Structural and spectroscopic studies on mercury(II) tribenzylphosphine complexes

Article abstract of DOI:10.1039/b419253b

The spectroscopic and structural study of mercury (II) tribenzylphosphine complexes were discussed by using single-crystal X-ray crystallography. The mercury coordination in [Hf(Pbz₃)₂]⁶⁺ complex. The mercury coordination in [Hf(PR₃] can be described as distorted tetrahedral, with a significant of the P-Hg-P angle from linearity as a result of coordination of the nitrate groups. The Hg and P NMR shielding tensors of PMe₃ models of the tribezylphosphine complexes were calculated using relativistic density functional theory.

Full text of DOI:10.1039/b419253b

View Online / Journal Homepage / Table of Contents for this issue
F U L L P A P E R
Structural and spectroscopic studies on mercury(II)
tribenzylphosphine complexes
Graham A. Bowmaker,*a Behnam Assadollahzadeh,†a Andrew M. Brodie,b Eric W. Ainscough,b
Graham H. Freemanb and Geoffrey B. Jamesonb
a Department of Chemistry, University of Auckland, Private Bag 92 019, Auckland,
New Zealand. E-mail: ga.bowmaker@auckland.ac.nz; Fax: +64 9 373 7422;
Tel: +64 9 373 7599
b Chemistry-Institute of Fundamental Sciences, Massey University, Palmerston North,
Private Bag 11 222, Palmerston North, New Zealand
Received 23rd December 2004, Accepted 9th March 2005
First published as an Advance Article on the web 31st March 2005
The tribenzylphosphine (PBz3) complexes of mercury(II), [Hg(PBz3)2](BF4)2, [Hg(PBz3)2(NO3)2] and
[HgX(NO3)(PBz3)] (X = Cl, Br, I and SCN), have been synthesised and their structures determined by single-crystal
X-ray crystallography. [Hg(PBz3)2](BF4)2 contains [Hg(PBz3)2]2+ cations with linear P–Hg–P coordination, the first
example of a truly two-coordinate [Hg(PR3)2]2+ complex. The mercury coordination in [Hg(PBz3)2(NO3)2] can be
described as distorted tetrahedral, with a significant deviation of the P–Hg–P angle from linearity as a result of
coordination of the nitrate groups. Nitrate coordination is also observed in [HgX(NO3)(PBz3)] (X = Cl, Br, I),
resulting in significantly non-linear P–Hg–X coordination. The thiocyanate complex is a centrosymmetric
thiocyanate-bridged dimer with distorted trigonal-pyramidal mercury coordination to the P atom of PBz3, to the S
and N atoms of two bridging thiocyanate groups, and to the O atom of one nitrate group. For all the nitrato
˚
complexes, secondary mercury-nitrate interactions (Hg–O 2.7–3.1 A) effectively raise the coordination number of the
Hg(II) centres to six. High-resolution 31P solid-state NMR spectra of the six tribenzylphosphine mercury(II)-
complexes, obtained by combining magic-angle spinning, proton dipolar decoupling and proton–phosphorus
cross-polarization (CP-MAS), have been recorded. The spectra of [Hg(PBz3)2](BF4)2 and [HgX(NO3)(PBz3)] (X = Cl,
Br, I and SCN) exhibit a single line, due to species that contain non-magnetic isotopes of mercury, and satellite lines,
due to 1J(31P–199Hg) coupling. The asymmetric unit of [Hg(PBz3)2(NO3)2] contains two molecules with four
phosphorus environments, resulting in two AB spectra with 2J(31P–31P) coupling, due to species that contain
non-magnetic isotopes of mercury, and satellite lines consisting of two ABX spectra, due to 1J(31P–199Hg) coupling.
These spectra have been analysed to yield all of the chemical shifts and coupling constants involved. A remarkable
increase in 1J(31P–199Hg) is observed from [Hg(PBz3)2](BF4)2 to [Hg(PBz3)2(NO3)2] as a consequence of the
incorporation of the nitrate group into the Hg coordination sphere in the latter case. Several of the spectra also
exhibit broader satellites due to the presence of scalar spin–spin coupling between 31P and the quadrupolar 201Hg
nucleus. Slow-spinning methods have been used to analyze the spinning-sideband intensities of the NMR spectra, in
order to obtain the 31P shielding anisotropy and asymmetry parameters Dr and g. The 199Hg and 31P NMR shielding
tensors of PMe3 models of the above six compounds have been calculated using relativistic density functional theory.
The 31P results are in good agreement with experiment and assist in the assignment of some of the signals.
In this paper we report the synthesis, single crystal X-ray
Introduction
structures and solid-state 31P CP-MAS NMR spectra of the
Although the mercury(II) ion has a propensity towards two-
coordination, the presence of secondary bonding interactions
commonly results in an expansion of its coordination sphere to
tribenzylphosphine mercury(II) complexes, [Hg(PBz3)2](BF4)2,
[Hg(PBz3)2(NO3)2] and [HgX(NO3)(PBz3)] (X = Cl, Br, I and
SCN), and correlate these results with calculations of the 31P and
give a higher effective coordination number.1 This is particularly
199Hg shielding tensors. As for [Cu(PBz3)2](BF4)2,8 the complex,
evident with tertiary phosphine complexes where the co-ligands
[Hg(PBz3)2](BF4)2, contains a linear two-coordinate cation. As
are oxygen donors such as nitrate, perchlorate or acetate
far as we are aware, this bis-tribenzylphosphine complex is
anions.2–7 Since we have previously found that tribenzylphos-
the first example of a Hg(II) complex with tertiary phosphine
phine (PBz3) affords a number of low-coordinate complexes
ligands that does not show secondary bonding interactions to
of another d10 ion, copper(I), e.g. the two-coordinate linear
the mercury that would raise its effective coordination number
[Cu(PBz3)2]2+ cation,8,9 it was of interest to examine the be-
to greater than two. In contrast, the anions in the nitrato
haviour of PBz3 towards mercury(II) and to compare this with
complex, [Hg(PBz3)2(NO3)2], are bound to the mercury to give
that of triphenylphosphine. Moreover, the relatively high natural
a ‘2 + 2 + 2’ coordination sphere. In the mono-phosphine
complexes, [HgX(NO3)(PBz3)] (X = Cl, Br and I), the nitrate
anions link Hg(II) centres to give a polymeric array, whereas the
thiocyanate complex (X = SCN) is a centrosymmetric dimer.
The low frequency FT-IR spectra of these latter complexes are
also included in this report.
abundance of 199Hg (16.84%) allows the observation of spin–spin
coupling to 31P nuclei as satellite peaks in the 31P NMR spectra.
Solid-state 31P CP-MAS NMR spectra have been reported for
a range of mercury(II) complexes with phosphine ligands,10–14
but to date no mercury(II) complexes with tribenzylphosphine
ligands have been investigated by this technique.
Experimental
All chemicals were reagent grade or better and solvents were
dried by the usual methods. Solvents were distilled before
† Present address: Institute of Fundamental Sciences, Massey Univer-
sity, Auckland, Private Bag 102 904, Auckland, New Zealand.
1 6 0 2
D a l t o n T r a n s . , 2 0 0 5 , 1 6 0 2 – 1 6 1 2
T h i s j o u r n a l i s
T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 5
©

View Online
use. Tribenzylphosphine (PBz3) was obtained from the Aldrich
Chemical Co. All complexes were routinely dried in vacuo.
Microanalyses were performed by the Campbell Microanalytical
Laboratory, University of Otago.
Preparation of compounds
[Hg(PBz3)2](BF4)2. A suspension of red mercuric oxide
(0.072 g, 0.33 mmol) in aqueous tetrafluoroboric acid (50% w/w
solution, 0.0825 cm3, 0.66 mmol) and ethanol (2.5 cm3) was
added to a solution of tribenzylphosphine (0.20 g, 0.66 mmol)
in ethanol (5 cm3), and the mixture stirred at room temperature
for 2 h. The resulting crude product (0.28 g) was filtered off and
redissolved in warm acetonitrile. Vapour diffusion with diethyl
ether yielded light-sensitive, white crystals (0.16 g, 49%). Found:
C, 51.41; H, 4.40. Calc. for C42H42B2F8HgP2: C, 51.32; H, 4.30%.
[Hg(PBz3 )2(NO3 )2 ].
A solution of tribenzylphosphine
(1.742 g, 5.72 mmol) containing a suspension of Hg(NO3)2·H2O
(0.992 g, 2.86 mmol) in 30 cm3 of chloroform was refluxed
under nitrogen for 2 h. Pentane (200 cm3) was added to the
filtered hot solution and the mixture left to stand overnight.
After decantation, the remaining solid was recrystallised from
ethanol to give light tan crystals (2.00 g, 75%). Found: C, 54.14;
H, 4.49; N, 2.93. Calc. for C42H42HgN2O6P2: C, 54.05; H, 4.54;
N, 3.00%.
[HgX(NO3)(PBz3)] (X = Cl, Br, I and SCN). These com-
plexes were prepared by warming
a
suspension of
[Hg(PBz3)2(NO3)2] (as prepared above) and the appropriate
mercury(II) halide or thiocyanate in a 1 : 1 molar ratio in
a mixture of absolute ethanol and acetonitrile. The prepa-
ration of [HgCl(NO3)(PBz3)] is given as a typical example.
[Hg(PBz3)2(NO3)2] (0.30 g, 0.32 mmol) and mercury(II) chloride
(0.088 g, 0.32 mmol) were suspended in 20 cm3 of absolute
ethanol. After the addition of acetonitrile (4 cm3) the mixture
was heated until all the solid had dissolved. The clear solution
was filtered and on cooling white crystals formed (0.33 g, 85%).
Found: C, 41.97; H, 3.42; N, 2.26. Calc. for C21H21ClHgNO3P:
C, 41.87; H, 3.51; N, 2.33%.
[HgBr(NO3)(PBz3)]. Yield 0.28 g (68%). Found: C, 39.13;
H, 3.14; N, 2.07. Calc. for C21H21NO3PBrHg: C, 38.99; H, 3.27;
N, 2.17%.
[HgI(NO3)(PBz3)]. Yield 0.35 g (79%). Found: C, 36.44; H,
2.78; N, 1.94. Calc. for C21H21HgINO3P: C, 36.35; H, 3.05; N,
2.02%
[Hg(SCN)(NO3)(PBz3)]. Yield 0.31 g (77%). Found: C,
42.49; H, 3.35; N, 4.45. Calc. for C22H21HgN2O3PS: C, 42.27;
H, 3.39; N, 4.48%.
X-Ray crystallography
Single crystals were grown as described above (except that
they were removed directly from the mother-liquor and not
dried in vacuo), mounted in inert oil (Paratone N, Exxon) and
transferred into the cold gas stream of a Siemens SMART CCD
area-detector diffractometer. Crystal data are given in Table 1.
Structure solution and refinement were performed as previously
described.15 A well defined solvent molecule (acetonitrile) was
located in the case of [HgCl(NO3)(PBz3)] but not in any of the
other complexes.
CCDC reference numbers 259223–259228.
See http://www.rsc.org/suppdata/dt/b4/b419253b/ for cry-
stallographic data in CIF or other electronic format.
NMR Spectroscopy
The spectra in this work were obtained at ambient temperature
on a 300 MHz Bruker Avance spectrometer operating at a
31P frequency of 121.5 MHz. Conventional cross polarization
from protons16 and magic-angle-spinning17 techniques were
D a l t o n T r a n s . , 2 0 0 5 , 1 6 0 2 – 1 6 1 2
1 6 0 3

View Online
implemented using a Bruker 7 mm double-air-bearing probe in
which MAS frequencies of up to 7 kHz were achieved. Typical
pulse widths for 31P were 4 ls, with contact times for the cross-
polarization sequence lasting 5–1 ms. All 31P chemical shifts were
referenced with respect to 85% H3PO4(aq) at 0 ppm, using the
31P signal of external triphenylphosphine (PPh3), which occurs
at −9.9 ppm.
Results and discussion
Synthesis
The reaction of PBz3 with Hg(BF4)2 (prepared in situ from
HgO and aqueous HBF4) or Hg(NO3)2·H2O in a 2 : 1 molar
ratio afforded crystals of the bis-complexes, [Hg(PBz3)2](BF4)2
or [Hg(PBz3)2(NO3)2], respectively. The mixed anion complexes,
[HgX(NO3)(PBz3)] (X = Cl, Br, I and SCN), were obtained from
the reaction of [Hg(PBz3)2(NO3)2] with the appropriate mercury
salt, HgX2, in a 1 : 1 molar ratio.
FT-IR Spectroscopy
Far-IR spectra were recorded with 2 cm−1 resolution at room
temperature as Nujol mulls between KBr plates or as petroleum
jelly mulls between polythene plates on a Digilab FTS-60
Fourier-transform infrared spectrometer employing an FTS-
60 V vacuum optical bench with a 5 lines mm−1 wire mesh beam
splitter, a mercury lamp source and a pyroelectric triglycine
sulfate detector.
Structural data
The single-crystal X-ray structure of [Hg(PBz3)2](BF4)2 (Fig. 1)
reveals that the complex contains discrete [Hg(PBz3)2]2+ cations
with a linear ‘P2’ donor set, the P–Hg–P angle being exactly
180◦ as a result of the mercury atom being located on a centre
of symmetry. This makes it the first example of a truly two-
coordinate Hg(II) complex of the type [Hg(PR3)2]2+. However,
the exact three-fold symmetry axis along P–Hg–P is lost because
Computational details
All electronic structure and properties calculations were
performed with the Amsterdam Density Functional (ADF
2002.03) program package.18,19 Relativistic effects use the ZORA
approximation.20 The VWN local density approximation (LDA)
and the Becke88 and Perdew8621 generalized gradient approx-
imation (GGA) were used throughout, if not stated otherwise.
The all-electron basis sets (ADF 2002.03), which use Slater type
orbitals (STOs), are triple f in both the core and in the valence
shells with two polarization functions added (TZ2P).
˚
of a network of hydrogen bonds (F · · · H 2.37–2.55 A) between
−
the BF4 anion and the benzyl hydrogen atoms on the PBz3
ligands. As for the bis-tribenzylphosphine copper(I) complexes,
[Cu(PBz3)2]X (X = PF6, CuCl2 or CuBr2),8,9,27 the phenyl groups
of the PBz3 ligand swing back over the metal atom to generate
a sixfold phenyl embrace (SPE)28 as the hydrogen atoms of each
phenyl group are directed towards the plane of the phenyl group
on the opposite ligand. The Hg–P bond length at 2.4032(5) A
is shorter than found for other bis-tertiary phosphine mercury
complexes of the type, [Hg(PR3)2]X2 (X = a monovalent anion).
˚
The principal components r11, r22, r33 of the nuclear magnetic
shielding tensor are defined such that
|r33 − riso| ≥ |r11 − riso| ≥ |r22 − riso|
(1)
where riso is the isotropic shielding constant, related to the
principal components of the shielding tensor by
riso = (r11 + r22 + r33)/3
(2)
and measured as
riso − rr = −diso
(3)
where diso is the isotropic chemical shift (the centreband shift)
and rr is the shielding constant for the reference compound. The
shielding anisotropy is defined as
Dr = r33 − (r11 + r22)/2
(4)
and the departure of the shielding tensor from axial symmetry
is described by the asymmetry parameter
g = (r22 − r11)/(r33 − riso)
(5)
The ADF calculations yield the principal components rxx,
ryy, rzz of the shielding tensor. The principal axis system is
defined in accordance with eqn. (1) (x = 1, y = 2, z = 3). The
calculations also yield the paramagnetic, diamagnetic and spin–
orbital contributions to the shielding tensor, which are labelled
as rP, rD and rSO, respectively.
The program Simpson22 was used for numerical simulation
of the solid-state 31P CP-MAS NMR spectra. The Minuit23,24
implementation of the Simplex algorithm25 has been used to fit
simulated to experimental spectra iteratively. However, due to
the J coupling and its sidebands, the fitted spectra contained
rather large errors and were not satisfactory. The values for
the shielding anisotropy and asymmetry parameter obtained
in the fitting procedure were finally optimized by hand to
yield satisfactory spectra. The errors in the resulting NMR
parameters are unknown. However, it should be noted in
particular that for small asymmetry parameters the spectrum
becomes relatively insensitive to small changes in g.26
Fig. 1 Two views of the [Hg(PBz3)2]2+ cation in [Hg(PBz3)2](BF4)2
(ellipsoids are drawn at the 50% probability level). Selected bond
◦
ꢀ
˚
lengths (A) and angles ( ): Hg(1)–P(1) 2.4032(5), Hg(1)–P(1 ) 2.4032(5);
P(1)–Hg(1)–P(1ꢀ) 180.0.
1 6 0 4
D a l t o n T r a n s . , 2 0 0 5 , 1 6 0 2 – 1 6 1 2

View Online
Table 2 Selected bond distances (A) and angles (◦) in
[HgX(NO3)(PBz3)] (X = Cl, Br and I)a
˚
For example, for [Hg(PCy3)2](ClO4)2 (PCy3 = tricyclohexylphos-
4
˚
phine), the Hg–P bond lengths are 2.443(2) and 2.442(2) A),
4
˚
for [Hg(PCy3)2](OAc)2, 2.439(2) and 2.440(2) A), and for
3
X
Cl
Br
I
˚
[Hg(PPh3)2(NO3)2], 2.451(1) A. However, in all these latter cases
the P–Hg–P bond angle is less than 180◦, being in the range
170.7(1)–131.76(3)◦, as a result of secondary interactions with
the anions that increase the effective coordination number of
the mercury to greater than two. Similarly, the P–Hg–P angle in
[Hg(tmpp)2]2+ [tmpp = tris(2,4,6-trimethoxyphenyl)phosphine]
is 166.51(9)◦, but in this case it is the ortho-methoxy oxygen
Hg(1)–X(1)
Hg(1)–P(1)
Hg(1)–O(1)
Hg(1)–O(l)
Hg(1)–O(m)
Hg(1)–O(n)
2.3375(7)
2.3713(8)
2.645(2)
2.717(2)
2.802(2)
3.081(2)
2.4436(7)
2.373(2)
2.506(5)
2.824(6)
2.814(6)
3.136(5)
2.6137(3)
2.392(1)
2.491(3)
2.862(4)
2.817(4)
3.140(4)
atoms on the phenyl groups that make contact with the
X(1)–Hg(1)–P(1)
X(1)–Hg(1)–O(1)
P(1)–Hg(1)–O(1)
X(1)–Hg(1)–O(l)
P(1)–Hg(1)–O(l)
O(1)–Hg(1)–O(l)
X(1)–Hg(1)–O(m)
P(1)–Hg(1)–O(m)
O(1)–Hg(1)–O(m)
161.77(3)
90.77(5)
98.50(5)
92.64(5)
97.34(5)
115.78(7)
82.88(5)
114.97(5)
70.68(6)
156.41(4)
91.6(1)
109.2(1)
100.7(1)
81.8(1)
113.1(2)
89.6(1)
107.7(1)
69.9(2)
152.66(3)
94.01(9)
109.73(8)
101.90(8)
81.84(8)
112.5(1)
92.21(7)
108.25(7)
69.9(1)
29
Hg [Hg–O 2.726(9)–3.044(9) A]. In [Hg(PMe3)2]2+ (PMe3 =
˚
trimethylphosphine), which occurs in the 2 : 3 adduct of PMe3
with HgI2, the P–Hg–P angle is exactly 180◦ but the Hg is
best regarded as six coordinate as a result of interactions
with iodine atoms in [Hg2I6]2− counterions.30 SPE interactions,28
which are considered to be strongly attractive with stabilisation
energies ranging between 60 and 85 kJ mol−1, would, along
with the weakly coordinating BF4− anion, explain the preference
for the formation of the two-coordinate [Hg(PBz3)2]2+ cation.
Although p-coordination of arenes to Hg(II) centres is known, it
is relatively weak and results in Hg–C(aromatic) distances in the
a For l read 2 for the Cl and 2ꢀ for the Br and I; for m read 3 for the Cl
and 3ꢀ for the Br and I; for n read 3ꢀ for the Cl and 3 for the Br and I:
primed atoms are generated by x + 1/2, −y + 3/2, −z (Br and I) and
−x + 3/2,y − 1/2, −z + 1/2 (Cl).
range 3–3.5 A.31 In the [Hg(PBz3)2]2+ cation, the nearest atoms to
˚
the Hg(II) centre are in fact two benzyl carbons, C(11) and C(21)
˚
[Hg(1)–C 3.434(2), and 3.460(2) A], with the nearest aromatic
analogue, [Hg(PPh3)2(NO3)2], although in this latter case the
P–Hg–P angle at 131.76(3)◦ is markedly smaller and the Hg–P
carbons being C(12) and C(17) [Hg(1)–C 3.504(2) and 3.453(2)
A], which suggests that there are no significant interactions of
the phenyl rings with the Hg(II) centres.
The single-crystal X-ray structure of [Hg(PBz3)2(NO3)2] con-
tains two distinct molecules in the asymmetric unit. In this
complex the coordination of the anions (Fig. 2) causes the
P–Hg–P bond angles [153.18(4) and 150.85(4)◦] to deviate
from linearity. Therefore, if only the four closest contacts are
considered, the Hg can be considered to be in a distorted
C2v ‘P2O2’ environment. However, each nitrate ion also makes
˚
3
˚
bond distance [2.451(1) A] longer.
Selected geometric parameters for the three halide complexes
[HgX(NO3)(PBz3)] (X = Cl, Br and I) are listed in Table 2.
The complexes all contain one molecule in the asymmetric unit
of the structure with [HgI(NO3)(PBz3)] and [HgBr(NO3)(PBz3)]
crystallising in the orthorhombic space group P212121, and
[HgCl(NO3)(PBz3)] in the monoclinic space group P21/n. This
latter compound also contains a molecule of acetonitrile (which
is readily lost) in the lattice. Despite these differences, all
three structures are characterized by a linking of successive
asymmetric units by the nitrate anions. In the case of the bromide
and iodide complexes this leads to a polymer generated by a 21
screw axis parallel to a and in the case of the chloride complex
to a polymer generated by the 21 screw axis parallel to b (Fig. 3).
˚
another weaker contact [Hg–O 2.868(4)–3.048(4) A] with the Hg
atoms compared to 2.479(4)–2.601(4) A for the more strongly
coordinated nitrate oxygen atoms. The weaker contacts are
less than the sum of the van der Waals radii for mercury and
˚
32
˚
oxygen (∼3.1 A) and are therefore indicative of Hg being six
coordinate. In accord with the nitrate coordination, the Hg–
˚
P bond lengths, which lie in the range 2.416(1)–2.435(1) A,
are slightly longer than the value of 2.4032(5) A observed for
˚
the two-coordinate [Hg(PBz3)2](BF4)2 compound. The structure
of the nitrato complex is similar to its triphenylphosphine
Fig. 2 View of [Hg(PBz3)2(NO3)2] (molecule 1) (ellipsoids are drawn
˚
at the 50% probability level). Selected bond lengths (A) and angles
(◦) for molecules 1 and 2: Hg(1)–P(1) 2.429(1), 2.435(1), Hg(1)–P(2)
2.416(1), 2.431(1), Hg(1)–O(1) 2.601(4), 2.510(3), Hg(1)–O(4) 2.479(4),
2.523(4), Hg(1)–O(2) 3.010(4), 3.048(4), Hg(1)–O(5) 3.031(4), 2.868(4);
P(1)–Hg(1)–P(2) 153.18(4), 150.85(4), P(2)–Hg(1)–O(1) 112.1(1),
112.93(8), P(1)–Hg(1)–O(1) 79.76(10), 88.12(8), P(2)–Hg(1)–O(4)
101.19(9), 91.32(9), P(1)–Hg(1)–O(4) 105.23(9), 113.43(9), O(1)–Hg(1)–
O(4) 74.6(1), 78.14(12).
Fig. 3 View of the polymeric chain in (a) [HgCl(NO3)(PBz3)], (b)
[HgBr(NO3)(PBz3)] and (c) [HgI(NO3)(PBz3)].
In each case, the mercury(II) atom environment (Fig. 4) is
dominated by close associations of the halide and phosphorus
D a l t o n T r a n s . , 2 0 0 5 , 1 6 0 2 – 1 6 1 2
1 6 0 5

View Online
Fig. 5 View of [Hg(SCN)(NO3)(PBz3)] (ellipsoids drawn at the
˚
50% probability level). Selected bond lengths (A) and angles
(◦): Hg(1)–P(1) 2.3958(8), Hg(1)–O(1) 2.464(3), Hg(1)–N(2)
2.486(3), Hg(1)–S(1) 2.4438(8), Hg(1)–O(3) 2.902(3), Hg(1)–O(3ꢀ)
3.011(3); P(1)–Hg(1)–S(1) 147.70(3), O(1)–Hg(1)–N(2) 87.14(10),
P(1)–Hg(1)–O(1) 116.13(6), S(1)–Hg(1)–O(1) 91.28(6), P(1)–Hg(1)–
N(2) 101.92(7), S(1)–Hg(1)–N(2) 95.87(7).
closer to a distorted trigonal-pyramidal than to a distorted
tetrahedral arrangement. However, two weaker nitrate oxygen
atom contacts indicate that the Hg has a ‘4 + 2’coordination
˚
sphere with the nitrate bidentate [Hg(1)–O(3) 2.902(3) A]. The
sixth position is occupied by an oxygen from a nitrate bound to
ꢀ
˚
an adjacent dimer [Hg(1)–O(3 ) 3.011(3) A]. Overall, the dimers
form a polymer-like structure, comparable to [HgI(NO3)(PBz3)]
and [HgBr(NO3)(PBz3)], along the crystallographic a-axis.
NMR Spectra
The 31P CP-MAS NMR spectrum of [Hg(PBz3)2](BF4)2 shows,
consistent with the crystal structure, one phosphorus environ-
ment with an isotropic chemical shift of 104.5 ppm (Fig. 6).
The expected A2X-spin system consists of three main features:
an intense central peak arising from 31P nuclei that are bound
to Hg isotopes which do not possess nuclear spin, two very
low-intensity and broad satellites arising from 31P nuclei that are
bound to 201Hg isotopes with I = 3/2 (these are explained in more
detail below) and two satellites, each with approximately 13% of
the intensity of the central feature, arising from 31P nuclei which
are adjacent to 199Hg nuclei. The intensity of these satellites
is approximately that expected from the natural abundance of
199Hg (16.84% natural abundance, I = 1/2). These satellites are
displaced to either side of the central feature due to the large
indirect spin–spin coupling between 31P and 199Hg.
The isotropic values for the 31P chemical shift and the 1J(31P–
199Hg) coupling constant can readily be determined from the
spectrum and are presented in Table 3. The 31P CP-MAS
NMR spectrum of [Hg(PBz3)2(NO3)2] is presented in Fig. 7.
This compound shows a much more complex spectrum than
[Hg(PBz3)2](BF4)2 and is consistent with a structure in which
there are two molecules of the complex in the asymmetric unit,
each with two inequivalent phosphorus atoms. The spectrum
consists of eight intense central peaks, arising from two AB
spectra, and of satellites with complicated structures, displaced
to either side of the large central features, which are the AB
parts of the two ABX spectra. The intensity of these satellites is
again much lower than that of the central AB spectra due to the
Fig. 4 View of the mercury environment in (a) [HgCl(NO3)(PBz3)], (b)
[HgBr(NO3)(PBz3)], (c) [HgI(NO3)(PBz3)] (ellipsoids are drawn at the
50% probability level).
atoms. Significant deviations from linear coordination of the
mercury centre [P(1)–Hg(1)–X(1) 161.77(3)–152.66(3)◦] arise
from contacts normal to the P(1)–Hg(1)–X(1) array with the
nitrate oxygen atoms. In all three structures the Hg(1)–O(1)
˚
bond length is shorter [2.645(2)–2.491(3) A] than the other
Hg–O(nitrate) contacts, thus creating an approximate T-shaped
coordination sphere for the Hg atom. However, the longer Hg–
low natural abundance of 199Hg compared to mercury isotopes
2
O distances are all within the sum of the van der Waals radii
with no nuclear spin. The central two AB spectra yield JAB
=
32
˚
˚
(∼3.1 A) although one contact, at 3.081(2)–3.140(4) A, is
192 Hz for both molecules. The two ABX spectra, which involve
1J(31P–199Hg) for every phosphorus site, have been analyzed
by simulation. The 31P NMR parameters obtained from these
spectra are listed in Table 3.
˚
weaker than the other two [2.717(2)–2.862(4) A]. The overall
result is that the nitrate anions bridge to adjacent mercury atoms
via bidentate interactions. Similar structures have been observed
for the triphenylphosphine complexes, [HgBr(NO3)(PPh3)] and
[HgI(NO3)(PPh3)].6
The solid-state 31P CP-MAS NMR spectrum of [HgI(NO3)-
(PBz3)] is shown in Fig. 8. The spectra of the bromide and
thiocyanate complexes are very similar and show the same
features. The 31P spectra all show a single central, intense
line due to molecules that contain the non-magnetic isotopes
of Hg, together with satellites (the intensity is approximately
[Hg(SCN)(NO3)(PBz3)] crystallizes as a centrosymmetric
dimer (Fig. 5) with thiocyanate ions bridging the mercury
˚
atoms. The shortest contacts to the mercury atom (2.4–2.5 A,
Fig. 5) suggest a four-fold ‘PONS’ coordination, which is
1 6 0 6
D a l t o n T r a n s . , 2 0 0 5 , 1 6 0 2 – 1 6 1 2

View Online
Fig. 6 31P CP-MAS NMR of [Hg(PBz3)2](BF4)2, at a rotor spinning rate of 6 kHz. Spinning sidebands are marked with asterisks. ↓ peaks assigned
as 201Hg satellites (see text).
Table 3 31P NMR parameters for some mercury(II) complexes
Complexa
Spin systemb
d(31P)/ppm
1J(31P–199Hg)/Hz
1J(31P–201Hg)c/Hz
2J(31P–31P)/Hz
Ref.
dA
104.5
dB
—
61.2
66.2
—
—
—
—
—
—
—
—
—
—
JAX
3283
JBX
JAB
—
192
192
—
—
—
—
—
—
—
—
—
—
154
164
180
g
g
[Hg(PBz3)2](BF4)2
[Hg(PBz3)2(NO3)2]
A2X
2ABX
—
1100
54.2
58.0
57.3
76.0
69.4
61.9
52.8
32.2
40.0
33.0
32.2
30.5
2.0
5770f
4760f
6406
7174
7856
7818
7851
9572
5550
6330
7400
8085
2644
4182
∼4600
5330f
5400f
g
g
g
g
g
h
i
[HgI(NO3)(PBz3)]
[HgBr(NO3)(PBz3)]
[HgCl(NO3)(PBz3)]d
[HgCl(NO3)(PBz3)]e
[Hg(SCN)(NO3)(PBz3)]
[Hg(PPh3)(NO3)2]2
[Hg(PPh3)(NO3)2]
[HgI(NO3)(PPh3)]
[HgBr(NO3)(PPh3)]
[HgCl(NO3)(PPh3)]
[Hg(PPh3)I2]
AX
AX
AX
AX
AX
AX
A2X
AX
AX
—
2400
2800
—
—
—
—
—
—
—
—
—
2850
2300
2700
3000
13
13
13
10
10
10
AX
ABX
ABX
ABX
6.8
18.5
∼27
2330
3183
∼4600
[Hg(PPh3)Br2]
[Hg(PPh3)Cl2]
13.0
∼27
a All chemical shifts are referenced with respect to 85% H3PO4(aq) at 0 ppm. b A, B = 31P; X = 199Hg. c Calculated from 201Hg satellite spacing/1.65,
see ref. 13 for further details ( 60 Hz). d Site 1, see text for discussion. e Site 2, see text for discussion. f 50 Hz; see the text for a discussion of the
assignments. g This work. h R. E. Wasylishen, M. D. Lumsden and W. P. Power, J. Am. Chem. Soc., 1991, 113, 8257. i R. E. Wasylishen, M. D. Lumsden
and J. F. Britten, J. Phys. Chem., 1995, 99, 16602.
1
13% of the central feature), which are assigned to 1J(31P–199Hg)
coupling. The linewidths of these satellites are similar to that
of the central line. The pattern of the spectra is the same as for
[Hg(PBz3)2](BF4)2. The 31P NMR spectra of the iodide, bromide
and thiocyanate compounds each contain peaks that cannot be
assigned and are considered as small amounts of impurity (see
peaks marked X in the spectrum of the iodide, Fig. 8). The iso-
tropic 31P NMR parameters obtained from these spectra are lis-
ted in Table 3. The observation of single 31P chemical shifts
and coupling constants for the iodide, bromide and thiocyanate
complexes is consistent with the results of the crystal structures,
which show that there is one molecule in each asymmetric unit
and therefore only one phosphorus environment. However, the
observed spectrum for the chloride complex is inconsistent
with the results of the crystal structure. The NMR spectrum
shows two central lines and thus two different J(31P–199Hg)
coupling constants, indicating two phosphorus environments.
The crystal structure, though, shows only one phosphorus
environment. The observation of two environments in the NMR
is attributed to partial loss of the lattice acetonitrile (see above),
generating another solid phase of the complex with different
NMR parameters.
The increase of the 1J(31P–199Hg) values on going from X = I
to X = Cl (Table 3) is similar to that observed for the
analogous triphenylphosphine series, [HgX(NO3)(PPh3)], as well
as for other phosphine–mercury(II) complexes (see Table 3).
Considering these observations, there is some indication that the
J coupling increases with decreasing electron-donating ability of
the halide donor ligand, as a result of the consequent increase
in the Hg 6s electron density.14 This argument can also be
D a l t o n T r a n s . , 2 0 0 5 , 1 6 0 2 – 1 6 1 2
1 6 0 7

View Online
Fig. 7 Simulated (upper) and experimental (lower) 31P CP-MAS NMR of [Hg(PBz3)2(NO3)2], at a rotor spinning rate of 6 kHz.
Fig. 8 31P CP-MAS NMR of [HgI(NO3)(PBz3)], at a rotor spinning rate of 6 kHz; X = impurity, ↓ peaks assigned as 201Hg satellites (see text).
Spinning sidebands are marked with asterisks. The upper spectrum is scaled by a factor of 6.
1
applied to account for the rather small J(31P–199Hg) coupling
constant in [Hg(PBz3)2](BF4)2 compared to [Hg(PBz3)2(NO3)2].
The mercury(II) centre in [Hg(PBz3)2](BF4)2 is only coordinated
to the tribenzylphosphine ligands, whereas the mercury(II)
centre in [Hg(PBz3)2(NO3)2] is additionally coordinated to the
nitrate anions. The electron-withdrawing character of the nitrate
anions yields a greater 6s electron density and hence a greater J
coupling constant.
In addition to the sharp 199Hg satellites, the spectra of the
iodide, bromide, thiocyanate and tetrafluoroborate complexes
each contain two much broader, less intense satellites, which are
also obtained in the 31P CP-MAS spectrum of [Hg(PBz3)2](BF4)2
1 6 0 8
D a l t o n T r a n s . , 2 0 0 5 , 1 6 0 2 – 1 6 1 2

View Online
Table 4 31P CSA parameters of [HgX(NO3)(PBz3)] (X = I, Br, SCN) and [Hg(PBz3)2](BF4)2; dii and Dr are given in ppm
Complex
[HgI(NO3)(PBz3)]
[HgBr(NO3)(PBz3)]
[Hg(SCN)(NO3)(PBz3)]
[Hg(PBz3)2](BF4)2
a
d33
d22
d11
diso
Dr
g
20.6
67.3
83.9
57.3
54.9
0.45
10.7
102.1
115.2
76.0
98.0
0.2
95.3
43.7
19.4
41.5
134.5
137.5
104.5
94.5
a
a
52.8
−63.8
0.57
0.05
a Calculated from diso, Dr and g, and assigned in accordance with eqn. (1).
Fig. 9 Spinning sideband analysis of the 31P spectrum of [Hg(PBz3)2](BF4)2 at a rotation rate of 2.5 kHz. Spinning sidebands are marked with
asterisks. The experimental 1J(31P–199Hg) value of 3282 Hz is used. The dashed spectrum represents the simulation.
(see peaks marked ↓ in Figs. 6 and 8). These satellites have also
been observed previously for [HgX(NO3)(PPh3)] (X = Cl, Br and
I), where they were shown to be due to 1J(31P–201Hg) coupling.13
The 31P CP-MAS NMR spectra of [Hg(PBz3)2](BF4)2 and
[HgX(NO3)(PBz3)] (X = Br, I and SCN) have been measured
at a rotor spinning speed of 2.5 kHz {3 kHz} and the spinning
sideband manifold simulated with Simpson.22 The chemical shift
anisotropy (CSA) parameters are presented in Table 4. The
experimentally obtained 1J(31P–199Hg) constants have been used
throughout the simulations, which were satisfactory for all four
cases. As an example, the simulation of the 31P spectrum of
[Hg(PBz3)2](BF4)2 is presented in Fig. 9. The bromide complex
gives the shortest Hg–P bond distance and the largest P–Hg–
X angle of the three complexes [HgX(NO3)(PBz3)] (X = Br, I
and SCN), which is reflected in the CSA parameters, yielding
the smallest shielding asymmetry and the largest shielding
anisotropy.
Fig. 10 Far-infrared spectra of [HgX(NO3)(PBz3)]; (a) X = Cl, (b) X =
Br, (c) X = I and (d) X = SCN. Bands due to m(HgX) and m(HgP) are
FT-IR Spectra
labelled with their wavenumbers.
The far-infrared spectra of [HgX(NO3)(PBz3)] (X = I, Br, Cl)
are shown in Fig. 10. These display intense, halogen-sensitive
bands at 316, 232 and 181 cm−1 (X = Cl, Br and I), which can
readily be assigned as m(Hg–X) modes. Assignments of m(Hg–X)
for these and for some PPh3 complexes of mercury(II) halides
are given in Table 5. The present assignments are supported
by the monotonic increase in m(Hg–X) with decreasing mass
of X. Such a relationship has been observed for a wide range
of mercury halide complexes and the present results fit in very
well with the correlations reported previously.13 Relatively broad
peaks at about 160–170 cm−1 are tentatively assigned to m(Hg–
P), in agreement with assignments for the corresponding PPh3
complexes.13
D a l t o n T r a n s . , 2 0 0 5 , 1 6 0 2 – 1 6 1 2
1 6 0 9

View Online
Table 5 IR absorption bands [wavenumbers (cm−1); Xt = termi-
nal halide] for [HgX(NO3)(PBz3)] and some related mercury halide
complexes
Table 6 Calculated 199Hg and 31P shielding parameters (ppm) in HgMe2
and PPh3
a
b
HgMe2
D3d
PPh3
C1
Complex
X
m(Hg–Xt)
m(Hg–P)
Ref.
Symmetry
Nucleus
[HgX(NO3)(PBz3)]
Cl
Br
I
Cl
Br
I
316
232
181
321
231
189
232
221
153
—
170
157
160
164
161
147
137
108
132
104
133
98
This work
This work
This work
13
13
13
199Hg
31P
riso
rxx
ryy
rzz
rP
7965.5c
5311.7
5311.7
13237.1
−4103.4
9614.1
2454.8
7961.3
0.0
354.3d
326.7
342.3
393.8
−624.3
966.1
12.4
[HgX(NO3)(PPh3)]
[HgX2(PPh3)2]
Cl
33, 34
rD
rSO
Dr
g
Br
I
33, 34
34, 35
59.3
0.4
127
—
a Experimental geometry from ref. 38, see text. b Experimental geometry
from ref. 42. c Reference for 199 Hg chemical shift calculations. d Reference
for 31P chemical shift calculations.
As expected, there is no significant discrepancy for
the assigned modes between the [HgX(NO3)(PBz3)] and
[HgX(NO3)(PPh3)] complexes, as the geometry of the mer-
cury(II) environment is very similar. The assignment of the HgS
stretching mode in the X = SCN complex is not so clear. The
peaks at 272 and 306 cm−1 are both in the region appropriate
for m(HgS). In solid Hg(SCN)2, m(HgS) occurs at 313 cm−1.36
In [Hg(SCN)(CH3CO2)], m(Hg–S) is assigned at 280 cm−1.37
Considering the SCN ion as a single pseudohalide X, there
should be four m(Hg–X) modes, two antisymmetric (IR active)
and two symmetric (Raman active), so the presence of two bands
in the IR is consistent with the observed structure. However,
bands at 272 and 306 cm−1 also appear in the spectrum of
[HgBr(NO3)(PBz3)], which casts some doubt on this assignment.
referenced to 85% H3PO4, all of the calculated 31P chemical
shifts are raised by 9.9 ppm, so that experimental and calculated
31P chemical shifts are directly comparable. Note that for
calculations of 31P shieldings with the TZ2P basis set, 364.2 ppm
is used as a reference (rr) in this work, as highlighted in Table 6.
In order to make the calculations of the 31P and 199Hg
shielding feasible, it is imperative to ‘scale down’ the large triben-
zylphosphine ligands of the introduced complexes. Otherwise
the calculations are computationally too demanding. However,
the greatest simplification in this work is that only one molecule
from the solid structure of the introduced compounds has been
considered in the calculations. To date no ab initio calculations of
the NMR shielding of a solid have been reported in the literature,
because the required calculation time is simply immense. As
noted before, all following calculations of the NMR shielding
are based on the experimental geometries obtained by X-ray
diffraction.
Calculation of 31P and 199Hg NMR shielding tensors
The aim of this part of the work was to calculate the 31P and
199Hg shielding tensors of the above mentioned compounds
and compare the 31P results with experimental data that have
been obtained for these compounds, in order to assist with
assignments of some of the NMR signals.
In principle, the tribenzylphosphine ligand can be simplified
to trimethylphosphine (PMe3) or simply phosphine (PH3). We
have calculated the 31P chemical shift for [Hg(PBz3)2(NO3)2]
using both ‘simplified’ ligands, PMe3 and PH3, respectively. As
expected, the simplification of the PBz3 by PH3 is inappropriate
and PMe3 gives far better results. In Table 7 the outcome of
the 31P and 199Hg chemical shifts for the two [Hg(PMe3)2(NO3)2]
molecules are presented. Here, the geometries of both molecules
of the asymmetric unit of [Hg(PBz3)2(NO3)2] have been adopted.
The calculated 31P chemical shifts agree very well with the
experimental data for the PBz3 complexes (Table 7). This very
satisfactory result enables the assignment of the chemical shifts
In order to obtain a relative shielding scale for 199Hg, the
shielding of dimethylmercury(II), which is a very common
standard for 199Hg, has been calculated using experimental
38
˚
˚
geometries (Hg–C 2.083 A, C–H 1.106 A). Wolff and Ziegler
have shown that the mercury NMR shieldings of HgCl2 depend
linearly on the bond length with a change of approximately
˚
50 ppm for each 0.01 A. A roughly linear relationship was also
found between the shift and the Cl–Hg–Cl bond angle with a
change of 100 ppm for each incremental decrease on the bond
angle by 10◦.38 Thus, the accuracy of the calculation depends
strongly on the geometry used and it is sensible to adopt fixed
geometries (e.g. from X-ray diffraction), rather than to optimize
the structure. The fact that using optimized geometries yields
insufficient accuracy in the calculation of the shielding tensor of
course limits the possibilities of a complete theoretical treatment
of NMR properties.
1
to each phosphorus environment and thus each J(31P–199Hg)
in [Hg(PBz3)2(NO3)2] (see Table 7). According to the present
assignment, the 1J(31P–199Hg) generally increase with decreasing
d(P–Hg) bond length, which has also been found for other
mercury(II)–phosphine compounds.14
Furthermore, it has been shown that, using experimental
geometries combined with the generalized gradient approxi-
mation (GGA) and the BP functional approach with TZ2P
basis sets, good agreement with experiment is obtained.38–41 The
results of this calculation are presented in Table 6. Note that
for calculations of 199Hg shieldings with the TZ2P basis set,
7965.5 ppm is used as a reference (rr) in this publication, as
highlighted in Table 6.
A relativistic formulation of the scalar nuclear spin–spin
coupling constant is implemented into the ADF program
package used in this work. Several attempts have been made to
obtain accurate calculations for the 1J(31P–199Hg). However, the
deviation from the experimental values was always significant.
Autschbach and Ziegler have reported calculations of the J
coupling constants in various compounds, including mercury(II)
complexes.43 Considering the mercury(II) complexes, their cal-
culations yield a relative accuracy of about 25% for binary com-
pounds. In cases of ‘light’ nuclei (e.g. 31P), however, the accuracy
is much better. The calculated 2J(31P–31P) in [Hg(PBz3)2(NO3)2]
matches indeed the experimental value (calculated 192.1 Hz,
experimental 192 Hz).
In order to obtain a relative shielding scale for 31P, the
shielding of triphenylphosphine (PPh3), which has a shift of
d −9.9 ppm with respect to 85% H3PO4, has been calculated
using C1 symmetry, based on the geometric parameters of
the triphenylphosphine molecule in crystal form, where the
molecular symmetry is reduced to C1 relative to the C3 symmetry
of the gas phase.42 Since the experimental 31P chemical shifts are
In Table 7 the calculated NMR parameters for [Hg(PMe3)2]-
(BF4)2 using the structural parameters of [Hg(PBz3)2](BF4)2
1 6 1 0
D a l t o n T r a n s . , 2 0 0 5 , 1 6 0 2 – 1 6 1 2

View Online
Table 7 Calculated 31P and 199Hg shielding parameters and chemical shifts (ppm)
b
Compound
Atom
rr − riso
diso,exp
Dr
33.3
−35.8
−59.9
−32.3
2778.4
2573.4
55.2
6152.1
58.0
3817.9
66.6
g
[Hg(PMe3)2(NO3)2]a
P(1)
P(2)
P(1A)
P(2A)
Hg(1)
Hg(1A)
P
Hg
P
Hg
P
69.2
56.1
62.6
66.2
54.2
61.2
58.0
0.96
0.76
0.50
0.74
0.20
0.15
0.14
0.02
0.27
0.11
0.58
0.09
0.47
0.20
0.87
60.6
−1256.7
−1247.3
102.3
—
—
[Hg(PMe3)2](BF4)2
[HgCl(NO3)(PMe3)]
[HgBr(NO3)(PMe3)]
[HgI(NO3)(PMe3)]
[HgSCN(NO3)(PMe3)]
104.5
—
69.4 (61.9)
—
76.0
—
57.3
—
−1036.5
52.9
−1442.9
55.3
Hg
P
Hg
P
−1594.5
3056.3
62.7
2140.3
−31.4
56.3
−2241.7
56.9
52.8
a Atom labelling as for the corresponding PBz3 complex; see Fig. 2 (shows labelling for molecule 1 only) b Experimental results for corresponding
PBz3 complexes.
are summarized. The experimentally observed 31P chemical shift
is 104.5 ppm, which deviates less than 3% from the calcu-
lated value. In the same manner as for [Hg(PBz3)2](BF4)2 and
[Hg(PBz3)2(NO3)2], the NMR parameters for [HgX(NO3)(PBz3)]
(X = I, Br, Cl and SCN) have been calculated by carrying
out calculations on the corresponding PMe3 compounds. The
results are presented in Table 7. The 31P calculations reveal
isotropic chemical shifts in the vicinity of 55 ( 2) ppm for
all four complexes. In the case of the halide compounds there
exists a trend, where the iodide complex shows the most shielded
chemical shift (56.3 ppm) and the chloride the least (52.9 ppm).
This trend is not observed in the experimental results for
[HgX(NO3)(PPh3)] (X = I, Br and Cl) (Table 7), where agreement
with the calculated value is good for X = I but not for the X =
Br, Cl complexes, for which significantly larger chemical shifts
are observed experimentally. The reason for this is not known at
present, but it can be noted that these shifts are quite sensitive to
the crystal environment, two quite different shifts being observed
for two different crystal forms for X = Cl (Table 7). This point is
discussed further below. Agreement between the calculated and
observed chemical shifts is also reasonable for the X = SCN
complex (Table 7).
As with the isotropic chemical shifts, agreement between the
calculated and experimental shielding anisotropy and asymme-
try parameters for [HgX(NO3)(PMe3)] is reasonable in the cases
X = I, SCN. It is noteworthy that the experimentally observed
change in the sign of Dr from X = I to X = SCN is reproduced in
the calculations. Experimental values of the shielding anisotropy
and asymmetry parameters could not be obtained for X =
Cl, and significantly poorer agreement between the calculated
and experimental values was observed for X = Br. A possible
reason for this is that the calculations only involve one molecule
and no further atoms or molecule fragments, which may be
sufficiently close to the phosphorus nucleus to affect these values,
are considered. Furthermore, the tribenzylphosphine ligand has
been simplified to trimethylphosphine, which also will have a
distinct effect on the asymmetry and anisotropy.
with ligand type and coordination environment that have been
observed previously for a range of mercury(II) complexes.14
HgMe2 shows the largest Dr value (Table 6), which is in
reasonable agreement with the experimentally observed value
of 7300 ppm for this molecule in a liquid crystal environment.44
The other complexes show lower Dr values because the ligands
involved are weaker r-donors than the methyl group, and show
a further decrease as the Hg bonding environment changes from
linear two-coordinate to four-coordinate with the incorporation
of the nitrate ions into the coordination sphere.
Conclusions
The tribenzylphosphine (PBz3) complexes of mercury(II),
[Hg(PBz3)2](BF4)2, [Hg(PBz3)2(NO3)2] and [HgX(NO3)(PBz3)]
(X = Cl, Br, I and SCN) have been synthesised and their
structures determined by single-crystal X-ray crystallography.
[Hg(PBz3)2](BF4)2 contains [Hg(PBz3)2]2+ cations with linear
P–Hg–P coordination, the first example of a truly two-
coordinate [Hg(PR3)2]2+ complex. The mercury coordination
in [Hg(PBz3)2(NO3)2] can be described as distorted tetrahedral,
with a significant deviation of the P–Hg–P angle from linearity
as a result of coordination of the nitrate groups. Nitrate coor-
dination is also observed in [HgX(NO3)(PBz3)] (X = Cl, Br, I),
resulting in significantly non-linear P–Hg–X coordination. The
thiocyanate complex is a centrosymmetric thiocyanate-bridged
dimer with distorted trigonal-pyramidal mercury coordination
to the P atom of PBz3, to the S and N atoms of two bridging
thiocyanate groups, and to the O atom of one nitrate group.
The 31P CP-MAS NMR spectra of [Hg(PBz3)2](BF4)2 and
[HgX(NO3)PBz3] (X = Cl, Br, I and SCN) exhibit a single
line due to species that contain non-magnetic isotopes of
1
mercury and satellite lines due to J(31P–199Hg) coupling. The
asymmetric unit of [Hg(PBz3)2(NO3)2] contains two molecules
with four phosphorus environments, resulting in two AB spectra
2
with J(31P–31P) coupling and satellite lines consisting of two
The presence of effects of ‘neighbouring’ molecule fragments
on Dr and g is shown and discussed for diammine–silver(I)
cations in diammine–silver(I) nitrate and sulfate.39 Hence, the
accuracy of the calculated Dr and g values can only be
surveyed by calculating another model that represents the actual
phosphorus environment more closely, and then compare the
difference between the results of these parameters for both
models.
ABX spectra due to 1J(31P–199Hg) coupling. A remarkable
increase in 1J(31P–199Hg) is observed from [Hg(PBz3)2](BF4)2
to [Hg(PBz3)2(NO3)2] as a consequence of the incorporation
of the nitrate groups into the Hg coordination sphere in the
latter case. Several of the spectra also exhibit broader satellites
due to the presence of scalar spin–spin coupling between 31P
and the quadrupolar 201Hg nucleus. The 199Hg and 31P NMR
shielding tensors of PMe3 models of the above six compounds
have been calculated using relativistic density functional theory.
The 31P results are in good agreement with the values obtained
experimentally by slow-spinning methods, and assist in the
assignment of some of the signals.
Considering the calculated 199Hg NMR parameters, the chlo-
ride complex shows the least shielded and the iodide complex
the most shielded isotropic chemical shift. The calculated
anisotropic 199Hg NMR parameters show the kinds of changes
D a l t o n T r a n s . , 2 0 0 5 , 1 6 0 2 – 1 6 1 2
1 6 1 1

View Online
21 A. D. Becke, Phys. Rev. A, 1988, 38, 3098–3100; J. P. Perdew, Phys.
Acknowledgements
Rev. B, 1986, 33, 8822–8824; J. P. Perdew, Phys. Rev. B, 1986, 34,
7406–7406.
We would like to thank Professor Peter Schwerdtfeger for
computer time and advice regarding the ADF calculations and
Dr Zoran Zujovic for helping with the NMR measurements.
22 M. Back, J. T. Rasmussen and N. C. Nielsen, J.Magn. Reson., 2000,
147, 296–330.
23 T. Vosegaard, A. Malmendal and N. C. Nielsen, Chem. Monthly,
2002, 133, 1555–1574.
24 F. James and M. Ross, Comput. Phys. Commun., 1975, 10, 343–367.
25 F. H. Walters and L. R. Parker, Sequential Simplex Optimization,
CRC Press, Boca Raton, FL, 1993.
26 R. K. Harris, P. Jackson, L. H. Merwin and B. J. Say, J. Chem. Soc.,
Faraday Trans. 1, 1988, 84, 3649–3672; N. J. Clayden, C. M. Dobson,
L.-Y. Lian and D. J. Smith, J. Magn. Reson., 1986, 69, 476–493.
27 P. D. Akrivos, P. P. Karagiannidis, C. P. Raptopoulou, A. Terzis and
S. Stoyanov, Inorg. Chem., 1996, 35, 4082–4083.
References
1 K. Brodersen and H.-U. Hummel, Comprehensive Coordination
Chemistry, ed. G. Wilkinson, Pergamon, Oxford, 1987, vol. 5, p.
1047.
2 P. J. Roberts, G. Ferguson, R. G. Goel, W. O. Ogni and R. J. Restivo,
J. Chem. Soc., Dalton Trans., 1978, 253–256.
3 H. B. Buergi, E. Fischer, R. W. Kunz, M. Parvez and P. S. Pregosin,
Inorg. Chem., 1982, 21, 1246–1256.
4 E. C. Alyea, S. A. Dias, G. Ferguson and M. A. Khan, J. Chem. Res.
(S), 1979, 360–361.
28 M. Scudder and I. Dance, J. Chem. Soc., Chem. Commun., 1995,
1039–1040; M. Scudder and I. Dance, J. Chem. Soc., Dalton Trans.,
1998, 3167–3175.
5 S. H. Whitlow, Can. J. Chem., 1974, 52, 198–202.
6 L.-J. Baker, G. A. Bowmaker, P. C. Healy, B. W. Skelton and A. H.
White, J. Chem. Soc., Dalton Trans., 1992, 989–997.
7 T. S. Lobana, M. K. Sandu, D. C. Povey and G. W. Smith, J. Chem.
Soc., Dalton Trans., 1988, 2913–2914.
8 E. W. Ainscough, A. M. Brodie, A. K. Burrell, J. V. Hanna, P. C.
Healy and J. M. Waters, Inorg. Chem., 1999, 38, 201–203.
9 E. W. Ainscough, A. M. Brodie, A. K. Burrell, G. H. Freeman, G. B.
Jameson, G. A. Bowmaker, J. V. Hanna and P. C. Healy, J. Chem.
Soc., Dalton Trans., 2001, 144–151.
10 G. A. Bowmaker, H. J. Clase, N. W. Alcock, J. M. Kessler, J. H.
Nelson and J. S. Frye, Inorg. Chim. Acta, 1993, 210, 107–204.
11 M. D. Lumsden, R. E. Wasylishen and J. F. Britten, J. Phys. Chem.,
1995, 99, 16602–16608.
29 L.-J. Baker, G. A. Bowmaker, B. W. Skelton and A. H. White, J. Chem.
Soc., Dalton Trans., 1993, 3235–3240.
30 N. A. Bell, L. A. March and I. W. Nowell, Inorg. Chim. Acta, 1991,
179, 73–76.
31 M. R. Haneline, J. B. King and F. P. Gabbai, Dalton Trans., 2003,
2686–2690, and references therein.
32 R. Haid, R. Gutmann, G. Czermak, C. Langes, W. Oberhauser, H.
Kopacka, K.-H. Ongania and P. Bruggeller, Inorg. Chem. Commun.,
¨
2003, 6, 61–67.
33 N. A. Bell, T. D. Dee, M. Goldstein and P. McKenna, Inorg. Chim.
Acta, 1983, 71, 135–140.
34 L. Falth, Chem. Scr., 1976, 9, 71–73.
¨
35 G. B. Deacon, J. H. S. Green and D. Harrison, Spectrochim. Acta.,
Part A, 1968, 24, 1921–1928.
12 W. P. Power, M. D. Lumsden and R. E. Wasylishen, J. Am. Chem.
Soc., 1991, 113, 8257–8262.
13 L. J. Baker, G. A. Bowmaker, P. C. Healy, B. W. Skelton and A. H.
White, J. Chem. Soc., Dalton Trans., 1992, 989–997.
14 G. A. Bowmaker, R. K. Harris and S.-W. Oh, Coord. Chem. Rev.,
1997, 167, 49–94.
36 R. P. Cooney and J. R. Hall, Aust. J. Chem., 1969, 22, 2117–2123.
37 G. A. Bowmaker, A. V. Churakov, R. K. Harris and S.-W. Oh,
J. Organomet. Chem., 1998, 550, 89–99.
38 S. K. Wolff, T. Ziegler, E. van Lenthe and E. J. Baerends, J. Chem.
Phys., 1999, 110, 7689–7698.
39 G. A. Bowmaker, R. K. Harris, B. Assadollahzadeh, D. C. Apperley,
P. Hodgkinson and P. Amornskchai, Magn. Reson. Chem., 2004, 42,
819–826.
15 E. W. Ainscough, A. M. Brodie, P. C. Healy and J. M. Waters,
Aust. J. Chem., 2000, 53, 971–973.
16 A. Pines, M. G. Gibby and J. S. Waugh, J. Chem. Phys., 1973, 59,
¨
40 J. Jokisaari, S. Jarvinen, J. Autschbach and T. Ziegler, J. Phys. Chem.
569–590.
A, 2002, 106, 9313–9318.
17 E. R. Andrew, A. Bradbury and R. G. Eades, Nature (London), 1958,
182, 1659.
¨
41 Diplomarbeit Behnam Assadollahzadeh, Westfalische Wilhelms-
Universitat zu Munster, Germany, 2003.
¨
¨
18 G. te Velde, F. M Bickelhaupt, E. J. Baerends, S. J. A. van Gisbergen,
C. Fonseca Guerra, J. G. Snijders and T. Ziegler, ‘Chemistry with
ADF’, J. Comput. Chem., 2001, 22, 931–967.
19 ADF2002.03, SCM, Theoretical Chemistry, Vrije Universiteit Ams-
terdam, The Netherlands, http://www.scm.com.
42 V. A. Naumov, M. A. Tafipolskii, A. V. Naumov, D. Y. Shorokhov
and S. Samdal, Russ. J. Gen. Chem., 2001, 8, 1225–1232.
43 J. Autschbach and T. Ziegler, T., J. Chem. Phys., 2000, 113, 936–
947; J. Autschbach and T. Ziegler, J. Chem. Phys., 2000, 113, 9410–
9418.
20 E. van Lenthe, E. J. Baerends and J. G. Snijders, J. Chem. Phys., 1993,
99, 4597–4610; E. van Lenthe, A. Ehlers and E. J. Baerends, J. Chem.
Phys., 1999, 110, 8943–8953.
44 J. D. Kennedy and W. McFarlane, J. Chem. Soc., Faraday Trans., 2,
1976, 72, 1653–1660; J. Jokisaari and P. Diehl, Org. Magn. Reson.,
1980, 13, 359–362.
1 6 1 2
D a l t o n T r a n s . , 2 0 0 5 , 1 6 0 2 – 1 6 1 2