Redox properties of the 17-electron sandwich (η5-C5H5)Mo(η6-C 6Ph6): Electrochemical characterization of the 16 e- Lewis acid monocation and its PF6- adduct

Article abstract of DOI:10.1016/S0022-328X(98)01038-9

Thermolysis of CpMo(CH₂C₆H₅)(η²-PhC ₂Ph)[P(OMe)₃] [Cp=(η⁵-C₅H₅)] unexpectedly gives the mixed-sandwich 17-electron metallocene CpMo(η⁶-C₆Ph₆), 1. The title compound undergoes reversible one-electron oxidation (E_1/2=-0.25 V vs. ferrocene) and reduction (E_1/2=-2.02 V) reactions which were characterized by voltammetry and coulometry. When the oxidation of 1 is accomplished in CH₂Cl₂/[Bu₄N][PF₆] electrolyte, the 16 e^- Mo(II) complex 1⁺ is in equilibrium with a thermodynamically-favored product formulated as the [PF₆]^- adduct, 1-FPF₅, 3. Adduct 3 remains intact upon further one-electron oxidation to the Mo(III) complex 3⁺, but it rapidly releases [PF₆]^- when reduced to the Mo(I) species 3^-. Other counterions such as [BF₄]^- and [ClO₄]^- do not show the same degree of interaction with the 16 e^- Mo(II) cation.

Full text of DOI:10.1016/S0022-328X(98)01038-9

Journal of Organometallic Chemistry 577 (1999) 189–196
Redox properties of the 17-electron sandwich (h5-C5H5)Mo(h6-C6Ph6):
electrochemical characterization of the 16 e− Lewis acid monocation
and its PF6− adduct
Ian C. Quarmby b,1, Richard C. Hemond a,2, Frank J. Feher a,3, Michael Green a,4
William E. Geiger b,*
,
a Department of Chemistry, King’s College, Strand, London WC2R 2LS, UK
b Department of Chemistry, Uni6ersity of Vermont, Burlingtion VT 05405, USA
Received 30 September 1998
Abstract
Thermolysis of CpMo(CH2C6H5)(h2-PhC2Ph)[P(OMe)3] [Cp=(h5-C5H5)] unexpectedly gives the mixed-sandwich 17-electron
metallocene CpMo(h6-C6Ph6), 1. The title compound undergoes reversible one-electron oxidation (E1/2= −0.25 V vs. ferrocene)
and reduction (E1/2= −2.02 V) reactions which were characterized by voltammetry and coulometry. When the oxidation of 1 is
accomplished in CH2Cl2/[Bu4N][PF6] electrolyte, the 16 e− Mo(II) complex 1+ is in equilibrium with a thermodynamically-fa-
vored product formulated as the [PF6]− adduct, 1-FPF5, 3. Adduct 3 remains intact upon further one-electron oxidation to the
Mo(III) complex 3+, but it rapidly releases [PF6]− when reduced to the Mo(I) species 3−. Other counterions such as [BF4]− and
[ClO4]− do not show the same degree of interaction with the 16 e− Mo(II) cation. © 1999 Elsevier Science S.A. All rights
reserved.
Keywords: Electrochemistry; 17-Electron; Metallocene; Molybdenum
1. Introduction
complexes, especially isolable ones, are of interest from
a variety of viewpoints [4], we have investigated the
electrochemical oxidation and reduction of 1 in non-
aqueous solvents. We find that the 16-electron complex
1+ is a strong Lewis acid which forms an adduct with
the [PF6]− anion of the supporting electrolyte. Voltam-
metry measurements give information about the in-
tegrity of this adduct in three different Mo oxidation
states [(I), (II), (III)] and the enthalpy of its formation
in the Mo(II) complex. The existence of the 18-electron
anion 1− was also confirmed.
In the course of pursuing studies of the roles of
a-hydrogen elimination reactions [1] and alkyne bond-
ing modes [2] in the synthesis of molybdenum p-com-
plexes, the complex 2 has been found to react with
diphenylacetylene to form the 17-electron mixed sand-
wich metallocene, CpMo(h6-C6Ph6), 1, where Cp=(h5-
C5H5), a complex that had been prepared earlier by
different routes [3]. Since paramagnetic organometallic
* Corresponding author.
1 Present address: Lord Corporation, Raleigh, NC.
2 Present address: Shipley Company, Inc, Marlboro, MA, USA.
3 Present address: Department of Chemistry, University of Califor-
nia at Irvine, Irvine, CA, USA.
4 Present address: School of Chemistry, University of Bath, Claver-
ton Down, Bath, UK.
0022-328X/99/$ - see front matter © 1999 Elsevier Science S.A. All rights reserved.
PII: S0022-328X(98)01038-9

190
I.C. Quarmby et al. / Journal of Organometallic Chemistry 577 (1999) 189–196
Table 1
Potentials (vs. Fc) of CpMo(h6-C6Ph6), 1, and its [PF6]− adduct, 3
Mo(II) complex
Electrolytea, T (K)
E1/2 (V)
Mo(III)/(II)
Mo(II)/(I)
Mo(I)/(0)
CpMo(C6Ph6), 1
CpMo(C6Ph6), 1
CpMo(C6Ph6), 1
[CpMo(C6Ph6)–PF6], 3
CH2Cl2/[Bu4N][PF6], 298
THF/[Bu4N][PF6], 298
CH2Cl2/[Bu4N][ClO4], 293
CH2Cl2/[Bu4N][PF6], 273
+0.92b
−0.25
−0.24
−0.20
−1.5b
−1.94c
−2.02
−2.0a
d
e
+0.35
a Electrolyte concentration, 0.1 M.
b Irreversible. Peak potential with w=0.20 V s−1 reported.
c T=233 K.
d Wave not observed owing to limited positive potential window of this solvent.
e Process not assigned. Irreversible anodic peaks at +0.78 and +1.00 V were observed.
2. Results and discussion
2.2. Electrochemistry
Complex 1 displays chemically reversible one-elec-
tron oxidation and reduction processes which establish
the electron-transfer series of Eq. (1), involving the
16-electron 1+, 17-electron 1 and 18-electron 1−:
2.1. Characterization of 1
Selective attack of the Grignard reagents
CH2ꢀCHMgBr and CH3MgI on the metal center of the
cation {CpMo(h2-PhC2Ph)–[P(OMe)3]2}[BF4] has been
demonstrated to afford reactive molecules which un-
dergo novel ring-forming reactions with dipheny-
lacetylene to give the h3-cyclopropenyl and
−
−
1+ X 1 eX 1−
(1)
e
The monoanion 1− was more persistent in THF than
in CH2Cl2. An essentially Nernstian [8] 1 e− wave was
observed in the former solvent at ambient temperatures
and slow cyclic voltammetry (CV) scan rates, E1/2= −
2.02 V versus Fc (Table 1). In CH2Cl2, however, the
reduction wave (designated as wave D in our discussion
below) was reversible only at reduced temperatures
(Fig. 1). Under ambient conditions, the cathodic wave
lacked chemical reversibility and was several times
h4-cyclopentadiene
complexes
CpMo(h3-
C3MePh2){P(OMe)3}2 and CpMoH{h4-C(Ph)C(H)–
C(Ph)C(Ph)C(Ph)(H)}P(OMe)3 ([1]b). Under similar
conditions, the addition of two equivalents of
C6H5CH2MgCl to a purple solution of {CpMo(h2-
PhC2Ph)[P(OMe)3]2}[BF4] [5] in THF at 273 K gave an
emerald-green
colored
mixture
from
which
CpMo(CH2C6H5)(h2-PhC2Ph){P(OMe)3}, 2, was iso-
lated as a deep green air-sensitive solid after low tem-
perature column chromatography on alumina.
Thermolysis of 2 in benzene at 373 K with an excess
of diphenylacetylene afforded a deep green/yellow solu-
tion, from which 1 was isolated as a bright yellow,
extremely air-sensitive, solid. Spectroscopic analysis of
the supernatant liquid from the preparation revealed
the presence of trimethylphosphite, diphenylacetylene,
and dibenzyl. The last compound presumably arises
through coupling of benzyl radicals formed during
the thermolysis. The room temperature 1H- and
13C{1H}-NMR spectra of 1 displayed resonances that
were assigned to a Cp ligand and to phenyl rings.
Each signal was broad, suggesting that 1 was a
paramagnetic complex. This was confirmed by
employing a variation of the Evans NMR method [6]
which revealed a magnetic moment indicative of one
unpaired electron (veff=1.58 BM). This complex was
finally confirmed as having the sandwich structure 1 by
Hursthouse and Mazid using X-ray diffraction methods
[7].
Fig. 1. Voltammetry of 0.5 mM 1 in CH2Cl2/0.1 M [Bu4N][PF]6 at
228 K at Pt electrodes. Top, rotating Pt electrode scan, w=2 mV s−1
;
bottom, CV scan, w=0.20 V s−1, at twice the current sensitivity.
Waves B and D refer to the labels from text, processes 1+/12+ and
1
+/1, respectively.

I.C. Quarmby et al. / Journal of Organometallic Chemistry 577 (1999) 189–196
191
1+/12+, even though the dication decomposes on the
CV time scale (Eq. (3)):
1
+ –e− X 12+ “decomp
(3)
2.4. Oxidation of 1 in CH2Cl2/[Bu4N][PF6]: bulk
electrolysis and e6idence for adduct formation
Exhaustive electrolysis of 1 in CH2Cl2/[Bu4N][PF6]
with Eappl=0 V (positive of wave B) resulted in a color
change from yellow to red with release of ca. 0.9 F/eq.
Judging from CV and rotating Pt electrode (RPE)
scans, only two products were formed. One was the
expected 16 e− monocation, 1+, which was slightly
dominant at 293 K. Referring to Fig. 2, waves B
(E1/2= −0.25 V) and D (E1/2=1.9 V) after electrolysis
are those of 1/1+and 1/1− (Eq. (1)), respectively. The
height of the cathodic wave D is much larger than that
of B, an anomaly explained below. The two additional
waves (anodic wave A (Fig. 2, inset) and cathodic wave
C) arise from a second (final) product. We reiterate that
this second product is formed only on the longer time
scale (ca. 20 min) of bulk electrolysis, rather than in CV
scans of 1, the slowest of which requires ca. 20 s.
A CV scan of the electrolyzed solution at 260 K (Fig.
3) shows that wave A is chemically reversible (E1/2= +
0.35 V) and that wave C (Epc= −1.5 V) is irreversible.
Also appearing in Fig. 3 is wave E (the second oxida-
tion wave of 1 (Eq. (3)) because the scan extends to a
sufficiently positive potential.
The combined RPE plateau currents of A and B were
exactly equal to the height of the anodic current for 1
prior to electrolysis, suggesting quantitative conversion
of the starting material to the mixture of 1+ and the
second product. As is justified below, the second
product (3) is assigned a structure in which [PF6]− is
coordinated to 1+, most likely through donation of a
lone pair from the fluorine atom to the electron-defi-
cient metal, as in 3a. A strong ion pair (3b) cannot be
ruled out.
Fig. 2. Rotating Pt electrode scans of anodically electrolyzed (Eappl
=
0 V) solution of 1 in CH2Cl2/[Bu4N][PF]6 (same solution as Fig. 1), in
both negative (upper) and positive (lower) scan directions, T=233 K.
Waves A through D refer to labels in text.
larger in height than a one-electron value. It is probable
that 1− decomposes in CH2Cl2 by slow reduction of
the solvent [9], and the reductive process was not
studied in detail.
Reduction of 1 to 1− allows the sandwich complex
to attain the 18-electron configuration. The oxidation
of 1 to 1+, involving formation of a 16-electron com-
plex, is more novel and is therefore emphasized below.
2.3. Oxidation of 1 in CH2Cl2/[Bu4N][PF6]: cyclic
6oltammetry
Two oxidation waves are observed in CV scans in
CH2Cl2/[Bu4N][PF6]. The first (E1/2= −0.25 V) is a
chemically reversible and diffusion-controlled [10] 1 e−
reaction (Eq. (2)) involving formation of the 16 e−
monocation, 1+:
1 X 1+ +e−
(2)
The fact that a unity value of ic/ia persisted for this
wave down to a low sweep rate, w, of 0.05 V s−1
established a minimum half-life for 1+ of 20 s. In
subsequent discussions and Figs. 1–3, this wave at
E1/2= −0.25 V is wave B.
A second anodic wave (later, wave E) at Ep= +0.92
V was diffusion controlled but chemically irreversible,
having twice the height of 1/1+over a range of w and T.
The ultimate electrode products of wave E are un-
known. Noting, however, that the oxidation of 1+ is
involved, we may use the potential of +0.9 V as an
approximation of the formal potential of the couple
Fig. 3. CV scan w=0.20 V s−1, of same solution as Fig. 2, T=260
K. scan The potential scan window is different from that of Fig. 2,
thus showing wave E, Eq. (3).

192
I.C. Quarmby et al. / Journal of Organometallic Chemistry 577 (1999) 189–196
Table 2
Ratios of [3]/[1+] as measured by voltammetry at various tempera-
tures in CH2Cl2/0.1 M [Bu4N][PF6]
a
[3/[1+
]
Keq
Temperature (K)
0.72
0.85
1.15
1.35
1.74
2.14
7.2
8.5
11.5
13.5
17.4
21.4
293
282
272
260
248
233
a Keq=[3]/[l+][PF−6 ].
Fig. 5. CV scan, w=0.20 V s−1, of electrolyzed 1 in CH2Cl2/
Bu4NPF6, same solution as Fig. 2, T=253 K. Arrows indicate
direction of potential sweep.
2.5. Proof of chemical re6ersibility of 3 and 1
When the anodically prepared solutions containing a
mixture of 3 and 1 (e.g. Fig. 2) are exhaustively reduced
at potentials slightly negative of wave C (ca. −1.6 V),
waves B and D of the regenerated starting material 1
are formed in amounts equal to those measured prior
to the original oxidation. Therefore both 1+ and 3 are
reduced by one electron to give 1. The fact that the
conversion of 3 to 1 (Eq. (5)) is not only quantitative
but extremely rapid is shown by the following analysis
of CV data on solutions containing 1+ and 3.
The electrolysis products 1+ and 3 are in equilibrium
in CH2Cl2/0.1 M [Bu4N][PF6], as shown by the effect of
temperature variations on the relative heights of waves
A and B in electrolyzed solutions. Table 2 gives the
measured ratios of [3]/[1+] and the corresponding equi-
librium constants (defined as in Eq. (4)), and Fig. 4
gives the plot of Keq as a function of temperature. The
adduct 3 is thermodynamically favored over 1+, with
3+e− “1+[PF6]−
(5)
Consider Fig. 5, a CV scan at 253 K, w=0.20 V s−1
,
DH= −2.9 kcal mol−1
+ +[PF6]− X 3
Keq=[3]/[1+][PF6−]
.
of an anodically prepared solution of 1+ and 3. The
initial positive-going scan through wave A (E1/2= +
0.35 V) involves the Mo(II)/Mo(III) oxidation of 3 to
3+ (Eq. (6)). Wave B is that arising from reduction of
1+ to 1 (reverse of Eq. (2), Mo(II)/Mo(I)).
1
(4)
3 X 3+ +e−
(6)
Wave C is the irreversible cathodic process of Mo(II)–
PF6 going to Mo(I), Eq. (5). Finally, wave D is the
reduction of Mo(I) to Mo(0), 1 to 1−. In this scenario,
the height of wave D reflects the amount of 1 produced
at the electrode from both waves B and C. Once the
irreversibility of wave C is taken into account, the
heights of the three waves are completely consistent
with this model. Irreversible wave C has a lower peak
height when compared to a reversible system; Eq. (7) is
required if ip(C) is to be directly compared with the
reversible peaks ip(B) and ip(D) in Fig. 5 [11].
ip(irrev)=1.1 [ip(rev)] [h1/2
]
(7)
This relationship predicts a peak height of approxi-
mately 0.77×ip(rev) for an irreversible system with a
charge transfer coefficient, h, of 0.5 (this value of h is
justified by analysis of the shape of wave C) [12]. The
Fig. 4. Plot of log Keq (=[3] [{PF6}−]/[1+]) as function of 1/T for
solution of Fig. 2. The lowest temperature sample may not have
achieved equilibration. Linear regression including all points gives
DH°= −2.51 kcal mol−1; that excluding the lowest temperature
˚
predicted height of wave D (see Eq. (8)) is 1.8 mA, close
point gives DH° −2.86 kcal mol−1
.
to the value of 2.0 mA measured experimentally.
˚

I.C. Quarmby et al. / Journal of Organometallic Chemistry 577 (1999) 189–196
193
ip(B)+ip(C)/0.77=ip(D)
(8)
must come from an outside source, either the solvent or
the supporting electrolyte anion.
This analysis and the aforementioned analysis of RPE
plateau currents demonstrate that the 17 e− complex 1
is generated immediately and quantitatively in the one-
electron reduction of 3 (Eq. (5)). Taken together, these
data support Schemes 1 and 2 for interconversion of
the various redox forms with and without [PF6]−
attachment:
When CH2Cl2 is used with a perchlorate electrolyte
anion (as either [Et4N][ClO4] or [Bu4N][ClO4]), the
wave for 1/1+ is observed in CV experiments, but
exhaustive anodic electrolysis of 1 gave neither a wave
for 3 nor any other recognizable product wave. Instead,
a series of small cathodic waves most likely arising from
decomposition products remained after electrolysis.
These control experiments show that the presence of
[PF6]− is necessary for the formation of 3.
Electrochemical results in THF/[Bu4N][PF6], how-
ever, mimicked those obtained in CH2Cl2/[Bu4N][PF6].
The couples 1/1+ (wave B) and 1/1− (wave D) were
apparent after anodic electrolysis. There is also elec-
troactivity in the range of −1.5 V (wave C), and the
height of wave D is again roughly that expected from
the combined products of waves B and C. There is also
a broad anodic wave near +0.4 V (Figure in Supple-
mentary Material) and an irreversible cathodic wave at
−0.95 V. The broad wave is also observed in CV scans
of the fresh solution of 1 prior to electrolysis, and
appears to arise from two or more overlapping anodic
processes. These results suggest that 3 forms in THF/
[PF6]− media as well as in CH2Cl2/[PF6]− media, albeit
not as cleanly in the former. The other waves in the
THF electrolysis solutions arise from decomposition
products or from a structure in which THF is metal-
coordinated.
Scheme 1.
2.6. Deduction of the identity of 3
The above electrochemical experiments do not in
themselves establish that the second product involves
association of 1+ with a [PF6−] counterion. Rather, this
deduction is made from observations of the changes
engendered by alterations in the solvent and supporting
electrolyte. The assumption that underlies our conclu-
sion is that 3 must be related to 1+ in one of three
ways: as an isomer, as a solvent adduct, or as a
counterion adduct (or ion pair).
The first of these is eliminated in part on the basis
that two isomers would be unlikely to differ as greatly
in redox potentials as is observed in the Mo(II)/Mo(III)
couples of the systems under study (0.35 V for 3, ca. 0.9
V for 1+) [13]. The stabilization of the Mo(III) oxida-
tion state for 3 (ca. 12 kcal mol−1) implies that the
metal center is significantly more electron rich in 3
compared to 1+. Since the two p ligands in 1+ have no
additional ligation sites, the added electron density in 3
3. Discussion
Reversible redox reactions involving 16 e−/17 e−/18
e− couples have been described for a number of metal
sandwich compounds [14], including those to the left of
iron in the periodic table. The latter include [Cp2*Mn]+/
0/− [15] (Cp*=h5-C5Me5), [(C6Me6)2Cr]2+/+/0 [16],
+/0/−
[(C6Me6)CrCp*]
, , [16]
[16] [(C6H6)2V]+/0/−
[(C7H7)VCp]+/0/−
,
[16,18] and [(C7H7)NbCp]+/0/−
[19]. Also relevant are the 15 e−/16 e−/17 e− couples
of [Cp2Cr]+/0/− and its analogues [20]. There is clearly
ample precedent for reversible redox processes involv-
ing both oxidation and reduction of 17 e− metal sand-
wich compounds of the early transition metals. The
difference in E1/2 values between the 1+/1 and 1/1−
couples (1.78 V in THF) is very similar to those re-
ported for [(C6Me6)CrCp]*+/0/− (1.8 V in THF) [17]
and [Cp*2 Mn]+/0/− (1.94 V in CH3CN) [15]. In this
context, the electron-transfer series 1+/1/1− is
unremarkable.
It is the equilibration of the 16 e− complex 1+ with
3 that makes this system of more than routine interest.
Metal complexes which incorporate the anion [PF6]−
into their coordination spheres are known [21]. For that
matter, studies have been reported in which the coordi-
Scheme 2.

194
I.C. Quarmby et al. / Journal of Organometallic Chemistry 577 (1999) 189–196
nation of halocarbons to metals was demonstrated [22].
The new complex 3 was most prominent in CH2Cl2/
[Bu4N][PF6] media, in which it is thermodynamically
favored over 1+. The facts that 1+ undergoes irre-
versible decomposition in CH2Cl2 when [ClO4]− is
substituted for [PF6]−, and that evidence was found for
3 in THF/[PF6]− media is suggestive that the Lewis
base in the adduct 3 is the hexafluorophosphate anion,
rather than CH2Cl2.
A troublesome aspect of this conclusion relates to the
fact that the perchlorate anion is normally considered
to have a coordination tendency superior to that of
hexafluorophosphate [21,23]. In this light, one must ask
why electrolysis in CH2Cl2/[R4N][ClO4] did not yield a
stable adduct. A possible answer is that the ‘normal’
coordination tendency might be reversed for molybde-
num. The majority of isolated perchlorate complexes
have contained the later transition metals [24], although
this circumstance may not arise from inherent stability
factors. It is also pertinent to note that the anion
also a reasonable candidate. Thermodynamic studies
have been used to characterize equilibria between cova-
lently-bonded and tight ion-paired species such as 3a
and 3b [26]. Thermodynamic data on the present system
do not, however, have the accuracy necessary for analy-
sis of the entropic trends that are diagnostic of the
various association modes [27].
In a broader sense, the present results are important
in demonstrating another class of electron-deficient
sandwich complexes of the early transition metals may
interact reversibly with weak anionic ligands. Interac-
tions of this nature are known to influence the activities
of metallocene catalysts in olefin polymerizations [28],
and could, in principle, form a basis for selective ana-
lytical detection of the anions.
4. Experimental
Standard Schlenck procedures were employed to as-
sure dry- and oxygen-free preparative procedures. Elec-
trochemical experiments were conducted with samples
within a controlled-atmosphere enclosure using meth-
ods previously detailed [29]. The complex {CpMo(h2-
PhC2Ph)[P(OMe)3]2}[BF4] was prepared as described in
Ref. [5].
[PF6]−
stabilizes
the
Cr(I)
radical
cation
[(arene)Cr(CO)3]+ in CH2Cl2, whereas others such as
[ClO4]− and [BF4]− fail to do so. Thus, only when
[PF6]− is present is the one-electron oxidation of
(C6H6)Cr(CO)3 chemically reversible by cyclic voltam-
metry [25].
The electronic effect of coordination of [PF6]− to
Mo(II) can be quantified by comparison of the redox
potentials of 3 and 1 in the various oxidation state
changes. To do this we must first assume that the peak
potentials of the irreversible processes of Eqs. (3) and
(5) (waves E and C, respectively) are reasonable ap-
proximations of the formal potentials of the couples
1+/12+ and 3/3−, respectively (Table 1). With these
values in hand we can then compare the potentials of
the Mo(II)/Mo(III) and Mo(II)/Mo(I) couples with and
without [PF6]− coordination.
Comparison of the potentials for waves A (E1/2= +
0.35 V) and E (Ep= +0.9 V) shows that [PF6]−-coor-
dination strongly stabilizes the Mo(III) oxidation state
relative to Mo(II). Similarly, the negative shift of wave
C compared to B argues for stabilization of Mo(II)
relative to Mo(I) by [PF6]−-coordination. Thus, in the
Mo(I)/Mo(II)/Mo(III) sequence, addition of a single
weakly basic ligand continuously stabilizes the next
higher oxidation state. Since the difference in E1/2 po-
tentials between the Mo(III)/Mo(II) and Mo(II)/Mo(I)
couples is considerably greater (ca. 1.9 V)in the [PF6]−-
bound complex than in the unbound complex (ca. 1.2
V), it can also be stated that Lewis-base coordination of
PF6− stabilizes the Mo(II) oxidation state with respect
to disproportionation into Mo(III) and Mo(I).
4.1. Preparation of complexes
CpMo(CH2C6H5)(p2-PhC2Ph)[P(OMe)3](2)
To
a
purple solution of {CpMo(h2-PhC2Ph)-
[P(OMe)3]2}[BF4] (0.49 g, 0.73 mmol) in THF (10 ml) at
273 K was added C6H5CH2MgCl (2 M in THF, two
equivalents). The solution was stirred at this tempera-
ture for 10 min and turned emerald green. The solvent
was removed at 273 K and the residue subjected to
column chromatography on alumina at 243 K. Elution
with Et2O/hexane gave a green band from which
CpMo(CH2C6H5)(h2-PhC2Ph)[P(OMe)3] was obtained
1
as a deep green, air-sensitive, solid (0.34 g, 84%). H-
NMR (C6D6, 293 K) l 7.7–6.8 (m, 15H, Ph), 4.84 (s,
3
5H, C5H5), 3.15 [d, 9H, JHP=10.7, P(OCH3)3], 2.80
(qt, 1H, 3JHP=6.6, 2JHHgem=11.4, CHHC6H5), 2.41
(qt, 1H, 3JHP=6.6, 2JHHgem=11.4, CHHC6H5;
13C{1H}-NMR (C6D6, 293 K) l 200.9 (d, JCP=16.3,
2
ꢁCPh), 200.7 (d, 2JCP=47.5, ꢁCPh), 139–121 (Ph), 92.3
2
2
(C5H5), 51.6 [d, JCP, P(OCH3)3], 21.7 (d, JCP=13.5,
CH2C6H5); 31P{1H}-NMR (C6D6, 293 K) l 197.0 (s).
4.2. (p5-C5H5)Mo(p6-C6Ph6) (1)
To
a
green solution of CpMo(CH2C6H5)(h2-
Finally we note that although we favor the Lewis
acid/base pair 3a as the structure of the adduct, we
cannot rule out other means of association between 1+
and [PF6]−. A non-solvent-separated ion pair, 3b, is
PhC2Ph)[P(OMe)3] (0.18 g, 0.32 mmol) in benzene (10
ml) was added PhCꢁCPh (0.43 g, 2.4 mmol). The closed
system (Young’s tube) was heated at ca. 375 K for 17 h,
whereupon the color changed to green–yellow. After

I.C. Quarmby et al. / Journal of Organometallic Chemistry 577 (1999) 189–196
195
[4] (a) W.C. Trogler, (Ed.), Organometallic Radical Processes, El-
sevier, Amsterdam, 1990. (b) D. Astruc, Electron-Transfer and
Radical Processes in Transition Metal Chemistry, VCH Publish-
ers, New York, 1995.
[5] M. Bottrill, M. Green, J. Chem. Soc. Dalton Trans. (1977) 2365.
[6] D.F. Evans, J. Chem. Soc. (1959) 2003.
reduction of the solvent to 2 ml, pentane (10 ml) was
added and the reaction mixture was stirred 5 min, then
filtered. The yellow precipitate was washed twice with
pentane (10 ml) and crystallized from THF at 243 K to
afford 1 as a bright yellow, extremely air-sensitive, solid
1
[7] The crystal structure of 1 was obtained at 173 K by M.A. Mazid
and M.B. Hursthouse at the University of Wales, Cardiff. The
(0.11 g, 50%). H-NMR (CD2Cl2, 293 K) l 11.40 (br),
10.35 (br), 8.84 (br), 7.84 (br); 13C{1H}-NMR (CD2Cl2,
293 K) l 134.8 (br), 132.4 (br), 130.9 (br), 117.6 (br);
Evans NMR (CHCl3), veff=1.58 BM.
˚
arene ring is planar to within 0.01 A.
[8] In the context of this study, Nernstian refers to a description in
which the CV peak separations of a wave are essentially equal to
those of ferrocene under the same experimental conditions,
including T, concentration, and scan rate, w. Typically, peak
separations of 70–90 mV were observed for ferrocene0/+ with
4.3. Electrochemistry
w=0.05–0.50 V s−1
.
CH2Cl2 and THF were reagent-grade solvents twice
distilled from CaH2, the second time under vacuum in a
bulb-to-bulb transfer. The supporting electrolytes,
[Bu4][NPF6], [Et4N][ClO4], and [Bu4N][ClO4], were vac-
uum dried at 350 K before use. An Ag/AgCl wire was
used as the experimental reference electrode. Ferrocene
was added to the solution near completion of each
experiment, and all potentials in this paper employ
Fc/Fc+ as the reference couple. Conversion to the
aqueous SCE reference requires addition of +0.46 V
for CH2Cl2 or +0.56 V for THF.
[9] The anion 1− likely reduces CH2Cl2 (to an unspecified product),
and in the process regenerates neutral 1, which is reduced again
by the electrode. This mechanism increases the cathodic current
of wave D over the 1 e− height and decreases or eliminates the
anodic component of the wave. It is an example of homogeneous
redox catalysis (see C.P. Andrieux, P. Hapiot, J.-M. Saveant,
Chem. Rev. 90 (1990) 723).
[10] The diagnostic criterion for diffusion control was the linear
dependence of ip versus w1/2 over the range of scan rates.
[11] R.N. Adams, Electrochemistry at Solid Electrodes, Marcel
Dekker, New York, 1969, p. 136.
[12] On the basis of the relationship Ep−Ep/2=48 mV/hn, a width
of 96 mV is predicted for wave C with n=0.5, close to the
measured value of 98 mV. Further verification of the value is
obtained from the relative height of C to the anodic branch of A,
both of which arise from the same concentration of 3 at the
electrode surface, assuming interconversion of 1+ and 3 which is
slow on the CV time scale. It is predicted that ip(C)/ip(A)=0.77,
and the measured value is 0.68.
[13] This comparison makes use of the peak potential of the irre-
versible oxidation of 1+ (Epa=0.92 V, scan rate 200 mV s−1) as
an approximation of the E1/2 of 1+/12+. There is also a less
obvious assumption involved for the reversible oxidation of 3
(E1/2=0.35 V). The Nernstian wave demonstrates that if there is
ligand ([PF6]−) loss and gain during the redox reaction, both
processes are fast on the CV time scale. Our analysis assumes
that any such process would not significantly change the E1/2
value. For more on the implications of a Nernstian wave see
W.E. Geiger, in: S.J. Lippard (Ed.), Progress in Inorganic Chem-
istry, vol. 33, Wiley, New York, 1985, pp. 277–281.
5. Supplementary material available
One voltammogram showing electrochemical behav-
ior of solution of 1 after bulk oxidation in THF/0.1 M
[Bu4N][PF6].
Acknowledgements
W.E. Geiger wishes to thank the National Science
Foundation for generous financial support. We are
grateful to M.B. Hursthouse for informing us of the
X-ray results on the title compound.
[14] (a) W.E. Geiger, in Ref. 4, p. 142 (b) V.V. Strelets, Coord.
Chem. Rev. 114 (1992) 1.
[15] J.C. Smart, J.L. Robbins, J. Am. Chem. Soc. 100 (1978) 3936.
[16] Ch. Elschenbroich, E. Bilger, B. Metz, Organometallics 10 (1991)
2823, and references therein.
References
[17] D.S. Richeson, J.F. Mitchell, K.H. Theopold, Organometallics 8
(1989) 2570.
[18] (a) W.M. Gulick, Jr., D.H. Geske, Inorg. Chem. 6 (1967) 1320.
(b) U. Koelle, W. Holzinger, J. Mueller, J. Naturforsch. 34b
(1979) 759.
[19] M.L.H. Green, P. Mountford, V.S.B. Mtetwa, P. Scott, S.J.
Simpson, J. Chem. Soc. Chem. Commun. (1992) 314.
[20] (a) J.D.L. Holloway, W.L. Bowden, W.E. Geiger, J. Am. Chem.
Soc. 99 (1979) 7089. (b) U. Koelle, J. Organomet. Chem. 157
(1978) 327. (c) J.L. Robbins, N. Edelstein, B. Spencer, J.C.
Smart, J. Am. Chem. Soc. 104 (1982) 1882. (d) M.P. Castellani,
S.J. Geib, A.L. Rheingold, W.C. Trogler, Organometallics 6
(1987) 1703. (e) J.D.L. Holloway, F.C. Senftleber, W.E. Geiger,
Anal. Chem. 50 (1978) 1010. (f) E.U. van Raaij, S. Monkeberg,
H. Kiesele, H. Brintzinger, J. Organomet. Chem. 356 (1988) 307.
[21] W. Beck, K. Suenkel, Chem. Rev. 88 (1988) 1405.
[1] (a) S.R. Allen, P.K. Baker, S.G. Barnes, M. Green, L. Trollope,
L. Manajlovic-Muir, K.W. Muir, J. Chem. Soc. Dalton Trans.
(1981) 873. (b) F.J. Feher, M. Green, A.G. Orpen, J. Chem. Soc.
Chem. Commun. (1986) 291.
[2] (a) S.R. Allen, R.G. Beevor, M. Green, A.G. Orpen, K.E.
Paddick, I.D. Williams, J. Chem. Soc. Dalton Trans. (1987) 591.
(b) J.L. Templeton, P.B. Winston, B.C. Ward, J. Am. Chem.
Soc. 103 (1981) 7713. (c) B.C. Ward, J.L. Templeton, J. Am.
Chem. Soc. 102 (1980) 1532. (d) L. Ricard, R. Weiss, W.E.
Newton, G.J. Chen, J.W. McDonald, J. Am. Chem. Soc. 100
(1978) 1318.
[3] (a) E.O. Fischer, F.J. Kohl, Chem. Ber. 98 (1965) 2134. (b) H.W.
Werner, E.O. Fischer, J. Mueller, Chem. Ber. 103 (1970) 2258.
(c) M.L.H. Green, J. Knight, J. Segal, J. Chem. Soc. Chem.
Commun. (1975) 283. (d) M.L.H. Green, J. Knight, J. Chem.
Soc. Dalton Trans. (1976) 213.

196
I.C. Quarmby et al. / Journal of Organometallic Chemistry 577 (1999) 189–196
[22] (a) T.D. Newbound, M.R. Colsman, M.M. Miller, G.P. Wulfs-
berg, O.P. Anderson, S.H. Strauss, J. Am. Chem. Soc. 111
(1989) 3762, and references therein. (b) J.M. Fernandez, J.A.
Gladysz, Organometallics 8 (1989) 207.
[23] (a) M.G. Hill, W.M. Lamanna, K.R. Mann, Inorg. Chem. 30
(1991) 4687. (b) A.M. Bond, S.R. Ellis, A.F. Hollenkamp, J.
Am. Chem. Soc. 110 (1988) 5293, and references therein.
[24] N.M.N. Gowda, S.B. Naikar, G.K.N. Reddy, Adv. Inorg.
Chem. Radiochem. 28 (1984) 255.
[26] (a) M. Feigel, H. Kessler, Chem. Ber. 111 (1978) 1659. (b) Ibid
112 (1979) 3715.
[27] We compute the relative concentrations of 3 and 1+ from the
heights of voltammetric waves. To do so we must assume that
the diffusion coefficients of the two species are identical, an
approximation that is difficult to verify without isolation of the
species involved.
[28] M.A. Giardello, M.S. Eisen, C.L. Stern, T.J. Marks, J. Am.
Chem. Soc. 117 (1995) 12114, and references therein.
[29] W.J. Bowyer, W.E. Geiger, J. Am. Chem. Soc. 107 (1985)
5657.
[25] N. Stone, D.A. Sweigart, A.M. Bond, Organometallics 5 (1986)
2553.
.