar cations. When a
single crystal of 1 is exposed to vapours of acetonitrile the vapours are sorbed without loss of single
crystallinity, as confirmed by crystal structure determinations of 1 and 1·CH3CN using the same single
crystal. Similarly, single crystals of 1 sorb vapours of methanol without loss of single crystallinity, as
confirmed by a crystal structure determination of 1·CH3OH. We also report the crystal structure of
1·(CH3)2CO; however, in this case the single crystal was grown directly from acetone. Compound 1 and
its solvates are all yellow. Nevertheless, there are differences between the emission spectra recorded for
1 and its solvates in the solid state. Thus, whereas 1 exhibits very weak multiple emission from 3MLCT
(MLCT = metal-to-ligand charge transfer) and excimeric 3π–π* excited states, 1·CH3CN and
1·(CH3)2CO both exhibit more intense 3MLCT emission; and the emission by 1·CH3OH is complicated
by the presence of metallophilic interactions in the crystal. We discuss the role of the solvent in causing
these differences.
Many of these exhibit a spectroscopic response to the inclusion
of small gaseous molecules, a recent example being the supra-
Introduction
Porous metal–organic framework (MOF) materials are con-
structed on the principle that metal ions linked by bridging
organic ligands provide a reticular framework that encapsulates
pores (or voids) in the solid.1 A host of materials of this type has
been investigated because of their potential applications in cata-
lysis,2 hydrogen (and other gas) storage,3 luminescent sensor
devices,4 magnetic materials,5 and non-linear optical materials.6
There are examples of materials of this type where the reticular
framework of covalent bonds is supported by (weaker) intermo-
lecular interactions such as hydrogen bonding,7 and extended
aromatic interactions;8 the latter are usually labeled as “π–π inter-
actions”.9,10 A second group of molecular materials relies on the
principles of self-assembly of organic or coordination com-
pounds to construct supramolecular systems that are porous.11
molecular luminescent system based on 2-cyano-3{4-(diphenyla-
mino)phenyl} acryclic acid, which acts as a selective acetonitrile
sensor.12 The system that we describe here is also molecular and
porous with sensing properties, but it has a crystal structure that
does not fit into any of the above categories. Instead, the com-
pound [Pt{4′-(Ph)trpy}(NCS)]SbF6 (1) where trpy is 2,2′:6′,2′′-
terpyridine, has an open microporous structure containing
2-dimensional stacks of planar [Pt{4′-(Ph)trpy}Cl]+ cations that
are stabilised solely by extended π–π interactions.
To test the availability of the microporous space in 1 for guest
molecule inclusion, we have exposed single crystals of the com-
pound to vapours of acetonitrile, methanol and acetone. Aceto-
nitrile and methanol are sorbed without loss of the single
crystallinity but the larger acetone molecule, though sorbed,
does disrupt the long range order in the crystal. We also report
the photoluminescent properties of 1 and its acetonitrile, metha-
nol and acetone solvates. These materials are all yellow in
colour, yet there are differences in their emission spectra,
an observation that is relevant to studies of the solid state
photophysical properties of terpyridyl ligand complexes of
platinum(II).13–23
School of Chemistry, University of KwaZulu-Natal, Private Bag X01,
Pietermaritzburg, 3201, South Africa. E-mail: fieldj@ukzn.ac.za
†Electronic supplementary information (ESI) available: X-ray structures
of 1_1st, 1_2nd, 1_3rd, 1·CH3CN, 1·CH3OH and 1·(CH3)2CO. CCDC
857048–857053. For ESI and crystallographic data in CIF or other elec-
tronic format see DOI: 10.1039/c2dt12398c
5486 | Dalton Trans., 2012, 41, 5486–5496
This journal is © The Royal Society of Chemistry 2012

View Article Online
step of 0.02° at a 1° per minute counting interval. Fourier trans-
form infrared (FTIR) spectra were measured using a Perkin
Elmer Spectrum One spectrometer, and the samples prepared as
KBr pellets. UV-vis absorption spectra were recorded at 22 °C
using a Perkin-Elmer Lambda 45 UV-vis spectrometer. 1H
(500 MHz) and 13C (125 MHz) NMR spectra were recorded at
30 °C in CD3CN on a Bruker Avance III 500 MHz spectrometer
with chemical shifts referenced to the solvent.
Experimental
Materials and methods
The acetonitrile for the syntheses, crystal growth and for generat-
ing vapours was of the Chromosolv™ HPLC grade from Aldrich
and used as received. The diethyl ether was from Saarchem and
of the uniLAB™ grade and also used as received. The acetone
and methanol used for generating vapours were of ACS grade
from Aldrich and used as received. The [Pt{4′-(Ph)trpy}Cl]SbF6
precursor was synthesised by the method described in ref. 22.
The AgSCN was obtained from Aldrich and used as received.
Emission measurements
[Pt{4′-(Ph)trpy}(NCS)]SbF6 (1). To a suspension of [Pt{4′-
(Ph)trpy}Cl]SbF6 (100 mg, 0.129 mmol) in acetonitrile (10 mL)
was added a suspension of a 10% molar excess of AgSCN
(24 mg) in acetonitrile (40 mL). The mixture was heated to
reflux for 24 h. After cooling to room temperature the resulting
AgCl precipitate was removed by cannular filtration, and the
volume of the solvent was reduced in vacuo. The concentrated
acetonitrile solution was transferred to the lower compartment of
a dual chamber purpose-built apparatus; diethyl ether was added
to the upper compartment. Slow vapour diffusion of diethyl
ether into the concentrated acetonitrile solution resulted in the
formation of yellow crystals over a period of ∼7 d. These are
crystals of the acetonitrile solvate, [Pt{4′-(Ph)trpy}(NCS)]
SbF6·CH3CN. The yellow crystals were isolated by decantation
of the mother liquor, and then washed with small amounts of
cold acetonitrile followed by diethyl ether. Allowing the crystal-
line material to stand in air affords the de-solvated compound,
which is also yellow. In order to ensure that all the acetonitrile
solvent is lost from the crystals, they were exposed to air for a
full week: however, the de-solvation process can be speeded-up
by the application of a vacuum. Analytical and spectroscopic
data are as follows. [Pt{4′-(Ph)trpy}(NCS)]SbF6: Yield 71 mg
(69%). Colour, yellow. Anal. calcd for C22H15N4F6PtSSb·H2O
(FW = 816.30 g mol−1): C, 32.37; H, 2.10; N, 6.86. Found: C,
32.21; H, 1.95; N, 6.57%. IR (KBr, cm−1): ν[SCuN]: 2095s;
ν[S–CN]: 866m; ν[4′-(Ph)trpy]: 1610s, 1557ms, 1479ms, 1418s,
881m; ν[SbF6−]: 656vs. ESI MS: m/z (relative intensity, ion):
562.0670 (100, M2+). 1H NMR (CD3CN):‡ δ 8.51 (4H, m,
H6/6′′, 3′/5′); 8.42 (4H, m, H3/3′′, 2′′′/6′′′); 8.05 (1H, m, H4′′′); 7.83
(2H, q, H4/4′′); 7.71 (4H, m, H5/5′′, 3′′′/5′′′). 13C NMR (CD3CN):‡
δ 157.8 (2C, s, quat.C2/2′′); 155.5 (2C, s, quat.C2′/6′); 154.7 (1C,
s, quat.C1′′′); 152.2 (2C, s, C6/6′′); 142.9 (2C, s, C2′′′/6′′′); 142.4
(1C, s, quat.C4′); 131.7 (2C, s, C3′′′/5′′′); 129.7 (4C, s, C5/5′′,4/4′′);
128.0 (1C, s, C4′′′); 125.8 (2C, s, C3/3′′); 122.1 (2C, s, C3′/5′).
UV-vis (1 μM in MeCN): λmax/nm (ε, M−1 cm−1): 409 (20 563);
387 (17 055); 333 (46 406); 319 (49 463); 304 (49 888); 290
(75 511); 282 (68 701); 273 (63 785); 259 (64 492).
The instrument used for the measurement of emission spectra
was a Photon Technologies Int. (PTI) fluorescence spectrometer
controlled by PTI’s Felix32© Version 1.1 software.26 Steady
state emission spectra were recorded using PTI’s XenoFlash™
300 Hz pulsed light source and gated emission scans with a
delay of 95 ms, an integration window time of 100 ms and 50
pulses per channel (shots). Detection was by means of PTI’s
Model 814 Analog/Photon-Counting Photomultiplier Detector.
The excitation wavelength was 420 nm for all the solid state
samples; with the scattered light being removed by means of a
suitable wavelength band-pass filter. The spectra are uncorrected
for instrument response. For the lifetime measurements, the exci-
tation source was again the Xenon flash lamp (set at 420 nm)
with the emission decay captured by the Photomultiplier Detec-
tor at a wavelength corresponding to the peak of interest, and
analysed by the FeliX32™ software.26 For the measurements at
77 K a PTI-supplied quartz cold finger filled with liquid nitrogen
was used to hold the sample tube.
The sample of 1 used for the emission measurements was
obtained by crushing single crystals of the compound and
thoroughly mixing the powder with finely divided anhydrous
KBr to give a 10% (w/w) solid solution. To ensure that any
vestiges of water in the sample were removed, the sample was
subjected to a high vacuum for several days. The sample was
transferred under dry argon to a quartz NMR tube with its top
half widened to fit a B14 joint, into which a stopcock was
inserted thus allowing for both purging of the sample with argon
and the introduction into the sample space of the organic
vapours via cannulae inserted through a septum. Solvent vapours
were generated by vigorously bubbling argon through a sealed
100 mL Erlenmeyer flask containing the anhydrous solvent, and
transferred to the sample tube via cannulae. Total exposure time
of the sample to the stream of solvent vapours was ∼60 min.
The stopcock was then closed and the sample kept in contact
with solvent vapours for a further 3 h. The envelope of solvent
vapours was then removed by gently purging with argon and the
emission spectrum of the solvated material recorded, at 77 K so
as to increase the intensity of the spectrum and to minimise any
loss of solvent. For the monitoring experiment the acetonitrile
solvate sample was warmed to room temperature and exposed to
ambient conditions: emission spectra were then recorded at
regular intervals, at 77 K, so ensuring that each spectrum is a
“freeze-frame” of the de-solvation process. The emission spec-
trum of 1 shown in Fig. 10 was obtained in this way, in particu-
lar by ensuring that the acetonitrile solvate had de-solvated fully
through application of a vacuum.
Characterisation. Elemental analyses were determined by
Galbraith Laboratories of Knoxville, Tennessee, USA. The
powder XRD spectrum of 1 was recorded at 295 K with a
Philips PW1050 diffractometer using monochromated CoKα
radiation (λ = 1.7902 Å) from 3 to 40° in 2θ with a scanning
‡The chemical shifts listed are for the N-bound isomer that dominates in
CD3CN solution:24,25 further details are available from the authors on
request.
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 5486–5496 | 5487

View Article Online
Single crystal preparations and structure determinations
on symmetry-equivalent and repeated reflections. The structures
were solved by Patterson methods and refined on F2 values for
all unique data; Table 1 gives further details. All non-hydrogen
atoms were refined anisotropically, and H atoms were con-
strained with a riding model; U(H) was set at 1.2 times (1.5
times in the case of a methyl group) Ueq for the parent carbon
atom. The programs used were Oxford Diffraction CrysalisRED
v. 170 (control and data reduction),27 SHELXS-97 and
SHELXL-97 (structure solution and refinement);28 for the mol-
ecular graphics the programs ORTEP329 and X-SEED30a were
used.
The first batch of single crystals was isolated by slow vapour
diffusion of diethyl ether into an acetonitrile solution of [Pt{4′-
(Ph)trpy}(NCS)]SbF6 (1): these were of the acetonitrile solvate,
1·CH3CN. Single crystals of 1 itself were obtained by allowing
the solvated crystals to lose solvent over a period of about 6 d
under ambient conditions: a single crystal was selected from this
batch and its structure determined at 200 K. This same crystal
(glued to a fibre attached to a brass pin) was transferred to a
glass platform by fixing the pin to the platform with Prestik™.
The platform was constructed by extending a B24 glass stopper,
designed such that when it is inserted through the mouth of a
two-necked round bottomed flask, the platform is positioned
roughly halfway between the bottom of the flask and the stopper.
The flask contained about 20 mL of pure acetonitrile. A gentle
vacuum was then applied to the flask and switched off as
soon as the solvent began to boil. In this way the crystal was
enveloped in vapours of acetonitrile at its vapour pressure at
room temperature; in this case the laboratory temperature was
∼20 °C and, therefore, the vapour pressure of the acetonitrile
was ∼43 Torr. The crystal was left exposed to the solvent’s
vapours for 48 h. After this period, the pin and mounted crystal
were removed from the flask and transferred immediately to the
diffractometer where it was cooled to 200 K. The temperature of
200 K is a compromise: a low temperature was chosen to mini-
mise solvent loss but dropping the temperature to below 200 K
caused the crystal to crack. The data collection for 1·CH3CN
was done with this crystal. After the data collection was com-
plete, the crystal was warmed to room temperature thus allowing
the acetonitrile to evaporate to afford a single crystal of 1. A
second intensity data set, again at 200 K, was collected and the
crystal structure of 1 re-determined: see CCDC 857049 in the
ESI.† With methanol the same procedure was followed in order
to obtain a single crystal and hence an intensity data set for the
solvate, 1·CH3OH. However, a different starting single crystal of
1 was used and, of course, the vapour pressure of the methanol
is different i.e. ∼125 Torr (with the methanol solvate a tempera-
ture of 180 K was used for the data collection). The same pro-
cedure was also used for the acetone solvate, using a different
starting single crystal of 1 that was exposed to acetone vapours
at a pressure of ∼230 Torr. Intensity data collected on this crystal
gave a structure solution that showed the inclusion of the acetone
molecule in the crystal of 1, see the CheckCIF report in the
ESI.† However, with a high final R factor of 10% and problems
with thermal motion, the structure determination is not suffi-
ciently accurate to allow for a detailed comparison of geometric
parameters between the three solvates and 1. For this reason we
grew a good quality crystal of 1·(CH3)2CO by slow vapour dif-
fusion of diethyl ether into an acetone solution of 1. This is the
structure, determined at 200 K, discussed in the text below. As a
final experiment, this same single crystal of 1·(CH3)2CO was
warmed to room temperature thus allowing the acetone to evap-
orate to afford a single crystal of 1: see CCDC 857050 in the
ESI† for details of the structure determination.
Results and discussion
Treatment of a solution of [Pt{4′-(Ph)trpy}Cl]SbF6 in refluxing
acetonitrile, with one equivalent of silver thiocyanate, afforded
via a metathesis reaction a precipitate of silver chloride and a
solution of [Pt{4′-(Ph)trpy}(NCS)]SbF6 (1). After removal of
the AgCl by filtration, the acetonitrile filtrate was concentrated,
cooled and transferred to a two compartment vapour diffusion
apparatus. Slow diffusion of diethyl ether into the concentrated
solution led to the precipitation of a mass of yellow, and mostly
single, crystals of the acetonitrile solvate, 1·CH3CN. Application
of a vacuum to the crystals extricates the solvent, finally afford-
ing compound 1, also as a mass of yellow, and mostly single,
crystals. The infrared spectrum recorded as a KBr pellet exhibits
a strong sharp peak at 2102 cm−1 assigned to the ν(SCuN)
stretching mode and a weaker, somewhat broader peak at
866 cm−1, due to the ν(S–CN) stretching mode. These stretching
frequencies are consistent with a SCN− ion bound to the Pt atom
through the N atom in the solid state i.e. as the isothiocyanate
ligand.31–33
Crystal structure of [Pt{4′-(Ph)trpy}(NCS)]SbF6 (1)
The single crystal of 1 was obtained by exposing crystals of
the acetonitrile solvate to ambient conditions for a period of
6 d. Note that the de-solvation process can be speeded up to 1 d
by application of a vacuum, again without loss of single crystalli-
nity. A temperature of 200 K was used for the data collection so
as to correlate with the temperatures used for the crystal structure
determinations of the solvates. Fig. 1 gives a perspective view of
the cation in 1. Selected geometric parameters for the cation are
given in Table 2.
The bond distances and angles associated with the Pt(trpy)
unit agree well with those obtained from crystal structure deter-
minations of other platinum terpyridine complexes reported in
Intensity data were collected on an Oxford Diffraction
Xcalibur 2 CCD diffractometer in the range 2 ≤ θ ≤ 25° using
graphite-monochromated Mo-Kα radiation (λ = 0.71073 Å).
Intensities were corrected semi-empirically for absorption, based
Fig. 1 ORTEP diagram (50% ellipsoids) of the cation in 1.
5488 | Dalton Trans., 2012, 41, 5486–5496
This journal is © The Royal Society of Chemistry 2012

View Article Online
Table 1 Crystal structure and refinement data
Compound
1a
1·CH3CNb
1·CH3OHc
1·(CH3)2COd
Empirical formula
Mr
C22H15F6N4PtSSb
798.28
C24H18F6N5PtSSb
839.33
C23H19F6N4OPtSSb
830.32
C25H21F6N4OPtSSb
856.36
Crystal size/mm
T/K
0.15 × 0.25 × 0.60
200
0.15 × 0.25 × 0.60
200
0.15 × 0.22 × 0.55
180
0.10 × 0.20 × 0.30
200
λ/Å
Crystal system
Space group
0.71073
Monoclinic
P21/n
0.71073
Monoclinic
P21/n
0.71073
Monoclinic
P21/n
0.71073
Monoclinic
P21/n
a/Å
b/Å
c/Å
α (°)
9.6158(2)
25.6118(7)
10.3103(3)
90
9.9301(2)
25.7960(9)
10.347(3)
90
9.7239(3)
25.4860(7)
10.4632(3)
90
9.7859(3)
25.3897(8)
10.9182(4)
90
β (°)
96.392(2)
90
2523.4(1)
4
2.101
6.753
100.060(2)
90
2609.7(8)
4
2.136
6.537
98.004(2)
90
2567.8(1)
4
2.148
6.644
98.355(3)
90
2684.0(2)
4
2.119
6.360
γ (°)
V/Å3
Z
Dc/g cm−3
μ/mm−1
F(000)
θ Range (°)
1496
2–25
1584
2–25
1568
2–25
1624
2–25
Reflections collected (independent)
Observed [I > 2σ(I)]
Rint
No. refined parameters (restraints)
Final R1 [I > 2σ(I)]
Final wR2 (all data)
Max, min Δρ/e Å−3
40 442 (9037)
6857
0.0317
316 (0)
0.0369
0.0821
1.86, −1.51
38 184 (8961)
5626
0.0423
344 (0)
0.0803
0.2107
5.66, −4.63
41 513 (9434)
6588
0.0395
339 (2)
0.0373
0.0862
3.02, −2.30
28 462 (5291)
5224
0.0338
352 (0)
0.0865
0.1643
3.36, −3.30
a Crystal selected from the batch of crystals obtained after allowing acetonitrile to evaporate from the crystals of 1·CH3CN that originally formed in
the crystal growth chamber: see 1_1st: CCDC 857048. b Crystal obtained by exposing 1_1st to vapours of acetonitrile. c Crystal obtained by exposing
a different single crystal of 1 to vapours of methanol. d Crystal grown directly from acetone.
the trpy moiety, as is reflected in a dihedral angle between the
phenyl ring and the central pyridine ring of only 3.1(1)°; in fact,
the cation as a whole is essentially planar. This is not an
expected result in view of a consequent very short intramolecular
H7⋯H17 non-bonded contact of 2.00 Å, a value that is ∼0.4 Å
less than the sum of the van der Waals radii for two hydrogen
atoms.35 Indeed, a crystal structure determination of the free 4′-
(Ph)trpy ligand shows that the phenyl group twists out of the
plane of the trpy moiety by 10.9°.36 On the other hand, two pre-
vious reports of crystal structure determinations of complexes
containing the 4′-(Ph)trpy ligand show that the out-of-plane
twist of the phenyl group is variable; thus whereas the relevant
torsion angle is 33.4° in [Pt{4′-(Ph)trpy}Cl]BF4·CH3CN,22 it
is only 1.9° in [Pt{4′-(Ph)trpy}(CN)]BF4·CH3CN.37 Clearly,
crystal packing forces play a key role in determining the exact
extent to which the phenyl group twists out of the plane of the
trpy moiety.
Table 2 Selected geometric parameters for the cations in 1 and its
solvates (Å/°)
Compound
Temperature/K
1
200
1·CH3CN
200
1·CH3OH
180
1·(CH3)2CO
200
Pt–N1
Pt–N2
Pt–N3
Pt–N4
N4–C22
C22–S
N1–Pt–N2
N1–Pt–N3
N1–Pt–N4
N2–Pt–N3
N2–Pt–N4
N3–Pt–N4
Pt–N4–C22
N4–C22–S
2.022(4)
1.925(3)
2.019(4)
2.023(4)
1.098(6)
1.629(5)
80.7(1)
162.1(1)
98.9(1)
81.3(1)
179.6(2)
99.0(2)
173.1(4)
179.5(5)
2.00(1)
1.920(8)
2.01(1)
1.996(9)
1.13(1)
1.63(1)
81.0(4)
162.8(4)
98.8(4)
81.9(4)
179.7(4)
98.4(4)
175.2(9)
177(1)
2.010(3)
1.923(3)
2.016(3)
2.014(3)
1.120(5)
1.623(5)
81.5(1)
162.4(1)
98.5(1)
80.9(1)
179.2(1)
99.1(1)
175.7(4)
179.2(5)
2.02(1)
1.93(1)
2.02(1)
2.01(1)
1.11(2)
1.64(2)
81.6(5)
162.6(5)
98.6(5)
81.0(5)
179.1(5)
98.8(5)
175.1(1)
180(2)
Aview down the [c]-axis of the unit cell contents in 1 is given
in Fig. 2; note that this view is chosen since it is approximately
perpendicular to the mean plane through the non-H atoms of the
cation, henceforth labeled the “cation plane”. This view shows
that the [Pt{4′-(Ph)trpy}(NCS]+ cations stack separately from the
the literature.13–23 For example, the irregular square-planar
geometry about the platinum atom is reflected in a N1–Pt–N3
‘trans’ angle of 162.1(1)°, and a platinum to central nitrogen dis-
tance that is significantly shorter than the distances of the plati-
num to the two outer pyridine nitrogen atoms. The SCN− ion is
bound to the platinum through the N atom with a Pt–N4–C22
angle of 173.1(4)° and a N4–C22–S angle of 179.5(5)°. These
angles, as well as the N4–C22 and C22–S distances of 1.098(6)
and 1.629(5) Å, respectively, are typical of a SCN− ion bound to
a metal through the N atom.34 Of particular interest is that the
phenyl ring is very nearly coplanar with the mean plane through
−
SbF6 anions, but not in one-dimensional columns parallel to
the [c]-axis. Instead, the cations slot together to give a stack of
constant thickness, but which extends along the [a]- and
[c]-axes; as shown in Fig. 2 these stacks repeat at half-intervals
along the [b]-axis and are separated by the anions. To discuss the
nature of the interactions between the cations within a stack
we refer to Fig. 3; in this view the cations within one stack
are viewed side-on, specifically along the longer [b]-axis. By
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 5486–5496 | 5489

View Article Online
Fig. 2 View down the [c]-axis of the unit cell contents in 1. The [a]-
axis points down the page and the [b]-axis across the page.
Fig. 4 Views perpendicular to the cation planes for each of the three
different types of cation–cation interaction in 1, see text. In (A) the
Pt⋯Pt distance is 3.553 Å, in (B) 11.757 Å, and in (C) 9.616 Å.
Fig. 3 Side view along the [b]-axis of the “brick-in-wall” packing
arrangement of the cations in 1. Hydrogen atoms have been omitted for
clarity. The “bricks” are coloured according the nature of the interaction
of a cation with the central “pink” cation, see text.
interplanar spacing of 3.34 Å is short enough to support a stabi-
lising π–π interaction: usually 3.8 Å is taken as the upper dis-
tance limit for this kind interaction between planar aromatic
species.9,10 Finally, in Fig. 4(C) (pink-to-blue) the cations are
displaced with respect to each other to such an extent that the
only overlap is one between the leading edges of the phenyl ring
of one cation and the outer pyridine ring of the adjacent cation.
This overlap geometry also supports a stabilising π–π interaction
as does the interplanar spacing of 3.27 Å:9,10 the Pt⋯Pt distance
is now 9.616(1) Å, clearly too long to support a metallophilic
focusing on one cation within the stack (colour coded pink) we
arrive at the following description of the various cation–cation
interactions. This cation is attracted to four nearest neighbour
cations, two above (colour coded green and blue) and two below
(colour coded blue and orange) with three unique pair-wise inter-
actions: pink-to-green, pink-to-blue and pink-to-orange. These
unique interactions are highlighted in Fig. 4, using the same
colour coding but applied to the borders and taking a view per-
pendicular to the cation planes. In Fig. 4(A) (pink-to-orange) the
two cations are laterally offset such that π–π overlap occurs at
the edges of both outer pyridine rings for the two adjacent trpy
moieties; as noted by Hunter and Sanders9 and by Janiak10 this
has as a result that π–σ attractive forces dominate between the
two aromatic rings. Consistent with this is that the perpendicular
distance between the cation planes (the interplanar spacing) is
3.38 Å, a value that has been calculated to be optimal for a stabi-
lising interaction between adjacent aromatic species.9,10 On the
other hand, the Pt⋯Pt distance of 3.553(1) Å is too long to
support a metallophilic dz2(Pt)–dz2(Pt) orbital interaction.38 In
Fig. 4(B) (pink-to-green) the cations are laterally and longitudin-
ally offset such that overlap now occurs (twice) between the
phenyl and central pyridine rings of the two cations. Again, this
overlap geometry favours a stabilising π–π interaction.9,10 The
Pt⋯Pt distance is now much longer at 11.757(1) Å, but the
d
z2(Pt)–dz2(Pt) orbital interaction.38 In summary, the cations
within a stack are held together by extended stabilising π–π inter-
actions, while the crystal as a whole is stabilised by attractive
electrostatic forces between the positively charged stacks and the
SbF6− anions that are interspersed between them.
As already noted, de-solvation of a single crystal of the aceto-
nitrile solvate, 1·CH3CN, proceeds without loss of the single
crystallinity to give the single crystal of 1 with the structure
described above. The net result is that the crystal structure of 1
contains solvent accessible voids. Using the MSROLL interface
of the program X-SEED,30 we estimate that the voids occupy a
volume of 320 Å3 in the unit cell, i.e. ca. 13% of the total unit
cell volume of 2523 Å3 measured at 200 K. In order to show the
location of the voids we have plotted the Connolly surface for 1
using a probe radius of 1.2 Å;39 three different views of the
surface are shown in Fig. 5. In views (A) and (B) the corkscrew
5490 | Dalton Trans., 2012, 41, 5486–5496
This journal is © The Royal Society of Chemistry 2012

View Article Online
Fig. 6 Plots showing two different views of the location of the solvent
molecules in 1·CH3CN, 1·CH3OH, and 1·(CH3)2CO in relation to the
Connolly surface (shown in green) obtained from the crystal structure
of 1.
This crystal was exposed to vapours of acetonitrile at its
vapour pressure in a closed container for a period of 48 h during
which time the solvent was sorbed by the crystal to afford
1·CH3CN, without loss of the single crystallinity. During the
sorption process the cation π-stacks remain intact with only very
Fig. 5 Three different views of the solvent accessible voids in 1 shown
as green Connolly surfaces drawn with a probe radius of 1.2 Å.39 In (A)
and (B) the “corkscrew” motif for the Connolly surface is emphasised
while view (C) shows the bulges that are connected by narrower
channels.
−
small changes to their dimensions; also the SbF6 anion
shape of the channels that extend through the crystal is apparent,
while view (C) shows that the channels are not of uniform cross-
section, but rather comprise narrow sections that connect larger
bulges. It is these larger bulges that accommodate molecules of
acetonitrile, methanol and acetone in the corresponding solvates,
1·CH3CN, 1·CH3OH and 1·(CH3)CO. We now discuss the
uptake of vapours of these solvents by single crystals of 1.
positions determined for 1 hardly change. As a result, the
packing of the cations and anions in 1·CH3CN (see Fig. S1 in
the ESI†) appears very similar to the crystal packing illustrated
in Fig. 2 for the de-solvated compound, 1. With regard to the
location of the solvent molecules in 1·CH3CN, we show in
Fig. 6 their positions in relation to the Connolly surface drawn
for 1; as expected, the acetonitrile molecules fit neatly into the
bulges of the Connolly surface implying little disruption of the
porous crystal structure of 1. As a result, the internal dimensions
and geometry of the cation are much the same as those deter-
mined from the crystal structure determination of 1, see Table 2.
More interesting is that there are no unusually short intermolecu-
lar contacts between atoms of the acetonitrile molecule and
atoms of the cation and anion, i.e. there is no evidence from the
crystal structure of 1·CH3CN for strong binding of the solvent in
the lattice, consistent with the ease with which acetonitrile mol-
ecules escape the crystal. Indeed, we have shown that this same
crystal can undergo loss of solvent for a second time without
loss of single crystallinity: see CCDC 857049.†
Crystal structures of 1·CH3CN, 1·CH3OH and 1·(CH3)2CO
The intensity data for the crystal structure determinations of the
three solvates were collected at low temperatures in order to
slow down the loss of solvent that occurs quite rapidly at room
temperature. The actual temperatures used were a compromise:
dropping the temperature too low caused the crystals to crack.
The single crystal used for the determination of the structure of
the acetonitrile solvate that we describe here was the same one
used for the crystal structure determination of 1, described
above.
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 5486–5496 | 5491

View Article Online
Table 3 Comparison of the three unique Pt⋯Pt distances (Å)
measured for 1 and its solvates
Compound
Temp (K)
1
200
1·CH3CN
200
1·CH3OH
180
1·(CH3)2CO
200
Pt⋯Pt(A)a
Pt⋯Pt(B)a
Pt⋯Pt(C)a
3.553(1)
11.757(1)
9.616(1)
3.519(1)
11.562(1)
9.930(1)
3.395(1)
11.754(1)
9.724(1)
3.443(1)
11.980(1)
9.786(1)
a A, B and C label the distances according to the nature of the cation–
cation interaction: see Fig. 4.
Single crystals of 1 have also been exposed to vapours of
acetone, which is the largest of the three solvent molecules
studied, containing four non-hydrogen atoms. We find that
acetone is sorbed, but not without some disruption of long range
order in the crystal. Evidence for this comes from a crystal struc-
ture determination of the solvate at 200 K. A better crystal of
1·(CH3)2CO was obtained by direct crystallization of the solvate
from a solution of 1 in acetone and a more reliable crystal struc-
ture determination made. As with the acetonitrile and methanol
solvates, a visual comparison of the relevant crystal packing dia-
grams (Fig. 2 with Fig. S3†) shows no obvious differences in the
positions of the cation and anions between the two crystal struc-
tures. Consistent with this observation is that the acetone mol-
ecules fit neatly into the bulges in the Connolly surface drawn
for 1, see Fig. 6. Noteworthy is that there is no evidence for
hydrogen bonding (potential or actual) of the O atom of the
acetone to the S atom of the isothiocyanate ligand; in fact, there
are no unusually short intermolecular contacts involving the
acetone molecules. However, because of the increased size of the
acetone molecule, its inclusion in the crystal does have a small
effect on the dimensions of the cation stack, in particular on the
Pt⋯Pt distances, see Table 3. The shortest (A) distance of 3.443
(1) Å is now just below the upper distance limit of ∼3.5 Å
usually taken to indicate a finite dz2(Pt)–dz2(Pt) orbital inter-
action,38 but is not short enough to have any discernable effect
on the emission measured for 1·(CH3)2CO in the solid state, see
below. In so far as the loss of acetone from a single crystal of
1·(CH3)2CO is concerned, this occurs relatively slowly under
ambient conditions (∼2 weeks) and there is no concomitant loss
of single crystallinity, as confirmed by a crystal structure deter-
mination: see CCDC 857050.† The longer de-solvation time is
unexpected in view of the high vapor pressure for acetone,
certainly higher than that for acetonitrile, which escapes more
quickly. We suggest that this has to do with the larger size of the
acetone molecule and that because it is larger, it becomes
more difficult for the acetone molecules to escape through the
narrow channels that link the larger bulges, see Fig. 5.
Fig. 7 ORTEP diagram (50% ellipsoids) showing the O–H⋯S inter-
molecular contact between the methanol and S atom of the isothiocya-
nate ligand: O–H = 1.00 Å, O⋯S = 3.31, S⋯H = 2.38 Å, O–H⋯S =
157°.
Turning to the sorption of methanol by single crystals of 1 we
find that, though single crystals of 1 can sorb methanol mol-
ecules from the vapour phase without loss of single crystallinity,
extrication of methanol from the crystal is much more difficult;
in fact, the application of a vacuum is required with a concomi-
tant loss of single crystallinity. We are not sure why this is the
case, especially as methanol has a higher vapour pressure than
acetonitrile, and is a smaller molecule (two non-hydrogen atoms
rather than three). Of course, the possibility of hydrogen
bonding exists between the methanol and the sulfur atom of the
isothiocyanate ligand; and such a bond would make the metha-
nol more difficult to extricate. To investigate this possibility we
have checked the shortest intermolecular contacts between the
methanol and the sulfur atom, see Fig. 7 and its caption, which
gives these parameters. In fact, the dimensions given do not
suggest the presence of an O–H⋯S hydrogen bond,40 for
example the O⋯S distance simply fits with the sum of the
van der Waals radii for the two atoms concerned.35 However,
the methanol is oriented with respect to the sulfur atom such
that even a very small shift of atom positions could lead to the
formation of a hydrogen bond – perhaps this happens when
attempts are made to remove the methanol. Turning to the
overall packing in this solvate we first note that there are no
obvious differences in the positions of the cation and anions as
compared to those in 1 and in the acetonitrile solvate; compare
Fig. S2† with Fig. 2 and Fig. S1.† Consistent with this ob-
servation is that the methanol molecules fit neatly into the bulges
in the Connolly surface drawn for 1, see Fig. 6. However, exam-
ination of the Pt⋯Pt distances shows that there is a small but sig-
nificant effect on these distances (Table 3) the most relevant
being on the Pt⋯Pt distance associated with the cation–cation
interaction labeled “A” in Fig. 4. The Pt⋯Pt(A) distance drops
from a value above 3.5 Å for 1 and 1·CH3CN to a value of
3.395(1) Å for 1·CH3OH, which is below the upper distance
Emission spectroscopy of solid 1, 1·CH3CN, 1·CH3OH
and 1·(CH3)2CO
Emission measurements and gas sorption studies were made
using a dry powder sample of 1. An XRD spectrum was
recorded on the powder and is compared in Fig. S4 in the ESI†
with the powder XRD spectrum calculated using the single
crystal data: the excellent agreement suggests that there is no
2
2
limit of ∼3.5 Å usually taken to indicate a finite dz (Pt)–dz (Pt)
orbital interaction.38 As expected, the presence of the metallophi-
lic interaction influences the photophysical properties of the
methanol solvate in the solid state, see below.
5492 | Dalton Trans., 2012, 41, 5486–5496
This journal is © The Royal Society of Chemistry 2012

View Article Online
phase change brought about by crushing the single crystals to
give the powder. This powder sample was then mixed with an-
hydrous KBr to give a 10% (w/w) solid solution – in order to
increase the surface area of the sample and so as to speed up gas
sorption and de-sorption. The emission spectrum recorded for 1
was found to be extremely weak with no observable signal at
room temperature – it was only by lowering the temperature to
77 K that a measurable spectrum could be obtained. In contrast,
the emission spectra recorded for the solvates were found to be
more intense, in particular those of the acetonitrile and acetone
solvates. These observations informed the following sequence in
which the gas sorption and emission measurements were done.
Dry samples of 1 were first exposed to vapours of acetonitrile
and acetone and the emission spectra of the corresponding sol-
vates recorded, at 77 K, in order to minimise any potential loss
of solvent and to maximise the intensity of the emission, see
Fig. 8 for the emission spectra of 1·CH3CN and 1·(CH3)2CO.
These are very similar, both comprising three bands at 550, 587
and 632 nm whose intensities decrease monotonically with an
increase in wavelength. This envelope of vibronic structure as
well as the vibrational spacing of ∼1200 cm−1 are a signature
for monomeric 3MLCT emission by a platinum terpyridine
complex.37,41–43 Similar emission lifetimes of ∼12 μs at 77 K
have been recorded for the emission by both solvates, but the
intensity of the emission by the acetonitrile solvate is higher
The following approach was used in order to obtain a reliable
emission spectrum of the de-solvated material, 1. The emission
spectrum of the acetonitrile solvate, 1·CH3CN, was recorded fol-
lowing which the sample was exposed to ambient conditions.
The loss of solvent was monitored by recording the emission
spectrum at regular intervals – at 77 K, so ensuring that each
spectrum is a “freeze-frame” of the de-solvation process. The
results are shown in Fig. 9. Initially, solvent loss causes the
3MLCT emission to gradually and systematically drop in inten-
sity, i.e. the relative intensities of the vibrational components
stay the same but the overall (integrated) intensity of the emis-
sion drops. However, when the de-solvation process is well-
advanced (see the insert spectra) it becomes apparent that while
the vibrational components at 550 and ∼590 nm continue to
drop in intensity, the component at ∼650 nm stays much the
same in intensity. Eventually, a very weak emission spectrum is
obtained that does not change even if the sample is pumped with
a high vacuum to remove any vestiges of solvent: this is the
77 K spectrum for the de-solvated compound, 1, and is shown
separately in Fig. 10. As the figure shows, the spectrum consists
of a broad band that maximises at 650 nm, with higher energy
shoulders at ∼550 and ∼590 nm. To aid in its interpretation we
have determined the emission lifetimes by monitoring the decay
in intensity with time of each of the three component peaks. The
decay curves determined at 550 and 590 nm are biphasic afford-
ing best fit lifetimes of 2.0 and 9.3 μs for the 500 nm peak, and
2.5 and 12.4 μs for the 590 nm peak. In contrast, the decay
curve for the dominant 650 nm peak is monophasic with a life-
time of 1.9 μs. These lifetime data suggest that 1 exhibits simul-
taneous emission from two different excited states, one with τ ∼
10 μs and the other with τ ∼ 2 μs. Since the bands at 550 and
3
than that for the acetone solvate. In commenting on the MLCT
emission exhibited by 1·CH3CN and 1·(CH3)2CO, we first note
that the crystal structures of the two solvates show very similar
arrangements for the cations and, therefore, similar emission
properties are to be expected. The enhanced intensity of the
emission by the acetonitrile solvate as compared to the acetone
solvate is presumably to do with slower non-radiative decay of
the excited state for the former as compared to the latter i.e. knr
for 1·CH3CN is <knr for 1·(CH3)2CO. The magnitude of a knr
value for a solid is usually linked to the extent of thermal
vibration by the atoms in the solid.44 On this basis, the atoms in
the acetone solvate vibrate on average more than the atoms in the
acetonitrile solvate, as might be expected since the larger
acetone molecules disrupt the efficiency of packing more than
the acetonitrile molecules.
3
590 nm coincide with the higher energy MLCT bands recorded
for the two solvates (as shown clearly by the monitoring exper-
iment, see Fig. 9) we assign one emitting state as 3MLCT in
origin. Consistent with this assignment is that the lifetime
3
measured independently for the MLCT emission by the aceto-
nitrile solvate (∼12 μs, vide supra) fits with the lifetimes of
Fig. 9 Emission spectra of various levels of solvation of 1 by aceto-
nitrile, starting with 1·CH3CN (red line on main plots) and ending at 1
(purple line on insert plots). The intensities of the spectra are such that
the least intense spectrum on the main plots is slightly more intense than
the most intense spectrum on the insert plots.
Fig. 8 Emission spectra recorded at 77 K on solid samples of
1·CH3CN (red) and 1·(CH3)2CO (blue). λex = 420 nm.
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 5486–5496 | 5493

View Article Online
Fig. 10 Emission spectrum of a crystalline powder sample of 1
recorded at 77 K. λex = 420 nm.
Fig. 11 Emission spectrum recorded at 77 K on a solid sample of
1·CH3OH. λex = 420 nm.
∼10 μs measured for the 550 and 590 nm bands in the emission
spectrum of 1. We assign the band at 650 nm to emission from
an excimeric 3π–π* excited state. Consistent with this assignment
is that the emission is broad, at lower energy, weak and poorly
resolved.13,14,17,18,20,45 In summary, irradiation of compound 1
leads to emission from two distinct excited states, one 3MLCT in
normal range of 1200–1400 cm−1 for the relevant stretching
vibrations of a terpyridyl ligand.37,41–43 These observations lead
to the inevitable conclusion that the methanol solvate, unlike the
acetonitrile and acetone solvates, does not exhibit vibronically
3
structured MLCT emission in the solid state, at least not solely
3
3MLCT emission. Yet, the peak at 550 nm does coincide with
origin and other excimeric π–π* in origin, with the latter more
intense than the former.
3
the 550 nm band assigned to the 0–0 component of the MLCT
To conclude the discussion of the emission spectra of the
acetonitrile and acetone solvates and of 1, we make the following
observations. As shown in Fig. 9, the 3MLCT emission that
characterises the acetonitrile and acetone solvates systematically
drops in intensity as solvent molecules are lost from the solid.
The reverse is also true: when a sample of 1 is exposed to
emission by 1·CH3CN and 1·(CH3)2CO, see Fig. 8. To aid in
the interpretation of the emission by the methanol solvate we
have determined the emission lifetimes by monitoring the decay
in intensity with time of each of the two bands. The decay
curves determined for the 550 and 607 nm bands are biphasic
affording best fit lifetimes of 3.3 and 14.2 ms for the 550 nm
band, and 2.3 and 15.5 ms for the 607 nm band, respectively.
The longer lifetimes fit approximately with the lifetimes of
3
vapours of acetonitrile or acetone the MLCT component of the
emission systematically intensifies. Note, this does not apply to
3
3
the excimeric π–π* component of the emission, which remains
∼12 ms measured for the MLCT emission by the acetonitrile
3
very weak whether solvent molecules are present or not. We do
not have an explanation for the different responses by the
and acetone solvates, further suggesting that there is a MLCT
contribution to the emission by the methanol solvate. We now
turn our attention to the peak at 607 nm. This peak is signifi-
cantly blue-shifted and narrower than the peak at 650 nm that
3
3MLCT and excimeric π–π* emission to small molecule sorp-
tion and de-sorption. However, the geometries of the two excited
states must be different, and on this basis also the mechanisms
of their de-activation: clearly the 3MLCT state is stabilised to de-
activation by the presence of solvent molecules whereas the exci-
3
was assigned to excimeric π–π* emission for the de-solvated
compound, 1, see Fig. 10. Therefore, it is highly unlikely that
3
excimeric emission π–π* is solely responsible for the peak at
3
meric π–π* emission is not. Whatever the precise mechanisms,
607 nm. Thus, the relatively intense peak at 607 nm is new and,
being new, leads one to conclude that the peaks at 550 and
607 nm have different origins, i.e. 1·CH3OH exhibits simul-
taneous emission from (at least) two separate excited states in the
solid state.
the result is that the sorption–de-sorption of acetonitrile and
acetone by 1 can be easily monitored by emission spectroscopy.
In fact, using this approach we have shown that the uptake and
loss of these two solvent vapours by a powder sample of 1 can
be cycled through indefinitely, specifically at least 10 times.
Whether this is in general true, i.e. whether sorption–de-sorption
of other small molecules will induce the same response, will
have to be the subject of further investigations.
The emission spectrum recorded at 77 K for the methanol
solvate, 1·CH3OH, is shown in Fig. 11; at room temperature the
emission is barely detectable. The emission is characterised by
two distinct peaks, one at 550 nm and another somewhat more
intense and rather broad peak at 607 nm. Clearly, the spectrum is
very different to the well-defined spectra measured for the aceto-
nitrile and acetone solvates, see Fig. 8. Also different is that the
peak separation is ∼1800 cm−1, a value which falls outside the
We now address the question of the origin of the multiple
emission by 1·CH3OH. The crystal structure determination of
1·CH3OH at 180 K is helpful in this regard, vide supra. This
structure determination shows that there is a short Pt⋯Pt dis-
tance of 3.395(1) Å linking the two platinum atoms of the pair
labeled “A” in Fig. 5, a distance well within the limit of ∼3.5 Å
for a finite dz2(Pt)–dz2(Pt) orbital interaction.38 On this basis, we
3
would anticipate that a MMLCT excited state would contribute
to the emission by 1·CH3OH and that, therefore, the “new” peak
3
at 607 nm in Fig. 10 has its origins in MMLCT emission. As
already implied, the peak at 550 nm is associated with the mono-
3
meric MLCT emission that typifies emission by 1·CH3CN and
5494 | Dalton Trans., 2012, 41, 5486–5496
This journal is © The Royal Society of Chemistry 2012

View Article Online
1·(CH3)2CO. To test this assumption we have de-convoluted the
spectrum in Fig. 11 into constituent Gaussian bands using
standard non-linear regression methods; the results are shown in
Fig. S5.† A total of five bands are mandatory in order to provide
a suitable fit of the experimental emission spectrum of the
sample. The Gaussian bands coloured blue, light blue and pink
are centred at 550, 585 and 631 nm and, moreover, decrease
monotonically in intensity as the wavelength increases. This
the difference – the cations stay in position and do not move
when the material is exposed to vapours of small molecules such
as acetonitrile. Or, to put it in another way, if acetonitrile is lost
from the solvate, 1·CH3CN, the cations stay fixed in position
and an open porous crystal structure is obtained.
Acknowledgements
3
profile, as well as the band energies, match that of the MLCT
The authors thank the South African National Research Foun-
dation and the University of KwaZulu-Natal for financial
support. B. P. W. thanks the Department of Labour for a Scarce
Skills bursary.
emission exhibited by 1·CH3CN and 1·(CH3)2CO, see Fig. 8.
The fourth band of interest, coloured green, is centred at 609 nm
with a fwhm value of 1590 cm−1, i.e. it is relatively narrow. This
band is, both in terms of energy and profile, typical of the kind
of emission exhibited by planar complexes of platinum(II)
that have crystal structures containing Pt2 dimers: see for
example [Pt(trpy)Cl]CF3SO3 (λem
References
max
= 625 nm at 77 K),13
1 (a) O. M. Yaghi, M. O’Keeffe, N. W. Ockwig, H. K. Chae, M. Eddaoudi
and J. Kim, Nature, 2003, 423, 705; (b) A. K. Cheetham, C. N. R. Rao
and R. K. Feller, Chem. Commun., 2006, 4780; (c) J. W. Steed and
J. L. Atwood, Supramolecular Chemistry, Wiley, Chichester, UK, 2009,
2nd edn, pp. 538–589; (d) S. Kitagawa and R. Matsuda, Coord. Chem.
Rev., 2007, 251, 2490.
2 P. M. Forster and A. K. Cheetham, Top. Catal., 2003, 24, 79.
3 (a) J. L. C. Roswell, A. R. Millward, K. S. Park and O. M. Yaghi, J. Am.
Chem. Soc., 2004, 126, 5666; (b) T. Duren, L. Sarkisov, O. M. Yaghi and
R. Q. Snurr, Langmuir, 2004, 20, 2684; (c) J. L. C. Roswell and
O. M. Yaghi, Angew. Chem., Int. Ed., 2005, 44, 4670; (d) J. A. Rood,
B. C. Noll and K. W. Henderson, Inorg. Chem., 2006, 45, 5521.
4 (a) A. de Bettencourt-Dias, Inorg. Chem., 2005, 44, 2734; (b) X. D. Gou,
G. S. Zhu, Q. R. Fang, M. Xue, G. Tian, J. Y. Sun, X. T. Li and
S. L. Qiu, Inorg. Chem., 2005, 44, 3580.
[(2,6-di-(2′-naphthyl)-4-pyridine)Pt2{μ-bis(diphenylphosphino)-
max
methane}] (λem
= 602 nm) and its solvates.46 We therefore
3
assign it as MMLCT in origin. Finally, the fifth band, coloured
red, is very broad and centered at 652 nm. Of course, this band
is reminiscent of the 650 nm band observed in the emission
spectrum of the de-solvated compound, 1, see Fig. 10. We con-
clude that the emission spectrum of 1·CH3OH, though compli-
cated, is understandable in terms of the simultaneous emission
from 3MLCT, 3MMLCT and excimeric 3π–π* excited states.
Conclusions
5 G. Férey, Chem. Mater., 2001, 13, 3084.
The crystal structure of 1 is unusual in that it is porous with an
open framework that is constructed from repeating 2-dimensional
π-stacks of planar [Pt{4′-(Ph)trpy}(NCS)]+ cations interspersed
6 (a) K. T. Holman, A. M. Pivovar, J. A. Swift and M. D. Ward, Acc.
Chem. Res., 2001, 34, 107; (b) O. R. Evans and W. B. Lin, Acc. Chem.
Res., 2002, 35, 511.
7 K. Uemura, K. Saito, S. Kitagawa and H. Kita, J. Am. Chem. Soc., 2006,
128, 16416.
−
with SbF6 anions; far more common are coordination com-
pounds that are porous but where the open structure comprises a
reticular framework of metal ions linked by bridging organic
ligands.1 The π-stacks are stabilised by extended π–π inter-
actions. These involve relatively weak electrostatic forces,9,10 yet
compound 1 reversibly sorbs and de-sorbs small molecules
without disruption of the open framework structure – provided,
of course, that the intermolecular forces between the molecules
and the cations (or anions) are of the weak van der Waals type.
Note that this is not true of the closely-related (parent) com-
pound, [Pt(trpy) (NCS)]SbF6, for which there is no substituent
in the 4′-position of the terpyridyl ligand. This compound is red
(not yellow) and also sorbs and de-sorbs small molecules, in par-
ticular acetonitrile, pyridine and dimethylformamide.24 However,
the sorption–de-sorption process is accompanied by a reversible
red-to-yellow colour change. This colour change can be attribu-
ted to the breaking and making of metallophilic Pt⋯Pt inter-
actions in the solid state.24 The point is that [Pt(trpy)(NCS)]
SbF6 has a space-filling crystal structure that is disrupted on
small molecule sorption: to be precise, the planar [Pt(trpy)
(NCS)]+ cations slide into new positions when molecules are
sorbed or desorbed.24 Clearly this does not apply to the planar
[Pt{4′-(Ph)trpy}(NCS)]+ cations in compound 1 – they hardly
move when molecules such as acetonitrile are reversibly sorbed.
The reason has to do with the 4′-phenyl substituent, in particular
that it participates in a stabilising π–π interaction with the outer
pyridine ring of an adjacent cation, see Fig. 4(C). This inter-
action, not possible for the [Pt(trpy)(NCS)]+ cation, makes all
8 S.-I. Noro, S. Kitagawa, T. Nakamura and T. Wada, Inorg. Chem., 2005,
44, 3960.
9 C. A. Hunter and J. K. M. Sanders, J. Am. Chem. Soc., 1990, 112, 5525.
10 C. Janiak, J. Chem. Soc., Dalton Trans., 2000, 3885.
11 (a) B. Chatterjee, J. C. Noveron, M. J. E. Resendiz, J. Liu, T. Yamamoto,
D. Parker, M. Cinke, C. V. Nguyen, A. M. Arif and P. J. Stang, J. Am.
Chem. Soc., 2004, 126, 10645; (b) T. Jacobs, J.-A. Gertenbach,
D. Dinabandhu and L. J. Barbour, Aust. J. Chem., 2010, 63, 573;
(c) A. W. Kleij, M. Kuil, D. M. Tooke, M. Lutz, A. L. Spek and
J. N. H. Reek, Chem.–Eur. J., 2005, 11, 4743; (d) S.-S. Sun and
A. J. J. Lees, J. Am. Chem. Soc., 2000, 122, 8956; (e) L. Dobrzanska,
G. O. Lloyd, H. G. Raubenheimer and L. J. Barbour, J. Am. Chem. Soc.,
2005, 127, 13134; (f) L. Dobrzanska, G. O. Lloyd, H. G. Raubenheimer
and L. J. Barbour, J. Am. Chem. Soc., 2006, 128, 698;
(g) L. Dobrzanska, G. O. Lloyd, C. Esterhuysen and L. J. Barbour,
Angew. Chem., Int. Ed., 2006, 45, 5856; (h) S. Belanger, J. T. Hupp,
C. L. Stern, R. V. Slone, D. F. Watson and T. G. Carrell, J. Am. Chem.
Soc., 1999, 121, 557; (i) M. Albrecht, M. Lutz, A. L. Spek and G. van
Koten, Nature, 2000, 406, 970; ( j) Z. Huang, P. S. White and
M. Brookhart, Nature, 2010, 465, 598; (k) C. D. Jones, J. C. Tan and
G. O. Lloyd, Chem. Commun., 2012, 48, 2110.
12 S. P. Anthony, C. Delaney, L. Varughese, L. Wang and S. M. Draper,
CrystEngComm, 2011, 13, 6706.
13 H.-K. Yip, L.-K. Cheng, K.-K. Cheung and C.-M. Che, J. Chem. Soc.,
Dalton Trans., 1993, 2933.
14 J. A. Bailey, M. G. Hill, R. E. Marsh, V. M. Miskowski, W. P. Schaefer
and H. B. Gray, Inorg. Chem., 1995, 34, 4591.
15 H.-S. Lo, S.-K. Yip, N. Zhu and V. W.-W. Yam, Dalton Trans., 2007,
4386.
16 J. S. Field, J.-A. Gertenbach, R. J. Haines, L. P. Ledwaba, N. T. Mashapa,
D. R. McMillin, O. Q. Munro and G. C. Summerton, Dalton Trans.,
2003, 1176.
17 J. S. Field, R. J. Haines, D. R. McMillin and G. C. Summerton, J. Chem.
Soc., Dalton Trans., 2002, 1369.
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 5486–5496 | 5495

View Article Online
18 J. S. Field, L. P. Ledwaba, O. Q. Munro and D. R. McMillin, CrystEng-
Comm, 2008, 10, 740.
19 T. J. Wadas, Q.-M. Wang, Y.-J. Kim, C. Flaschenreim, T. N. Blanton and
R. Eisenberg, J. Am. Chem. Soc., 2004, 126, 16841.
20 R. Büchner, J. S. Field, R. J. Haines, C. T. Cuninngham and
D. R. McMillin, Inorg. Chem., 1997, 36, 3952.
21 P. Du, J. Schneider, W. W. Brennessel and R. Eisenberg, Inorg. Chem.,
2008, 47, 69.
22 R. Büchner, C. T. Cunningham, J. S. Field, R. J. Haines, D. R. McMillin
and G. C. Summerton, J. Chem. Soc., Dalton Trans., 1999, 711.
23 R. Büchner, J. S. Field, R. J. Haines, L. P. Ledwaba, R. McGuire Jr.,
D. R. McMillin and O. Q. Munro, Inorg. Chim. Acta, 2007, 360, 1633.
24 J. S. Field, C. D. Grimmer, O. Q. Munro and B. P. Waldron, Dalton
Trans., 2010, 39, 1558.
25 B. P. Waldron, MSc thesis, University of KwaZulu-Natal, 2009.
26 Photon Technologies International, Birmingham, NJ 08011, USA,
Felix32© Analysis version 1.1 (build 51 Beta 8), program for the analysis
of emission spectra, 2003.
34 (a) C. Mantel, C. Baffert, I. Romero, A. Deronzier, J. Pecaut,
M.-N. Collomb and C. Duboc, Inorg. Chem., 2004, 43, 6455;
(b) S. A. Cotton, V. Franckevicius, R. E. How, B. Ahrens, L. L. Ooi,
M. F. Mahon, P. R. Raithby and S. J. Teat, Polyhedron, 2003, 22, 1489;
(c) D. V. Naik and W. R. Scheidt, Inorg. Chem., 1973, 12, 272;
(d) J. Rall, F. Weingart, D. M. Ho, M. J. Heeg, F. Tisato and E. Deutsch,
Inorg. Chem., 1994, 33, 3442; (e) R. C. T. Rojo, L. Lezama,
M. I. Arriortua, K. Urtiaga and G. Villenueve, J. Chem. Soc., Dalton
Trans., 1991, 1779.
35 F. A. Cotton, G. Wilkinson and P. L. Gaus, in Basic Inorganic Chemistry,
Wiley Press, New York, 3rd edn, 1995, p. 98.
36 E. C. Constable, J. Lewis, M. C. Liptrot and P. R. Raithby, Inorg. Chim.
Acta, 1990, 178, 47.
37 J. S. Field, R. J. Haines, L. P. Ledwaba, R. McGuire, Jr., O. Q. Munro,
M. R. Low and D. R. McMillin, Dalton Trans., 2007, 192.
38 D. S. Martin, in Extended Interactions between Metal Ions, ed.
L. V. Interannte, ACS Symp. Ser. 5, American Chemical Society,
Washington, DC, 1974, p. 254.
27 Oxford Diffraction Ltd., Abingdon, Oxford, OX14 1RL, UK, 2003.
28 G. M. Sheldrick, SHELXS-97 and SHELXL-97, University of Göttingen,
1997; G. M. Sheldrick, Acta Crystallogr., Sect A, 2008, 64, 112.
29 L. J. Farrugia, ORTEP3 for Windows, Department of Chemistry, Univer-
sity of Glasgow, Glasgow, G12 8QQ, UK, 2001; L. J. Farrugia, J. Appl.
Crystallogr., 1997, 30, 565.
30 (a) L. J. Barbour, X-SEED Version 2.0, Program for the Analysis of
Single Crystal Structure Determinations, University of Stellenbosch,
South Africa, 1999; (b) M. L. Connolly, Molecular Surface Package
Version 3.9.3, Program for the Determination of Atom (Solvent) Accessi-
ble Areas and Volumes: see: M. L. Connolly, J. Mol. Graphics, 1993, 11,
139.
31 J. L. Burmeister and F. Basolo, Inorg. Chem., 1964, 3, 1587.
32 (a) J. L. Burmeister, Coord. Chem. Rev., 1966, 1, 205;
(b) J. L. Burmeister, Coord. Chem. Rev., 1968, 3, 225.
33 (a) M. J. Coyer, M. Croft, J. Chen and R. H. Herber, Inorg. Chem., 1992,
31, 1752; (b) M. J. Coyer, R. H. Herber, J. Chen, M. Croft and S. P. Szu,
Inorg. Chem., 1994, 33, 716.
39 M. L. Connolly, Science, 1983, 221, 709.
40 J. Hartung, I. Svoboda and M. T. Duarte, Acta Crystallogr., Sect. C,
1997, 53, 1631.
41 D. R. McMillin and J. J. Moore, Coord. Chem. Rev., 2002, 229,
113.
42 J. F. Michaelec, S. A. Bejune, D. G. Cuttell, G. C. Summerton,
J.-A. Gertenbach, J. S. Field, R. J. Haines and D. R. McMillin, Inorg.
Chem., 2001, 40, 2193.
43 M. H. Wilson, L. P. Ledwaba, J. S. Field and D. R. McMillin, Dalton
Trans., 2005, 2754.
44 A. Islam, N. Ikeda, K. Nozaki, Y. Okamoto, B. Gholamkhass,
A. Yoshimura and T. Ohno, Coord. Chem. Rev., 1986, 171, 355.
45 (a) K.-T. Yam, C.-M. Che and K.-C. Cho, J. Chem. Soc., Dalton Trans.,
1991, 1077; (b) H. Kunkely and A. Vogler, J. Am. Chem. Soc., 1990,
112, 5625; (c) V. M. Miskowski and V. H. Houlding, Inorg. Chem., 1989,
28, 1529.
46 S. C. F. Kui, S. S.-Y. Chui, C.-M. Che and N. Zhu, J. Am. Chem. Soc.,
2006, 128, 8297.
5496 | Dalton Trans., 2012, 41, 5486–5496
This journal is © The Royal Society of Chemistry 2012