Activation parameters in flash photolysis studies of Mo(CO)6

Article abstract of DOI:10.1016/S0020-1693(03)00384-0

Reported is a combined time-resolved optical (TRO) and infrared (TRIR) spectroscopic investigation of the flash photolysis of Mo(CO)₆ in cyclohexane solution. TRIR studies using 308 nm excitation led to transient bleaching of the strong ν_CO band at 1987 cm^-1 of Mo(CO)₆ and appearance of new bands at 1931 and 1964 cm^-1 attributed to Mo(CO)₅(Sol). Using a high pressure/variable temperature flow cell, the kinetics of back reaction with CO (k_CO) to regenerate the hexacarbonyl was studied over the P_CO range 1-20 atm and at five temperatures. These data gave k_CO=4.6±0. 2×10⁶ M^-1 s^-1 (298 K) and the activation parameters ΔH_CO^?=32.6±3 kJ/mol and ΔS_CO^?=-7.3±11 J mol^-1 K^-1 from which an interchange mechanism was proposed. The analogous species seen in the TRO experiment displayed a transient absorbance at 420 nm and analogous kinetics properties although at lower P_CO self-trapping with Mo(CO)₆ (to give Mo₂(CO)₁₁) is a competitive process. The Mo(CO)₅(Sol) transient could also be trapped by ⁿPrBr (k_RBr=5.3??0.7×10 ⁷ M^-1 s^-1).

Full text of DOI:10.1016/S0020-1693(03)00384-0

Inorganica Chimica Acta 357 (2004) 143–148
www.elsevier.com/locate/ica
Activation parameters in flash photolysis studies of Mo(CO)6
1
Ceyhan Kayran , Margaret Richards, Peter C. Ford
*
Department of Chemistry and Biochemistry, University of California, Santa Barbara, CA 93106, USA
Received 30 May 2003; accepted 7 June 2003
Abstract
Reported is a combined time-resolved optical (TRO) and infrared (TRIR) spectroscopic investigation of the flash photolysis of
Mo(CO)6 in cyclohexane solution. TRIR studies using 308 nm excitation led to transient bleaching of the strong mCO band at 1987
cmꢀ1 of Mo(CO)6 and appearance of new bands at 1931 and 1964 cmꢀ1 attributed to Mo(CO)5(Sol). Using a high pressure/variable
temperature flow cell, the kinetics of back reaction with CO (kCO) to regenerate the hexacarbonyl was studied over the PCO range
1–20 atm and at five temperatures. These data gave kCO ¼ 4:6 ꢁ 0:2 ꢂ 106 Mꢀ1 sꢀ1 (298 K) and the activation parameters
z
z
DHCO ¼ 32:6 ꢁ 3 kJ/mol and DSCO ¼ ꢀ7:3 ꢁ 11 J molꢀ1 Kꢀ1 from which an interchange mechanism was proposed. The analogous
species seen in the TRO experiment displayed a transient absorbance at 420 nm and analogous kinetics properties although at lower
CO self-trapping with Mo(CO)6 (to give Mo2(CO)11) is a competitive process. The Mo(CO)5(Sol) transient could also be trapped by
P
nPrBr (kRBr ¼ 5:3 ꢁ 0:7 ꢂ 107 Mꢀ1
s
ꢀ1).
Ó 2003 Elsevier B.V. All rights reserved.
Keywords: Molybdenum carbonyl; TRIR; Kinetics
1. Introduction
processes such as the oxidative carbonylation of urea and
the polymerization of alkynes [5].
This laboratory has had a long-term interest in ap-
plying time-resolved infrared (TRIR) and optical (TRO)
spectroscopy to characterize reactive intermediates of
various homogeneous catalysts [1]. For examples these
studies have probed the reaction dynamics of interme-
diates in H–H and C–H activation by rhodium phosphine
complexes [2] and the formation of carbon–carbon bonds
via CO migratory insertion into metal–alkyl bonds of
manganese, cobalt and iron carbonyl models for hydro-
formylation catalysts [3]. Another homogeneous car-
bonylation catalyst is one based upon Mo(CO)6 as the
metal bearing precursor active for the selective ‘‘single
step’’ hydroxycarbonylation of ethene under relatively
mild conditions [4]. Notably, the Group VI carbonyls
have also drawn attention as homogeneous catalysts in
The pentacarbonyl species Mo(CO)5(Sol) (Sol ¼ sol-
vent) is proposed to be an intermediate in the formation
of the chain carrying complexes responsible for the eth-
ene hydroxycarbonylation pathways [4]. Under catalytic
conditions reactive intermediates are present only in
small steady state concentrations even at elevated tem-
peratures. However, the pentacarbonyl is readily gener-
ated and its reactivity investigated in using laser flash
photolysis techniques. Indeed the flash photolysis of
Mo(CO)6 and its chromium and tungsten analogs have
been favorite topics for testing new photophysical tech-
niques (for example [6,7]). These studies have demon-
strated that photoexcitation is followed by very fast (<1
ps) reactive decay to give a vibrationally excited penta-
coordinate intermediate followed by cooling and trap-
ping by solvent to give the solvated pentacarbonyl
complex Mo(CO)5(Sol) (Eq. (1)). A slower, second-order
reaction with added ligands L (among them CO) leads to
formation of the product Mo(CO)5L (Eq. (2)). The
relative stabilities of various Mo(CO)5(Sol) have been
determined and the kinetics of Eq. (2) have been inves-
tigated for several ligands [8]. However, the reaction with
* Corresponding author. Tel.: +1-805-893-2443; fax: +1-805-893-
4120.
E-mail address: ford@chem.ucsb.edu (P.C. Ford).
1
Present address: Middle East Technical University, Department of
Chemistry, 06531 Ankara, Turkey.
0020-1693/$ - see front matter Ó 2003 Elsevier B.V. All rights reserved.
doi:10.1016/S0020-1693(03)00384-0

144
C. Kayran et al. / Inorganica Chimica Acta 357 (2004) 143–148
L ¼ CO is complicated by the competition with Mo(CO)6
itself to generate an intermediate dinuclear species, pre-
sumably Mo2(CO)11 (for example [6b,9]).
then transferred from the glovebox and loaded into the
high-pressure flow systemÕs sample reservoir of the HP/
VT flow system. For TRO studies, solutions were pre-
pared on a vacuum line using Schlenk techniques. The
sample cell was a 1.0 cm path length square quartz
fluorescence cell, which had been modified for attach-
ment to vacuum lines. The solutions were frozen using
liquid nitrogen and a freeze, pump and thaw (f–p–t)
procedure was repeated three times to deaerate the so-
lutions before equilibrating with the desired CO pres-
sure. The concentrations of CO at various temperatures
were calculated from published data [11].
The work described here is a reexamination the re-
activity of the pentacarbonyl intermediate(s) generated
by flash photolysis and is part of broader investigation
of these speciesÕ roles in relevant catalysts. We have
utilized TRIR and TRO spectroscopy to determine
the kinetics of the decay of the solvento species
Mo(CO)5(Sol) (A) in cyclohexane solution in the pres-
ence of excess CO to give Mo(CO)6 and the competitive
reactions of A with added CO and with nPrBr. A unique
feature is the use of a high pressure/variable temperature
(HP/VT) cell and flow system [9] that allows investiga-
tion of detailed kinetics and activation parameters for
decay of A over a very wide range of CO pressures and
reaction temperatures.
2.4. Flash photolysis instrumentation
For TRO studies, the pump source was Nd:YAG
pulsed laser operating in the third harmonic with the
power attenuated to ꢄ20 mJ/pulse and a pulse width of
ꢄ10 ns. For kinetics traces and point-by-point transient
spectra, the probe beam was the output form ILC
Technology Model R300-5 300W Xenon lamp passed
through an IR filter and a high throughput Spex
monochromator. The probe and pump beams were ap-
proximately collinear and were passed through a Spex
Model 1680 Doublemate dual-grating monochromator
after the sample to discriminate against the latter before
detection at the RCA 8852 PMT. The PMT output was
recorded by Tektronix TDS 540 digital oscilloscope and
transferred to a computer for data analysis and storage.
The TRIR apparatus using tunable single frequency
detection (2150–1550 cmꢀ1) has described in detail
elsewhere [3]. In the present study the excitation source
was a Lambda Physiks XeCl excimer laser (308 nm
operating at 1 Hz).
hm ꢀCO
þSol
MoðCOÞ ! ! fMoðCOÞ gꢃ ! MoðCO5ÞðSolÞ
ð1Þ
ð2Þ
6
5
MoðCOÞ ðSolÞ þ L ! MoðCOÞ L þ Sol
5
5
2. Experimental
2.1. Materials
Cyclohexane used for flash photolysis was reagent
grade (Aldrich) and was distilled by published proce-
dures [10]. All gases used for laser flash photolysis ex-
periments were passed through an Oxyclear oxygen
scrubber and indicating oxygen trap prior to use. Gases
ꢀ
used in synthesis were passed through a column of 4 A
molecular sieves, Drierite and a reduced chromium silica
gel column to remove adventitious water and oxygen.
Prepurified chromatographic grade (99.998%) argon
was used in the glovebox. Reagent-grade CO was used
for synthesis, and research grade (99.5%) CO (Liquid
Carbonics or Spectra Gas) was used to prepare solutions
for laser flash photolysis studies.
3. Results and discussion
3.1. TRO studies
Flash photolysis studies with optical detection were
carried out using 355 nm excitation. When a Mo(CO)6
solution (3.2 ꢂ 10ꢀ4 M) in 298 K cyclohexane under 1.03
atm CO (PCO) was flashed, transient absorbanceat 420 nm
immediately formed that can be attributed to the solvento
species, Mo(CO)5Sol (A, Sol ¼ cyclohexane) (for example
[6a]). The decay of this transient absorbance could be fit to
a single exponential to give kobs ¼ ð5:1 ꢁ 0:1Þ ꢂ 104 sꢀ1
although such fit required a residual absorbance about
23% that of the initial DAbs, implying that not all of the
absorbance change was reversible. A better quality fit
requiring only a small residual was obtained when a
double exponential function was used to give
kobsð1Þ ¼ 4:8 ꢂ 104 sꢀ1 and kobsð2Þ ¼ 1:2 ꢂ 103 sꢀ1. When
2.2. Instrumentation
Electronic absorption spectra were recorded for so-
lutions in 1.0 cm path length quartz cells using a Hew-
lett–Packard 8452A diode array spectrophotometers.
FTIR spectra were recorded with a Bio-Rad FTS-60
FTIR spectrophotometer.
2.3. Photolysis solutions
For TRIR experiments, Mo(CO)6 solutions (0.23
mM) were prepared under an argon atmosphere in the
glovebox and were loaded into foil-wrapped 50 ml gas-
tight Hamilton syringes, equipped with shut-off valves
and adapters for Leur-Lock connections. These were
analogous experiments were carried out at lower PCO
,
single exponential fits of the fast process left increasingly

C. Kayran et al. / Inorganica Chimica Acta 357 (2004) 143–148
145
larger residuals implying that the species responsible for
the residual absorbance is formed in competition with the
reaction with CO to reform Mo(CO)6. Double exponen-
tial functions with small residuals gave consistently good
fits with both kobsð1Þ (3.5 ꢂ 104, 2.4 ꢂ 104 and 1.4 ꢂ 104
s
ꢀ1) and kobs(2) (0.96 ꢂ 103, 0.89 ꢂ 103 and 0.53 ꢂ 103 sꢀ1
)
responding to the change in PCO (0.75, 0.49 and 0.25 atm,
respectively). Plots of kobs versus [CO] gave
kCOð1Þ ¼ 5:3 ꢁ 0:3 ꢂ 106 Mꢀ1 sꢀ1 and kCOð2Þ ¼ 8:8 ꢁ
1 ꢂ 104 Mꢀ1 sꢀ1
.
We interpret these results as being consistent with
earlier reports (for example [6b,8e]) whereupon photol-
ysis labilizes CO to form the pentacarbonyl solvento
species A, which undergoes competitive trapping by CO
or by the hexacarbonyl present in solution to give the
dimeric species Mo2(CO)11 (B) (Scheme 1). The fast
absorbance change is a combination of these two pro-
cesses, but since B also absorbs at the monitoring
wavelength, it is responsible for the residual absorption
after the fast process. The slower change can be attrib-
uted to the reaction of the dimer with CO to regenerate
Mo(CO)6. Consistent with this argument, the residual
absorbance after the fast process is quite large when the
reaction is carried out under an argon rather than a CO
atmosphere. On the other hand, when the photolysis
was carried out at 1 atm CO but when a lower initial
concentration of [Mo(CO)6] was utilized, the residual
due to B was smaller.
At the higher concentration of CO, the dominant
contribution to this decay is the reaction with CO, and
the second-order rate constant kCOð1Þ is a reasonable
estimate of the kCO in Scheme 1. The 5.3 ꢂ 106 Mꢀ1 sꢀ1
value determined agrees well with the values reported by
other workers (for example [6a]). Furthermore, since the
absorption spectra of the Mo(CO)5(Sol) and Mo2(CO)11
are likely to be similar, the residual absorbances after
the fast process should indicate the amount of the dimer
formed. From the magnitude of the residual, which in-
creased as [CO] decreased, and the concentration of
Mo(CO)6 we estimate the value of kd to be ꢄ3 ꢂ 107
Scheme 2.
Mꢀ1 ꢀ1, based on the assumption that the molar ab-
s
sorbances of A and B are nearly the same at 420 nm.
TRO studies were also carried out with another ad-
ded solution ingredient, namely the halocarbon 1-
n
bromopropane, PrBr. This was chosen in part because
halocarbons are used as promoters in the molybde-
num catalyzed hydroxycarbonylation mechanism men-
tioned above. When the flash photolysis of Mo(CO)6
(3.2 ꢂ 10ꢀ4 M in cyclohexane solution) was carried out
under CO (1 atm) in the presence PrBr (3.2 ꢂ 10ꢀ4 M),
n
the decay of the prompt transient absorbance attributed
to Mo(CO)5(Sol) was markedly accelerated. This decay
could be fit to a single exponential (kobs ¼ 7:4 ꢁ 0:2 ꢂ
104 sꢀ1) but with an increased residual relative to the
observation in the absence of added RBr. This residual
proved to be very long-lived; there was minimal decay
on the longest time-scale (10 ms) of the TRO instru-
ment. The kobs values increased with [nPrBr] as expected
if the halocarbon were competing directly with CO and
with Mo(CO)6 for the pentacarbonyl intermediate
(Scheme 2). A plot of kobs versus [nPrBr] proved to be
linear with a slope (kRBr) of 5.3 ꢁ 0.7 ꢂ 107 Mꢀ1 sꢀ1 and
a non-zero intercept (5.2 ꢁ 0.6) ꢂ 104 sꢀ1 equal to the
kobs determined in the absence of nPrBr for PCO ¼ 1 atm
(Fig. 1).
3.2. TRIR studies
A key advantage of the TRIR technique relative to the
TRO method when working with compounds having
strong IR chromophores is the greater opportunity to find
detection frequencies where the absorption changes are
specific to a single species, even though several may be
present. For example, the mCO modes of Mo(CO)6 occur at
different frequencies than those of the pentacarbonyl in-
termediates. When a Mo(CO)6 solution in cyclohexane
was subjected to flash photolysis (308 nm excitation) in a
Scheme 1.

146
C. Kayran et al. / Inorganica Chimica Acta 357 (2004) 143–148
tributed to Mo(CO)5(Sol) (Sol ¼ cyclohexane). Fig. 2
shows temporal absorbance data from such an experi-
ment with [Mo(CO)6] ¼ 2.3 ꢂ 10ꢀ4 M and PCO ¼ 1 atm.
The decay of the transient absorbance at 1931 cmꢀ1 fits
well to a single exponential giving kobs ¼ 5:0 ꢁ 0:1 ꢂ 104
s
ꢀ1 in excellent agreement with the rates seen for the TRO
experiment under analogous conditions. The transient
bleaching at 1987 cmꢀ1 shows a fast component with es-
sentially the same rate constant (kobs ¼ 4:7 ꢁ 0:1 ꢂ 104
s
ꢀ1) but a large residual bleaching that appeared to con-
tinue decaying but at a rate more than an order of mag-
nitude slower (the TRIR system is not well suited for
studying these slower decays).
Another substantial advantage is the ability to control
temperature and pressure (up to ꢄ100 atm PCO) in the
HP/VT flow cell of the TRIR system. In contrast, the cells
used in the TRO system are limited to pressures not sig-
nificantly greater than 1 atm. With a system such as
Mo(CO)6, the ease of using much higher pressures allows
one to evaluate CO dependent processes over a much
wider range of PCO thereby avoiding some of the com-
plications of the Mo(CO)6 trapping of reactive interme-
diates. The HP/VT cell also provides excellent control of
temperature and is well suited for determining kinetics
activation parameters. The subsequent experiments in
this series were carried out at much higher PCO (3–20 atm)
under which conditions the residual bleaching was largely
suppressed in accord with the predictions of Scheme 1.
Activation parameters for the reaction of CO with
Mo(CO)5(Sol) (Eq. (3)) were determined by measuring
kobs values for at least 6 different PCO over the range 3–
20 atm at each T. The kCO values were determined from
the slopes of linear kobs versus [CO] plots at 25, 30, 35,
40 and 45 °C as shown in Fig. 3 and these values were
used to generate the linear Eyring plot shown in Fig. 4.
From the latter, the activation parameters
Fig. 1. Plot of kobs vs. [RBr] for the decay of transient, Mo(CO)5(Sol)
in 298 K cyclohexane under PCO ¼ 1:0 atm.
standing CaF2 cell (0.5 mm pathlength), transient
bleaching of the strong mCO band at 1987 cmꢀ1 charac-
teristic of Mo(CO)6 was observed, accompanied by the
appearance of new mCO bands at 1931 and 1964 cmꢀ1 at-
z
z
DHCO ¼ 32:6 ꢁ 3 kJ/mol and DSCO ¼ ꢀ7:3 ꢁ 11 J molꢀ1
Kꢀ1 were extracted.
MoðCOÞ ðcyclohexaneÞ þ COk!COMoðCOÞ
5
6
þ cyclohexane
ð3Þ
Considerable effort has been expended in attempts to
evaluate quantitatively the strength of the bonding be-
tween unsaturated metal carbonyls and alkane solvents
[12]. The most reliable studies would appear to be those
based on photo-acoustical calorimetry (PAC), and PAC
has been utilized to probe the Mo(CO)5–solvento in-
teraction. The accuracy of the PAC determination of the
metal–solvent interaction is dependent on the accuracy
of experimental parameters such as the quantum yield
for CO photodissociation as well as corrections for
volume changes upon reactions. This value is further
dependent upon an accurate value for the M(CO)5–CO
bond dissociation energy. A recent study by Burkey and
co-workers [12c] summarizes the possible uncertainties
Fig. 2. Temporal absorbance changes at IR frequencies characteristic
)
of (a) Mo(CO)6 (1987 cmꢀ1) and of (b) Mo(CO)5(Sol) (1931 cmꢀ1
subsequent to 308 nm flash photolysis of Mo(CO)6 in cyclohexane
solution. T ¼ 298 K, PCO ¼ 1:0 atm.

C. Kayran et al. / Inorganica Chimica Acta 357 (2004) 143–148
147
Fig. 3. Plots of kobs vs. [CO] for the decay of the transient Mo(CO)5(Sol) as monitored at 1931 cmꢀ1 in cyclohexane solution.
Acknowledgements
This research was supported by a grant (DE-FG03-
85ER13317) to PCF from Division of Chemical Sciences,
Office of Basic Energy Research, US Department of
Energy. C.K. thanks TUBITAK for financial support.
References
[1] (a) P.C. Ford, J.S Bridgewater, S. Massick, J. Marhenke, Catal.
Today 49 (1999) 419;
(b) P.C. Ford, S. Massick, Coord. Chem. Rev. 226 (2002) 39.
[2] J.S. Bridgewater, J.R. Schoonover, T.L. Netzel, S.M. Massick,
P.C. Ford, Inorg. Chem. 40 (2001) 1466.
Fig. 4. Eyring plot of kCO data in Fig. 3 for the reaction of CO with
Mo(CO)5(Sol) in cyclohexane solution.
[3] (a) W.T. Boese, P.C. Ford, J. Am. Chem. Soc. 117 (1995) 8381;
(b) K.L. McFarlane, B. Lee, W.F. Fu, R. van Eldik, P.C. Ford,
Organometallics 17 (1998) 1826;
in this determination and concludes that the Mo(CO)5–
alkane bond energy in alkane solution is 45 ꢁ 4 kJ
molꢀ1. Notably this value is about 12 kJ molꢀ1 l larger
(c) S. Massick, J. Rabor, S. Elbers, J. Marhenke, S. Bernhard, J.
Schoonover, P.C. Ford, Inorg. Chem. 39 (2000) 3098;
(d) S.M. Massick, V. Mertens, J. Jon Marhenke, P.C. Ford,
Inorg. Chem. 41 (2002) 3553;
¼
than the DHCO determined above for Eq. (3). In this
€
(e) S.M. Massick, T. Buttner, P.C. Ford, Inorg. Chem. 42 (2003)
context, one might safely argue that the rate limiting
step of this substitution reaction is unlikely to be a
limiting dissociative pathway since the activation energy
of such a step would be expected to have a minimum
value equal to or greater than the equilibrium bond
energy. Similarly, one would expect a more positive
575.
[4] J.R. Zoeller, N.L. Buchanan, T.J. Dickson, K.K. Ramming,
Catal. Today 49 (1999) 431.
[5] (a) J.R. McCusker, F. Qian, L.J. McElwee-White, Mol. Catal. A:
Chem. 159 (2000) 11;
(b) H.C.M. Vosloo, J.A.K. Du Plessis, J. Mol. Catal. 79 (1993) 7.
[6] (a) H. Hermann, F.W. Grevels, A. Henne, K. Schaffner, J. Phys.
Chem. 86 (1982) 5151;
¼
entropy of activation than the DSCO seen here. The near
¼
(b) J.M. Kelly, C. Long, R. Bonneau, J. Phys. Chem. 87 (1983)
3344;
zero value of DSCO would to be appear more consistent
with a concerted or interchange substitution mechanism.
In this context, we favor an interchange mechanism
where there is considerable Mo–Sol bond breaking at
R. Bonneau, J.M. Kelly, J. Am. Chem. Soc. 102 (1980) 1220;
(c) J.A. Ganske, R.N. Rosenfeld, J. Phys. Chem. 93 (1989)
1959.
¼
the transition state given that DHCO is about 75% that of
the metal–alkane bond.
[7] (a) A.G. Joly, K.A. Nelson, Chem. Phys. 152 (1991) 69;
(b) T.P. Dougherty, E. Heilweil, J. Chem. Phys. Lett. 227 (1994) 1.

148
C. Kayran et al. / Inorganica Chimica Acta 357 (2004) 143–148
[8] (a) F.W. Grevels, J. Jacke, S. Ozkar, J. Am. Chem. Soc. 109
(1987) 7536;
[10] J.A. Riddick, W.B. Bunger, T.K. Sakano, Organic Solvents:
Physical Properties and Methods of Purification, John Wiley &
Sons, New York, 1986.
(b) D.E. Marx, A. Lees, J. Inorg. Chem. 26 (1987) 620;
(c) S. Wieland, R. Van Eldik, J. Phys. Chem. 94 (1990)
5865;
[11] IUPAC Solubility Data Series: Vol. 43. Carbon Monoxide,
Pergamon Press, Oxford, 1990.
(d) A. Shagal, R.H. Schultz, Organometallics 21 (2002) 5657;
(e) C. Breheny, J.M. Kelly, C. Long, S. OÕKeeffe, M.T. Pryce, G.
Russell, M.M. Walsh, Organometallics 17 (1998) 3690;
(f) D.J. Darensbourg, J.D. Draper, B.J. Frost, J.H. Reibenspies,
Inorg. Chem. 38 (1999) 4705.
[12] (a) G.K. Yang, V. Vaida, K.S. Peters, Polyhedron 7 (1988)
1619;
(b) J. Morse, G. Parker, T.J. Burkey, Organometallics 8 (1989)
2471;
(c) T. Jiao, G.-L. Leu, G.J. Farrell, T.J. Burkey, J. Am. Chem.
Soc. 123 (2001) 4960.
[9] S.M. Massick, P.C. Ford, Organometallics 18 (1999) 4362.