Solution calorimetric and stopped-flow kinetic studies of the reaction of ?Cr(CO)3C5Me5 with PhSe-SePh and PhTe-TePh. Experimental and theoretical estimates of the Se-Se, Te-Te, H-Se, and H-Te bond strengths

Article abstract of DOI:10.1021/ic048321p

The kinetics of the oxidative addition of PhSeSePh and PhTeTePh to the stable 17-electron complex ?Cr(CO)₃C₅Me5 have been studied utilizing stopped-flow techniques. The rates of reaction are first-order in each reactant, and the enthalpy of activation decreases in going from Se (ΔH^? = 7.0 ± 0.5 kcal/mol, ΔS ^? = -22 ± 3 eu) to Te (ΔH? = 4.0 ± 0.5 kcal/mol, ΔS^? = -26 ±3 eu). The kinetics of the oxidative addition of PhSeH and ?Cr(CO)₃C₅Me ₅ show a change in mechanism in going from low (overall third-order) to high (overall second-order) temperatures. The enthalpies of the oxidative addition of PhE-EPh to ?Cr(CO)₃C₅Me₅ in toluene solution have been measured and found to be -29.6, -30.8, and -28.9 kcal/mol for S, Se, and Te, respectively. These data are combined with enthalpies of activation from kinetic studies to yield estimates for the solution-phase PhE-EPh bond strengths of 46, 41, and 33 kcal/mol for E = S, Se, and Te, respectively. The corresponding Cr-EPh bond strengths are 38, 36, and 31 kcal/mol. Two methods have been used to determine the enthalpy of hydrogenation of PhSeSePh in toluene on the basis of reactions of HSPh and HSePh with either ?Cr(CO)₃C₅Me₅or 2-pyridine thione. These data lead to a thermochemical estimate of 72 kcal/mol for the PhSe-H bond strength in toluene solution, which is in good agreement with kinetic studies of H atom transfer from HSePh at higher temperatures. The reaction of H-Cr(CO)₃C5Me₅ with PhSe-SePh is accelerated by the addition of a Cr radical and occurs via a rapid radical chain reaction. In contrast, the reaction of PhTe-TePh and H-Cr(CO)₃C₅Me ₅ does not occur at any appreciable rate at room temperature, even in the presence of added Cr radicals. This is in keeping with a low PhTe-H bond strength blocking the chain and implies that H-TePh ≤ 63 kcal/mol. Structural data are reported for PhSe-Cr(CO)₃C₅Me₅ and PhS-Cr(CO)₃C₆Me₅. The two isostructural complexes do not show signs of an increase in steric strain in terms of metal-ligand bonds or angles as the Cr-EPh bond is shortened in going from Se to S. Bond strength estimates of the PhE-H and PhE-EPh derived from density functional theory calculations are in reasonable agreement with experimental data for E = Se but not for E = Te. The nature of the singly occupied molecular orbital of the ?EPh radicals is calculated to show increasing localization on the chalcogenide atom in going from S to Se to Te.

Full text of DOI:10.1021/ic048321p

Inorg. Chem. 2005, 44, 3127−3136
Solution Calorimetric and Stopped-Flow Kinetic Studies of the Reaction
of Cr(CO)3C5Me5 with PhSe SePh and PhTe TePh. Experimental and
Se, Te Te, H Se, and H Te Bond
•
−
−
Theoretical Estimates of the Se
Strengths
−
−
−
−
James E. McDonough, John J. Weir, Matthew J. Carlson, and Carl D. Hoff*
Department of Chemistry, UniVersity of Miami, 1301 Memorial DriVe,
Coral Gables, Florida 33146
Olga P. Kryatova and Elena V. Rybak-Akimova*
Department of Chemistry, Tufts UniVersity, 62 Talbot AVenue, Medford, Massachusetts 02155
Christopher R. Clough and Christopher C. Cummins*
Department of Chemistry, Massachusetts Institute of Technology, Rm. 2-227,
77 Massachusetts AVenue, Cambridge, Massachusetts 02139
Received November 28, 2004
The kinetics of the oxidative addition of PhSeSePh and PhTeTePh to the stable 17-electron complex •Cr(CO)3C5Me5
have been studied utilizing stopped-flow techniques. The rates of reaction are first-order in each reactant, and the
enthalpy of activation decreases in going from Se ( 7.0 0.5 kcal/mol, 3 eu) to Te (
H‡ S‡ ) −22 H‡
4.0 0.5 kcal/mol, 3 eu). The kinetics of the oxidative addition of PhSeH and Cr(CO)3C5Me5 show
S‡ ) −26
a change in mechanism in going from low (overall third-order) to high (overall second-order) temperatures. The
enthalpies of the oxidative addition of PhE EPh to Cr(CO)3C5Me5 in toluene solution have been measured and
found to be 29.6, 30.8, and 28.9 kcal/mol for S, Se, and Te, respectively. These data are combined with
enthalpies of activation from kinetic studies to yield estimates for the solution-phase PhE EPh bond strengths of
46, 41, and 33 kcal/mol for E S, Se, and Te, respectively. The corresponding Cr EPh bond strengths are 38,
36, and 31 kcal/mol. Two methods have been used to determine the enthalpy of hydrogenation of PhSeSePh in
toluene on the basis of reactions of HSPh and HSePh with either Cr(CO)3C5Me5 or 2-pyridine thione. These data
lead to a thermochemical estimate of 72 kcal/mol for the PhSe H bond strength in toluene solution, which is in
good agreement with kinetic studies of H atom transfer from HSePh at higher temperatures. The reaction of
Cr(CO)3C5Me5 with PhSe SePh is accelerated by the addition of a Cr radical and occurs via a rapid radical
chain reaction. In contrast, the reaction of PhTe TePh and H Cr(CO)3C5Me5 does not occur at any appreciable
rate at room temperature, even in the presence of added Cr radicals. This is in keeping with a low PhTe H bond
strength blocking the chain and implies that H TePh 63 kcal/mol. Structural data are reported for PhSe
Cr(CO)3C5Me5 and PhS Cr(CO)3C5Me5. The two isostructural complexes do not show signs of an increase in
steric strain in terms of metal ligand bonds or angles as the Cr EPh bond is shortened in going from Se to S.
Bond strength estimates of the PhE H and PhE EPh derived from density functional theory calculations are in
reasonable agreement with experimental data for E Se but not for E Te. The nature of the singly occupied
∆
)
±
∆
±
∆
)
±
∆
±
•
−
•
−
−
−
−
)
−
•
−
H
−
−
−
−
−
−
e
−
−
−
−
−
−
)
)
molecular orbital of the
•
EPh radicals is calculated to show increasing localization on the chalcogenide atom in
going from S to Se to Te.
Introduction
dissociation energies for H2E are available and represent the
most accurate data available for the heavier chalcogenides,
The availability of thermodynamic data on the RE-ER
and RE-H bonds drops off rapidly in going from E ) O to
S and is virtually absent for Se and Te. Gas-phase first bond
* Authors to whom correspondence should be addressed. E-mail:
c.hoff@miami.edu (C.D.H.).
10.1021/ic048321p CCC: $30.25
Published on Web 03/29/2005
© 2005 American Chemical Society
Inorganic Chemistry, Vol. 44, No. 9, 2005 3127

McDonough et al.
as shown in eq 1:11
H-EH f •H + •EH
recently by Armstrong of 49.2 kcal/mol.8 Additional bond
strength estimates quoted by Armstrong are MeS-SMe )
65.5 and HS-SH ) 64.5.
∆H (kcal/mol)
(1)
E
S
Se
Te
91
80
66
The weaker bond strength of the aryl disulfides is
attributed to the stabilization of the •SPh radical relative to
•SR radicals by delocalization in the π system. This
stabilization energy was estimated9 by Benson to be ∼10
kcal/mol, a significant decrease compared to the •OPh radical
where a stabilization energy of ∼17 kcal/mol was estimated.
It seems reasonable to assume that this stabilization would
continue to decrease while descending group 16; however,
the authors could not find literature data for the estimation
of the additional stabilization of the heavier congeners •SePh
and •TePh. Theoretical calculations of the MeSe-SeMe bond
strength ranging from 30 to 61 kcal/mol have been reported.10
Aside from an early report for PhSe-SePh,3 experimental
data for enthalpies of formation of alkyl or aryl diselenides
or ditellurides could not be found.
It is worth noting that although hydrogenation of elemental
sulfur is exothermic (∆H°f of H2S ) -4.9 kcal/mol), it is
endothermic for both H2Se (∆H°f ) +7.1 kcal/mol) and
H2Te (∆H°f ) +23.8 kcal/mol)2.
These periodic trends would be expected to show up in
the thermodynamics of the oxidative addition/reductive
elimination of hydrogen to organic dichalcogenides.
RE-ER + H-H f 2H-ER
(2)
Except for sulfur, experimental data of use in eq 2 is severely
limited.3 The enthalpy of reaction 2 for E ) S, with all
species in toluene solution, is -11 kcal/mol.4 This provides
a method to derive the PhS-SPh bond strength provided
the H-ER bond strength is known. Bordwell and co-workers
have utilized pKa and electrochemical data to estimate the
H-SPh bond strength as 79 kcal/mol.5 We have previously
used these data together with enthalpies of formation and
solution to estimate the PhS-SPh bond dissociation enthalpy
to be 43 kcal/mol in toluene solution.4 Photoacoustic
calorimetric (PAC) studies6 have recently yielded a higher
estimate of the PhS-H bond strength in solution of 83.5
kcal/mol, which would yield an estimate of 51 kcal/mol for
the PhS-SPh bond strength in toluene solution (see ref 6b).
This compares to an earlier estimate of 40.6 kcal/mol made
by Griller and co-workers7 and to a gas-phase value quoted
Data for these bond strengths are important in understand-
ing the oxidative addition reactions of transition-metal
complexes and the heavier dichalcogenides. We have recently
studied synthetic, kinetic, and thermodynamic aspects11 of
oxidative addition reactions to the metal complexes and
metal-bound ligands such as those shown in eqs 3 and 4
Mo(N[R]Ar)3 + 1/2PhE-EPh f Mo(N[R]Ar)3(EPh) (3)
Mo(N[R]Ar)3[NdC(•)-Ph] + 1/2PhE-EPh f
Mo(N[R]Ar)3[NdC(EPh)(Ph)] f
Mo(N[R]Ar)3(EPh) + PhCN (4)
where R ) t-Bu and Ar ) 3,5-C6H3Me2. Prior to under-
standing the thermochemistry of reactions of complexes of
selenides and tellurides, however, it is essential to obtain
relevant bond strength estimates for the main-group PhE-
EPh compounds themselves.
A thermodynamic study of reaction 5 indicated that the
Cr-SPh bond was ∼8 kcal/mol weaker than the S-S bond
in PhS-SPh.12
(1) (a) Blanksby, S. J.; Ellison, G. B. Acc. Chem. Res. 2003, 36, 255. (b)
Gibson, S. T.; Greene, J. P.; Berkowitz, J. J. Chem. Phys. 1986, 85,
4815. (c) Sablier, M.; Fujii, T. Chem. ReV. 2002, 102, 2855. (d) CRC
Handbook of Chemistry and Physics, 84th edition; Lide, D. R., Ed.;
CRC Press: Boca Raton, FL, 2003-2004.
(2) (a) Wagman, D. D.; Evans, W. H.; Parker, V. B.; Schumm, R. H.;
Halow, I.; Bailey, S. M.; Churney, K. L.; Nuttall, R. R. J. Chem. Phys.
Ref. Data 1982, 11 (Suppl. 2). (b) NIST Chemistry WebBook. http://
webbook.nist.gov
(3) A value of 66.9 kcal/mol has been reported for the PhSe-SePh bond
strength on the basis of enthalpy of combustion data; however, it seems
unlikely that this bond is stronger than the PhS-SPh bond. The authors
could not find data for PhTe-TePh. Mortimer, C. T.; Waterhouse, J.
J. Chem. Thermodyn. 1980, 12, 961.
2•Cr(CO)3C5Me5 + PhS-SPh f 2PhS-Cr(CO)3C5Me5
(5)
(4) Mukerjee, S. L.; Gonzalez, A. A.; Nolan S. P.; Ju, T. D.; Lang, R. F.;
Hoff, C. D. Inorg. Chim. Acta 1995, 240, 175.
(5) Bordwell, F. G.; Zhang, X. M.; Satish, A. V.; Cheng, J. P. J. Am.
Chem. Soc. 1994, 116, 6605.
Kinetic studies showed an enthalpy of activation of ∼10 kcal/
mol for this reaction.12 This and other observations indicated
that reaction 5 occurred by rate-determining radical scission,
as shown in eq 6:
(6) (a) Santos, R. M. B. D.; Muralha, V. S. F.; Correia, C. F.; Guedes, R.
C.; Cabral, B. J. C.; Simoes, J. A. M. J. Phys. Chem. 2002, 106, 9883.
(b) Estimates of the PhS-H and PhS-SPh bonds interrelate through
enthalpy of hydrogenation values as noted in ref 4. The value of 83.5
kcal/mol for the PhS-H bond in ref 6a leads to an estimate of 51.8
kcal/mol for the PhS-SPh bond strength in solution using the method
described in ref 4. The value of PhS-SPh reported later in this work
of 46 kcal/mol for the PhS-SPh bond in toluene solution would lead
to an estimate of 80.5 kcal/mol for the solution phase PhS-H bond
strength. This is roughly between the value of Bordwell (ref 5) of 79
and the PAC value of 83.5 (ref 6a). At this time, an estimate of 81 (
3 kcal/mol seems reasonable for the solution phase PhS-H bond
strength estimate. A more precise analysis will require a better
understanding of the solvation enthalpy of the PhS radical than is
currently available. In addition, there are experimental errors in all of
the measurements utilized in these estimates.
•Cr(CO)3C5Me5 + PhS-SPh f
•SPh + PhS-Cr(CO)3C5Me5 (6)
The low overbarrier to reaction 6 (∼2 kcal/mol) is in keeping
with literature data for related radical reactions. Franz and
co-workers have shown that •SPh radicals abstract H atoms
(8) Armstrong, D. A. In S-Centered Radicals; Alfassi, Z. B., Ed.; John
Wiley & Sons: New York, 1999.
(9) Benson, S. W. Chem. ReV. 1978, 78, 23.
(10) Muang, N.; Williams, J. O.; Wright, A. C. THEOCHEM 1998, 453,
181.
(7) Griller, D.; Barclay, L. R. C.; Ingold, K. U. J. Am. Chem. Soc. 1975,
97, 6151.
3128 Inorganic Chemistry, Vol. 44, No. 9, 2005

Reaction of ‚Cr(CO)3C5Me5 with PhSe-SePh and PhTe-TePh
in 210 mL of freshly distilled toluene is prepared in a Schlenk tube.
This solution is filtered into a clean flask, taken from the glovebox,
and loaded into the calorimeter by cannula transfer. The calorimeter
system is of a special design for air sensitive work and was
constructed by Ace Glass. It contains a sealed Dewar flask with
six threaded glass joints as the only openings. The entire calorimeter
is mounted to the rocking mechanism of an autoclave modified to
give a 180° rotation of the entire apparatus. Once loaded, three of
the six threaded joints are used to seal Teflon rods that contain
sealed ampules of between 0.100 and 0.200 g of recrystallized
PhE-EPh. Two of the remaining joints house a rapid response
thermistor obtained from Omega Scientific and a calibrated electric
heater. The final threaded joint is attached to a stopcock, which
can be connected to a Schlenk line during filling or sample removal.
This design allows measurement of the thermal response of breaking
three ampules in sequence without opening or exposing the
calorimeter to air. The rotating shaft of the autoclave rocker is fed
through the wall of a thermostated chamber held at 20 °C. The
entire apparatus was calibrated first and found to be accurate on
the basis of the measurement of the enthalpy of solution of tris-
(hydroxymethyl) aminomethane in 0.1 N hydrochloric acid. The
response time of the calorimeter system is on the order of 10 s.
During operation, electrical calibrations are run before and after
ampule breaking, and the calorimeter is momentarily stopped from
rocking during the ampule breaking procedure. Measured data based
on the reaction of the solid dichalcogenides was corrected for the
enthalpy of solution of the solid dichalcogenide, which was
measured separately. At the concentration and temperature of the
runs, small amounts of [Cr(CO)3(η5-C5Me5)]2 are present in
solutions that are predominantly •Cr(CO)3(η5-C5Me5). Concomitant
with the reaction with dichalcogenide, a small amount of residual
dimer will dissociate to a radical monomer. Reported data is the
average of at least six independent measurements and is corrected
using reported Keq data29 so that all of the species are as reported
in the reaction.
from transition-metal hydrides with very low activation
energies.13 Organic radicals react with H-ER with enthalpies
of activation of less than 2 kcal/mol.14 The •SCH3 radical is
reported to react in the gas phase with MeS-SMe with an
activation energy of ∼1.5 kcal.mol.15
In the present work, we combine kinetic and thermody-
namic data to estimate the bond strengths themselves, relying
on the assumption that for selenium and tellurium, ∆H‡ -
∆H° ≈ 2 kcal/mol for radical reactions analogous to that
shown in eq 6 for sulfur. These data are extended to include
an experimental estimate of the PhSe-H bond strength as
well as an estimation of an upper limit for the PhTe-H bond
dissociation energy (BDE). Theoretical calculations provide
additional insight and are in reasonable accord with the
experimental data. The crystal structures of Ph-S-Cr(CO)3-
(η5-C5Me5) and Ph-Se-Cr(CO)3(η5-C5Me5) are reported.
Experimental Section
All preparative work and physical measurements (in particular,
stopped-flow kinetic measurements) were performed under an
atmosphere of purified argon using standard inert atmosphere
techniques strictly analogous to those reported in previous stud-
ies.11,12 The compounds PhE-EPh were obtained from Aldrich
Chemical and recrystallized from a toluene/heptane solution prior
to use. The complexes [Cr(CO)3(η5-C5Me5)]2 and H-Cr(CO)3(η5-
C5Me5) were prepared and purified as described previously.11,28
Infrared data were obtained in toluene solution in CaF2 cells
obtained from Harrick Scientific on a Perkin-Elmer System-2000
FTIR spectrometer. UV-vis data were obtained on a Shimadzu
UV-vis spectrometer of solutions loaded in quartz cells obtained
from Wilmad Scientific. The cells were loaded and capped in a
glovebox from freshly prepared solutions under an argon atmo-
sphere.
Crystal Growth of Ph-S-Cr(CO)3(η5-C5Me5) and Ph-Se-
Cr(CO)3(η5-C5Me5). Following solution calorimetric measurements
as described above, additional phenyl disulfide to give the correct
stoichiometry to yield Ph-S-Cr(CO)3(η5-C5Me5) was weighed into
a Schlenk flask. The calorimetric solution was transferred by
cannula to this flask, the solid dissolved by shaking the vessel, and
the solution left at room temperature for about 2 h. Following that
time, it was filtered into a clean Schlenk vessel, reduced in volume
Enthalpy of Reaction of PhE-EPh and •Cr(CO)3(η5-C5Me5).
A typical experimental procedure is described. In a glovebox under
an argon atmosphere, a solution of 1.43 g of [Cr(CO)3(η5-C5Me5)]2
(11) (a) Mendriatta, A.; Cummins, C. C.; Kryatova, O. P.; Rybak-Akimova,
E. V.; McDonough, J. E., Hoff, C. D. Inorg. Chem. 2003, 42, 8621.
(b) McDonough, J. E.; Wier, J. J.; Sukcharoenphon, K.; Hoff, C. D.;
Kryatova, O. P.; Rybak-Akimova, E. V.; Scott, B.; Kubas, G. J.;
Stepehens, F. H.; Mendriatta, A.; Cummins, C. C. Manuscript in
preparation.
(12) Ju, T. D.; Capps, K. B.; Lang, R. F.; Roper, G. C.; Hoff, C. D. Inorg.
Chem. 1997, 36, 614.
(13) Franz, J. A.; Linehan, J. C.; Birnbaum, J. C.; Hicks, K. W.; Alnajjar,
M. S. J. Am. Chem. Soc. 1999, 121, 9824.
(14) (a) Newcomb, M.; Choi, S. Y.; Horner, J. H. J. Org. Chem. 1999, 64,
1225. (b) Renaud, P. Top. Curr. Chem. 2000, 208, 81.
(15) Urbanski, S. P.; Wine, P. H. In S-Centered Radicals; Alfassi, Z. B.,
Ed.; John Wiley & Sons: New York, 1999.
(16) te Velde, G.; Bickelhaupt, F. M.; van Gisbergen, S. J. A.; Fonseca
Guerra, C.; Baerends, E. J.; Snijders, J. G.; Ziegler, T. J. Comput.
Chem. 2001, 22, 931.
(17) Fonseca Guerra, C.; Snijders, J. G.; te Velde, G.; Baerends, E. J. Theor.
Chem. Acc. 1998, 99, 391.
(18) ADF2004.01; Scientific Computing & Modeling: Amsterdam, The
Netherlands. http://www.scm.com.
(19) Vosko, S. H.; Wilk, L.; Nusair, M. Can. J. Phys. 1980, 58, 1200.
(20) Becke, A. Phys. ReV. A: At., Mol., Opt. Phys. 1988, 38, 3098.
(21) Perdew, J. P. Phys. ReV. B: Condens. Matter 1986, 34, 7406.
(22) Perdew, J. P. Phys. ReV. B: Condens. Matter 1986, 33, 8822.
(23) See, for example: Deng, L.; Schmid, R.; Ziegler, T. Organometallics
2000, 19, 3069.
(26) The current data agrees, within overlap of experimental error, with
earlier data reported in ref 11. The earlier enthalpy of reaction with
PhS-SPh was reported as ∆H ) -26.6 ( 3.0 kcal/mol, which
compares to the current value of ∆H ) -29.6 ( 2.2 kcal/mol. It is
known, as described in ref 4, that the loss of CO and formation of
dinuclear S-bridged complexes are endothermic for analogous com-
plexes of Mo. Despite the overlap, the current measured value is
considered more accurate because the earlier measurement was made
at a higher temperature and over a 1-2 h time period, which may
have allowed the dinuclear complex to be formed in an amount too
small to be detected spectroscopically. As a result of the highly air-
sensitive nature of the radical, however, it may simply be that the
independent measurements reflect the statistical error in the measure-
ment process. Because one goal of this work was a side-by-side
comparison of the S, Se, and Te derivatives, it was deemed necessary
to repeat this measurement.
(27) (a) Goh, L. Y.; Tay, M. S., Mak, T. C. W.; Wang, R. J. Organo-
metallics 1992, 11, 1711. (b) Goh, L. Y.; Tay, M. S.; Wei, C.
Organometallics 1994, 13, 1813. (c) Goh, L. Y.; Lim, Y. Y.; Tay, M.
S.; Mak, T. C. W., Zhou, Z. Y. J. Chem. Soc., Dalton Trans. 1992,
1239.
(24) van Lenthe, E.; Baerends, E. J.; Snijders, J. G. J. Chem. Phys. 1993,
99, 4597.
(25) van Lenthe, E. The ZORA Equation. Thesis, Vrije Universiteit,
Amsterdam, Netherlands, 1996.
(28) Sukcharoenphon, K.; Moran, D.; Schleyer, P. V. R.; McDonough, J.
E.; Abboud, K. A.; Hoff, C. D. Inorg. Chem. 2003, 42, 8494.
(29) Watkins, W. C.; Jaeger, T.; Kidd, C. E.; Fortier, S.; Bard, M. C.; Kiss,
G.; Roper, G. C.; Hoff, C. D. J. Am. Chem. Soc. 1992, 114, 907.
Inorganic Chemistry, Vol. 44, No. 9, 2005 3129

McDonough et al.
from approximately 200 mL to approximately 50 mL, and placed
in the freezer. Over a period of 1 month, dark red crystals of Ph-
S-Cr(CO)3(η5-C5Me5) formed in the flask and were isolated.
Similar procedures were used to prepare Ph-Se-Cr(CO)3(η5-
C5Me5). Full structural data are reported in the Supporting Informa-
tion.
was found to not significantly interfere with kinetic studies at
temperatures above 10 °C. Additional information as well as
spectroscopic and kinetic data are provided as supporting informa-
tion.
Results and Discussion
Reaction of PhSe-SePh and PheTe-TePh with H-Cr(CO)3-
(η5-C5Me5). In a glovebox, a solution of 0.06 g of [Cr(CO)3(η5-
C5Me5)]2 was prepared in 3 mL of C6D6. This was then stirred
under 1.5 atm of H2 for 2 h to produce a solution of H-Cr(CO)3-
(η5-C5Me5). To this solution was added a solution of 0.07 g of
PhSe-SePh in 2 mL of C6D6. A FTIR spectrum run within twenty
seconds of the addition showed a complete conversion to Ph-Se-
Cr(CO)3(η5-C5Me5). NMR data showed the formation of H-SePh.
Reactions done in an analogous fashion with PhTe-TePh showed
no reaction over a period of 1 h. The addition of ∼10 mol %
•Cr(CO)3(η5-C5Me5) resulted in the stoichiometric conversion of
the radical complex to Ph-Te-Cr(CO)3(η5-C5Me5) but no con-
version of H-Cr(CO)3(η5-C5Me5) over an additional 1 h time
period.
Reaction of •Cr(CO)3C5Me5 and PhE-EPh. Reac-
tions of •Cr(CO)3C5R5 (R ) H, CH3) and dichalcogenides
have been extensively studied from a synthetic point of
view. The initially formed PhE-Cr(CO)3C5R5 adduct is
known to lose CO and form dimeric chalcogen-bridged
complexes:
2•Cr(CO)3C5R5 + PhE-EPh fast8
2PhE-Cr(CO)3C5R5 slow8 [(µ-PhE)-Cr(CO)2C5R5]2 + 2CO
(7)
Because of the fact that slow formation of these dimers
occurs, calorimetric measurements were performed using a
more rapid response isoperibol calorimeter as opposed to
the Calvet techniques used originally on the reaction of PhS-
SPh.26 To compare data on all dichalcogenides under the
same conditions, the PhS-SPh measurements were repeated
using this method as well. Experimental data in eq 8 refers
to the enthalpy of reaction with all of the components in
toluene solution as shown.
Computational Details. All density functional theory (DFT)
calculations were carried out using the Amsterdam Density
Functional (ADF) program package,16,17 version 2004.01.18 The
local exchange-correlation potential of Vosko et al.19 (VWN) was
augmented self-consistently with gradient-corrected functionals
for electron exchange according to Becke20 and for electron
correlation according to Perdew.21,22 This nonlocal density func-
tional is termed BP86 in the literature and has been shown to give
excellent results for both the geometries and energetics of transition-
metal systems.23 Relativistic effects were included using the zero-
order regular approximation (ZORA).24,25 The basis set used was
the all-electron ADF ZORA/TZ2P (triple ú with two polarization
functions) basis.
2•Cr(CO)3C5Me5 + PhE-EPh f 2PhE-Cr(CO)3C5Me5
E
S
∆H
(8)
-29.6 ( 2.2
Se -30.8 ( 0.6
Te -28.9 ( 0.8
Stopped-Flow and UV-Vis Study of Reactions. Stopped-flow
kinetic studies were done using a Hi-Tech Scientific (Salisbury,
Wiltshire, U.K.) SF-43 Multi-Mixing CryoStopped-Flow instrument
equipped with stainless steel plumbing, a stainless steel mixing cell
with sapphire windows, and an anaerobic gas-flushing kit. The
instrument was connected to an IBM computer with IS-2 Rapid
Kinetics software by Hi-Tech Scientific. The temperature in the
mixing cell was maintained to (0.1 K, and the mixing time of the
instrument is 2 ms. Time-resolved spectra were acquired with a
diode array attachment. Because of the relatively rapid nature of
the reactions, a pseudo-first-order study of the reactions of
•Cr(CO)3C5Me5 and PhE-EPh was not attempted. In a typical
procedure, solutions that, prior to mixing, were exactly matched
and equal to 1 mM concentrations of [Cr(CO)3C5R5]2 and PhE-
EPh were prepared in toluene under strictly anaerobic conditions
in a glovebox, loaded into gastight syringes equipped with three-
way valves, and then transferred to the stopped-flow system, which
was purged with argon. Upon mixing, these solutions are diluted
to 0.5 mM, and at temperatures above 10 °C, the concentration of
the radical monomer •Cr(CO)3C5Me5 corresponds to >94% of the
total Cr concentration, as calculated on the basis of published data.28
Dissociation of the Cr-Cr dimer is rapid33 compared to all of the
other reactions studied, and the absorption spectrum of the dimer
The measured enthalpies of reaction are remarkably constant
for reaction 8 and imply that changes in the PhE-EPh bond
strength are roughly 1/2 the changes in the PhE-Cr bond as
a function of E throughout this series.
Structures of PhE-Cr(CO)3C5Me5 for E ) S, Se. Are
There Signs of Steric Strain in the Cr-EPh Bond? The
crystal structures of the derivatives where R ) H and E )
Se and Te have been reported by Goh and co-workers,27
whereas for the E ) S derivatives, only bridged dimers such
as [(µ-PhS)-Cr(CO)2C5H5]2 have been structurally character-
ized. As part of an investigation of pyridine thione reactivity,
we recently reported28 the structures of (η2-2-mp)-Cr(CO)2C
5Me5 and (η1-4-mp)-Cr(CO)3C5Me5. Following solution
calorimetric studies performed in this work, we obtained
X-ray quality crystals of PhS-Cr(CO)3C5Me5 and PhSe-
Cr(CO)3C5Me5 and determined their structures. Complete
structural details are available as supporting information.
Figure 1 shows an overlay of the two “piano stool” structures
for the isostructural S and Se derivatives. It might be
expected that in going from the Se to S derivative that the
shorter Cr-S bond length might result in some distortion of
the CO or Cp* ligands away from the E-Ph group. There
is no sign of that occurring, and little ligand rearrangement
around Cr is found in changing from S to Se. A lengthening
of the Cr-EPh bond from 2.4526 Å (E ) S) to 2.5890 Å
(30) Adams, R. D.; Collins, D. E.; Cotton, F. A. J. Am. Chem. Soc. 1974,
96, 749.
(31) McClain, S. L. J. Am. Chem. Soc. 1988, 110, 643.
(32) Yao, Q.; Bakac, A.; Espenson, J. H. Organometallics 1993, 12, 2010.
(33) Richards, T. C.; Geiger, W. E.; Baird, M. C. Organometallics 1994,
13, 4494.
3130 Inorganic Chemistry, Vol. 44, No. 9, 2005

Reaction of ‚Cr(CO)3C5Me5 with PhSe-SePh and PhTe-TePh
Data from eq 8 can be used to calculate the enthalpy of eq
12:
2PhSe-Cr(CO)3C5Me5 + PhS-SPh f
2PhS-Cr(CO)3C5Me5 + PhSe-SePh
∆H ) 1.2 ( 2.8 kcal/mol (12)
Addition of eqs 11 and 12 leads to a calculated enthalpy of
reaction 13:
2PhSe-H + PhS-SPh f PhSe-SePh + 2H-SPh
∆H ) -13.8 ( 4.0 kcal/mol (13)
A second method is based on the enthalpy of reaction of
2,2′ pyridine disulfide, as shown in eqs 14 and 15:
2,2′-PyS-SPy(solid) + 2H-SPh f
2(2-PySH) + PhS-SPh
∆H ) -11.9 ( 1.2 kcal/mol (14)
2,2′-PyS-SPy(solid) + 2H-SePh f
2(2-PySH) + PhSe-SePh
∆H ) -22.7 ( 1.4 kcal/mol (15)
Subtraction of eq 14 from eq 15 leads to an independent
measurement of reaction 13 and yields a value of ∆H )
-10.8 ( 2.6 kcal/mol. An average value of ∆H ) -12.3 (
4.0 kcal/mol appears to be a reasonable value for reaction
13 based on these two independent methods. It should be
emphasized that the measured enthalpies leading to this
estimate each combine several measured enthalpies, all of
which are for rather difficult reactions. Because the enthalpy
of hydrogenation of phenyl disulfide in toluene solution is
-11 kcal/mol, it is, therefore, straightforward to conclude
that
Figure 1. Overlay of PhS-Cr(CO)3C5Me5 (red) and PhSe-Cr(CO)3C5Me5
(turquoise) structures showing little distortion of ancillary metal-ligand
bonds. The Cr-EPh bond vector is omitted for clarity. Separated structural
pictures and full crystallographic data are available in the Supporting
Information.
(E ) Se) and a narrowing of the Cr-E-Ph bond angle from
114.44° (E ) S) to 109.85° (E ) Se) are the dominant
structural differences.
It is also worth noting that the Cr-SePh bond length for
PhSe-Cr(CO)3C5Me5 (2.5890) reported here is essentially
the same as that reported earlier by Goh for PhSe-
Cr(CO)3C5H5 (2.588).27 That is in contrast to the metal-
metal bonded dimers (C5R5)(CO)3Cr-Cr(CO)3(C5R5) where
PhSe-SePh + H2 f 2PhSe-H
∆H ) +1.3 ( 4.0 kcal/mol (16)
29
the Cr-Cr bond is longer for R ) CH3 (3.3107 Å) than
for R ) H30 (3.288 Å). This also implies no significant
change in steric factors in going from S to Se in these
complexes.
Using the estimate for PhSe-SePh of 41 kcal/mol derived
below yields a first estimate of PhSe-H of 72 ( 3 kcal/
mol.
Reaction of H-Cr(CO)3C5Me5/•Cr(CO)3C5Me5 and
PhE-EPh. In earlier work,12 we had investigated the
reaction of phenyl disulfide with the chromium hydride
complex, as shown in eq 17:
The Enthalpy of Hydrogenation of PhSe-SePh. En-
thalpy of hydrogenation data can provide a valuable method
to estimate element-hydride bond strengths. Two approaches
to determine this were used for PhSeH. The first is based
on measurements of eqs 9 and 10:
H-Cr(CO)3C5Me5 + PhS-SPh f
H-SPh + PhS-Cr(CO)3C5Me5 (17)
[Cr(CO)3C5Me5]2(solid) + PhSH f
H-Cr(CO)3C5Me5 + PhS-Cr(CO)3C5Me5
∆H ) -0.7 ( 0.3 kcal/mol (9)
The rate of reaction 17 was found to depend greatly on the
amount of residual radical present in the hydride solution.
The two-step radical chain mechanism shown in eqs 18 and
19 was shown to operate for E ) S.
[Cr(CO)3C5Me5]2(solid) + PhSeH f
H-Cr(CO)3C5Me5 + PhSe-Cr(CO)3C5Me5
∆H ) -8.2 ( 0.3 kcal/mol (10)
•Cr(CO)3C5Me5 + PhE-EPh f
PhE-Cr(CO)3C5Me5 + •EPh (18)
The subtraction of twice eq 9 from twice eq 10 yields eq
11:
•EPh + H-Cr(CO)3C5Me5 f H-EPh + •Cr(CO)3C5Me5
(19)
2PhS-Cr(CO)3C5Me5 + 2H-SePh f
2PhSe-Cr(CO)3C5Me5 + 2H-SPh
∆H ) -15.0 ( 1.2 kcal/mol (11)
These earlier studies were extended to the heavier dichal-
cogenides.
Inorganic Chemistry, Vol. 44, No. 9, 2005 3131

McDonough et al.
In the case of PhSe-SePh, rapid and quantitative oxidative
addition of H-Cr(CO)3C5Me5 occurred in the presence of
only trace amounts of •Cr(CO)3C5Me5. The rapid nature of
the reaction of phenyl diselenide is in keeping with a radical
chain mechanism similar to that observed earlier for PhS-
SPh. A key aspect of this radical chain reaction is eq 19, in
which in situ •EPh must be competent to abstract an H atom
from H-Cr(CO)3C5Me5 to continue the chain. As discussed
above, thermodynamic data indicate the H-SePh bond
enthalpy to be ∼72 kcal/mol as compared with the Cr-H
bond enthalpy of ∼63 kcal/mol; thus, reaction 19 is both
thermodynamically favored and expected to have a low
enthalpy of activation, allowing the chain reaction of 18 and
19 to occur rapidly.
Figure 2. Time-resolved (20 s scan, 1 s intervals) spectral changes and a
kinetic trace at 600 nm (overlapped with a second-order fit) obtained upon
mixing toluene solutions of [Cr(CO)3C5Me5]2 (1 mM) and Ph2Se2 (1 mM)
in a stopped-flow cell at 10 °C. An Eyring equation plot yielding activation
parameters ∆H‡ ) 7.0 ( 0.5 kcal/mol and ∆S‡ ) -22 ( 3 eu (R2 ) 0.996)
is also shown.
In contrast, attempts to study analogous reaction chemistry
with PhTe-TePh showed no reaction over a 2 h period, even
when the chromium radical was deliberately injected into
the reaction solution:
the formation of a meta-stable, blood-red adduct between
oxygen and •Cr(CO)3C5Me5 with λmax at 512 nm (Figure SI-
2, Supporting Information). It seems likely that the literature
data32 analysis based on the monitoring of the peak at λmax
) 512 nm corresponded to the oxygen adduct and not the
radical. That could explain the apparent disagreement
between rate constants derived for this radical and the same
rate constants derived by the electrochemical results of Baird
and co-workers.33
•Cr(CO)3C5Me5
H-Cr(CO)3C5Me5 + PhTe-TePh
8 no reaction
(20)
The failure of reaction 20 to proceed to the right implies
that either reaction 18 or reaction 19 is not competent for E
) Te. As we will discuss later, the attack of the chromium
radical on the (reaction 18) is even faster for Te than it is
for Se, and so it must be H atom transfer (reaction 19) that
is responsible for the failure of reaction 20 to proceed as
written. Because kinetic barriers for such a H atom transfer
are low, this implies that the PhTe-H bond strength is e
63 kcal/mol.
UV-Vis Spectroscopy of •Cr(CO)3C5Me5, PhE-
Cr(CO)3C5Me5, and PhE-EPh. A qualitative kinetic study
of the rates of reactions 7 and 8 by FTIR indicated that they
were too rapid to be studied by the same techniques
employed for PhS-SPh12 and that stopped-flow kinetics
would be needed to obtain quantitative data. Prior to these
studies, it was necessary to investigate the UV-vis spec-
troscopy of these species, in particular that of •Cr(CO)3C5-
Me5, which exists in rapid equilibrium with its dimer,29 as
shown in eq 21:
Selection of the spectral window to study was also dictated
by the UV-vis spectrum of the dichalcogenides and their
complexes (Figures SI-3 and SI-4, Supporting Information).
Of particular note is the need to avoid photochemical
activation and radical formation of the dichalcogenides
themselves, which change from white (PhS-SPh) to yellow
(PhSe-SePh) to orange (PhTe-TePh). In addition, it is well-
known that the product complexes are photolabile. For
example, Abrahamson and co-workers34 have shown that a
loss of CO and a subsequent reactivity or dimerization occurs
photochemically for ArS-W(CO)3C5H5 complexes at 436
nm. The product complexes PhE-Cr(CO)3C5Me5 were found
to have absorptions in this area, which trail off to higher
wavelengths (Figure SI-4, Supporting Information). On the
basis of these observations, the majority of the reactions were
followed by monitoring the band at 550 nm for decay of the
radical. Reactions were performed in both single wave-
length and diode array modes (the latter allows time-resolved
spectra to be acquired). In some cases, as described below,
decomposition attributed to secondary photochemistry was
observed when either high-energy light was employed or
monitoring was performed for long reaction times. In these
cases, single wavelength monitoring of the reaction was
followed.
2•Cr(CO)3C5Me5 S C5Me5(CO)3Cr-Cr(CO)3C5Me5 (21)
UV-vis spectral data (Figure SI-1, Supporting Information)
as well as stopped-flow kinetic spectral data, discussed later,
support that although the metal-metal bonded dimer has a
strong λmax near 465 nm, the radical complex •Cr(CO)3C5Me5
shows only a weak absorption, which is a broad band from
450 to 600 nm. These results are qualitatively similar to the
spectral data reported by McClain31 for the related but less
extensively dissociated radical/dimer pair •Cr(CO)3C5H5/
C5H5(CO)3Cr-Cr(CO)3C5H5. For •Cr(CO)3C5Me5/C5Me5-
(CO)3Cr-Cr(CO)3C5Me5, a strong peak with λmax at 512 nm
has been reported in the literature32 and was assigned to
[Cr(CO)3C5Me5]2. However, under anaerobic conditions, no
such absorption was seen in our work. The deliberate
injection of air into a solution of •Cr(CO)3C5Me5 results in
Stopped-Flow Kinetic Studies. Reaction of •Cr(CO)3-
C5Me5 and PhSe-SePh. This reaction was studied under
conditions of matched stoichiometry between the reac-
tants. Typical kinetic data as well as an Eyring plot for the
reaction of PhSe-SePh with •Cr(CO)3C5Me5 are shown in
Figure 2.
(34) (a) Weinman, D. J.; Abrahamson, H. B. Inorg. Chem. 1987, 26, 2133.
(b) Brandenburg, K. L.; Heeg, M. J.; Abrahamson, H. B. Inorg. Chem.
1987, 26, 1064.
3132 Inorganic Chemistry, Vol. 44, No. 9, 2005

Reaction of ‚Cr(CO)3C5Me5 with PhSe-SePh and PhTe-TePh
Table 1. Second-Order Rate Constants,a Activation Parameters,b
Enthalpies of Reaction, and Estimated Bond Strength Datac for
Reactions of PhE-EPh and • Cr(CO)3C5Me5
E
k
∆H‡
∆S‡
∆Hrxn
PhE-EPh
PhE-Cr
S
Se
Te
1.3
1400
19000
10.2
7.0
4.0
-24.4
-22
-26
-29.6
-30.8
-28.9
46
41
33
38
36
31
a Rate constants are extrapolated to 298 K from activation parameters
and are in M-1s-1 b Enthalpy of activation in kcal/mol and entropy of
.
activation in cal/mol K. Data for S is taken from ref 12. c Bond strength
estimates (kcal/mol) are in toluene solution and were generated as discussed
in the text. Error limits on absolute bond strengths are in the range (3
kcal/mol.
Figure 3. Kinetic trace at 550 nm obtained upon mixing toluene solutions
of [Cr(CO)3C5Me5]2 (1 mM) and Ph2Te2 (1 mM) in a stopped-flow cell at
10 °C using a visible spectrum lamp at the single wavelength mode. No
photochemical decomposition was observed under these conditions. A plot
of the temperature dependence of the second-order rate constant for the
reaction of [Cr(CO)3C5Me5]2 (1 mM) and Ph2Te2 (1 mM) in toluene is
also shown. The fit to the linear form of the Eyring equation gives activation
parameters of ∆H‡ ) 4.0 ( 0.5 kcal/mol and ∆S‡ ) -26 ( 3 eu (R2 )
0.99).
These data led to reproducible kinetics under all conditions
on a short time scale. When studied with polychromatic light,
however, photochemistry was observed leading to the
decomposition of PhSe-Cr(CO)3C5Me5 (Figure SI-5, Sup-
porting Information). This did not interfere with the kinetic
runs. A study of this reaction by FTIR spectroscopy in the
absence of light showed that PhSe-Cr(CO)3C5Me5 is formed
rapidly and quantitatively and does not decompose in the
dark over time scales longer than those used in the stopped-
flow experiments. A photochemical loss of CO, as observed
earlier by Abrahamson et al.34 for analogous W complexes,
is proposed to occur in the stopped-flow cell.
Reaction of •Cr(CO)3C5Me5 and PhTe-TePh. The
rather slow photodecomposition of PhSe-Cr(CO)3C5Me5
referred to in the preceding section was found to be much
more significant for PhTe-Cr(CO)3C5Me5 (Figure SI-6,
Supporting Information); a net increase in absorbance was
observed when the sample was exposed to polychromatic
light and, in this case, did influence the stopped-flow kinetic
studies. Because of instability under these conditions, the
reaction of PhTe-TePh was studied at the single wavelength
mode at 550 nm. As shown in Figure 3, this allowed stable
kinetic traces, which could be fit accurately to a second-
order kinetic mechanism over the entire temperature range
studied.
Figure 4. Reaction profile for the two-step cleavage of PhE-EPh by •Cr.
The first step corresponds to the formation of PhE-Cr and •EPh, and the
second step corresponds to the combination of •EPh and •Cr. An overbarrier
of 2 ( 1 kcal/mol is assumed for step i, as discussed in the text. A lower
barrier of 1.5 ( 1 kcal/mol is assumed for the radical combination in
step ii.
Complete data from these studies, together with
those reported earlier for PhS-SPh (as well as the bond
strength estimates discussed later), are summarized in
Table 1.
Combined Reaction Profile for the Oxidative Addition
of PhE-EPh. The kinetic and thermodynamic data for the
oxidative addition of the dichalcogenides provide a basis for
the estimation of both the PhE-EPh and PhE-Cr bond
strengths. This only requires the assumption of a low
overbarrier for these reactions. Data reported earlier by us
on phenyl disulfide12 were in agreement with that assumption,
and there is considerable literature precedent in that regard.13-15
Data on enthalpies of the oxidative addition of the dichal-
solution. Assuming a low overbarrier of 2 ( 1 kcal/mol in
which the metallo radical attacks the dichalcogenide means
that ∆H‡ ) DPhE-EPh - DPhE-Cr + 2. These two equations
allow the calculation of bond strengths based as collected
in Table 1, which are used to construct the reaction profile
shown in Figure 4.
Factors Influencing the PhE-Cr(CO)3(C5Me5) Bond
Strength. An examination of the data in Table 1 shows that
although both the PhE-EPh and PhE-Cr bonds show the
expected decrease in bond strength in going from S to Se to
Te, this is somewhat attenuated for the chalcogenide-metal
bond strengths. Thus, the ratio of DPhE-EPh/DPhE-Cr goes from
0.83 (S) to 0.88 (Se) to 0.94 (Te). One possible explanation
would be that, for steric reasons, the smaller -SPh group
could not come into its favored bonding distance as a result
cogenides in toluene solution obey the relationship ∆Hrxn
)
DPhE-EPh - 2DPhE-Cr, where it is understood that the bond
dissociation energies and all of the data refer to toluene
Inorganic Chemistry, Vol. 44, No. 9, 2005 3133

McDonough et al.
Figure 5. FTIR data in toluene for •Cr(CO)3C5Me5 (red) and PhE-Cr(CO)3C5Me5 [E ) S (blue), Se (green), Te (black)], showing that while oxidation
moves the M-CO bands to higher wavenumbers, it does so in the order S > Se > Te.
of interligand repulsion and that, for the larger chalcogenides,
this steric repulsion is not a factor. However, as shown in
Figure 1, X-ray data give no indication of steric “pressure”
changing, at least for the S and Se derivatives. Bond strengths
in a transition-metal complex depend not only on the M-X
bond but on how this bond influences the ancillary M-L
bonds. In the case of metal-carbonyls, oxidation of the metal
is well-known to decrease the M-CO bond strength. As
shown in the FTIR data in Figure 5, following oxidation by
the dichalcogenides, the E band of the tricarbonyl moves to
higher wavenumbers and is split to varying degrees. The A
band is also moved to higher wavenumbers for S, but less
so for Se, and it actually moves to lower wavenumbers for
the Te derivative. This effect is well-known in metal-
carbonyl chemistry, and the degree of back bonding and,
hence, the M-CO bond strength itself is perturbed by the
M-X bond. On the basis of the FTIR data, the M-CO bonds
would be strongest for the Cr-TePh derivative and weakest
for the Cr-SPh bond. This opposing trend may explain why
there is a slower rate of decrease in bond strength in the
M-EPh bond versus that of the H-EPh bond for these
complexes.
In addition, these differences in relative bond strengths
for PhE-EPh and PhE-Cr result in the nearly constant
enthalpy of oxidative addition in this series, as shown in the
energy diagram in Figure 2. Studies of enthalpies of oxidative
addition and derived bond strength data for the complexes
in reaction 3, which do not contain carbonyl ligands, do not
show this trend.10
Stopped-Flow Kinetic Study of Reaction of •Cr(CO)3-
C5Me5 and H-SePh. Two possible reaction mechanisms
for the reaction of the chromium radical and benzeneselenol
(eq 10) were considered. The first would involve the
thermodynamically uphill H atom transfer reaction shown
in eq 22 followed by the radical combination reaction shown
in eq 23:
H-SePh + •Cr(CO)3C5Me5 S H-Cr(CO)3C5Me5 + •SePh
(22)
•SePh + •Cr(CO)3C5Me5 f PhSe-Cr(CO)3C5Me5 (23)
The competing mechanism would be an overall third-order
process, such as the one shown in eq 24:
•Cr + H-SePh S (HSePh)(•Cr) + •Cr f
H-Cr + PhSe-Cr (24)
The order of bond strengths plays a vital role in determining
which of these two mechanisms is followed: [H-Cr (∼63
kcal/mol) < H-SePh (∼72 kcal/mol) < H-SR (∼79-89
kcal/mol)].35 For both alkyl and aryl thiols, an overall third-
order rate law was followed: d[P]/dt ) kthird[RSH][•Cr]2 with
∆H‡ ≈ 0 kcal/mol-1 and ∆S‡ ≈ -50 cal/mol K.
The large negative entropy of activation is due to the
formation of a termolecular transition state. The ∼9 kcal/
mol gap in bond strengths between H-Cr and H-SePh
indicates that, particularly at higher temperatures, a mech-
anism according to reactions 22 and 23 might be expected
to compete with that in eq 24.
An initial study of the reaction of •Cr with benzeneselenol
showed it to be qualitatively much faster than that observed
for thiophenol in infrared studies. The reaction is clean, and
as described earlier, calorimetric data could be obtained.
Because of the fact that benzeneselenol has no significant
absorption in the spectral window and no photochemistry at
these wavelengths, rate studies were done under pseudo-first-
order conditions of a constant excess benzeneselenol con-
centration.
(35) Hoff, C. D. Coord. Chem. ReV. 2000, 206, 451.
3134 Inorganic Chemistry, Vol. 44, No. 9, 2005

Reaction of ‚Cr(CO)3C5Me5 with PhSe-SePh and PhTe-TePh
Time-resolved spectroscopic data for this reaction from
-20 to +60 °C were obtained. Because of the spectroscopic
presence of the Cr-Cr dimer, only data obtained above 10
°C are considered for kinetics. The data between -20 and
+10 °C did, however, confirm the UV-vis assignment of
the Cr-Cr dimer band as discussed in an earlier section.
An analysis of the original spectroscopic traces at T ) 10-
60 °C showed that, in the range of 10-30 °C, there was
very little change in the apparent rate of reaction but, from
40-60 °C, the initial rate of reaction increased (Figure SI-
7, Supporting Information). The temperature-insensitive
nature of the data at 10, 20, and 30 °C recalled the data
observed for the third-order activation of thiophenol, which
also shows no temperature dependence on the rate.24 Reason-
able fits to a second-order dependence in the Cr radical could
be obtained in the temperature range 10-30 °C for an overall
third-order reaction (Figure SI-8, Supporting Information).
Estimates of the third-order rate constant based on these data
are kobs ≈ 3900 M-2s-1. That compares to the previously
reported third-order rate constants for the reaction of the
chromium radical and PhSH, kobs ≈ 25 M-2s-1. Because the
temperature dependence is low, this must be primarily due
to a reduction in the entropy of activation for the reaction
of benzeneselenol, by ∼R ln[3900/25] ≈ 10 cal/mol K, in
comparison to that of PhSH, for which ∆S‡ ≈ -52 cal/mol
deg was determined.35
Because a minimum amount of correction for zero-point
energy effects is anticipated, data for the nearly isodesmic
reaction 25 do not include corrections for zero-point energy
and are not corrected to 298 K.36 In the case of the open-
shell systems, calculations were carried out using the spin-
unrestricted formalism. The values of S2 (exact, expectation
value) calculated for these radicals were: SPh, (0.750 00,
0.758 60); SePh (0.750 00, 0.756 00); TePh (0.750 00,
0.754 16).37
Using a value for the PhS-H bond strength in toluene
solution of 81 ( 3 kcal/mol (as discussed in ref 6b) and
assuming negligible solvation energy effects for reaction 25
leads to a theoretical estimate of the PhSe-H BDE of 75 (
3 kcal/mol and an estimate of the PhTe-H BDE of 70 ( 3
kcal/mol. The experimental data reported above led to an
estimate of the PhSe-H bond strength of 72 ( 3 kcal/mol,
which overlaps with the theoretically derived value within
experimental error. The kinetic observations described above
led to the conclusion that the PhTe-H bond strength was e
63 kcal/mol. That conclusion appears valid in view of the
well-established H-Te bond strength in H2Te of 66 kcal/
mol, as shown in eq 1, and the expected weakening of the
H-Te-Ph bond relative to H-Te-H (as discussed below).
Therefore, it seems likely that the BDE calculated by our
DFT methods is too high. Wright and co-workers (see ref
36) have shown that accurate H-X bond energies can be
determined by these techniques for X ) C, N, O, and S.
Facile extension to heavier elements remains a challenge to
both theory and experiment.
Consistent with a crossover in mechanism from third order
to second order at higher temperatures, plots of ln[A] versus
time were found to be curved at 10-30 °C but they begin
to approach linearity at 40 °C and can be fit with straight
lines at 50 and 60 °C (Figure SI-9, Supporting Information).
That is consistent with the overall second-order process
having a higher activation energy and overtaking the slower
third-order process at the higher temperatures used. Second-
order rate constants of 7.2 M-1s-1 at T ) 50 °C and 12.7
M-1s-1 at T ) 60 °C lead to an estimated activation energy
of 12 kcal/mol, in excellent agreement with an endothermic
first step calculated to be ∼9 kcal/mol uphill, based on the
bond strength data discussed above, and calculated to have
a small overbarrier to H atom transfer. These data provide
additional confirmation of our assignment of the H-SePh
bond strength as being 72 ( 3 kcal/mol.
Computational Studies. The goals of the theoretical
calculations were to investigate relative PhE-H and PhE-
EPh bond strengths as well as the nature of the singly
occupied molecular orbital (SOMO) for the •EPh radicals.
In the PAC study of the PhS-H bond strength discussed
earlier,6 several levels of theory were used to determine this
bond strength. Composite ab initio procedures G3(MP2) led
to a value of 82.9 kcal/mol, in excellent agreement with the
experimental PAC studies.
Calculation of homolytic cleavage of the dichalcogenide
bond, as shown in eq 26, was also studied computationally.
PhE-EPh + 2•SPh f 2•EPh + PhS-SPh
(26)
The calculated enthalpies of reaction 26 are surprising low:
-1.3 kcal/mol for Se and 0.0 kcal/mol for Te. As shown in
Table 1, the experimental data reported in this paper yield
estimates of 46, 41, and 33 kcal/mol for the PhE-EPh bond
in going from S to Se to Te. The computational value is
again in reasonable accord with experimental data for Se;
however, the calculational result that the PhTe-TePh bond
has essentially the same strength as the PhS-SPh bond
disagrees with the experimentally derived results.
A complicating factor in these calculations is the extent
to which the chalcogenide radicals are stabilized by de-
localization into antibonding orbitals of the aryl group. As
mentioned in the Introduction, Benson has proposed a value
of about 17 kcal/mol for O and about 10 kcal/mol for S.
These factors tend to make the H-E-Ph bond weaker than
the H-E-H bond as a result of stabilization of the •EPh
(36) The electronic energy ∆E° (0 K) referred to herein can potentially be
corrected to yield ∆H° (298 K) according to methods discussed in
the following reference: DiLabio, G. A.; Pratt, D. A.; LoFaro, A. D.;
Wright, J. S. J. Phys. Chem. A 1999, 103, 1653. The required zero-
point energy calculations were deemed too computationally intensive
for our present purposes, and our presentation in terms of isodesmic
reactions should lead to a maximum of error cancellation in any case.
(37) See note 29 of the following reference for the definition of S2 as
implemented in ADF: Bulo, R. E.; Ehlers, A. W.; Grimme, S.;
Lammertsma, K. J. Am. Chem. Soc. 2002, 124, 13903.
DFT calculations in our work were obtained as described
in more detail in the Experimental Section and used to
calculate the enthalpies of reaction 25:
H-EPh + •SPh f •EPh + H-SPh
(25)
∆H ) -6.4 kcal/mol, E ) Se
∆H ) -11.3 kcal/mol, E ) Te
Inorganic Chemistry, Vol. 44, No. 9, 2005 3135

McDonough et al.
Additional work aimed at better defining the theoretical
underpinnings to the chemistry of the heavier chalcogenides
is called for.
Conclusion
The main goal of this work was to generate PhE-EPh
bond strengths, which can be of use in further studies of the
reactivity of the chalcogenides. The PhE-EPh bond, as
shown in Table 1, was found to decrease in strength: S )
46, Te ) 41, Te ) 33 kcal/mol. Experimental errors in these
absolute bond strengths are high, on the order of (3 kcal/
mol, but represent our best central data for these important
bonds. In addition, estimates of the H-EPh bond were also
found to decrease in this series: S) 81,6b Se ) 72, Te < 63
kcal/mol. These data roughly parallel the accurate literature
data for H-EH, as shown in eq 1. The low strength of the
H-TePh bond has important kinetic consequences in that it
is even weaker than the H-Cr bond strength, thus, inter-
rupting the normal direction of H atom transfer from
transition-metal hydride to main-group radical. The Cr-EPh
bond strength also follows this order (S ) 38, Se ) 36, Te
) 31), but the decrease is less marked; in particular, the gap
in the Cr-SPh and Cr-SePh bond strengths is within
experimental error. Initially, it was felt that this might be
due to steric factors destabilizing the Cr-SPh bond; however,
as shown in Figure 1, there is little sign of perturbation of
the metal-ligand bonding framework in this molecule. As
discussed earlier, the “soft” metal system, most notably the
Cr-CO bonds, may act as a buffer and tend to stabilize the
heavier chalcogenide bonds because they are less electro-
negative and less electron withdrawing, resulting in some-
what stronger Cr-CO bonds, as indicated in the FTIR spectra
in Figure 5. This is not believed to be a dominant effect;
rather it is believed to be an attenuating effect. Additional
kinetic, thermodynamic, and computational studies with other
“harder” complexes will shed additional light on this question
and are in progress.
Figure 6. The calculated SOMO orbitals for SPh, SePh, and TePh from
top to bottom, respectively. Note the increase in %E character on descending
the column.
radical by donation from chalcogen to the π orbitals of the
pendant arene. Results of our calculations show that the
percentage of the SOMO that resides on the E atom decreases
from 69.4 (Te) to 58.7(Se) to 51.4(S). Thus, the •SPh radical
shows the largest delocalization into the arene ring, and this
presumably results in greater stabilization for it relative to
the analogous Se and Te radicals. An additional observation
from the computational study is that the SOMO is also the
HOMO-1 (HOMO ) highest occupied molecular orbital) for
each of the three radicals; this non-Aufbau electron config-
uration is a consequence of the partial C-C π-bonding nature
of the SOMO (see Figure 6). The HOMO, in contrast, is an
in-plane lone pair of electrons residing in a nearly pure p
orbital of the E atom, perpendicular to the C-E bond.
The chemical consequences of the changing nature of the
SOMO calculated for these radicals are not clear at this time.
Acknowledgment. For support of this work, the authors
are grateful to the National Science Foundation (C.R.C.;
Grant -0209977).
Supporting Information Available: Tables with bond lengths,
bond angles, atomic coordinates, anisotropic displacement factors
for Ph-S-Cr(CO)3(η5-C5Me5) and Ph-Se-Cr(CO)3(η5-C5Me5),
tables of calculated geometries for PhEEPh and PhEH, UV-vis
and stopped-flow kinetic data, and graphs of fits to kinetic data for
the reaction of PhSeH and Ph-•Cr(CO)3(η5-C5Me5). This informa-
tion is available free of charge via the Internet at http://pubs.acs.org.
IC048321P
3136 Inorganic Chemistry, Vol. 44, No. 9, 2005