Experimental observation of spin delocalisation onto the aryl-alkynyl ligand in the complexes [Mo(CCAr)(Ph2PCH2CH 2PPh2)(η-C7H7)]+ (Ar = C6H5, C6H4-4-F; C 7H7 = cycloheptatrienyl): An EPR and ENDOR investigation

Article abstract of DOI:10.1039/c0dt00642d

The paramagnetic aryl-alkynyl complexes [Mo(CCAr)(dppe)(η-C ₇H₇)]⁺ (dppe = Ph₂PCH ₂CH₂PPh₂; Ar = C₆H₅, [1]⁺; C₆D₅, [2]⁺; C ₆H₄-4-F, [3]⁺; C₆H₄-4-Me, [5]⁺) and [Mo(CCBu^t)(dppe)(η-C₇H ₇)]⁺ [4]⁺, have been investigated in a combined EPR and ENDOR study. Direct experimental evidence for the delocalisation of unpaired spin density over the framework of an aryl-alkynyl ligand has been obtained. The X-band solution EPR spectrum of the 4-fluoro derivative, [3] ⁺, exhibits resolved hyperfine coupling to the remote para position of the aryl group [a_iso(¹⁹F) = 4.5 MHz, (1.6 G)] in addition to couplings attributable to ^95/97Mo, ³¹P and ¹H of the C₇H₇ ring. A full analysis of the ¹H ENDOR spectra is restricted by the low g anisotropy of the system which prevents the use of orientation selection. However, inter-comparison of the ¹H cw-ENDOR frozen solution spectra of [1]⁺, [2] ⁺, [4]⁺ and [5]⁺, combined with spectral simulation informed by calculated values derived from DFT investigations, has facilitated estimation of the experimental a_iso(¹H) hyperfine couplings of [1]⁺ including the ortho, ±3.7 MHz (±1.3 G) and para, ±3.9 MHz (±1.4 G) positions of the C₆H₅ substituent of the aryl-alkynyl ligand.

Full text of DOI:10.1039/c0dt00642d

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/dalton | Dalton Transactions
Experimental observation of spin delocalisation onto the aryl-alkynyl ligand
in the complexes [Mo(C CAr)(Ph2PCH2CH2PPh2)(g-C7H7)]+ (Ar = C6H5,
C6H4-4-F; C7H7 = cycloheptatrienyl): an EPR and ENDOR investigation†
Emma Carter,a David Collison,*b Ruth Edge,b Emma C. Fitzgerald,c Hannah N. Lancashire,c
Damien M. Murphy,*a Joseph J. W. McDouall,c Joseph Sharplesc and Mark W. Whiteley*c
Received 11th June 2010, Accepted 23rd September 2010
DOI: 10.1039/c0dt00642d
The paramagnetic aryl-alkynyl complexes [Mo(C CAr)(dppe)(h-C7H7)]+ (dppe = Ph2PCH2CH2PPh2;
Ar = C6H5, [1]+; C6D5, [2]+; C6H4-4-F, [3]+; C6H4-4-Me, [5]+) and [Mo(C CBut)(dppe)(h-C7H7)]+ [4]+,
have been investigated in a combined EPR and ENDOR study. Direct experimental evidence for the
delocalisation of unpaired spin density over the framework of an aryl-alkynyl ligand has been obtained.
The X-band solution EPR spectrum of the 4-fluoro derivative, [3]+, exhibits resolved hyperfine coupling
to the remote para position of the aryl group [aiso(19F) = 4.5 MHz, (1.6 G)] in addition to couplings
attributable to 95/97Mo, 31P and 1H of the C7H7 ring. A full analysis of the 1H ENDOR spectra is
restricted by the low g anisotropy of the system which prevents the use of orientation selection.
However, inter-comparison of the 1H cw-ENDOR frozen solution spectra of [1]+, [2]+, [4]+ and [5]+,
combined with spectral simulation informed by calculated values derived from DFT investigations, has
facilitated estimation of the experimental aiso(1H) hyperfine couplings of [1]+ including the ortho,
3.7 MHz ( 1.3 G) and para, 3.9 MHz ( 1.4 G) positions of the C6H5 substituent of the aryl-alkynyl
ligand.
carbon chain ligand has hitherto been limited by the poorly
resolved spectra obtained for many of the leading MLx end-
Introduction
cap groups.6a,6d,7b,14 An exception is provided by the stable, 17-
electron radicals of the cycloheptatrienyl molybdenum system
[Mo{(C C)nC CR}(dppe)(h-C7H7)]+, which exhibit isotropic
EPR solution spectra with well resolved hyperfine coupling to
95/97Mo, 31P and 1H of the cycloheptatrienyl ring.13,15,16
Complexes of unsaturated carbon chain ligands supported by
redox active metal end-cap units1,2 are the focus of intense
current research activity with potential applications in electronic,3
magnetic4 and optical5 devices. Key to these investigations is an un-
derstanding of the electronic interaction between the metal centre
and the carbon chain ligand and the examination of spin density
distribution in paramagnetic species [{MLx}{(C C)nC CR)}]+
constitutes an important theoretical and experimental approach
to this issue.6,7 DFT calculations indicate that significant delocal-
isation of unpaired spin density onto the carbon chain ligand
may occur,6c,6d,7 consistent with large through bridge exchange
coupling in related ligand-bridged assemblies8–11 [{MLx}(C C–
X–C C){MLx}]+ but direct experimental confirmation of these
theoretical predictions extends a significant challenge.
In selected cases, 17-electron, paramagnetic complexes
[{MLx}{(C C)nC CR}]+ are sufficiently stable to enable direct
spectroscopic characterisation and this presents the possibility
of probing unpaired spin density distribution experimentally
using the techniques of EPR, ENDOR and paramagnetic NMR
spectroscopy.12,13 However the use of EPR spectroscopy to ob-
serve directly, hyperfine coupling to remote positions on the
DFT
calculations
on
the
aryl-alkynyl
radical
significant
[Mo(C CPh)(dppe)(h-C7H7)]+ reveal16 that
a
proportion of the unpaired spin density is localised at the
2
metal centre in an orbital with substantial metal dz character,
consistent with the isotropic character of the observed EPR
spectra. Nevertheless, there is a degree of spin delocalisation
onto the aryl-alkynyl ligand, and therefore direct experimental
observation of hyperfine coupling to the remote o- and p-
positions of the aryl substituent might be feasible given
the excellent resolution of the EPR spectra. However, the
calculated hyperfine couplings to the protons of the arene ring
(approximately 4.5 MHz, (1.5 G) for o- and p- positions)16
are close to the limit of resolution of a standard X-band,
EPR solution investigation and therefore modified strategies
to facilitate the direct experimental observation of spin density
on the aryl-alkynyl ligand are required; this paper reports
two approaches. First we examined the substitution of arene
hydrogens by fluorine with the expectation that the resultant
increased isotropic interaction for unit spin density17 would
lead to larger, observable hyperfine couplings in the solution
EPR spectrum. Secondly a 1H cw-ENDOR study has been
carried out on [Mo(C CC6X5)(dppe)(h-C7H7)]+ (X = H or
D) to identify and evaluate the authentic hyperfine parameters
for the hydrogens of the aryl-alkynyl substituent. The results
of these investigations, to our knowledge, comprise the first
aSchool of Chemistry, Cardiff University, Main Building, Park Place, Cardiff,
UK CF10 3AT. E-mail: MurphyDM@cardiff.ac.uk
bEPSRC National Service for EPR Spectroscopy, School of Chemistry,
University of Manchester, Manchester, UK M13 9PL
cSchool of Chemistry, University of Manchester, Manchester, UK M13 9PL.
E-mail: Mark.Whiteley@manchester.ac.uk; Tel: 0161 275 4634
† Electronic supplementary information (ESI) available: Fig. S1–S5.
Additional EPR/ENDOR spectra includingcw X-band31P and cw Q-band
1H-ENDOR spectra, X-band FSED EPR spectra and X-band MIMS
ENDOR spectra. See DOI: 10.1039/c0dt00642d
11424 | Dalton Trans., 2010, 39, 11424–11431
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
direct experimental observations of spin delocalisation over
the molecular framework of an aryl-alkynyl ligand using the
techniques of EPR and ENDOR spectroscopy and complement
the results of previous work on [Fe(C CAr)(dppe)Cp*]+ in
which corresponding hyperfine parameters were accessed via
paramagnetic NMR contact shifts.12
solid following purification by column chromatography. Simi-
larly [Mo(C CC6H4-4-F)(dppe)(h-C7H7)], 3, was prepared from
the reaction of [Mo(C CH)(dppe)(h-C7H7)] with 1-bromo-4-
fluorobenzene. Although the yields from these reactions were low
(ca. 10–15%), the one-step, small scale synthesis has the potential
for extension to a wide range of aryl-alkynyl ligands.
The new complexes 2 and 3 were characterised by microanalysis,
IR, mass spectroscopy, H and 31P{ H} NMR spectroscopy (see
experimental section). The successful incorporation of a per-
deuterated aryl substituent in 2 is confirmed by the peak at m/z =
693 in the mass spectrum, 5 units higher than the protio analogue,
1
1
Results and discussion
Synthetic studies
An overview of the synthetic work undertaken is given in
Scheme 1.
1
and in the H NMR spectrum, the absence of resonances at [d
(CD2Cl2) 6.03, 2H, C6H5-o; 6.74, 1H, C6H5-p; 6.84, 2H, C6H5-m]
assigned to the alkynyl phenyl substituent of 1.16 Both 2 and 3
exhibit characteristic n(C C) stretches in the solution IR spectra
in CH2Cl2; the deuteration of the phenyl substituent results in a
small decrease in n(C C) from 2045 cm-1 for 1 to 2042 cm-1 for
2.
Redox chemistry
The investigation of the 17-electron radicals [Mo(C CAr)-
(dppe)(h-C7H7)]+ ([1]+, Ar = C6H5; [2]+, Ar = C6D5; [3]+, Ar =
C6H4-4-F) by EPR and ENDOR spectroscopy requires redox
generation from the 18-electron precursors, 1, 2 and 3 respectively.
We have previously described in some detail the synthetic redox
chemistry,16,22 spectroscopic and structural23 characterisation of
[1]+ as a typical example of this complex type and therefore
the discussion here is limited to essential data to confirm that
[2]+ and [3]+ are accessible for study. To establish that [2]+ and
[3]+ are reversibly generated from 2 and 3, we examined the
cyclic voltammetry of 2 and 3 and the effect on the IR n(C C)
stretching frequency of chemical oxidation with the ferrocenium
ion [FeCp2][PF6]. The results of these studies are presented in
Table 1 alongside data for 1.
Scheme 1 Routes to [Mo(C CR)(dppe)(h-C7H7)] complexes. Reagents
and conditions: (i)/(ii) HC CPh, K[PF6], KOBut, MeOH, reflux 3 h. (iii)
[PdCl2(PPh3)2], CuI, C6D5Br or 1-bromo-4-fluorobenzene, NHiPr2, reflux
6 h.
The parent complex [Mo(C CC6H5)(dppe)(h-C7H7)], 1, was
prepared, as described previously by reaction of [MoBr(dppe)(h-
C7H7)] with HC CC6H5 in MeOH in the presence of KOBut
(Method 1); this route involves the initial formation of the
vinylidene [Mo(C CHC6H5)(dppe)(h-C7H7)]Br, which is de-
protonated in situ by KOBut to afford 1.16 Syntheses of the
substituted arene complexes [Mo(C CC6D5)(dppe)(h-C7H7)], 2,
and [Mo(C CC6H4-4-F)(dppe)(h-C7H7)], 3, might be similarly
effected using HC CC6D5 and HC CC6H4-4-F respectively,
however Method 1 has some limitations in reactions where
the arene substituent is strongly electron withdrawing and also
depends upon the availability of precursor terminal alkynes.18,19
We therefore examined the development of an alternative route
to 2 and 3, which combines enhanced flexibility in the selection
of arene substituents with the potential to carry out small scale
preparations. A reported synthesis of the aryl-alkynyl complexes
[Fe(C CAr)(dppe)Cp*] involves direct Sonogashira coupling of
a functionalised aryl-halide with the unprotected metal alkynyl
complex20 [Fe(C CH)(dppe)Cp*] and our recent synthesis21 of
[Mo(C CH)(dppe)(h-C7H7)] made possible investigation of an
analogous approach (Method 2) for the molybdenum system.
Reaction of [Mo(C CH)(dppe)(h-C7H7)] with C6D5Br under
Sonogashira coupling conditions ([PdCl2(PPh3)2], CuI, NHiPr2)
resulted in the formation of a deep brown solution from which
[Mo(C CC6D5)(dppe)(h-C7H7)], 2, was isolated as a brown
Under the conditions given in Table 1, each of the complexes
1, 2 and 3 undergoes a diffusion controlled, chemically reversible,
one-electron oxidation with the separation between cathodic and
anodic peak potentials comparable to that determined for the
internal ferrocene standard. The redox potentials demonstrate that
the substitution of C6H5 by C6D5 or C6H4-4-F has only a very small
effect on E 1 values. We have shown previously that E 1 is sensitive
2
2
to the identity of the arene para-substituent X in [Mo(C CC6H4-
4-X)(dppe)(h-C7H7)] with values ranging from -0.81 V (X =
NH2) to -0.65 V (X = CO2Me).16 However the very small change
Table 1 Cyclic voltammetric and IR data for [Mo(C CR)(dppe)(h-
C7H7)]n+
Infra Redb n(C C) cm-1
Compound
E1/2/Va
n = 0
n = 1
Dn(C C)
1c
2
3
-0.72
-0.72
-0.71
2045
2042
2048
2032
2022
2024
-13
-20
-24
a By cyclic voltammetry at a glassy carbon electrode. All potentials
+
+
are reported vs. FeCp2/FeCp2 (FeCp2/FeCp2 = 0.00 V) from 0.2 M
NBu4PF6/CH2Cl2 solutions at ambient temperature. b Solution spectra in
CH2Cl2. c Data from ref. 16
This journal is
The Royal Society of Chemistry 2010
Dalton Trans., 2010, 39, 11424–11431 | 11425
©

View Article Online
comparison, equivalent spectra for [1]+, recorded under identical
conditions, are illustrated in Fig. 2.
Table 2 EPR data for the 17-electron radicals [Mo(C CR)(dppe)(h-
C7H7)]+a
19
a
iso(Mo)
aiso (31P)
a
iso(1H) (C7H7)
a
( F)
giso
iso
[1]+
[2]+
[3]+
87.7 (31.3)
91.1 (32.5)
87 (31)
63.3 (22.6)
63.3 (22.6)
63.6 (22.7)
12.1 (4.3)
12.3 (4.4)
12.3 (4.4)
1.996
1.993
1.996
4.5 (1.6)
a X-band solution spectra in CH2Cl2 at 243 K; hyperfine couplings in MHz
and (in parentheses) Gauss. 1 G ª 2.802 MHz.
resulting from substitution of H by F is consistent with equivalent
data reported for [Fe(C CC6H4-4-X)(dppe)Cp*].6b
The chemical oxidation of brown solutions of 2 and 3 with
[FeCp2][PF6] resulted in the formation of intense blue solutions.
Monitoring by solution IR spectroscopy reveals a decrease in the
n(C C) stretching frequency by 15–25 cm-1 as summarised in
Table 1. The magnitude of Dn(C C) is typical of that observed
on generation of the 17-electron radicals [Mo(C CC6H4-4-
X)(dppe)(h-C7H7)]+ from their 18-electron precursors16 and these
data therefore provide good evidence for the formation of the
required radicals. Addition of CoCp2 to the solutions of oxidised
complexes resulted in reformation of the n(C C) bands of
the 18-electron systems confirming that, in common with other
17-electron radicals in the series [Mo(C CC6H4-4-X)(dppe)(h-
C7H7)]+, [2]+ and [3]+ are sufficiently stable for in situ generation
and spectroscopic observation.
Fig. 2 Fluid solution (CH2Cl2, 243 K), second derivative, X-band EPR
spectra of [Mo(C CC6H5)(dppe)(h-C7H7)][PF6], [1][PF6]; top: experi-
mental; bottom: simulated.
The EPR parameters and spectral patterns for complexes [1]+
and [3]+ are similar, however comparison of Fig. 1 and 2 reveals
that the fluoro-arene complex, [3]+, clearly exhibits an additional
small hyperfine coupling of 4.5 MHz, (1.6 G). This coupling
was observed for two independently prepared samples of 3/[3]+
recorded on separate occasions; EPR spectra of a sample of [1]+
recordedimmediatelyafter thespecimenof[3]+ andunderidentical
instrumental conditions did not show this additional splitting. We
therefore attribute the splitting of 4.5 MHz, (1.6 G) detected in the
EPR spectrum of [3]+ to direct observation of hyperfine coupling to
the remote p-position of the aryl alkynyl ligand. In support of this
assignment, we note that for the related complex [Fe(C CC6H4-
4-F)(dppe)Cp*]+, aiso(19F) was estimated as 4.0 MHz, (1.44 G)
on the basis of paramagnetic 19F NMR parameters.12 Further,
indirect evidence for the delocalisation of spin density onto the
aryl-alkynyl ligand in the radicals [1]+, [2]+ and [3]+ is provided by
the magnitudes of aiso (Mo), aiso (31P) and aiso (1H) in Table 2. These
are all diminished by comparison with corresponding parameters
for [MoCl(dppe)(h-C7H7)]+ [aiso (Mo), 112.1 MHz, (40.0 G); aiso
(31P), 65.0 MHz, (23.2 G); aiso (1H), 12.9 MHz, (4.6 G)]15 indicating
a reduction of spin density at the Mo(dppe)(h-C7H7) centre in the
alkynyl complexes.
EPR investigations
The 17-electron radicals, [1]+, [2]+ and [3]+ were generated by in situ
oxidation of 1, 2 and 3 respectively with [FeCp2][PF6] in solution in
CH2Cl2. EPR solution spectra were recorded at 243 K as first and
second derivatives and experimental solution isotropic g values
and hyperfine coupling parameters (confirmed by simulation) are
summarised in Table 2. The experimental and simulated, second
derivative EPR spectra of [3]+ are presented in Fig. 1 and, for
ENDOR investigations
The X-band solution EPR investigations provide good exper-
imental evidence for the delocalization of spin density to the
para-position of the arene ring of the aryl-alkynyl ligand in the
favourable case of [3]+ where F is located at this site. However
the calculated hyperfine couplings for aryl-alkynyl hydrogen sub-
stituents in [1]+ indicate16 that these may be close to the resolution
limits of the EPR technique. To address this constraint, we utilised
ENDOR (Electron Nuclear DOuble Resonance) to study [1]+
Fig. 1 Fluid solution (CH2Cl2, 243 K), second derivative, X-band
EPR spectra of [Mo(C CC6H4-4-F)(dppe)(h-C7H7)][PF6], [3][PF6]; top:
experimental; bottom: simulated.
11426 | Dalton Trans., 2010, 39, 11424–11431
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
together with the related complexes [Mo(C CC6D5)(dppe)(h-
C7H7)][BF4], [2]+, [Mo(C CBut)(dppe)(h-C7H7)][BF4], [4]+,16 and
[Mo(C CC6H4-4-Me)(dppe)(h-C7H7)][BF4], [5]+. The ENDOR
technique has substantially enhanced resolution over the EPR
method and therefore hyperfine couplings of the order of 2.8 MHz
(1 G) should be easily observable.24 However, the low g anisotropy
of complexes of the type [Mo(C CAr)(dppe)(h-C7H7)]+, key to
the excellent resolution of the fluid solution EPR spectra, presents
some limitations to the ENDOR method because orientation
selection is not available and therefore all three components of
the hyperfine matrix are present in the resulting frozen solution
(polycrystalline) spectra.25 As a consequence, the spectral pattern
consists of a series of overlapping signals which complicate
interpretation.
It should be emphasised that in the analysis of orientation selec-
tive ENDOR spectra, the orientations of the g matrix contributing
to the EPR resonance positions at the selected magnetic fields
(B0) in the ENDOR experiment must first be determined. The
ENDOR spectra are then recorded at fixed magnetic fields, so
that the response from a polycrystalline sample arises only from
the subset of molecules having orientations that contribute to the
EPR intensity at that particular B0 field. The EPR resonance
positions (Br) for given (q, f) orientations of the paramagnetic
species in frozen solution are then calculated as a function of A(q,
f) and g(q, f). The resonant fields thus obtained are convolved
with a lineshape function (Gaussian) to obtain their intensity at
the magnetic field (B0), used for the ENDOR experiment, and
the transition frequencies for each nucleus calculated using the
following ENDOR resonance expression for the solid state:26
1/ 2
2
Fig. 3 cw-EPR spectra of a) [1]+, b) [2]+, c) [4]+ and d) [5]+ recorded
at X-band (a,b,c,d) and Q-band (a¢,c¢,d¢) frequencies. X-band spectra
recorded at 10 K, Q-band data recorded at 50 K. The static magnetic
field position used for measurement of the X-band (9.476 GHz; 3413 G)
and Q-band (34.127 GHz; 12248 G) ENDOR spectra are indicated by the
symbol ᭹.
⎡
⎢
⎢
⎤
⎥
⎥
⎥
⎥
⎦
3
3
⎡
⎢
⎢
⎣
⎤
⎥
⎥
⎥
⎦
⎛
⎜
⎜
⎜
⎞
⎟
ms
⎟
⎟
n =
gjhj Aji −hnN
i
⎟
⎢∑ ∑
⎜
⎟
⎟
⎠
g
⎢
⎜
(q,f)
⎝
i=1
j=1
⎢
⎣
where hi are the direction cosines of Br in the molecular axis system,
Aji is the orientation dependent value of the hyperfine coupling, nN
is the nuclear Larmor frequency and ms the spin quantum number.
The ENDOR spectra can then be simulated at the different field
positions, providing information on the principal components of
the various hyperfine matrices as well as their orientation relative to
g. In the present case a similar approach, using the above equation,
was taken to simulate the ENDOR data. However, as stated above,
owing to the low g anisotropy of the system at both X- and Q-
band, all information on the relative orientations of g and ligand
A are lost. Nevertheless the analysis provides useful insights into
the magnitudes of the experimental couplings.
The frozen solution X- and Q-band cw-EPR spectra for [1]+,
[2]+, [4]+, and [5]+, are shown in Fig. 3 where the large 31P hyperfine
coupling is clearly visible. Hyperfine couplings to 95/97Mo are also
visible in the wings of the spectra, supporting the data from the
fluid solution. Very little additional information on the g matrix
is available from the Q-band spectra, which are now dominated
by the anisotropy in PA, whilst the 95/97Mo hyperfines are now
no longer resolved owing to the minor g strain occurring at Q-
band. Most importantly the field positions at X- and Q-band
used for ENDOR measurement are shown in Fig. 3. ENDOR
measurements were also performed at other field positions and the
profile of the resulting spectra was essentially the same (although
with weaker intensity) particularly at X-band. This can be seen for
example in the 31P ENDOR spectra recorded at various fields (see
supplementary data†).
In the complex [Mo(C CC6H5)(dppe)(h-C7H7)]+, [1]+, hyper-
fine couplings potentially accessible in the ENDOR spectrum
comprise those arising from o-, m- and p- sites of the aryl-
alkynyl ligand and couplings to the methylene protons of the C2H4
bridge of the dppe ligand. The magnitudes of these hyperfines
were expected to lie intermediate between the aiso values of 8–
14 MHz, (3–5 G) determined by EPR for the hydrogens of
the cycloheptatrienyl ring and couplings of less than 0.5 MHz
attributable to the remote aryl hydrogens of the dppe ligand.
To address the issues resulting from the absence of orientation
selection, and assist in the delineation of multiple components
of the spectrum, a series of complexes related to [1]+ was inves-
tigated consisting of [Mo(C CC6D5)(dppe)(h-C7H7)][BF4], [2]+,
[Mo(C CBut)(dppe)(h-C7H7)][BF4], [4]+, and [Mo(C CC6H4-
4-Me)(dppe)(h-C7H7)][BF4], [5]+. These were chosen to remove
selected couplings from the spectral pattern; for example, in
[5]+, the largest predicted hyperfine to the arene p-H should be
absent whereas in [2]+ only couplings attributable to the methylene
hydrogens of the dppe ligand will contribute to the spectral pattern
(with minor couplings from remote protons of the Ph substituents
This journal is
The Royal Society of Chemistry 2010
Dalton Trans., 2010, 39, 11424–11431 | 11427
©

View Article Online
Table 3 1H couplings for [1]+ from simulated cw-ENDOR spectraa
of dppe). The X-band 1H cw-ENDOR spectra of complexes [1]+,
[2]+, [4]+ and [5]+, are shown in Fig. 4(a–d) respectively; the
spectra are clearly different from each other and therefore the
hyperfine couplings intrinsic to the discrete alkynyl substituents
must contribute to the form of the spectra.
Centre
A1
A2
A3
aiso
a
iso(1H) calculated (DFT)b
CH2(1)
(dppe)
CH2(2)
(dppe)
3.8 0.9
3.4 1.9
0.6 1.7 ( 0.6) -0.5 (-0.19), -0.8 (-0.27),
+1.3 (+0.48), -1.9 (-0.67)
1.0 2.1 ( 0.7)
C6H5-p-H
C6H5-o-H
7.0 3.5
6.0 3.0
1.1 3.9 ( 1.4) -4.8 (-1.7)
2.0 3.7 ( 1.3) -4.2 (-1.5)
a Hyperfine values are given in MHz, values in parentheses in G, 1 G ª
2.802 MHz. Estimated errors in experimental Ai values are 0.4 MHz.
b See ref. 16 For the dppe ligand, calculated hyperfine values are given for
the four individual methylene hydrogens. Estimated errors in calculated
a
iso(1H) values are 1.5 MHz.
1
Fig. 4 X-band H cw-ENDOR spectra (10 K) of a) [1]+, b) [2]+, c) [4]+
and d) [5]+, as [BF4] salts with spectral wings expanded. The spectra were
measured at the static field position of 3413 G (see Fig. 3).
Further examination of the spectra allows a provisional identi-
fication of peaks associated with specific groups in the complexes.
For example, the intense peak labelled ** in Fig. 4(c) most probably
arises from the aliphatic protons of the C CBut fragment, since
it appears to be absent in the spectra of the aryl derivatives [1]+,
[2]+ and [5]+; similarly the feature labelled * in Fig. 4(b) ([2]+) is
common to all spectra and must therefore represent one of the
hyperfine components attributable to the methylene protons of
the dppe backbone. The very broad resonance (labelled ***) in
the spectral wings of Fig. 4(a) is assigned to the large o- and p-
H couplings (effectively the A1 component of the experimental
hyperfine) of the aryl C6H5 group in [1]+, since it is absent in the
deuterated analogue {[2]+, Fig. 4(b)} where there is no alkynyl
ligand component to the spectrum. The differences between the
spectral wings of Fig. 4(a, b) for [1]+ and [2]+ are not readily
clear at first [owing to the dominant contribution from the
cycloheptatrienyl ring protons (see Q-band ENDOR data in
Fig. 6), which distort the baseline]; nevertheless the differences
can be easily detected in other regions of the spectrum by spectral
subtraction yielding the A2 and A3 components of the hyperfine.
It should be recalled that the frozen solution ENDOR spectra of
organic p radicals frequently produce powder patterns where the
magnitude of the observed anisotropy at low temperature can be
ª50% of the magnitude of the isotropic hyperfine coupling.27 It is
not unreasonable therefore to expect an appreciable anisotropy in
the hyperfine couplings to the o- and p-H protons (Table 3). The
consequence of this is that one component (A1) of the hyperfine
is larger than the other two and therefore contributes to the peak
intensity marked ***.
Fig. 5 a) Experimental (10 K), and b) simulated 1H cw-ENDOR
spectra of [Mo(C CC6H5)(dppe)(h-C7H7)][BF4], [1][BF4]. The individual
components of the composite simulation are shown in c) CH2 (dppe), d)
C6H5-o-H and e) C6H5-p-H, combined in ratio of 1 : 1 : 1 contribution.
A minor isotropic background signal of 1.7 MHz was also added to the
simulation to improve overall fit.
1
Fig. 6 Q-band H cw-ENDOR spectra (10 K) of a) [4]+, b) [1]+, and c)
[5]+. The spectra were measured at the static field position of 12248 G
(see Fig. 3). The h-C7H7 ring protons are responsible for broad resonances
around 45 MHz and 58 MHz. A 19F matrix peak appears at 48 MHz due
to BF4.
Although some useful insights can be gained from this cursory
examination of the data, the complexity of the experimental
spectra, which in the region selected involve multiple hyperfine
couplings each with three matrix components, militates against
the extraction of precise hyperfine data. Q-band 1H cw-ENDOR
spectra were also recorded (see Fig. 6), but whilst the large
cycloheptatrienyl couplings were better resolved at the higher
11428 | Dalton Trans., 2010, 39, 11424–11431
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
frequencies, no improved resolution was obtained for the smaller
couplings to the aryl-alkynyl protons. However, when the exper-
imental data were used in conjunction with calculated values for
aiso it was possible to perform preliminary ENDOR simulations
which were then refined to give the best fit to the experimental
spectra.
The experimental and simulated spectra for [1]+ are shown
in Fig. 5(a) and 5(b) respectively; the combined simulation
[Fig. 5(b)] is constructed from the contributions of four individual
components, each with three discrete interaction matrices, as
presented in Fig. 5(c)–5(e) and Table 3. Simulations for the
methylene protons of the C2H4 bridge of the dppe ligand are
extensively informed by the spectrum of deuterated [2]+; two
independent sets of hyperfine values were extracted from the
spectra but these are presented in combined form in Fig. 5(c).
The simulations for the o- and p- aromatic protons of the aryl
substituent are given in Fig. 5(d) and 5(e) respectively; the
simulated centre with the larger coupling (labelled C6H5-p-H: 7.0,
should be examined. The magnitudes of the hyperfine couplings
determined for the two series of complexes [Fe(C CC6H4-4-
X)(dppe)Cp*]+ and [Mo(C CC6H4-4-X)(dppe)(h-C7H7)]+ are in
reasonable agreement and suggest that in these systems coupling
to the p-F substituent is not significantly larger than that to p-
H. This may be rationalised by the more complex mechanism of
spin transfer from the arene p-system to the orbitals of an F-atom29
such that relative magnitudes of hyperfine coupling to F and H
cannot be reliably predicted on the basis of isotropic interactions
for unit spin density.
Whilst peripheral to the current investigation, we note for
completeness that the very large couplings due to the protons of
the h-C7H7 ring, are easily visible in the Q-band proton ENDOR
spectra (Fig. 6). At X-band, the peaks were broad and difficult to
observe, however they were readily observed in the Mims ENDOR
spectra (see supplementary data). The measured experimental
couplings were between 8-14 MHz (in agreement with DFT).
Couplings to the 31P nuclei were also readily observed in
the X-band cw-ENDOR spectra (Fig. 7). The profile of these
polycrystalline 31P ENDOR spectra, with distinctive parallel and
perpendicular components, has been discussed elsewhere.15 Simu-
lated 31P couplings for [4]+ were determined as 70.55 and 61.7 MHz
(see supplementary data for simulations of [4]+) analogous to the
68.6 and 64.7 MHz aiso(31P) values predicted by DFT for [1]+.16
The slight difference between theory and experiment is easily
accounted for by the very large Fermi contact term for 31P (4747.9
G).30 It is also interesting to note that the 31P nuclei are very slightly
inequivalent in [4]+, as revealed by the two parallel components
in the hyperfine (labelled * in Fig. 7a). This inequivalency is less
pronounced in [5]+ and is absent in [1]+.
16
3.5, 1.1 MHz; aiso
= 3.9 MHz) was assigned to the p-proton,
based on the observation that its component seems to disappear in
[Mo(C CC6H4-4-Me)(dppe)(h-C7H7)][BF4], [5]+. The hyperfine
coupling of the m-protons of the aryl-alkynyl ligand could not be
reliably identified and therefore this contribution is not included
in the simulated spectra; likewise no attempt was made to simulate
the inner ‘matrix’ lines associated with very remote couplings (a
< 0.5 MHz). For this reason, the ENDOR simulations of [2]+
are not shown, since in this case the spectrum is dominated by
weak ligand couplings which cannot be reproduced accurately.
Furthermore, we were unable to resolve any structured 2H signals
in the cw-ENDOR spectra of [2]+; this is not uncommon for nuclei
possessing small nuclear Larmor frequencies in cw-ENDOR,
particularly in polycrystalline media.
It was not possible to extract the signs of the components of
the hyperfine matrices from the experimental double resonance
ENDOR spectra and therefore all experimental values given in
Table 3 are presented in the format (although some indication
of sign is provided by the DFT data). Whilst the g anisotropy is
very small at this frequency (X-band), a small degree of variation
(intensity) of the ENDOR powder pattern was observed due to
the anisotropy of the 1H hyperfine matrix itself.
The key outcome from the ENDOR studies is the determination
of the magnitude of aiso values for the o- and p- hydrogens of
the aryl-alknyl ligand of [1]+. The values reported in Table 3
which, allowing for experimental error, lie in the approximate
range 3.3–4.3 MHz (1.2–1.5 G) are slightly smaller than predicted
by DFT calculations but larger than the corresponding values
for [Fe(C CPh)(dppe)Cp*]+ (o-and p-H: 1.7 MHz, 0.6 G)
determined from paramagnetic NMR contact shifts. However,
given the small magnitude of these hyperfine couplings in relation
to estimated errors in both experimental and calculated28 values
(see Table 3), the agreement is relatively good. Finally, the
relative magnitudes of the hyperfine couplings to F and H,
located at the p-position of the arene ring of the alkynyl ligand
merit comment. For [Fe(C CC6H4-4-X)(dppe)Cp*]+, these are
reported as: (X = F, 4.0 MHz; X = H, 1.7 MHz) whilst in the current
work, the respective values for [Mo(C CC6H4-4-X)(dppe)(h-
C7H7)]+ are (X = F, 4.5 MHz; X = H, 3.9 MHz). These values
have been obtained by complementary techniques (paramagnetic
NMR, EPR and ENDOR) and therefore their self-consistency
Fig. 7 X-band 31P cw-ENDOR spectra (10 K) of a) [4]+, b) [1]+, and c)
[5]+. The spectra were measured at the static field position of 3413 G (see
Fig. 3).
Conclusions
Sonogashira coupling of [Mo(C CH)(dppe)(h-C7H7)] with func-
tionalised aryl-halides provides a convenient, small scale syn-
thesis of aryl-alkynyl complexes [Mo(C CAr)(dppe)(h-C7H7)]
including the derivatives (Ar = C6D5, 2; C6H4-4-F, 3). EPR
and ENDOR investigations on the 17-electron radical cations
[Mo(C CC6H5)(dppe)(h-C7H7)]+, [1]+ and [Mo(C CC6H4-4-
F)(dppe)(h-C7H7)]+, [3]+ confirm that unpaired spin density is
This journal is
The Royal Society of Chemistry 2010
Dalton Trans., 2010, 39, 11424–11431 | 11429
©

View Article Online
delocalised onto the remote arene ring of the aryl-alkynyl ligand.
In the case of [3]+, aiso(19F) was observed directly as an additional
hyperfine coupling of 4.5 MHz (1.6 G) in the X-band solution
EPR spectrum. Interpretation of the 1H cw-ENDOR powder
spectrum of [1]+ is complicated by the low g anisotropy of the
system which prevents the use of orientation selection. However,
analysis of the spectral features was assisted by comparison with
data for substituted derivatives such as [2]+ (in which all hyperfines
attributable to the o-, m- and p- protons of the aryl-alkynyl
ligand are absent) and by spectral simulation guided by calculated
hyperfine values derived from DFT investigations. The ENDOR
simulation shown in Fig. 5(b), represents a reasonably good fit to
the experimental powder ENDOR spectrum for [1]+ and indicates
that the form of the experimental spectrum is consistent with
calculated values aiso(1H) hyperfines of the aryl-alkynyl ligand
(Table 3). On the basis of the above analysis, the experimental
aiso(1H) hyperfine couplings for the aryl-alkynyl protons of [1]+
were estimated as: ortho, 3.7 MHz ( 1.3 G), para, 3.9 MHz
( 1.4 G). Well resolved hyperfine couplings to the protons of the
h-C7H7 ring and the phosphorus nuclei of the dppe ligand were
Bruker ER5106QT resonator employing 12 dB RF power from
a 3200L RF amplifier, 250 kHz RF modulation depth and 150mW
1
power. Cw H-ENDOR simulations were performed using an in
house program based on the resonance expressions of Hurst et al.26
for ENDOR resonances in disordered systems.
Preparations
Preparation of [Mo(C CC6D5)(dppe)(g-C7H7)], 2. A mix-
ture of [Mo(C CH)(dppe)(h-C7H7)] (373 mg, 0.61 mmol),
[PdCl2(PPh3)2] (67 mg, 0.10 mmol) and CuI (35 mg, 0.18 mmol)
was placed in a 25 cm3 flask and a solution of C6D5Br (740 mg,
4.57 mmol) in 10 cm3 of freshly distilled NHPri2 added. The
mixture was heated gently at reflux for 6 h. then solvent removed
to give a deep brown residue which was dissolved in CH2Cl2 (ca.
5 cm3) and loaded onto a hexane/alumina column. Elution with
CH2Cl2–hexane/acetone (45 : 45 : 10 v:v:v) gave a major brown
band which was collected and evaporated to dryness to give a
deep-brown solid which was reprecipitated from CH2Cl2–hexane;
yield: 68 mg (16%). 1H NMR: d 2.09 (m, 2H, CH2), 2.47 (m, 2H,
1
CH2), 4.80 (s, 7H, C7H7), 7.31 – 7.83 (m, 20H, Ph). 31P{ H} NMR
1
also detected in the respective Q-band H and X-band 31P cw-
(162 MHz): d 64.1. IR (CH2Cl2) n(C C) 2042 cm-1. MALDI-MS
(m/z): 693, [M]+; 587 [(M – C CC6D5)]+. HR ES+-MS (m/z):
693.1664 (MoC41H31D5P2 requires 693.1655).
ENDOR spectra.
Experimental
Preparation of [Mo(C CC6H4-4-F)(dppe)(g-C7H7)], 3. The
preparative procedure was identical to that described for complex
2, starting from [Mo(C CH)(dppe)(h-C7H7)] (300 mg, 0.49
mmol), [PdCl2(PPh3)2] (69 mg, 0.10 mmol), CuI (38 mg, 0.20
mmol) and 1-bromo-4-fluorobenzene (550 mg, 3.14 mmol) in
NHPri2 (10 cm3). The product was isolated as a brown solid, yield
30 mg (9%). 1H NMR: d 2.07 (m, 2H, CH2), 2.47 (m, 2H, CH2),
General Procedures
The preparation, purification and reactions of the complexes
described were carried out under dry nitrogen using stan-
dard Schlenk techniques. The complexes [Mo(C CR)(dppe)(h-
C7H7)] (R = Ph, But, C6H4-4-Me) were prepared by published
procedures.16 1H, 31P{ H} and 19F{ H} NMR spectra were
recorded from CD2Cl2 solutions containing trace amounts of
CoCp2 to prevent accumulation of MoI species by aerial oxidation,
on a Bruker AV-400 spectrometer at room temperature and refer-
enced against solvent resonances. Infrared spectra were obtained
on a Perkin Elmer FT RX1 spectrometer and MALDI mass
spectra were recorded using a Micromass/Waters TOF Spec 2E
instrument. Microanalyses were by the staff of the Microanalytical
Services of the School of Chemistry, University of Manchester.
Cyclic voltammograms were recorded (n = 100 mV s-1) from 0.2 M
NBu4PF6/CH2Cl2 solutions, ca. 1 ¥ 10-4 M in analyte using a
gas-tight single compartment three-electrode cell equipped with a
glassy carbon working electrode, Pt wire counter electrode and Pt
wire pseudo-reference electrode, and data collected on an Autolab
PG-STAT 30 potentiostat. All redox potentials are reported with
reference to an internal standard of the ferrocene/ferrocenium
1
1
4.80 (s, 7H, C7H7), 5.94 (m, 2H, C CC6H4-4-F), 6.53 (m, 2H,
1
C
CC6H4-4-F), 7.31 – 7.81 (m, 20H, Ph). 31P{ H} NMR (162
MHz): d 64.3. 19F{ H} NMR (376 MHz): d -120.0. IR (CH2Cl2)
1
n(C C) 2048 cm-1. MALDI-MS (m/z): 706, [M]+; 587 [(M –
C
CC6H4-4-F)]+. Anal. Calcd. (%) for C41H35FMoP2: C, 69.9%;
H, 5.0%. Found, C, 69.5%; H, 5.0%.
Preparation of [Mo(C CR)(dppe)(g-C7H7)]+. Samples of the
17-electron radicals [1]+, [2]+, [3]+, [4]+ and [5]+ were prepared
as deep blue ([1]+, [2]+, [3]+ and [5]+) or orange ([4]+) solutions
by treatment of the 18-electron precursors with [FeCp2][BF4] or
[FeCp2][PF6] in CH2Cl2.16 For ENDOR measurements, samples
were prepared from approximately 3 mg of isolated complex, as a
[BF4]- salt, dissolved in CD2Cl2/toluene-D-8.
+
couple (FeCp2/FeCp2 = 0.00 V). EPR experiments were con-
Acknowledgements
ducted on a Bruker BioSPin EMX microspectrometer at X-band
(9 GHz); spectra were recorded at 243 K and are the average
of 16 scans. Spectral analysis and simulation was carried out
We thank the EPSRC for support of this work through grants
EP/E025544/1 and EP/E030122.
1
using Bruker WinEPR software (Bruker Biospin Ltd.). Cw H-
ENDOR spectra were recorded at 10 K on an X-band Bruker
ESP 300E series spectrometer equipped with an ESP360 DICE
ENDOR unit, operating at 12.5 kHz field modulation in an X-
band ENDOR cavity (Bruker EN 801). The ENDOR spectra
were obtained using 8 dB RF power from an ENI A-300 RF
amplifier and 50 or 250 kHz RF modulation depth with 3.2 mW
microwave power. Q-band cw EPR/ENDOR measurements were
performed on a Bruker ESP 300e series spectrometer using a
References
1 P. J. Low and M. I. Bruce, Adv. Organomet. Chem., 2001, 48, 71.
2 M. I. Bruce and P. J. Low, Adv. Organomet. Chem., 2004, 50, 179.
3 (a) N. Gauthier, G. Argouarch, F. Paul, M. G. Humphrey, L. Toupet, S.
Ababou-Girad, H. Sabbah, P. Hapiot and B. Fabre, Adv. Mater., 2008,
20, 1952; (b) P. J. Low, Dalton Trans., 2005, 2821; (c) B. Kim, J. M.
Beebe, C. Olivier, S. Rigaut, D. Touchard, J. G. Kushmerick, X. Y. Zhu
and C. D. Frisbie, J. Phys. Chem. C, 2007, 111, 7521.
11430 | Dalton Trans., 2010, 39, 11424–11431
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
4 (a) C. Lapinte, J. Organomet. Chem., 2008, 693, 793; (b) F. Paul and C.
Lapinte in Unusual Structures and Physical Properties in Organometallic
Chemistry, M. Gielen, R. Willem and B. Wrackmeyer (Eds), John Wiley
and Sons Ltd, New York, 2002.
14 P. H. Rieger, Coord. Chem. Rev., 1994, 135–136, 203.
15 G. M. Aston, S. Badriya, R. D. Farley, R. W. Grime, S. J. Ledger, F. E.
Mabbs, E. J. L. McInnes, H. W. Morris, A. Ricalton, C. C. Rowlands,
K. Wagner and M. W. Whiteley, J. Chem. Soc., Dalton Trans., 1999,
4379.
5 (a) M. Samoc, G. T. Dalton, J. A. Gladysz, Q. Zheng, Y. Velkov, H.
˚
Agren, P. Norman and M. G. Humphrey, Inorg. Chem., 2008, 47, 9946;
16 N. J. Brown, D. Collison, R. Edge, E. C. Fitzgerald, M. Helliwell,
J. A. K. Howard, H. N. Lancashire, P. J. Low, J. J. W. McDouall, J.
Raftery, C. A. Smith, D. S. Yufit and M. W. Whiteley, Organometallics,
2010, 29, 1261.
17 J. R. Morton and K. F. Preston, J. Magn. Res., 1978, 30, 577.
18 E. Negishi, M. Kotora and C. Xu, J. Org. Chem., 1997, 62, 8957.
19 T. K. Dougherty and K. S. Y. Lau, J. Org. Chem., 1983, 48, 5273.
20 R. Denis, T. Weyland, F. Paul and C. Lapinte, J. Organomet. Chem.,
1997, 545–546, 615.
21 E. C. Fitzgerald, N. J. Brown, R. Edge, M. Helliwell, H.N. Lancashire,
D. Collison, P. J. Low, and M. W. Whiteley, in preparation.
22 J. S. Adams, C. Bitcon, J. R. Brown, D. Collison, M. Cunningham and
M. W. Whiteley, J. Chem. Soc., Dalton Trans., 1987, 3049.
23 R. L. Beddoes, C. Bitcon and M. W. Whiteley, J. Organomet. Chem.,
1991, 402, 85.
(b) M. Samoc, N. Gauthier, M. P. Cifuentes, F. Paul, C. Lapinte, M. G.
Humphrey and G. T. Dalton, Angew. Chem., Int. Ed., 2008, 47, 629;
(c) J. P. Morrall, G. T. Dalton, M. G. Humphrey and M. Samoc, Adv.
Organomet. Chem., 2007, 55, 61.
6 (a) N. G. Connelly, M. P. Gamasa, J. Gimeno, C. Lapinte, E. Lastra,
J. P. Maher, N. Le Narvor, A. L. Rieger and P. H. Rieger, J. Chem.
Soc., Dalton Trans., 1993, 2575; (b) R. Denis, L. Toupet, F. Paul and
C. Lapinte, Organometallics, 2000, 19, 4240; (c) K. Costuas, F. Paul,
L. Toupet, J.-F. Halet and C. Lapinte, Organometallics, 2004, 23, 2053;
(d) F. Paul, L. Toupet, J.-Y. The´pot, K. Costuas, J.-F. Halet and C.
Lapinte, Organometallics, 2005, 24, 5464.
7 (a) O. F. Koentjoro, R. Rousseau and P. J. Low, Organometallics, 2001,
20, 4502; (b) F. Paul, B. G. Ellis, M. I. Bruce, L. Toupet, T. Roisnel,
K. Costuas, J.-F. Halet and C. Lapinte, Organometallics, 2006, 25, 649;
(c) M. A. Fox, R. L. Roberts, W. M. Khairul, F. Hartl and P. J. Low,
J. Organomet. Chem., 2007, 692, 3277.
8 (a) N. Le Narvor, L. Toupet and C. Lapinte, J. Am. Chem. Soc., 1995,
117, 7129; (b) S. I. Ghazala, F. Paul, L. Toupet, T. Roisnel, P. Hapiot
and C. Lapinte, J. Am. Chem. Soc., 2006, 128, 2463.
9 M. I. Bruce, P. J. Low, K. Costuas, J.-F. Halet, S. P. Best and G. A.
Heath, J. Am. Chem. Soc., 2000, 122, 1949.
24 H. Kurreck, B. Kirste and W. Lubitz, Angew. Chem., Int. Ed. Engl.,
1984, 23, 173.
25 D. M. Murphy and R. D. Farley, Chem. Soc. Rev., 2006, 35,
249.
26 (a) G. C. Hurst, T. A. Henderson and R. W. Kreilick, J. Am. Chem.
Soc., 1985, 107, 7294; (b) T. A. Henderson, G. C. Hurst and R. W.
Kreilick, J. Am. Chem. Soc., 1985, 107, 7299.
10 (a) M. Brady, W. Weng, Y. Zhou, J. W. Seyler, A. J. Amoroso, A. M.
Arif, M. Bo¨hme, G. Frenking and J. A. Gladysz, J. Am. Chem. Soc.,
1997, 119, 775; (b) H. Jiao, K. Costuas, J. A. Gladysz, J.-F. Halet, M.
Guillemot, L. Toupet, F. Paul and C. Lapinte, J. Am. Chem. Soc., 2003,
125, 9511.
11 K. Venkatesan, O. Blacque and H. Berke, Dalton Trans., 2007, 1091.
12 F. Paul, G. da Costa, A. Bondon, N. Gauthier, S. Sinbandhit, L. Toupet,
K. Costuas, J.-F. Halet and C. Lapinte, Organometallics, 2007, 26, 874.
13 N. J. Brown, D. Collison, R. Edge, E. C. Fitzgerald, P. J. Low, M.
Helliwell, Y. T. Ta and M. W. Whiteley, Chem. Commun., 2010, 46,
2253.
27 (a) H. Kurreck, B. Kirste and W. Lubitz, in Electron Nuclear Double
Resonance Spectroscopy of Radicals in Solution, VCH Publishers, 1988,
page 40; (b) K. Mobius and W. Lubitz, Biological Magnetic Resonance,
L. J. Berliner and J. Reuben, ed., Plenum Press, New York, 1986, 7,
129-245.
28 I. Ciofini, F. Illas and C. Adamo, J. Chem. Phys., 2004, 120,
3811.
29 F. Gerson and W. Huber, in Electron Spin Resonance Spectroscopy of
Organic Radicals, John Wiley and Sons, 2003, page 71.
30 J. A. Weil, J. R. Bolton and J. E. Wertz, in EPR: Elementary Theory
and Practical Applications, John Wiley and Sons, 1994.
This journal is
The Royal Society of Chemistry 2010
Dalton Trans., 2010, 39, 11424–11431 | 11431
©