Phosphorescent, liquid-crystalline complexes of platinum(ii): Influence of the β-diketonate co-ligand on mesomorphism and emission properties

Article abstract of DOI:10.1039/c2dt31525d

Two series of rod-like ortho-platinated complexes bound to a range of β-diketonate co-ligands are reported of the form [Pt(N^C)(O^O)]. The liquid crystal and luminescent properties are influenced through the introduction of different β-diketonates (O^O) as well as through the presence of a fused cyclopentene ring on the N^C-coordinated 2-phenylpyridine ligand. For metal complexes, smectic phases are dominant but no mesomorphism is exhibited when the β-diketonate is hexafluorinated acetylacetonate (1,1,1-5,5,5-hexafluoro-2, 4-pentandione). The introduction of an unsymmetrical trifluoroacetyl acetonate (1,1,1-trifluoro-2,4-pentandione) ligand is particularly interesting due to the discovery of the β-diketonate as a dynamic system that readily isomerises to a 1:1 ratio under thermal conditions, from an initial ratio of 3:1 formed during the synthesis. As expected, the presence of the 3,5-heptanedionato co-ligand decreases the transition temperatures due to the introduction of a larger lateral substituent. Unfortunately complexes based on 2-phenylpyridine ligands with a fused cyclopentene ring are, with one exception, not mesomorphic. With the exception of the hexafluoroacetyl acetonate complexes, the materials are brightly luminescent and have excited state lifetimes between 13 and 30 μs with emission quantum efficiencies exceeding 0.5, with one as high as 0.7. A time-dependent density functional theory (TD-DFT) study suggests that the lack of emission from the hexafluoroacetyl acetonate complexes may be associated with large structural distortion upon formation of the excited state, as well as to poorer overlap of orbitals leading to a lower radiative rate constant. The same analysis provides an explanation for the apparently quite different emission efficiencies of the two isomers of the trifluoroacetyl acetonate complexes.

Full text of DOI:10.1039/c2dt31525d

View Article Online / Journal Homepage / Table of Contents for this issue
Dalton
Transactions
An international journal of inorganic chemistry
www.rsc.org/dalton
Volume 41 | Number 47 | 21 December 2012 | Pages 14213–14440
ISSN 1477-9226
COVER ARTICLE
Bruce et al.
Phosphorescent, liquid-crystalline complexes of platinum(II): influence of the
β-diketonate co-ligand on mesomorphism and emission properties
1477-9226(2012)41:47;1-H

Dynamic ArtVicielweALritnicklesOnline
Dalton
Transactions
Cite this: Dalton Trans., 2012, 41, 14244
www.rsc.org/dalton
PAPER
Phosphorescent, liquid-crystalline complexes of platinum(II): influence of the
β-diketonate co-ligand on mesomorphism and emission properties†‡
Matthew Spencer,a Amedeo Santoro,a Gemma R. Freeman,b Álvaro Díez,a Paul R. Murray,c Javier Torroba,a
Adrian C. Whitwood,a Lesley J. Yellowlees,c J. A. Gareth Williamsb and Duncan W. Bruce*a
Received 11th July 2012, Accepted 30th August 2012
DOI: 10.1039/c2dt31525d
Two series of rod-like ortho-platinated complexes bound to a range of β-diketonate co-ligands are
reported of the form [Pt(N^C)(O^O)]. The liquid crystal and luminescent properties are influenced
through the introduction of different β-diketonates (O^O) as well as through the presence of a fused
cyclopentene ring on the N^C-coordinated 2-phenylpyridine ligand. For metal complexes, smectic phases
are dominant but no mesomorphism is exhibited when the β-diketonate is hexafluorinated acetylacetonate
(1,1,1-5,5,5-hexafluoro-2,4-pentandione). The introduction of an unsymmetrical trifluoroacetyl acetonate
(1,1,1-trifluoro-2,4-pentandione) ligand is particularly interesting due to the discovery of the β-diketonate
as a dynamic system that readily isomerises to a 1 : 1 ratio under thermal conditions, from an initial ratio
of 3 : 1 formed during the synthesis. As expected, the presence of the 3,5-heptanedionato co-ligand
decreases the transition temperatures due to the introduction of a larger lateral substituent. Unfortunately
complexes based on 2-phenylpyridine ligands with a fused cyclopentene ring are, with one exception, not
mesomorphic. With the exception of the hexafluoroacetyl acetonate complexes, the materials are brightly
luminescent and have excited state lifetimes between 13 and 30 μs with emission quantum efficiencies
exceeding 0.5, with one as high as 0.7. A time-dependent density functional theory (TD-DFT) study
suggests that the lack of emission from the hexafluoroacetyl acetonate complexes may be associated with
large structural distortion upon formation of the excited state, as well as to poorer overlap of orbitals
leading to a lower radiative rate constant. The same analysis provides an explanation for the apparently
quite different emission efficiencies of the two isomers of the trifluoroacetyl acetonate complexes.
that liquid-crystal order, and an organised supramolecular struc-
Introduction
ture in general, is expected to improve charge mobility,2 while
Sagara and Kato,3 and Kozhevnikov et al.4 have independently
There are several different devices where luminescence or charge
transport is one of the key points for development required for
profitable application in the marketplace. Indeed, in the last ten
years, significant effort has been invested in research into new
materials for, inter alia, organic light emitting diodes (OLEDs),
thin-film transistors (TFT devices), organic solar cells, new
photo-active materials, bio-active molecules, charge-transport
devices, active components for imaging and data treatment.1 The
cooperativity of physical properties could improve the efficiency
of these devices greatly and it is in this environment that our
research is mainly focused. For example, it has been reported
demonstrated in recent studies that liquid-crystal order can
influence the emission of materials under mechanical stimuli,
thermal phase transitions or both. From this point of view and in
particular because of the greater efficiency of emission from
many metals complexes that can access the triplet manifold,
phosphorescent, metal-containing liquid crystals represent an
interesting target and challenge.5
Square planar ortho-metallated platinum complexes are a
group of molecules that have been studied widely in several
different research fields.6,7 These complexes have proved
popular in part because their chemical and physical properties
can be tuned by systematic change or modification of the ligands
around the metal centre. In particular, the fact that the HOMO
and LUMO tend to be localised on different parts of the
molecule allows their energies to be controlled relatively in-
dependently.8 Although numerous examples of brightly emissive
platinum-containing materials have emerged over recent years,6,7
only a few reports describe a combination of liquid crystal and
luminescent properties.3,5c,9 In this context, we recently reported
very efficient luminescent phenylpyridine platinum complexes
bearing acetylacetonate as the co-ligand [Pt(N^C)(acac)],9d with
aDepartment of Chemistry, University of York, Heslington, York YO10 5DD,
UK. E-mail: duncan.bruce@york.ac.uk; Tel: +44 (0)1904 324085
bDepartment of Chemistry, University of Durham, University Science
Laboratories, South Road, Durham, DH1 3LE, UK
cSchool of Chemistry, Joseph Black Building, University of Edinburgh,
West Mains Road, Edinburgh, EH9 3JJ, UK
†Dedicated to Professor David Cole-Hamilton on the occasion of his
retirement and to recognise his outstanding contributions to chemistry.
‡Electronic supplementary information (ESI) available. CCDC 885300
and 884157. For ESI and crystallographic data in CIF or other electronic
format see DOI: 10.1039/c2dt31525d
14244 | Dalton Trans., 2012, 41, 14244–14256
This journal is © The Royal Society of Chemistry 2012

View Article Online
emission quantum efficiencies > 0.5, the highest yet reported for
a ppy derivative of platinum; the materials were also liquid crys-
tals. As part of this study, we also demonstrated the influence of
a cyclopentane ring fused onto the pyridine ring on the lumines-
cence and liquid crystallinity.
while the molecular structure of the one of the complexes from
the asymmetric unit (prepared using a cif file with the disorder
removed) is shown as Fig. 1.
Comparison of the four bond lengths to Pt (Table 2) shows
that, within the confidence level of the ESDs, there are no differ-
ences between the two complexes in the asymmetric unit and,
while it is a close-run thing in one case, there is also no statisti-
cally significant difference between any of the Pt–N and Pt–C
distances. However, this does not extend to the Pt–O bond
lengths where in each case, the bond length trans to the Pt–C
bond is longer, as expected on the basis of trans-influence.
Analysis of intermolecular contacts at the level of sum of the
van der Waals’ radii minus 0.1 Å shows none that could be
regarded as structure directing.
We have now extended this investigation into the influence of
the O^O-coordinated β-diketonate co-ligand on liquid crystal and
luminescent properties. Specifically, we have examined hexa-
fluoroacetylacetonate (hfac), trifluoroacetylacetonate (tfac) and
heptanedionate (hdn) in place of the acetylacetonate (acac)
ligand of the previous studies. Two series of cyclometallated
platinum complexes have been synthesised, with and without
cyclopentane rings fused to the pyridyl ring, bearing these three
different O^O-diketonate ligands groups. The synthesis, liquid
crystal and luminescence properties of these new compounds are
described here, together with the crystal structure of two of the
compounds and an EPR study used to rationalise the behaviour
of the hexafluorinated complexes. Density functional theory
(DFT) and time-dependent DFT (TD-DFT) are used to shed
light on the significant effect of the O^O co-ligand on the
ground and excited state properties of the complexes.
Liquid crystal properties of the ligands and complexes
The detail of the mesomorphic behaviour of the ligands was
reported earlier.9d Thus, ligands without the fused five-mem-
bered ring showed a variety of smectic phases with clearing
points up to almost 240 °C, while when the five-membered ring
was present, only a nematic phase was seen with all clearing
points lower than 115 °C. Such a change in behaviour is entirely
consistent with the introduction of a bulky, lateral substituent. In
terms of the complexes, binding of Pt and acac simplified the
mesomorphism a great deal so that only SmA phases were
observed.
In the current work, use of the heptanedionato co-ligand with
the unsubstituted phenylpyridine ligands also gave complexes
that showed a SmA phase, but the extra flexibility of the lateral
ligand affected the transition temperatures (Table 3). Thus,
melting points were typically suppressed by around 50 °C, while
clearing points dropped by about 25 °C. For example, with acac,
the C10 homologues showed Cr ·168 · SmA · 244 · Iso, while
with hdn the transitions were Cr · 116 · SmA · 217 · Iso. The
greater drop in the melting point meant that the liquid crystal
range also increased (76 to 101 °C in this example).
However, the situation for the tfac co-ligands was more com-
plicated owing to the presence of isomers (Fig. 2). Previously,
we had studied ortho-metallated benzylidene aniline complexes
of Pd and Pt with a tfac co-ligand and had come across the same
issue of isomers, but in each case, the synthesis led to a
1 : 1 mixture which behaved as a single material implying, coin-
cidentally, that the 1 : 1 mixture was almost eutectic in nature.11
However, with the present materials, the situation was somewhat
different.
Thus, it was found that simple synthesis of the complexes led
to a 3 : 1 mixture of two isomers, represented in Fig. 2, as evi-
denced by the appearance in the 1H NMR spectrum of two sing-
lets at δ = 5.80 (intensity 3) and 5.84 (intensity 1) from the
γ (methine) hydrogen. When the mixtures were heated, the C8,
C10 and C12 homologues showed a SmA phase, while the
C6 material showed a nematic phase. However, what was inter-
esting was that it was clear that the two isomers interconverted in
the melt so that after some time in the mesophase the ratio of the
two was 1 : 1. This was investigated in a little more detail for
complex 6–8, which melted totally to the SmA phase at 147 °C.
Thus, a dry sample (3 : 1 ratio of isomers) was heated at various
Results and discussion
Synthesis
The complexes were prepared according to the scheme shown as
Scheme 1, which follows exactly the methodology reported
earlier.9d Thus, the ligands (1) were prepared from 1,2,4-triazine
precursors using an inverse electron demand Diels–Alder proto-
col and were then reacted with K2[PtCl4] in HOAc to give the
rather insoluble di-μ-chloro dimer (2). Heating the dimers in
dmso led to the intermediate complexes with coordinated dmso
(3), which were very soluble and gave good conversion to the
target complexes (4 to 11) when reacted with the sodium salts of
the β-diketone. It is noted in passing that attempts to prepare the
complex containing bound dmso directly from the ligand and
cis-[PtCl2(S-dmso)2] resulted in oxidation to PtIII when carried
out in HOAc.10 The range of complexes synthesised is shown in
Scheme 1.
Crystal and molecular structure of complex 5–6
In the case of 5–6, it was possible to obtain crystals suitable for
a single crystal diffraction study, the results of which are now
described.
Thus, the complex crystallised in the monoclinic space group
P21/n with two complexes in the asymmetric unit (Fig. S1‡).
There was evidence of twinning in the crystal which led to sig-
nificant residual peaks in the difference map. These were close
to C9, C33, C45 and C69 and were believed to correspond to the
sites of the platinum atoms in the twins. Identical twinning was
observed for a data set from a different crystal, which means that
the twinning is a systematic process inherent in formation of the
crystals. The twinning caused the ADP of C9, C33, C45 and
C69 to be distorted and so these atoms were restrained to be
approximately isotropic. One of the methyl groups of the acac
(C66) was disordered and modelled in two positions with a
refined occupancy of 66 : 34. Key data are found in Table 1,
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 14244–14256 | 14245

View Article Online
Scheme 1 Preparation of the platinum β-diketonate complexes (n refers to the number of carbons in the alkoxy chain) of diphenylpyridine ligands
and structures of complexes prepared (4-n and 8-n were prepared and reported previously9d). Conditions: (i) K2[PtCl4]/HOAc/reflux/24 h; (ii) dmso/
reflux/30 min; (iii) Na+(β-diketonate)−/CH2Cl2/rt/8 h. Procedures are the same for ligands/complexes with or without fused C5 ring. The structures of
6-n and of 10 imply the two different isomers of the coordinated tfac ligand.
1
temperatures for an hour after which the H NMR spectrum was
recorded. The spectra revealed that up to 140 °C there was no
isomerism, while at 160 °C the ratio changed to 1.3 : 1, while at
170 °C it reached 1 : 1. More rapid isomerisation could be
achieved by intermittent heating of small amounts of the solid
with a heat gun.
resulted from the use of tfac. Thus, the most significant obser-
vation was the destabilisation of both the crystal and mesophase.
The crystal phase was destabilised by between 30 and 80 °C,
while the clearing points were destabilised by up to 90 °C. Of
course, these temperatures are those of 1 : 1 mixtures and so
there will be a destabilisation associated simply with the fact
that, as far as melting and clearing points are concerned, the
complexes can be regarded as ‘impure’. However, previous
studies comparing mesomorphism of acac and tfac complexes of
PdII complexes showed destabilisation of clearing points by
around 60 °C and in this and these previous cases, it is the
greater size of CF3 compared to CH3 that is held responsible.11
In the complex incorporating hfac as the O^O co-ligand,
it was found that all mesomorphism was suppressed and the
However, despite this isomerisation, each component under-
went phase changes independently so that it was clear that a
eutectic was not formed and the two forms appeared immiscible,
leading to the appearance of two, co-existing states (e.g. Cr and
N or N and Iso) at some temperatures. Thus, the data in Table 2
are for 1 : 1 mixtures that have been isomerised thermally. The
isomerism will be discussed in greater detail below, but it is
important to note the changes in transition temperature that
14246 | Dalton Trans., 2012, 41, 14244–14256
This journal is © The Royal Society of Chemistry 2012

View Article Online
Table 1 Crystallographic parameters for complexes 5–6 and 12
5–6
12
CCDC code
885300
C36H47NO4Pt
752.84
884157
C17H14F3NO2Pt·(CHCl3)0.5
576.07
Empirical formula
Formula weight/g mol−1
T/K
Crystal system
Space group
110(2)
Monoclinic
P21/n
110.00(10)
Triclinic
ˉ
P1
a/Å
b/Å
13.1603(11)
12.8623(11)
11.988(2)
12.8040(14)
c/Å
α/°
38.458(3)
90
13.095(2)
86.305(12)
β/°
93.993(2)
90
6494.0(9)
69.598(16)
77.406(12)
1838.5(5)
γ/°
V/Å3
Z
8
4
ρcalc/Mg m−3
1.540
2.081
F(000)
3040
0.26 × 0.16 × 0.04
1.60 to 25.04
−15 ≤ h ≤ 15, −15 ≤ k ≤ 15, −45 ≤ l ≤ 45
49 929
11 460 [Rint = 0.0446]
11 460/32/785
1.115
R1 = 0.0418, wR2 = 0.0907
R1 = 0.0524, wR2 = 0.0947
2.95 and −3.04
1092
Crystal size/mm3
θ range for data collection/°
Index ranges
0.1581 × 0.1275 × 0.0673
5.64 to 64.36°
−13 ≤ h ≤ 17, −19 ≤ k ≤ 18, −17 ≤ l ≤ 19
17 553
11 514[R(int) = 0.0264]
11 514/18/481
1.029
R1 = 0.0459, wR2 = 0.1031
R1 = 0.0622, wR2 = 0.1141
11.81/−1.83
Reflections collected
Independent reflections
Data/restraints/parameters
Goodness-of-fit on F2
Final R indices [I > 2σ(I)]
R indices [all data]
Largest diff. peak/hole/e Å−3
Isomerisation in the tfac complexes
The possibility that the two isomers of the tfac complexes might
have different photophysical properties from one another (vide
infra) prompted us to investigate whether the isomers could be
separated, despite their lability in the melt. First, attempts were
made selectively to prepare one isomer by changing either
solvent or temperature, the results of which are collected in
Table 4.
Fig. 1 Molecular structure of one of the complexes present in the
asymmetric unit of complex 5–6. An ORTEP representation is found as
Fig. S1(b).‡
Evidently one isomer can be favoured over the other, but even
from these preliminary experiments it seemed evident that regio-
selective synthesis was unlikely. Then, given that isomerisation
was possible in the melt, the analogous process was investigated
in solution, so that a 3 : 1 mixture of isomers was heated under
reflux for one hour in CH2Cl2, CHCl3 and CH3CN, but within
experimental error, no change in ratio could be observed.
More success in isolating a single isomer was obtained by
fractional crystallisation from acetone, where the less soluble
isomer (that with the methine hydrogen at δ = 5.80) could be
concentrated, after three successive crystallisations, to 36 : 1
Table 2 Pt-ligand bond lengths in complexes 5–6 and 12
5–6
5–6
12
12
(complex 1) (complex 2) (complex 1) (complex 2)
dPt–C/Å
dPt–N/Å
dPt–O(trans to C)
Å
1.973(7)
1.991(5)
2.089(4)
1.968(7)
2.001(5)
2.082(5)
1.966(6)
1.997(6)
2.083(5)
1.965(6)
1.990(5)
2.093(4)
/
dPt–O(trans to N)
/
2.009(4)
1.994(4)
2.017(5)
1.996(5)
1
(based on the H NMR integration). With this isomer in hand,
Å
exhaustive attempts were made to grow single crystals in order
to characterise the isomer obtained, but none was successful.
Similarly, none of the NOE experiments carried out was able to
provide any evidence as to the identity of the isomer present.
complex simply melted to an isotropic fluid at 151 °C. Precedent
suggests that the reduction in anisotropy consequent on the intro-
duction of two CF3 groups accounts for this large mesophase
destabilisation.11
The complexes of the ligand with the fused five-membered
ring were then studied, but it was little surprise that, with two
increments of lateral broadening, neither the complex with tfac
nor hfac was mesomorphic and that with heptanedionate showed
only a monotropic SmA phase at 87 °C. These complexes were
therefore not studied further.
Molecular structure of the single isomer of a tfac complex of
Pt-2-(4-tolyl)pyridine
Frustrated by the failure of these attempts to allow characteris-
ation of the isomer, a different strategy was employed in which
a simpler complex, incorporating 2-tolylpyridine as the N^C
ligand, was prepared. Fractional crystallisation of this complex
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 14244–14256 | 14247

View Article Online
Table 3 Thermal behavior of complexes from 4-n to 11
Compounda Transition T/°C Compound Transition
Table 4 Isomer ratios obtained on preparation of complex 6–6 in
different solvents at different temperatures. All reactions were run
overnight
T/°C
4–6
Cr–SmA
SmA–Iso
Cr–SmA
200
237.5
190
6–6b
Cr–Cr + N
Cr + N–N
N–N + Iso
N + Iso–Iso
Cr–Cr + SmA
Cr + SmA–SmA 147
SmA–SmA + Iso 149
SmA + Iso–Iso
Cr–SmA
SmA–SmA + Iso 144
SmA + Iso–Iso
Cr–Cr + SmA
Cr + SmA–SmA 114
SmA–SmA + Iso 128
SmA + Iso–Iso
Cr–Iso
134
143
145
150
135
Solvent
T/°C
Isomer ratioa
4–8
CH2Cl2
CH2Cl2
CH2Cl2
CHCl3
0
5 : 1
3 : 1
1 : 1
6 : 1
SmA–dec 250
Ambient
40
Ambient
Ambient
4–10
4–12
5–6
Cr–SmA
SmA–Iso
Cr–SmA
SmA–I
Cr–SmA
SmA–Iso
Cr–SmA
SmA–I
Cr–SmA
SmA–Iso
Cr–SmA
SmA–Iso
Cr–Iso
168
244
170
231
169
228
148
225
116
217
100
205
151
161
(101)
141
6–8b
Acetone
No reaction
166
133
a Identified as the ratio of the complex giving rise to the signal at
δ = 5.80 and that at δ = 5.84.
6–10b
6–12b
5–8
158
90
5–10
5–12
164
108
(87)
105
115
126
9
7
8–6
(SmA–Iso)
Cr–Cr + Iso
Cr + Iso–Iso
Cr–Iso
Cr–Iso
(N–Iso)
Cr–Iso
10b
11
8–8
(SmA–N) (110)
(N–Iso)
Cr–Iso
(SmA–Iso) (115)
Cr–Iso 121
(SmA–Iso) (119)
(115)
123
8–10
8–12
Fig. 3 Molecular structure of complex 12. An ORTEP representation
using thermal ellipsoids is found as Fig. S2(b).‡
a Data for complexes 4-n and 8-n from ref. 9d. b Behaviour is of a pre-
equilibrated 1 : 1 isomer mixture (see text).
hydrogen resonates at δ = 5.80 that is concentrated.
The structure was solved with one chloroform solvent mole-
cule of crystallisation in the asymmetric unit, which was mod-
elled isotropically as disordered over three positions in a
39.5 : 37 : 23.5 ratio. All C–Cl bond lengths were constrained to
be similar and all Cl–C–Cl 1,3-bond distances were constrained
to be similar. Several large peaks of residual electron density
remained which could not be explained satisfactorily by disorder
models and attempts to find a twin law were similarly unsuccess-
ful. The two large peaks of residual density close to the platinum
are believed to be due to non-merohedral twinning of the crystal,
although a suitable method could not be obtained to account for
this.
The asymmetric unit thus contains two complex molecules
along with one (disordered) CHCl3 of crystallisation (Fig. S2‡)
and there were no statistically significant differences in the bond
lengths between the two complexes in the unit cell (Table 2).
The two Pt–O bonds were different in length reflecting the
different trans influences of carbon and nitrogen, with the longer
Pt–O bond being that trans to carbon as expected (Table 2). The
Fig. 2 The tfac complexes 6-n and 10 were found to comprise of a
mixture of the two isomers illustrated.
was attempted, this time from chloroform. In this case and poss-
ibly because of the lower anisometry of the complex, the best
that could be obtained was a 10 : 1 mixture, but here it proved
possible to obtain single crystals suitable for X-ray diffraction
analysis. The molecular structure is shown as Fig. 3, accompanied
by key structural parameters in Table 1. It should be noted that
the chemical shifts of the methine hydrogens in these two
isomers are identical to those in the di(alkoxyphenyl)pyridine
complexes and that, in this case also, it is the isomer where this
14248 | Dalton Trans., 2012, 41, 14244–14256
This journal is © The Royal Society of Chemistry 2012

View Article Online
bands (ε > 10 000 M−1 cm−1) in the region < 330 nm are due to
π–π* transitions of the conjugated ligands, whilst the bands at
longer wavelength tailing into the visible region have no counter-
parts in the free ligands and are associated with charge-transfer
bands involving the combination of metal and ligand. These
low-energy bands are blue-shifted in the complexes incorporat-
ing the fused cyclopentane ring compared to those without,
which probably reflects the steric influence of the ortho-CH2
groups of the cyclopentene ring which will serve to inhibit con-
jugation between the two rings.9d On the other hand, the identity
of the O^O co-ligand is seen to have little influence on the
absorption energies (Table 4).
Fig. 4 Side-on view of the dimeric, face-to-face arrangement of the Pt
complexes.
All of the complexes incorporating heptanedionate or trifluor-
oacetylacetonate are strongly luminescent in degassed solution at
room temperature, emitting in the green-yellow region of the
visible spectrum (Fig. 6). The spectra are similar to those with
acac as the co-ligand reported previously.9d They are red-shifted
compared to that of [Py(ppy)(acac)], reflecting the expected
influence of the electron-donating methoxy group in raising the
HOMO level, which is typically localised primarily on the cyclo-
metallating aryl unit and the metal, and the more extended conju-
gation arising from the aryl ring appended on the pyridyl unit.
The higher energy of emission of the complexes incorporating
the fused cyclopentane ring compared to those without (Fig. 6)
mirrors the effect in absorption described above, due to the
likely inhibition of attainment of a planar conformation that
maximises conjugation in the ligand. Although the spectra
display some evidence of vibrational structure, it is less pro-
nounced than in [Pt(ppy)(acac)], perhaps reflecting the larger
number of vibrational and rotational modes introduced with the
additional aryl ring.
Fig. 5 UV-visible absorption spectra of the pair of complexes with
heptanedionate as the co-ligand, with (blue) and without (red) the fused
cyclopentane ring.
The quantum yields of luminescence are impressively high,
in the range 0.5 to 0.7, amongst the highest reported for
platinum(II) complexes with bidentate cyclometallating ligands.6c
The luminescence lifetimes are around 25 μs, similar to those
recorded previously for the acac complexes, and an order of
magnitude longer than for [Pt(ppy)(acac)].6c Formerly, we inter-
preted this elongation of the lifetime in terms of a decrease in
both the non-radiative and radiative rate constants.9d The former
probably arises because of the delocalisation of the excited state
over the pendent aryl ring at position 5 of the pyridyl ring, ren-
dering it less well coupled to deactivating vibrational modes
associated with the acac. The latter effect may reflect the
decrease in metal character in the excited state with more
extended delocalisation over the ligand, and hence less efficient
spin–orbit coupling to promote the formally forbidden radiative
decay of the triplet state.
Despite the existence of two isomeric complexes in the
samples of the tfac complexes, there is no evidence of any
additional bands in their emission spectra compared to those
with the hdn co-ligand, nor do the spectra appear to be notice-
ably broader [compare spectra (b) with spectra (a) in Fig. 6].
This could indicate either that both isomers of the tfac complexes
emit at essentially the same energy, or that one of the isomers is
significantly more luminescent and dominates the spectrum.
We chose to try to probe this point in a little more detail using
complex 6–6, given that it had proved possible to isolate an
almost pure sample of one isomer in this case, as described
earlier. First, considering the sample of this complex comprising
complexes arranged themselves as dimers through a formal
centre of inversion (see Fig. 4 for a side-on view of the arrange-
ment and Fig. S3‡ for a top view). The distance between the two
planes defined by Pt and the two oxygens and three carbons of
the tfac ligand framework is 3.39 Å, while the distance between
the two, offset Pt atoms is 4.684 Å.
It is postulated, therefore, that the isomer that it is possible to
isolate in the liquid-crystalline derivatives is also that found here,
in which the CF3 group is positioned anti with respect to the
Pt–N bond. It is interesting that this is consistent with a qualita-
tive proposal that this isomer is less polar owing to the anti dis-
position of the more polar Pt–N bond (wrt Pt–C) and the CF3
group and so would be expected to crystallise preferentially from
acetone (then later chloroform) as a polar solvent, which was the
working hypothesis employed until the single crystal was
obtained.§
Photophysics of the new complexes
The UV-visible absorption spectra of the complexes in solution
display profiles that are typical of cyclometallated PtII complexes
of arylpyridines6 (e.g. Fig. 5 and Table 5). Intense absorption
1
§Note that we did attempt to record a H NMR spectrum of the single
crystal used in the diffraction study. However, the crystal was obtained
from an oily residue and the resulting spectrum could not be used with
any certainty to draw useful conclusions.
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 14244–14256 | 14249

View Article Online
Table 5 Absorption and emission data for the platinum(II) complexes in CH2Cl2 at 298 K
Complexes without cyclopentane ring
Complexes with pyridyl-fused cyclopentane ring
11 hfac 10 tfac 9 hdn 8–12 acac
Absorption – λmax/nm 391 (15.7), 314 392 (13.2), 318 397 (10.5), 323 400 (8.2), 322 379 (15.4), 304 379 (13.3), 386 (10.4), 360 400 (8.9), 386
7 hfac
6–12 tfac
5–12 hdn
4–12 acac
(ε/103 M−1 cm−1
)
(32.5), 262
(24.2), 241
(20.8)
(34.5), 303
(25.2), 266
(24.9), 250
(24.0)
(31.8), 306
(26.4), 270
(22.7), 256
(23.1), 235
(21.2)
(30.5), 305
(24.7), 272
(18.3), 255
(20.4)
(28.2), 332
(12.9), 293
(26.3), 263
(28.9), 244
(23.5),
307 (28.1), (9.3), 309
265 (29.8), (28.6), 298
252 (27.5) (27.5), 271
(27.6), 256
(10.3), 298
(28.2), 271
(27.1), 256
(25.0)
(26.0), 241
(22.9)
525, 555 526, 554
Emission – λmax/nm
—
544, 581
26 (0.56)
0.70
545, 583
30 (0.50)
0.62
541, 582,
628 (sh)
27 (0.55)
—
—
—
523, 553,
608 (sh)
27 (0.38)
Lifetime – τ degassed —
(air equilibrated)/μs
Quantum yield – ϕlum
23 (0.35) 26 (0.37)
—
0.49
0.53
0.61
0.57
Fig. 6 (a) Left: Emission spectra of the pair of complexes with heptanedionate as the co-ligand, with (blue) and without (red) the fused cyclopentane
ring, λex = 400 nm. (b) Corresponding spectra for the complexes with trifluoroacetylacetonate as the co-ligand.
a 1 : 1 mixture of isomers, only one set of bands was observed in
the emission spectrum at room temperature. However, for a
sample in EPA glass (EPA = diethyl ether–isopentane–ethanol,
2 : 2 : 1 v/v) at 77 K, the emission spectrum revealed a clear
additional shoulder to high energy of the main bands (Fig. 7).
Moreover, when the sample comprising one isomer concentrated
to 36 : 1 was examined under the same conditions, the spectrum
was clearly shifted to high energy, with the position of the 0–0
band matching well with the higher-energy shoulder observed
in the 1 : 1 sample (Fig. 7). Thus, we tentatively conclude that
the isomer that had been concentrated (viz. that with the CF3
trans to Pt–N) emits at higher energy than the other, and with
a significantly lower quantum yield, such that, in the 1 : 1
samples, it is the emission of the other isomer that dominates the
spectrum.
In contrast to the complexes incorporating hdn and tfac co-
ligands, those containing hexafluoroacetyl acetonate showed no
luminescence at room temperature. The absence of emission
from these complexes is at first sight surprising, given that the
excited state in [Pt(N^C)(O^O)] complexes is normally based
largely on the metal and the cyclometallating ligand, with the
β-diketonate acting essentially as an ‘innocent’ auxiliary ligand.
Time-dependent density functional theory provides insight into
the origin of the difference between the hfac and the tfac/acac
complexes.
Density functional theory calculations
In order to probe the influence of the co-ligand (i.e. acac, tfac or
hfac) on the emissive excited state in the three types of complex,
and also the influence of the isomerism in the tfac systems,
time-dependent density functional theory has been applied. To
simplify the calculations, the four representative unsubstituted
complexes shown in Fig. 8 were considered. The two tfac
isomers are labeled cis and trans, on the basis of the relative
disposition of the CF3 and pyridyl groups.
Calculations were carried out with the LANL2DZ basis set for
PtII and cc-PVDZ basis set for the ligands, using both PBE0 and
CAM-B3LYP functionals. The use of Coulomb-attenuated func-
tionals, of which CAM-B3LYP is a well-established example,12
is frequently superior to more conventional functionals when
excited states with significant charge-transfer character are under
investigation.13,14 The calculations were applied to the gas-phase
molecules and also in dichloromethane (using the polarisable
continuum model) to check for any significant solvent effect.
14250 | Dalton Trans., 2012, 41, 14244–14256
This journal is © The Royal Society of Chemistry 2012

View Article Online
The ground-state geometries were first optimised without
symmetry constraints, and harmonic frequency calculations per-
formed to confirm the attainment of energy minima on the
potential energy surface. The key bond lengths and angles invol-
ving the metal, predicted by DFT in this way for cis and trans-
[Pt(ppy)(tfac)] in the ground state, are illustrated in Fig. 9. There
is good agreement between the calculated data and the experi-
mentally determined bond lengths in the crystal of complex 12
listed in Table 2. However, it may be noted that there are no sig-
nificant differences between the two isomers.
Plots of the frontier orbitals (HOMO − 2 to LUMO + 2)
obtained by TD-DFT are shown in the ESI (Fig. S4–S7‡) with
energies of orbitals listed in Table S1.‡ Similar results were
obtained using CAM-B3LYP as for PBE0, probably due to a
reasonably high degree of orbital overlap despite the element of
charge-transfer character to the excited state.14 Since the lowest-
energy triplet excited states comprise varying proportions of
excitations involving different orbitals, and are not exclusively
HOMO → LUMO in character, it is more appropriate and infor-
mative to inspect the density difference plots, which display the
net rearrangement of the electron density in the excited state
compared to the ground state. The density difference plots of
the lowest triplet excitation – most relevant to the emission – are
shown in Fig. 10. They reveal that the two [Pt(ppy)(tfac)]
isomers feature excited states that are similar to that of [Pt(ppy)-
(acac)]. There is depletion of electron density on the metal and
cyclometallating aryl ring, with an increase on the pyridyl ring,
consistent with the usual formulation of the excited states in
such complexes as dPt/πN^C → π*N^C (mixed MLCT/LLCT). In
contrast, the triplet excited state of [Pt(ppy)(hfac)] has very
different character: it involves the movement of the electron from
the metal and aryl ring of the ppy to the hfac co-ligand. Appar-
ently, this arises because the presence of the two electron-with-
drawing CF3 groups stabilises the π* orbitals associated with the
co-ligand to such an extent that the LUMO becomes based on
the hfac rather than on the ppy (Fig. S4–S7‡). Indeed, the
LUMO for [Pt(ppy)(hfac)] is seen to be similar to LUMO + 2
for the acac complex. Quite similar conclusions were drawn by
Ghedini et al. on related cyclometallated PtII and PdII complexes
with hfac co-ligands, where an inversion of the ordering of
LUMO and LUMO + 1 was observed on changing from acac to
hfac.15
The lack of emission from the hfac complexes might then be
attributable – at least in part – to the poor overlap of the orbitals
involved in the excitation, the orbitals being localised on differ-
ent parts of the molecule, leading to a low radiative rate constant.
Absence of emission has likewise been observed and rationalised
Fig. 7 The emission spectra of 6–6 in diethyl ether–isopentane–
ethanol (2 : 2 : 1 v/v) at 77 K. Green line: the spectrum of the
1 : 1 mixture of isomers. Purple line: corresponding spectrum of the
sample in which one isomer has been concentrated to 36 : 1 by fractional
recrystallisation.
Fig. 8 Structures of the four complexes studied by TD-DFT.
Fig. 9 Bond lengths (Å) and angles (°) determined by TD-DFT for the ground state of the trans- (left) and cis- (right) isomers of [Pt(ppy)(tfac)].
This journal is © The Royal Society of Chemistry 2012 Dalton Trans., 2012, 41, 14244–14256 | 14251

View Article Online
Fig. 11 Superimposed structures of the S0 ground state (red) and T1
excited state (green) for the four complexes of Fig. 8, optimised using
PBE0 with the polarisable continuum model for CH2Cl2. (a) [Pt(ppy)-
(acac)]; (b) [Pt(ppy)(hfac)]; (c) trans-[Pt(ppy)(tfac)]; (d) cis-[Pt(ppy)-
(tfac)].
stabilised in the tfac complexes relative to the acac, but not suffi-
ciently so to make it the LUMO, as occurs in the case of the
hfac complex.
Triplet state geometries
Although it is commonplace to assume, in calculations of this
sort, that the triplet state of metal complexes will have the same
geometry as the ground state, there is clearly no reason why this
should necessarily be the case. For the purpose of understanding
the emission properties, it should, therefore, be useful to con-
sider the extent to which the structure of the molecule in its
triplet state is distorted compared to the ground state, since large
distortions typically promote non-radiative decay with a corre-
spondingly detrimental effect on quantum yields. The triplet-
state geometries of the two isomers were therefore calculated by
direct minimisation of the SCF triplet state, again using the
polarisable continuum model for the dichloromethane solvent.
The triplet-state structures obtained in this way are shown in
Fig. 11. Strikingly, it can be seen that [Pt(ppy)(acac)] remains
essentially planar in the excited state, whereas the tfac and hfac
complexes show some distortion. The distortion is quite clearly
largest for [Pt(ppy)(hfac)]. This feature could help account for
the lack of emission observed for the hfac complexes: greater
distortion in the excited state would increase non-radiative decay,
compounding the lower radiative rate constant predicted on the
basis of the poorer frontier orbital overlap discussed above.
Moreover, the analysis may provide insight into the origin of
the difference in emission efficiencies of the two isomers of
[Pt(ppy)(tfac)]. If the extent of distortion is quantified in terms
of the root-mean-square displacement of the atoms between the
S0 and T1 structures, or indeed the maximum displacement
of any atom (Table 6), the degree of distortion can be seen
to increase in the order [Pt(ppy)(acac)] < cis-[Pt(ppy)(tfac)] <
trans-[Pt(ppy)(tfac)] < [Pt(ppy)(hfac)]. The weaker emission
tentatively attributed to the trans compared to the cis isomer of
[Pt(ppy)(tfac)] might thus be accounted for in terms of the
greater triplet excited state distortion in the former, leading to
enhanced non-radiative decay.
Fig. 10 Density difference plots for the lowest triplet excitations of the
four complexes at the ground-state geometry: (a) [Pt(ppy)(acac)], (b) [Pt-
(ppy)(hfac)], (c) trans-[Pt(ppy)(tfac)], (d) cis-[Pt(ppy)(tfac)]. Purple and
yellow represent zones of depletion and augmentation of electron
density, respectively. Calculated using PBE0 with the PCM model for
CH2Cl2 as solvent.
in the same way for cyclometallated iridium(III) complexes in
which the frontier orbitals are localised on different ligands, with
little overlap.16 However, substantial distortion in the triplet
excited state of the hfac complexes may also play a role in dis-
favouring emission (vide infra).
Although the calculations help to understand the different be-
haviour of the hfac complexes, there is little difference between
the density difference plots of the two isomers of [Pt(ppy)(tfac)],
even though the experimental data suggest that one form is sig-
nificantly more emissive than the other, as discussed earlier. The
energies of the orbitals calculated for the two isomers are very
similar to one another (Table S1‡). It is notable, however, that
the presence of one CF3 group leads to LUMO + 1 being the
orbital most similar to LUMO in the hfac complex, but to
LUMO + 2 in the acac complex (Fig. S4–S7‡). In other words,
the unoccupied orbital based on the O^O co-ligand is clearly
Cyclic voltammetry and EPR spectroscopy
Interested by the observation that the hfac complexes studied
(7 and 11) were not emissive and the possible explanation in
14252 | Dalton Trans., 2012, 41, 14244–14256
This journal is © The Royal Society of Chemistry 2012

View Article Online
terms of the differing localisation of the LUMO, a comparative
study was made of the cyclic voltammograms of representative
hfac and tfac complexes.
6–0.24 mT, 1–0.35 mT). Whilst this is a simplification of the ani-
sotropic processes, it does confirm that the splitting parameters
reported are consistent with the experimental spectrum observed.
The spectrum is consistent with the radical character being that
of the radical di-anion of hfac with an additional coupling to the
platinum, and the splittings are similar to those observed for
the alkali metal and tetraalkylammonium salts of the free radical
di-anion.18
Complexes of the Pt(N^C)(O^O) type that are emissive are
typically characterised by a chemically irreversible reduction
on population of the LUMO, which is mainly based on the
2-phenylpyridine unit.8 This is indeed the case here and the
cyclic voltammogram (Fig. 12) of 6–12 shows the reduction
wave with Epeak ≈ −1.0 V (relative to the Fc/Fc+ couple at
0.64 V). However, when the hfac complex 7 was studied, the
reduction process was found to be reversible (Fig. 12).
This simulation therefore requires that the electron be localised
on the hfac ligand. This conclusion is fully consistent with the
Keen to explore further the nature of the species formed on
reduction of 7, the reduction was then performed in an epr cavity
and the spectrum shown in Fig. 13 was obtained. The spectrum
was simulated assuming a binomial septet (aF = 2.16 mT) due to
coupling to the six equivalent fluorines, which is further split
into a doublet (aH = 0.23 mT) due to coupling to the methine
hydrogen. In addition, there are platinum satellites (a195
=
Pt
2.43 mT). The spectrum shows considerable anisotropy, which
manifests itself as a significant increase in the linewidth with
increasing magnetic field. The spectrum was simulated using
Easyspin17 by allowing the linewidths to vary with the major
septet component from low field to high field as follows
(1–0.07 mT, 6–0.08 mT, 15–0.09 mT, 20–0.12 mT, 15–0.17 mT,
Table 6 Root-mean-square displacement of atoms and maximum
displacement of any atom on going from the S0 to T1 structures of the
complexes, as determined from the optimised geometries obtained by
TD-DFT using PBE0 and PCM for CH2Cl2 solvent
Root-mean-square
displacement/Å
Maximum
displacement/Å
Compound
[Pt(ppy)(acac)]
[Pt(ppy)(hfac)]
trans-[Pt(ppy)(tfac)]
cis-[Pt(ppy)(tfac)]
0.0261
0.4185
0.3101
0.2496
0.0531
0.8176
0.7382
0.6478
Fig. 12 Cyclic voltammogram of complex 6–12 (red lines), showing
an irreversible reduction, and of complex 7 (black lines) showing a
reversible reduction. Scan rate was 100 mV s−1 and the reference elec-
trode was Ag/AgCl; [FeCp2]0/+ is found at 0.64 V in this cell.
Fig. 13 Real (left) and simulated (right) EPR spectra of the monoanion of 7. Simulation parameters are given in the text. The superimposed spectrum
and simulation are given as Fig. S8.‡ Spectrum was recorded at 25 °C.
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 14244–14256 | 14253

View Article Online
results of the TD-DFT calculations discussed in the previous
section, where the LUMO for the hfac complex was found to be
localised on the hfac ligand (e.g. Fig. S5b‡). This contrasts with
the results for the tfac complexes, for which the LUMO was
found to be similar to that of the acac complex, being located
primarily on the ppy ligand (e.g. Fig. S5a, c and d‡).
Hamamatsu R928 photomultiplier tube; the spectra shown are
corrected for the wavelength dependence of the detector, and the
quoted emission maxima refer to the values after correction.
Samples for emission measurements were contained within
quartz cuvettes of 1 cm pathlength modified with appropriate
glassware to allow connection to a high-vacuum line. Degassing
was achieved via a minimum of three freeze–pump–thaw cycles
whilst connected to the vacuum manifold; final vapour
pressure at 77 K was <10−2 mbar, as monitored using a Pirani
gauge. Luminescence quantum yields were determined using
[Ru(bpy)3]Cl2 in aqueous solution (Φ = 0.028)26 as the standard;
estimated uncertainty in Φ is 20% or better.
The luminescence lifetimes of the complexes in aerated solu-
tion were measured by time-correlated single-photon counting
(TCSPC), following excitation at 374.0 nm with an EPL-375
pulsed diode laser. The emitted light was detected at 90° using a
Peltier-cooled R928 PMT after passage through a monochroma-
tor. The lifetimes in degassed solution were determined by multi-
channel scaling using the same detector following excitation
with a pulsed xenon lamp.
Experimental
1H NMR spectra were recorded at 270 MHz using a JEOL
ECX270 instrument, while optical microscopy employed an
Olympus BX50 Optical Microscope equipped with a Link-Am
HFS91 hot stage, TMS92 controller and LNP2 cooling unit.
DSC data were collected using a Mettler Toledo DSC 822,
equipped with a TSO801R0 sample robot and calibrated using
indium; samples were run at heating and cooling rates of 10 °C
min−1. Elemental analysis was carried out on an Exeter Analyti-
cal Inc CE 440 Elemental Analyser and a Sartorius SE2 analyti-
cal balance at the University of York.
X-band EPR data were recorded at 25 °C on an Bruker ER
200-D SRC X-band spectrometer connected to a datalink
486DX desktop PC running EPR Acquisition System version
2.42. A specially designed19 variable-temperature, in situ cell
was used with a platinum gauze working electrode placed in
the flat part of the cell. Platinum wire counter and Ag/AgCl
reference electrodes were placed on the reservoir at the top of
the cell. Electrogeneration was performed using a Potentiostat
BASCV-27 Voltammograph. All g-values were referenced to the
g-value of 2,2′-diphenyl-1-picrylhydrazyl (DPPH) at 2.0036
0.002.20
Diffraction data were collected at 110 K on a Bruker Smart
Apex diffractometer with Mo-Kα radiation (λ = 0.71073 Å)
using a SMART CCD camera. Diffractometer control, data col-
lection and initial unit cell determination was performed using
‘SMART’.21 The crystal was cooled with an Oxford Cryostream.
Frame integration and unit-cell refinement was carried out with
‘SAINT+’.22 Absorption corrections were applied by
SADABS.23 Structures were solved by ‘direct methods’ using
SHELXS-97 (Sheldrick, 1997)24 and refined by full-matrix least
squares using SHELXL-97 (Sheldrick, 1997).25 All non-hydro-
gen atoms were refined anisotropically, whereas the disordered
chloroform was refined isotropically. O–H hydrogens were
located by difference map after all other atoms had been located
and refined. All other hydrogen atoms were placed using a
‘riding model’ and included in the refinement at calculated
positions.
TD-DFT calculations
DFT calculations were carried out using the Gaussian 0927 suite
of programmes to predict energy-minimized structures of both
the singlet ground state and lowest triplet excited state of the
complexes. The LANL2DZ basis set was used for PtII, with
the inner core electrons replaced by a relativistic core potential;
the all-electron cc-PVDZ basis set for the ligands. Calculations
were carried out using two different functionals, PBE028 and
CAM-B3LYP,12 in both the gas phase and with the polarizable
continuum model (pcm) solvent model for dichloromethane
(as specified) using the default Gaussian 09 cavity parameters.
Geometries were fully optimized without symmetry constraints.
The triplet-state structures were calculated by direct minimisation
of the SCF triplet state. Harmonic vibrational wavenumber calcu-
lations were performed to confirm that the structures obtained
correspond to minima of the potential energy surface. Time-
dependent calculations were carried out on the optimised struc-
tures to investigate the relevant transitions and for the generation
of density difference plots.
Synthesis
Ligands and complexes were prepared as described in ref. 9d, as
outlined in Scheme 1. The final step was reaction of [Pt(N^C)-
Cl(S-dmso)] with the sodium salt of the β-diketonate and con-
ditions did not depend on the diketone used. Preparations of
these sodium salts is given below. Analytical data for the new
complexes are found in the ESI as Table S2.‡
A Voltalab 50, PST050, produced by Radiometer Analytical
was used to record cyclic voltammograms. The reference elec-
trode was Ag/AgCl and the working electrode was platinum. The
voltammograms were recorded at 50 and 100 mV s−1 in DMF
containing [Bu4N][PF6] (0.1 mol dm−3) using [FeCp2]0/+ at
0.64 Vas internal standard.
Synthesis of sodium hexafluoroacetylacetonate
Photophysical measurements
Absorption spectra were measured on a Biotek Instruments XS
spectrometer, using quartz cuvettes of 1 cm pathlength. Steady-
state luminescence spectra were measured using a Jobin Yvon
FluoroMax-2 spectrofluorimeter, fitted with a red-sensitive
Hexafluoroacetylacetone (14.4 mmol, 3.00 g) was added to
sodium ethoxide (14.4 mmol, 0.98 g) in petroleum ether
14254 | Dalton Trans., 2012, 41, 14244–14256
This journal is © The Royal Society of Chemistry 2012

View Article Online
(50 cm3). The mixture was heated under reflux for 4 h. After
cooling to room temperature, the colourless precipitate was
recovered by filtration and washed with petroleum ether
(10 cm3). Yield: 1.83 g, 55.3%.
2010, 21, 3270–3274; (e) K. Binnemans, in Molecular Materials, ed.
D. W. Bruce, D. O’Hare and R. I. Walton, Wiley, Chichester, 2010,
pp. 61–141; (f) K. Binnemans, J. Mater. Chem., 2009, 19, 448;
(g) D. Pucci, Liq. Cryst., 2011, 38, 1451; (h) E. Terazzi, S. Suarez,
S. Torelli, H. Nozary, D. Imbert, O. Mamula, J. P. Rivera, E. Guillet,
J. M. Benech, G. Bernardinelli, R. Scopelliti, B. Donnio, D. Guillon,
J.-C. G. Bünzli and C. Piguet, Adv. Funct. Mater., 2006, 16, 157;
(i) Y. T. Yang, K. Driesen, P. Nockermann, K. Van Hecke, L. Van Meer-
velt and K. Binnemans, Chem. Mater., 2006, 18, 3698;
( j) Y. Galyametdinov, A. A. Knyazev, V. I. Dzhabarov, T. Cardinaels,
K. Driesen, C. Görller-Walrand and K. Binnemans, Adv. Mater., 2008,
20, 252; (k) R. Bayón, S. Coco and P. Espinet, Chem.–Eur. J., 2005, 11,
1079.
Synthesis of sodium trifluorofluoroacetylacetonate
6 (a) W. Paw, S. D. Cummings, M. A. Mansour, W. B. Connick,
D. K. Geiger and R. Eisenberg, Coord. Chem.Rev., 1998, 171, 125;
(b) M. Hissler, J. E. McGarrah, W. B. Connick, D. K. Geiger,
S. D. Cummings and R. Eisenberg, Coord. Chem. Rev., 2000, 208, 115;
(c) J. A. G. Williams, Top. Curr. Chem., 2007, 281, 205; (d) J. Reedijk,
Eur. J. Inorg. Chem., 2009, 10, 1303; (e) M. D. Perez, P. I. Djurovich,
A. Hassan, G. Y. Chang, T. J. Stewart, K. Aznavour, R. Bau and
M. E. Thompson, Chem. Commun., 2009, 4215; (f) W.-Y. Wong and
C.-L. Ho, Coord. Chem. Rev., 2008, 252, 2596; (g) Z. M. Hudson,
C. Sun, M. G. Helander, H. Amarne, Z. Lu and S. Wang, Adv. Funct.
Mater., 2010, 20, 3426; (h) D. N. Kozhevnikov, V. N. Kozhevnikov,
M. Z. Shafikov, A. M. Prokhorov, D. W. Bruce and J. A. G. Williams,
Inorg. Chem., 2011, 50, 3804; (i) Z. M. Hudson, M. G. Helander, L. Zu
and S. Wang, Chem. Commun., 2011, 47, 755; ( j) Z. M. Hudson,
B. A. Blight and S. Wang, Org. Lett., 2012, 14, 1700; D. A. K. Vezzu,
J. C. Deaton, J. S. Jones, L. Bartolotti, C. F. Harris, A. P. Marchetti,
M. Kondakova, R. D. Pike and S. Huo, Inorg. Chem., 2010, 49, 5107.
7 (a) W. Lu, B.-X. Mi, M. C.-W. Chan, Z. Hui, C.-M. Che, N. Zhu and
S.-T. Lee, J. Am. Chem. Soc., 2004, 126, 4958; (b) S. Develay,
O. Blackburn, A. L. Thompson and J. A. G. Williams, Inorg. Chem.,
2008, 47, 11129; (c) L. Murphy and J. A. G. Williams, Top. Organomet.
Chem., 2010, 28, 75; (d) J. Kalinowski, V. Fattori, M. Cocchi and
J. A. G. Williams, Coord. Chem. Rev., 2011, 255, 2401; (e) A. Y.-Y. Tam,
D. P.-K. Tsang, M.-Y. Chan, N. Zhu and V. W.-W. Yam, Chem. Commun.,
2011, 47, 3383; (f) C. A. Strassert, C.-H. Chien, M. D. Galvez Lopez,
D. Kourkoulos, D. Hertel, K. Meerholz and L. De Cola, Angew. Chem.,
Int. Ed., 2011, 50, 946; (g) E. Rossi, A. Colombo, C. Dragonetti,
D. Roberto, F. Demartin, M. Cocchi, P. Brulatti, V. Fattori and
J. A. G. Williams, Chem. Commun., 2012, 48, 3182.
1,1,1-Trifluoro-2,4-pentanedione (3.24 mmol, 0.50 g) was added
to sodium hydride (3.24 mmol, 77.8 mg) in petroleum ether
(15 cm3), the solution was heated under reflux for 1 h. After
cooling to room temperature, the colourless precipitate was
recovered by filtration and washed with petroleum ether
(10 cm3). Yield: 0.54 g, 94.5%.
Synthesis of sodium 3,5 heptanedionate
3,5-Heptanedione (3.90 mmol, 0.50 g) was added to sodium
hydride (3.90 mmol, 93.6 mg) in petroleum ether (15 cm3), the
solution was stirred at room temperature 1 h. The colourless
precipitate was recovered by filtration and washed with
petroleum ether (10 cm3). Yield: 0.48 g, 82.5%.
8 J. Brooks, Y. Babayan, S. Lamansky, P. I. Djurovich, I. Tsyba, R. Bau and
M. E. Thompson, Inorg. Chem., 2002, 41, 3055.
Acknowledgements
9 (a) K. Venkatesan, P. H. J. Kouwer, S. Yagi, P. Müller and T. M. Swager,
J. Mater. Chem., 2008, 18, 400; (b) S. W. Thomas, S. Yagi and
T. M. Swager, J. Mater. Chem., 2005, 15, 2829; (c) C. Damm, G. Israel,
T. Hegmann and C. Tschierske, J. Mater. Chem., 2006, 16, 1808;
(d) A. Santoro, A. C. Whitwood, J. A. G. Williams, V. N. Kozhevnikov
and D. W. Bruce, Chem. Mater., 2009, 21, 3871.
We thank EPSRC (AD, GRF), the University of York (AS), the
University of Edinburgh (PRM) and the Comunidad Autonoma
de la Rioja (JT) for funding, Johnson Matthey for generous
loans of K2[PtCl4] and Dr Victor Chechik (University of York)
for valuable discussions concerning the epr spectrum.
10 A. Santoro, M. Wegrzyn, A. C. Whitwood, B. Donnio and D. W. Bruce,
J. Am. Chem. Soc., 2010, 132, 10689.
11 L. Omnès, B. A. Timimi, T. Gelbrich, M. B. Hursthouse, G. R. Luckhurst
and D. W. Bruce, Chem. Commun., 2001, 2248.
12 T. Yanai, D. P. Tew and N. C. Handy, Chem. Phys. Lett., 2004, 393, 51.
13 M. J. G. Peach, P. Benfield, T. Helgaker and D. J. Tozer, J. Chem. Phys.,
2008, 128, 044118.
14 P. Wiggins, J. A. G. Williams and D. J. Tozer, J. Chem. Phys., 2009, 131,
091101.
15 M. Ghedini, T. Puglise, M. La Deda, N. Godbert, I. Aiello, M. Amati,
S. Belviso, F. Leji, G. Accorsi and F. Barigeletti, Dalton Trans., 2008, 32,
4303.
16 A. J. Wilkinson, H. Puschmann, J. A. K. Howard, C. E. Foster and
J. A. G. Williams, Inorg. Chem., 2006, 45, 8685.
17 S. Stoll and A. Schweiger, J. Magn. Reson., 2006, 178, 142 (http://www.
easyspin.org).
18 K. S. Chen and J. K. S. Wan, J. Chem. Soc., Perkin Trans. 2, 1982, 419.
19 P. Murray and L. Yellowlees, in Spectroelectrochemistry, ed. W. Kaim and
A. Klein, Royal Society of Chemistry, Cambridge, 2008.
20 S. A. Al’tshuler and B. M. Kozyrev, Electron Paramagnetic Resonance,
Academic Press, New York, 1964.
References
1 (a) S. Laschat, A. Baro, N. Steinke, F. Giesselmann, C. Hägele,
G. Scalia, R. Judele, E. Kapatsina, S. Sauer, A. Schreivogel and
M. Tosoni, Angew. Chem., Int. Ed., 2007, 46, 4832; (b) S. Xiao,
M. Myers, Q. Miao, S. Sanaur, K. Pang, M. L. Steigerwald and
C. Nuckolls, Angew. Chem., Int. Ed., 2005, 44, 7390; (c) J. Nelson,
Science, 2001, 293, 1059; (d) L. Schmidt-Mende, A. Fechtenkötter,
K. Müllen, E. Moons, R. H. Friend and J. D. Mackenzie, Science, 2001,
293, 1119; (e) S. R. Forrest, Nature, 2004, 428, 911.
2 M. O’Neill and S. M. Kelly, Adv. Mater., 2011, 23, 566; M. O’Neill and
S. M. Kelly, Adv. Mater., 2003, 15, 1135.
3 Y. Sagara and T. Kato, Angew. Chem., Int. Ed., 2008, 47, 5175.
4 V. N. Kozhevnikov, B. Donnio and D. W. Bruce, Angew. Chem., Int. Ed.,
2008, 47, 6286.
5 (a) A. M. Prokhorov, A. Santoro, J. A. G. Williams and D. W. Bruce,
Angew. Chem., Int. Ed., 2012, 51, 95; (b) A. Santoro, A. M. Prokhorov,
A. C. Whitwood, B. Donnio, J. A. G. Williams and D. W. Bruce, J. Am.
Chem. Soc., 2011, 133, 5248; (c) Y. Wang, Q. Chen, Y. Li, Y. Liu,
H. Tan, J. Yu, M. Zhu, H. Wu, W. Zhu and Y. Cao, J. Phys. Chem. C,
2012, 116, 5908; (d) E. L. Szerb, A. M. Talarico, I. Aiello, A. Crispini,
N. Godbert, D. Pucci, T. Pugliese and M. Ghedini, Eur. J. Inorg. Chem.,
21 ‘SMART’ – Control Software Bruker SMART Apex X-ray Diffractometer.
v5.625, Bruker-AXS GMBH, Karlsruhe, Germany.
22 ‘SAINT+’ – integration software for Bruker SMART detectors. v6.45,
Bruker-AXS GMBH, Karlsruhe, Germany.
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 14244–14256 | 14255

View Article Online
23 G. M. Sheldrick, ‘SADABS’ – Program for Absorption Correction. v2.10
Bruker AXS Inc., Madison, Wisconsin, USA, 2007.
24 G. M. Sheldrick, SHELXS-97’ – Program for Structure Solution Univer-
sity of Göttingen, Göttingen, Germany, 1997.
25 G. M. Sheldrick, SHELXL-97′ – Program for the Refinement of Crystal
Structures, University of Göttingen, Göttingen, Germany, 1997.
26 K. Nakamaru, Bull. Chem. Soc. Jpn., 1982, 55, 2697.
27 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb,
J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson,
H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov,
J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota,
R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao,
H. Nakai, T. Vreven, J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro,
M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov,
R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant,
S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene,
J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts,
R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli,
J. W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski,
G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels,
Ö. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski and D. J. Fox,
GAUSSIAN 09, 2009.
28 J. P. Perdew, K. Burke and M. Ernzerhof, Phys. Rev. Lett., 1996, 77,
3865.
14256 | Dalton Trans., 2012, 41, 14244–14256
This journal is © The Royal Society of Chemistry 2012