Synthesis, spectroscopy, and theoretical studies of platinum(II) phosphate complexes

Article abstract of DOI:10.1039/b111460n

A method for the synthesis of [Pt(dpp)₂(en)] (dpp = diphenyl phosphato, en = ethane-1,2-diamine) is reported, and the resulting complex characterised spectroscopically. The synthesis involves displacement of chloride ligands using silver(I), an

Full text of DOI:10.1039/b111460n

View Article Online / Journal Homepage / Table of Contents for this issue
Synthesis, spectroscopy, and theoretical studies of platinum(II)
phosphate complexes†
Andrew R. Battle,a James A. Platts,*b Trevor W. Hambley*a and Glen B. Deaconc
a Centre for Heavy Metals Research, School of Chemistry, University of Sydney, NSW 2006,
Australia
b Department of Chemistry, Cardiff University, P.O. Box 912, Cardiff, UK CF10 3TB.
E-mail: platts@cf.ac.uk
c School of Chemistry, Monash University, Vic. 3800, Australia
Received 17th December 2001, Accepted 22nd February 2002
First published as an Advance Article on the web 9th April 2002
A method for the synthesis of [Pt(dpp)2(en)] (dpp = diphenyl phosphato, en = ethane-1,2-diamine) is reported, and
the resulting complex characterised spectroscopically. The synthesis involves displacement of chloride ligands using
silver(), and requires non-aqueous media. Density functional studies suggest typical square-planar coordination
geometry, with rather weak Pt–O bonds and very strong N–H ؒ ؒ ؒ O hydrogen bonds. Comparison with model
complexes suggests that phosphates are weaker ligands to platinum() than nitrates, and are therefore likely to be
highly reactive.
nucleophilic sites on DNA, particularly the N7 site of guanine.4
Introduction
Cisplatin has several properties that make it less than ideal as a
therapeutic agent. It has a limited spectrum of activity against
different types of cancer, it can be subject to inherent or
acquired resistance to treatment, and is also highly toxic.2
Many analogous PtII complexes have been synthesized and
tested as potential drugs, including many direct analogues of
the general form cis-[PtX2A2]5 and more recently PtIV com-
pounds.6,7 Many of these adhere to the structure–activity rules
devised by Cleare and Hoeschele,8 which stipulate neutral com-
plexes containing cis-leaving groups and inert N–H bearing
nitrogen donors. Despite concerted efforts to find improved
drugs, only carboplatin (2) has found widespread use as an
alternative.9 Other drugs containing oxygen ligands as leaving
groups include oxaliplatin (3), which is licensed across Europe,
and nedaplatin, licensed in Japan.10 Both show promise in treat-
ing cancers not amenable to cisplatin therapy, but are yet to find
worldwide approval. There is therefore great interest in develop-
ing new compounds that might show different characteristics to
known compounds.
Since the serendipitous discovery1 of the biological activity of
cisplatin (cis-[PtCl2(NH3)2], 1) (see Fig. 1), platinum-containing
Although phosphates are ubiquitous in the human body, and
are an integral part of DNA, there has been little interest in
investigating platinum phosphato complexes for anti-tumour
activity. Bose and colleagues11 have investigated the kinetics of
formation of phosphonato- and phosphato-platinum() com-
plexes using 31P NMR. Appleton and co-workers12 have also
investigated a variety of platinum phosphato complexes. In
these studies, the platinum complexes had the ability to either
chelate or form dimeric species with the phosphato ligands, the
formation of the complexes being highly dependent on pH.
Kozelka and Barre13 have studied the binding of nuclear bases
Fig. 1 Molecular structures of cisplatin 1, carboplatin 2, oxaliplatin 3
and [Pt(dpp)2(en)], 4.
anti-tumour drugs have become the focus of intense chemical,
biological, and medicinal research.2 The “platinums” are one of
the two most widely used anti-tumour drugs classes in the
world, with proven effectiveness in treating testicular and
ovarian cancers as well as activity against several other forms.2
The accepted mechanism of cisplatin activity involves initial
aquation in intra-cellular fluid to give the highly reactive
cis-[PtCl(NH3)2(OH2)]ϩ species,3 followed by binding to
to platinum() through both N and P᎐O groups using 31P
᎐
NMR, reporting that phosphate groups do indeed show some
affinity for PtII. In order to minimise the possibility of
oligomeric binding we have chosen to use phosphato diester
ligands, and embarked on synthetic and theoretical studies of
platinum() phosphato complexes, to investigate their
chemical, spectroscopic, structural, and electronic properties.
In this study, we report the results of such investigations on
[Pt(dpp)2(en)], where dpp = diphenyl phosphato and en =
ethane-1,2-diamine, along with theoretical studies of several
† Electronic supplementary information (ESI) available: B3LYP opti-
mised Cartesian coordinates of [Pt(dmp)2(en)]. See http://www.rsc.org/
suppdata/dt/b1/b111460n/
1898
J. Chem. Soc., Dalton Trans., 2002, 1898–1902
This journal is © The Royal Society of Chemistry 2002
DOI: 10.1039/b111460n

View Article Online
analogous compounds. Of particular interest here is the
strength of the Pt–O bond in phosphato complexes, which is
implicated in the rate of in vivo aquation and hence is related to
biological activity. We envisage that “hard” phosphate ligands
such as dpp might interact relatively weakly with the softer PtII
centre, resulting in faster aquation and possibly greater activity
(and toxicity).
Results and discussion
a) Synthesis
Silver salts such as silver nitrate and silver sulfate have been
extensively used for the conversion of platinum()-halo com-
plexes into the complex of choice, via the highly reactive
diaquaplatinum() complex which is formed. Initial attempts to
prepare [Pt(dpp)2(en)] (4) in water using this method were
unsuccessful. [Pt(OH)(en)(OH2)]ϩ was introduced to a solution
of diphenyl phosphate (dppH) in water at a pH of 1 and stirred
at room temperature overnight, after which time no reaction
had occurred. While it was not expected that the high acidity
would prevent ligation of the diphenyl phosphato ligand to
platinum, the sodium salt of diphenyl phosphate was neverthe-
less prepared and the procedure was repeated. Once again, this
was unsuccessful, the diaquaplatinum() salt being formed. A
further attempt using triethylamine to increase the pH to 7 also
failed to yield the desired phosphato complex.
It was expected that this method would be successful in pre-
paring [Pt(dpp)2(en)] as it is successful in preparing other
platinum() complexes containing oxygen donors such as
carboxylato ligands.26 For example, oxalato-platinum() species
can be prepared successfully by this method. However, despite
several attempts using different reaction conditions and pH
values the conversion of [PtCl2(en)] to [Pt(dpp)2(en)] was not
successful. The starting complex, [PtCl2(en)] is insoluble in
most organic solvents, and only slightly soluble in water.
However, due to the high solubility of diphenyl phosphate and
moderate solubility of [PtCl2(en)] in DMF, the reaction was
carried out in this solvent. Silver nitrate effectively removes halo
ligands from platinum complexes in DMF,27 but silver() oxide
was chosen as the agent to remove the halo ligands from the
platinum() complexes as (1) it is a moderate base and hence
will deprotonate dppH, which can then bind to the platinum,
and (2) there is no counter ion to act as a potential competing
ligand, as is the case with nitrate. Using these conditions, the
complex [Pt(dpp)2(en)] could be synthesised in moderate yield.
This represents the first successful preparation of a platinum()
bisphosphato-diester complex, with no evidence for dimeric or
oligomeric binding.
Experimental
Synthesis of [Pt(dpp)2(en)] (4) is in part based on the method
described by Kemmitt et al.14 for the synthesis of platinum()
phosphonato and arsonato complexes. [PtCl2(en)] (0.21 g,
0.64 mmol) was suspended in 10 mL of DMF. Diphenyl phos-
phate (0.32 g, 1.28 mmol) and silver() oxide (0.18 g, 0.78 mmol)
were added to this suspension, the reaction being protected
from light and stirred at room temperature overnight. The
black colour imparted from the silver() oxide was replaced by
the pale yellow silver() chloride. The suspension was passed
through Celite, affording a pale yellow, clear filtrate. The
solvent was removed under reduced pressure and the resulting
oily solid, now containing some black material, was once
again dissolved in the minimum volume of DMF and filtered.
Dichloromethane was introduced to the pale yellow filtrate and
the solution was placed in a freezer at Ϫ18 ЊC overnight. Over
this time a white, crystalline solid precipitated. This solid was
collected at the pump, washed with dichloromethane and air-
dried. Yield of [Pt(dpp)2(en)] 0.32 g, 0.425 mmol, 66%, IR/cmϪ1
ν(NH) 3265m, ν(CH) 3058m, ν(PO) 1200s, δ(NH) 1592m,
ν(Pt–O) 532m.15 Microanalysis: Calculated for [Pt(dpp)2(en)] C
41.44%, H 3.76%, N 3.72%. Found C 41.45%, H 3.57%, N
3.85%.
Diffuse reflectance infrared Fourier transform spectra
(DRIFTS) of [Pt(dpp)2(en)] were obtained on a Bio-Rad FTS-7
spectrophotometer over the range 400–4000 cmϪ1 using a KBr
background and matrix. A Bio-Rad FTS-40 spectrophotometer
was used over the range 200–400 cmϪ1 with polyethylene as the
background and matrix. 1H NMR spectroscopy was carried out
1
using a Bruker AC 200 MHz spectrometer. 31P, 13C, and H
(where necessary) were carried out on a Bruker AMX 400 MHz
spectrometer. Solvents used were of at least 99.6% purity and
were available commercially.
b) Characterization
As solvent peaks from DMF occur at 8.01 (broad), 2.91 and
All theoretical calculations were carried out using Gaussian
9816 running on a Compaq XP1000 workstation. Initial
HF/LANL2DZ17 optimisations were performed assuming no
symmetry: harmonic frequency calculations at this level con-
firmed each species as a true minimum, and gave estimates of
the vibrational contributions to binding energies. Subsequent
B3LYP18 optimisations employed the Stuttgart–Dresden19
basis set and effective core potential on platinum, with the all-
electron 6-31ϩG* basis20 on all other atoms. In the cases of
[PtCl2(en)], [Pt(ox)(en)], [Pt(OH)2(en)] and [Pt(en)(OH2)2]2ϩ,
initial optimisations converged to C2 symmetry — this was
enforced during subsequent optimisations. Atomic partial
charges were calculated using the natural bonding orbital
(NBO) scheme,21 which has been widely used in the study of
transition metal complexes.22
1
2.74 ppm in the H NMR spectrum, d6-DMSO was chosen as
the NMR solvent, showing a solvent peak at 2.49 ppm only.
Due to the strongly coordinating nature of DMSO, the spec-
trum was acquired within 30 minutes of dissolution of the
compound. The region of the proton NMR spectrum between
6.2 and 7.6 ppm is shown in Fig. 2. The signal due to the phenyl
protons occurs as expected in the region 7.3 to 6.9 ppm. The
peak at 6.93 ppm corresponds to the para proton on the phenyl
carbon and is half the intensity of the peaks at 7.23 and 7.11
For two compounds with published X-ray crystallographic
structures, namely [PtCl2(en)]23 and [Pt(ox)(en)],24 the above
method resulted in Pt–N bond lengths that were on average
0.07 Å too long, Pt–O lengths 0.02 Å too short, and Pt–Cl
lengths 0.01 Å too long. Similar overestimations of Pt–N bond
lengths have been reported in a recent study of cisplatin,25 such
that this appears to be a systematic effect in DFT calculations.
Comparison of the properties of the free ligands supports the
use of dimethyl phosphate (dmp) as a model for diphenyl
phosphate: charges on O and P (not shown) differ by less than
5%, or around 0.05 electrons, while the geometry about P is
almost identical.
Fig. 2 1H NMR spectrum of 4 in the region 6.2–7.6 ppm.
J. Chem. Soc., Dalton Trans., 2002, 1898–1902
1899

View Article Online
the N–H of the en ligand and the (formally sp2 hybridised)
phosphate oxygen. This contact is very short (O ؒ ؒ ؒ H =
1.796 Å, O ؒ ؒ ؒ N = 2.747 Å) and is also relatively linear, as
measured by the O ؒ ؒ ؒ H–N angle of 148.4Њ. The ‘angle of
Table 1 Selected geometrical parameters of [Pt(dmp)2(en)]
Bond lengths/Å
Bond angles/Њ
Torsion angles/Њ
Pt–N
Pt–O
N–C
C–C
O–P
2.062
N–Pt–N
O–Pt–O
N–Pt–O
Pt–N–C
N–C–C
Pt–O–P
N–H ؒ ؒ ؒ O
83.7
88.5
93.9
109.2
108.4
119.8
148.4
110.8
Pt–N–C–C 38.8
N–C–C–N
N–Pt–O–P
attack’ of H-bonding to the O᎐P bond is similar to a value of
᎐
2.064
1.490
1.530
1.554
1.510
1.796
2.747
Ϫ50.9
Ϫ45.5
131.5Њ found recently for a complex of trimethyl phosphate
with H–F.28 We surmise that the hydrogen bonds in this struc-
ture are rather strong, and must contribute substantially to the
binding of the phosphate ligands to the Pt(en) fragment — we
will return to this point below. Evidence for these intra-
molecular hydrogen bonds is also seen in the IR spectrum of
the complex. The ν(N–H) stretch at 3265 cmϪ1 is broader and
less intense than that seen for [PtCl2(en)], and is most likely a
consequence of these hydrogen bonds. Also, the crystal struc-
ture of [Pt(5Ј-CMP)(en)]2, where CMP = 5Ј-cytidine mono-
phosphate), determined by Louie and Bau29 shows a strong
hydrogen bond between a phosphato oxygen and the ethane-
1,2-diamine nitrogen, the O–N distance being 2.75 Å which is in
good agreement with the analogous distance seen in our calcu-
O᎐P
᎐
O ؒ ؒ ؒ H
O ؒ ؒ ؒ N
P᎐O ؒ ؒ ؒ H
᎐
ppm corresponding to the equivalent protons on the two ortho
and meta positions respectively. The amine protons are present
as a broad peak at 6.63 ppm; the broadness of this peak in a
non-exchanging solvent such as DMSO appears to support our
theoretical assignment of strong N–H ؒ ؒ ؒ O᎐P intramolecular
᎐
hydrogen bonding (see below). Outside the range of Fig. 2, the
methylene protons are present at 2.51 ppm, while peaks at 2.01,
2.91 and 3.28 ppm are due to trace acetone, DMF and water
respectively. Due to the C2 axis in the complex, the 13C spectrum
shows only five peaks corresponding to the carbon atoms. The
peak at 153 ppm is the carbon directly attached to the oxygen,
while peaks at 119, 129 and 122 ppm are the carbons ortho
and meta and para to the phosphate group, respectively. The
aliphatic carbon atoms are present at 48 ppm, along with
solvent at 40 ppm.
lated structure. Further evidence for strong P᎐O ؒ ؒ ؒ H–N
᎐
hydrogen bonding comes from Appleton et al.,12b who inferred
their presence from broad peaks in the 31P NMR spectra of
cis-[Pt(NH3)2(PO4H2)(H2O)]ϩ.
The optimum conformation found contains hydrogen bonds
on either side of the coordination plane in a ‘trans’ arrange-
ment. A second conformation was identified at the Hartree–
Fock level, in which the phosphates form hydrogen bonds on
the same side of the coordination plane. However, this was
6.2 kJ molϪ1 less stable than that shown for the trans form, and
was not pursued further. The energy difference between these
forms, and the likely barrier to interconversion, is so small that
at room temperature all four hydrogens would appear
equivalent.
To further characterise the complex, the 31P NMR spectrum
was measured. The compound was initially dissolved in DMSO
and was allowed to stand before a NMR spectrum was
obtained. The spectrum showed two peaks, due to partial lig-
and exchange of the phosphate with DMSO, hence a saturated
solution of the compound was prepared in CD2Cl2 and showed
a singlet at Ϫ2.54 ppm, indicating phosphate binding to the
platinum (diphenyl phosphate in CD2Cl2 shows a singlet at
Ϫ8.64 ppm). Despite several attempts using different solvents
and methods, crystals suitable for X-ray diffraction were not
forthcoming, the complex persistently forming feathers.
Another feature of note is that the phosphate ligand shows
some evidence of localisation. The shortest P–O bond is that
involved in H-bonding, while the O coordinated to Pt has a
longer bond to P, and bonds to OMe groups are longer still.
Thus, it seems that the presence of PtII effectively localises the
negative charge on the anion on to a single oxygen. The opti-
mised structure also shows substantial deviation from square
planarity, with N–Pt–N = 83.7Њ and O–Pt–O somewhat larger at
88.5Њ. The ethane-1,2-diamine fragment lies out of the coordin-
ation plane, such that the Pt–N–C–C torsion angle is 38.8Њ,
similar to previously published values23,24 of 37.1 and 32.2Њ.
Finally, the conformations adopted by the phosphato ligands
appear to be driven by attempts to minimise steric repulsions
whilst maintaining the stabilising N–H ؒ ؒ ؒ O H-bonds. Each
phosphate ligand has one methoxy group approximately
‘equatorial’ (i.e. roughly in the coordination place) and one
‘axial’. These axial methoxy groups are sited roughly above and
below the Pt, i.e. the positions one would expect axial ligands to
be placed in the analogous PtIV complex. The small size of
the methoxy groups used in this model compound mean the
axial sites are not particularly crowded, but this crowding
would certainly be exacerbated with the larger phenyl groups
used in 4. This may have implications for the reactivity and
redox chemistry of such complexes — this will be investigated
in subsequent studies.
c) Theoretical studies
B3LYP optimisation of cis-[Pt(dmp)2(en)] resulted in the
expected square-planar geometry about Pt, with Pt–N bond
lengths of 2.062 Å and Pt–O lengths of 2.064 Å. Comparison
with X-ray coordinates above suggest these lengths might be
overestimated by around 0.03 Å. However, there is no doubt
that PtII forms stable complexes with η1-phosphato ligands, and
hence the postulated molecular structure of 4 is a reasonable
one. The optimised structure of [Pt(dmp)2(en)] is shown in
Fig. 3, with selected geometrical parameters in Table 1.
There are several features of this structure worthy of note:
firstly, two intramolecular hydrogen bonds are evident between
The difficulties encountered in attempts to synthesise 4 in
aqueous media suggests that the Pt–O bond to phosphate is
relatively weak, such that the ligand is rather labile and easily
displaced by nucleophiles. We have explored the strength of Pt–
ligand bonds by considering the simple dissociation process
[PtL2(en)]
Pt(en)2ϩ ϩ 2LϪ for a total of seven complexes. The
resulting estimates of Pt–ligand bond energies are reported in
Table 2, and appear to support the hypothesis that phosphates
make rather weak ligands. Our model ligand dmp yields the
second lowest bond energies of all the anionic ligands con-
sidered with a Pt–O bond energy of 838.8 kJ molϪ1. Only neu-
tral H2O and nitrate are weaker ligands — several studies have
Fig. 3 B3LYP optimised geometry of [Pt(dmp)2(en)] (all H’s except
NH removed for clarity).
1900
J. Chem. Soc., Dalton Trans., 2002, 1898–1902

View Article Online
Table 2 DFT calculated properties of [Pt(dmp)2(en)] and some analogues
Total energy/au
Pt–O energya/kJ molϪ1
Pt–O length/Å
µ/D
qNBO Pt
qNBO en
qNBO ligand
[Pt(dmp)2(en)]
[Pt(OH)2(en)]
[Pt(ac)2(en)]
[PtCl2(en)]
[Pt(ox)(en)]
[Pt(NO3)2(en)]
[Pt(OH2)2(en)]
Ϫ1754.33724
Ϫ461.59821
Ϫ766.98733
Ϫ1230.42481
Ϫ687.10913
Ϫ870.59894
Ϫ462.23181
838.8
1018.8
894.0
2.064
2.001
2.047
2.331
2.005
2.050
2.132
6.0
9.2
6.8
13.0
18.2
13.6
0.3
0.786
0.720
0.757
0.533
0.724
0.765
0.815
0.563
0.426
0.509
0.489
0.480
0.554
0.823
Ϫ0.675
Ϫ0.573
Ϫ0.633
Ϫ0.511
Ϫ1.205
Ϫ0.659
0.181
853.3
1013.5b
800.5
207.1
a Including a scaled (0.89) vibrational energy correction from HF/LANL2DZ harmonic frequency calculation. b Calculated as half the Pt(en)oxalate
Pt(en)2ϩ ϩ oxalate2Ϫ energy.
shown that nitrates form very reactive/toxic complexes and are
easily displaced by many reagents.8 Carboxylate groups (acetato
and oxalato) form much stronger complexes, while chlorides are
intermediate between these extremes. It should be noted that
these values take no account of desolvation or entropic effects,
such that their absolute values are unlikely to be accurate. How-
ever, it is a reasonable assumption that these effects are
approximately constant across the range of complexes con-
sidered and hence that trends can be drawn from these data.
The main apparent trend in Table 2 is that phosphates form
rather weak bonds to PtII. There also appears to be a rough
inverse correlation (R2 = 0.91) between Pt–O bond energy and
bond length, such that the strongest bound complexes, i.e. the
hydroxo and oxalato complexes, have the shortest bonds.
This conclusion appears to be in accordance with various
experimental observations. Appleton and co-workers12b have
examined the reactions between cis-[Pt(NH3)2(OH2)2]2ϩ and a
variety of O-donor ligands, including nitrato, carboxylato
and phosphato. Acetate was found to react more rapidly than
phosphate (under nitrogen) with the diaqua species, however
in the presence of oxygen addition of the phosphate to the
diaqua species resulted in the rapid formation of blue solutions,
attributed to be oligomers of platinum(). A further study12a
compared the reaction between amp, (amp = (aminomethyl-
)phosphonic acid) and the above platinum complex, as well as
cis-[Pt(OH)2(NH3)2]. The ligand was found to react much more
readily with the aqua species than the hydroxo species, which
may indicate that the hydroxo species is more resistant to attack
from the ligand. In a recent publication, Davies et al.30 showed
that phosphate buffer at a concentration of 9 mM did not inter-
fere with the rate of aquation of cisplatin, further indicating the
weak donor ability of phosphate to platinum().
calculation of the tetramethylethane-1,2-diamine (tmeda)
analogue [Pt(dmp)2(tmeda)] Pt–O bond energy, which is found
to be 737 kJ molϪ1 using the same methods. Although these
weak Pt–O bonds suggest high activity, the diphenyl phosphato
complex is not stable with respect to aquation and in biological
media will yield the diaqua complex. This complex is known to
be highly anticancer active but is too toxic for systemic adminis-
tration.31 However, lipophilic complexes such as 4 may have
application in local delivery: subsequent studies will attempt to
address this possibility.
Insight into these weak Pt–phosphate bonds is gained from
the NBO charge information also reported in Table 2. For
clarity, these charges have been summed into values for metal,
ethane-1,2-diamine, and ligand fragments. It is apparent that
the phosphato complex has larger charges (in absolute terms)
than all other neutral complexes — the Pt(en) fragment here
carries a charge of ϩ1.37, compared with just ϩ1.02 in
[PtCl2(en)], ϩ1.15 in [Pt(OH)2(en)], ϩ1.27 in [Pt(ac)2(en)], and
ϩ1.32 in [Pt(NO3)2(en)]. We interpret this charge separation as
evidence of the ‘hardness’ of the phosphato ligand — unlike
carboxylate or hydroxyl groups, phosphate does not share elec-
tron density with metals, but instead simply has electrostatic
attraction for the positive metal centre. Nitrate is almost as hard
as phosphate by this measure, which accords with its known
reactivity and low calculated bond strength.
Conclusions
We have shown that cis-[Pt(dpp)2(en)] can be synthesised from
[PtCl2(en)] by using a combination of diphenyl phosphato and
silver() oxide in non-aqueous solvents. It has also been estab-
lished that the “standard” method of preparation, involving
direct replacement of aqua ligands by phosphate fails in this
case. The complex has been characterised through infrared and
1H, 13C, and 31P NMR spectroscopy, which reveals that the
molecule has C2 symmetry. Analysis of the model complex
cis-[Pt(dmp)2(en)] using density functional methods indicates a
typical square-planar arrangement, with pronounced hydrogen
The estimate of Pt–O bond energy in Table 2 inevitably con-
tains a contribution from the strong N–H ؒ ؒ ؒ O hydrogen
bonds noted above, since separating the ligand from the Pt(en)
fragment breaks both interactions. It is not a simple matter to
decompose the overall figure of 838.8 kJ molϪ1 into values for
Pt–O and N–H ؒ ؒ ؒ O interactions. However, we can obtain an
estimate of the strength of such hydrogen bonds from consider-
ation of model systems. The simple model system of cisplatin
(cis-[PtCl2(NH3)2]) complexed with the phosphate anion
bonding between N–H of the en ligand and O᎐P of the phos-
᎐
phate. Comparison with several model complexes indicates that
the Pt–O bonds in this complex are very weak, less stable even
than those formed in Pt-nitrato complexes. Population analysis
suggests that this is due to the hardness of the phosphato lig-
and, which prevents any significant sharing of electron density.
H2PO4 gives a hydrogen bond energy of 123.7 kJ molϪ1 with
Ϫ
the same functional and basis set as used above. This is reduced
to 119.3 kJ molϪ1 when corrected for basis set superposition
error (BSSE) and changes in zero point energy (ZPE). This
complex could equally be described as a hydrogen bond or an
ion–molecule complex, but the conformation found here is very
similar to that described above, with H ؒ ؒ ؒ O = 1.765 Å and
N–H ؒ ؒ ؒ O = 158.8Њ. The value of 119 kJ molϪ1 is probably an
overestimate of the value in [Pt(dmp)2(en)], since our inter-
molecular model system can fully relax without the constraints
imposed by coordination to Pt. Despite this, it seems certain
that the hydrogen bonds in [Pt(dmp)2(en)] contribute at least
100 kJ molϪ1 to the stabilisation of the complex, and thus that
the ‘true’ Pt–O bond energy is very low indeed, in the region
of 720–750 kJ molϪ1. This estimate is further supported by
Acknowledgements
A. R. B. would like to thank the Australian Research Council
for a studentship. J. A. P. thanks the Royal Society for a travel
grant.
References
1 B. Rosenberg, L. van Camp and T. Krigas, Nature, 1965, 205, 698.
2 E. Wong and C. M. Giandomenico, Chem. Rev., 1999, 99, 2451.
J. Chem. Soc., Dalton Trans., 2002, 1898–1902
1901

View Article Online
3 (a) J. W. Reishus and D. S. Martin, J. Am. Chem. Soc., 1961, 83,
2457; (b) M. C. Lim and R. B. Martin, J. Inorg. Nucl. Chem., 1976,
38, 1911; (c) S. J. Berners-Price, T. A. Frenkiel, U. Frey, J. D. Ranford
and P. J. Sadler, J. Chem Soc., Chem. Commun., 1992, 789; (d ) J. L.
R. Van der Veer, Chem. Br., 1988, 775.
4 (a) J. M. Pascpe and J. J. Roberts, Biochim. Pharmacol., 1974, 23,
1345; (b) J. Brouwer, P. v. d. Putte, A. M. J. Fichtinger-Schepman
and J. Reedijk, Proc. Natl. Acad. Sci. USA, 1981, 78, 7010.
5 T. A. Connors, M. J. Cleare and K. R. Harrap, Cancer Treat. Rep.,
1979, 63, 1499.
6 L. R. Kelland, S. Y. Sharp, C. F. O’Niell, F. I. Raynaud, P. J. Beale
and I. R. Judson, J. Inorg. Biochem., 1999, 77, 111.
7 T. W. Hambley, A. R. Battle, G. B. Deacon, E. T. Lawrenz, G. D.
Fallon, B. M. Gatehouse, L. K. Webster and S. Rainone, J. Inorg.
Biochem., 1999, 77, 3.
8 M. J. Cleare and J. D. Hoeschele, Bioinorg. Chem., 1973, 2, 187.
9 M. Gordon and S. Hollander, J. Med., 1993, 24, 209.
10 D. Lebwohl and R. Canetta, Eur. J. Cancer, 1998, 34, 1522.
11 (a) R. N. Bose, N. Goswami and S. Moghaddas, Inorg. Chem., 1990,
29, 3461; (b) L. L. Slavin and R. N. Bose, J. Inorg. Biochem., 1990,
40, 339.
12 (a) T. G. Appleton, R. D. Berry and I. J. McMahon, Inorg. Chem.,
1986, 25, 720; (b) T. G. Appleton, R. D. Berry, C. A. Davis, J. R. Hall
and H. A. Kimlin, Inorg. Chem., 1984, 23, 3514; (c) T. G. Appleton,
J. R. Hall and I. J. McMahon, Inorg. Chem., 1986, 25, 726.
13 J. Kozelka and G. Barre, Chem. Eur. J., 1997, 3, 1405.
14 R. D. W. Kemmitt, S. Mason, J. Fawcett and D. R. Russell, J. Chem.
Soc., Dalton Trans., 1992, 851.
P. Y. Ayala, Q. Cui, K. Morokuma, D. K. Malick, A. D. Rabuck,
K. Raghavachari, J. B. Foresman, J. Cioslowski, J. V. Ortiz, A. G.
Baboul, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz,
I. Komaromi, R. Gomperts, R. L. Martin, D. J. Fox, T. Keith,
M. A. Al-Laham, C. Y. Peng, A. Nanayakkara, M. Challacombe,
P. M. W. Gill, B. Johnson, W. Chen, M. W. Wong, J. L. Andres,
C. Gonzalez, M. Head-Gordon, E. S. Replogle and J. A. Pople,
Gaussian, Inc., Pittsburgh PA, 1998.
17 P. J. Hay and W. R. Wadt, J. Chem. Phys., 1985, 82, 270.
18 (a) A. D. Becke, J. Chem. Phys., 1993, 98, 5648; (b) C. Lee, W. Yang
and R. G. Parr, Phys. Rev. B, 1988, 37, 785.
19 D. Andrae, U. Haussermann, M. Dolg, H. Stoll and H. Preuss,
Theor. Chim. Acta, 1990, 77, 123.
20 (a) R. Ditchfield, W. J. Hehre and J. A. Pople, J. Chem. Phys., 1971,
54, 724; (b) T. Clark, J. Chandrasekhar, G. W. Spitznagel and P. v. R.
Schleyer, J. Comput. Chem., 1983, 4, 294.
21 A. E. Reed, L. A. Curtiss and F. Weinhold, Chem. Rev., 1988, 88,
899.
22 G. Frenking and N. Frohlich, Chem. Rev., 2000, 100, 717.
23 (a) L. T. Ellis and T. W. Hambley, Acta Crystallogr., Sect. C, 1994,
50, 1888; (b) M. A. Bruck, R. Bau, M. Noji, K. Inagaki and
Y. Kidani, Inorg. Chim. Acta, 1984, 92, 279.
24 A. R. Battle, R. Dolman, G. B. Deacon and T. W. Hambley, Aust. J.
Chem., in press.
25 P. N. V. Pavankumar, P. Seetharamulu, D. Yao, J. D. Saxe,
D. G. Reddy and F. H. Hausheer, J. Comput. Chem., 1999, 20, 365.
26 A. R. Kokhar, Y. Deng, Y. Kido and Z. H. Siddik, J. Inorg.
Biochem., 1993, 50, 79.
15 K. Nakomoto, Infrared Spectra of Inorganic and Coordination
Compounds, John Wiley & Sons, New York, 1963.
27 B. E. Bowler, K. J. Ahmed, W. I. Sundquist, L. S. Hollis,
E. E. Whang and S. J. Lippard, J. Am. Chem. Soc., 1989, 111, 1299.
28 O. J. Lamarche and J. A. Platts, Chem. Eur. J., 2002, 8, 457.
29 S. Louie and R. Bau, J. Am. Chem. Soc., 1977, 99, 3874.
30 M. S. Davies, S. J. Berners-Price and T. W. Hambley, Inorg. Chem.,
2000, 39, 5603.
16 Gaussian 98, Revision A.9, M. J. Frisch, G. W. Trucks, H. B.
Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, V. G.
Zakrzewski, J. A. Montgomery, Jr., R. E. Stratmann, J. C. Burant,
S. Dapprich, J. M. Millam, A. D. Daniels K. N. Kudin, M. C. Strain,
O. Farkas, J. Tomasi, V. Barone, M. Cossi, R. Cammi, B. Mennucci,
C. Pomelli, C. Adamo, S. Clifford, J. Ochterski, G. A. Petersson,
31 J. Kozelka, F. Legendre, F. Reeder and J. C. Chottard, Coord. Chem.
Rev., 1999, 192, 61.
1902
J. Chem. Soc., Dalton Trans., 2002, 1898–1902