Investigation of the mechanism of alkane reductive elimination and skeletal isomerization in Tp′Rh(CNneopentyl)(R)H complexes: The role of alkane complexes

Article abstract of DOI:10.1021/ja003944x

Experiments are described that provide indirect evidence for the involvement of alkane σ-complexes in oxidative addition/reductive elimination reactions of Tp′Rh(L)(R)H complexes (Tp′ = tris-3,5-dimethylpyrazolylborate, L = CNCH₂CMe₃). Reductive elimination rates in benzene-d₆ were determined for loss of alkane from Tp′Rh(L)(R)H, where R = methyl, ethyl, propyl, butyl, pentyl, and hexyl, to generate RH and Tp′Rh(L)(C₆D₅)D. The isopropyl hydride complex Tp′Rh(L)(CHMe₂)H was found to rearrange to the n-propyl hydride complex Tp′Rh(L)(CH₂CH₂CH₃)H in an intramolecular reaction. The sec-butyl complex behaves similarly. These same reactions were studied by preparing the corresponding metal deuteride complexes, Tp′Rh(L)(R)D, and the scrambling of the deuterium label into the α- and ω-positions of the alkyl group monitored by ²H NMR spectroscopy. Inverse isotope effects observed in reductive elimination are shown to be the result of an inverse equilibrium isotope effect between the alkyl hydride(deuteride) complex and the σ-alkane complex. A kinetic model has been proposed using alkane complexes as intermediates and the selectivities available to these alkane complexes have been determined by kinetic modeling of the deuterium scrambling reactions.

Full text of DOI:10.1021/ja003944x

J. Am. Chem. Soc. 2001, 123, 7257-7270
7257
Investigation of the Mechanism of Alkane Reductive Elimination and
Skeletal Isomerization in Tp′Rh(CNneopentyl)(R)H Complexes: The
Role of Alkane Complexes
Todd O. Northcutt, Douglas D. Wick, Andrew J. Vetter, and William D. Jones*
Contribution from the Department of Chemistry, UniVersity of Rochester, Rochester, New York 14627
ReceiVed NoVember 13, 2000
Abstract: Experiments are described that provide indirect evidence for the involvement of alkane σ-complexes
in oxidative addition/reductive elimination reactions of Tp′Rh(L)(R)H complexes (Tp′ ) tris-3,5-dimeth-
ylpyrazolylborate, L ) CNCH2CMe3). Reductive elimination rates in benzene-d6 were determined for loss of
alkane from Tp′Rh(L)(R)H, where R ) methyl, ethyl, propyl, butyl, pentyl, and hexyl, to generate RH and
Tp′Rh(L)(C6D5)D. The isopropyl hydride complex Tp′Rh(L)(CHMe2)H was found to rearrange to the n-propyl
hydride complex Tp′Rh(L)(CH2CH2CH3)H in an intramolecular reaction. The sec-butyl complex behaves
similarly. These same reactions were studied by preparing the corresponding metal deuteride complexes, Tp′Rh-
(L)(R)D, and the scrambling of the deuterium label into the R- and ω-positions of the alkyl group monitored
by 2H NMR spectroscopy. Inverse isotope effects observed in reductive elimination are shown to be the result
of an inverse equilibrium isotope effect between the alkyl hydride(deuteride) complex and the σ-alkane complex.
A kinetic model has been proposed using alkane complexes as intermediates and the selectivities available to
these alkane complexes have been determined by kinetic modeling of the deuterium scrambling reactions.
Introduction
Reductive elimination of alkane from an alkyl hydride
complex is the microscopic reverse of oxidative addition of an
alkane C-H bond to a coordinatively unsaturated metal center.
Consequently, if formation of an alkane σ-complex precedes
oxidative addition of a C-H bond, then this same σ-complex
must also be an intermediate in the reductive elimination of an
alkane from an alkyl hydride complex. The structures proposed
for these alkane complexes include η2-C,H, η2-H,H, and η3-
H,H,H. Among the many pieces of evidence for the existence
of alkane σ-complexes are the observations of hydrogen
exchange between a metal deuteride complex and its alkyl
ligands prior to reductive elimination,7 the observation of inverse
kinetic isotope effects,8 and more recently, the direct observation
Numerous transition metal complexes are known to oxida-
tively add alkanes to form alkyl hydride complexes.1 Alkane
σ-complexes are increasingly implicated as significant inter-
mediates along the reaction pathway toward oxidative addition
(eq 1).2 Direct study of the C-H bond insertion step by a
coordinatively unsaturated transition-metal fragment has been
examined using fast spectroscopic techniques in liquified or
supercritical rare gases,3 low-temperature gas matrices,4 hydro-
carbon solution,5 and in the gas phase.6
1
of such complexes by TRIR and H NMR spectroscopies.9,10
Thermochemical studies also provide evidence for the interac-
tions of alkanes with coordinatively unsaturated metals.11
(1) (a) Janowicz, A. H.; Bergman, R. G. J. Am. Chem. Soc. 1982, 104,
352. (b) Hoyano, J. K.; Graham, W. A. G. J. Am. Chem. Soc. 1982, 104,
3723. (c) Janowicz, A. H.; Bergman, R. G. J. Am. Chem. Soc. 1983, 105,
3929. (d) Hoyano, J. K.; McMaster, A. D.; Graham, W. A. G. J. Am. Chem.
Soc. 1983, 105, 7190. (e) Jones, W. D.; Feher, F. J. Organometallics 1983,
2, 562. (f) Seidler, P. F.; Wenzel, T. T.; Bergman, R. G. J. Am. Chem. Soc.
1985, 107, 4358. (g) Baker, M. V.; Field, L. S. J. Am. Chem. Soc. 1987,
109, 2825. (h) Ghosh, C. K.; Graham, W. A. G. J. Am. Chem. Soc. 1987,
109, 4726. (i) Hackett, M.; Whitesides, G. M. J. Am. Chem. Soc. 1988,
110, 1449. (j) Harper, T. G. P.; Shinomoto, R. S.; Deming, M. A.; Flood,
T. C. J. Am. Chem. Soc. 1988, 110, 7915. (k) Belt, S. T.; Grevels, F. W.;
Klotzbu¨cher, W. E.; McCamley, A.; Perutz, R. N. J. Am. Chem. Soc. 1989,
111, 8373. (l) Kiel, W. A.; Ball, R. G.; Graham, W. A. G. J. Organomet.
Chem. 1993, 259, 481. (m) Jones, W. D.; Hessell, E. T. J. Am. Chem. Soc.
1993, 115, 554. (n) Wick, D. D.; Goldberg, K. I. J. Am. Chem. Soc. 1997,
117, 110235.
Norton and co-workers have shown that Cp2W(D)(CH3) and
Cp2W(H)(CD3) exhibit intramolecular hydrogen exchange be-
tween the hydride and methyl ligands prior to methane reductive
elimination.7c Mass spectroscopy of the reductive elimination
(2) Hall, C.; Perutz, R. Chem. ReV. 1996, 96, 3125.
(5) (a) Lian, T.; Bromberg, S. E.; Yang, H.; Proulx, G.; Bergman, R.
G.; Harris, C. B. J. Am. Chem. Soc. 1996, 118, 3769. (b) Osman, R.;
Pattison, D. I.; Perutz, R. N.; Bianchini, C.; Peruzzini, M. J. Chem. Soc.,
Chem. Commun. 1994, 513. (c) Lees, A. J.; Purwoko, A. A. Coord. Chem.
ReV. 1994, 132, 155.
(3) (a) Schultz, R. H.; Bengali, A. A.; Tauber, M. J.; Weiller, B. H.;
Wasserman, E. P.; Kyle, K. R.; Moore, C. B.; Bergman, R. G. J. Am. Chem.
Soc. 1994, 116, 7369. (b) Bengali, A. A.; Schultz, R. H.; Moore, C. B.;
Bergman, R. G. J. Am. Chem. Soc. 1994, 116, 9585.
(4) Haddleton, D. M.; McCamley, A.; Perutz, R. N. J. Am. Chem. Soc.
1988, 110, 1810.
(6) Wasserman, E. P.; Moore, C. B.; Bergman, R. G. Science 1992, 255,
315.
10.1021/ja003944x CCC: $20.00 © 2001 American Chemical Society
Published on Web 07/04/2001

7258 J. Am. Chem. Soc., Vol. 123, No. 30, 2001
Northcutt et al.
products from crossover experiments further support the in-
tramolecular nature of the rearrangements. A number of other
metal systems show similar exchanges in M(CH3)D com-
pounds.7 Girolami has recently observed dynamic exchange
between the hydrogens of the methyl group and the hydride
ligand in (C5Me5)Os(dmpm)(CH3)(H)+.12
Intramolecular rearrangement of an alkyl ligand has been
observed by Periana and Bergman, in addition to hydrogen/
deuterium exchange, in Cp*Rh(PMe3)(D)(13CH2CH2) at low
temperature.7b Migration of deuterium into the R position of
the ethyl group occurs more rapidly than migration to the â
position. Throughout the rearrangement the deuterium label
remains attached to the 13C-labeled carbon. Both migra-
tion processes are competitive with reductive elimination of
13C,2H-ethane.
and oxidative addition are measured, and determined to both
be normal kinetic isotope effects (kH/kD > 1). All of these
reactions are interpreted in terms of alkane σ-complexes as
intermediates.
The mechanism by which the isomerizations occur can be
broken down into several key steps: (1) C-H bond activation
and its reverse, reductive elimination without dissociation of
the alkane (these processes will be referred to herein more
specifically as reductiVe bond formation and oxidatiVe bond
cleaVage), (2) migration from one σ-alkane bond to another,
and (3) dissociation of alkane from the alkane σ-complex. The
relative rates of these processes, namely, dissociation vs
oxidative bond cleavage of alkane (eq 3), migration vs oxidative
bond cleavage (eq 4), and migration vs dissociation (eq 5), will
Direct evidence for alkane complexes first came from the
study of photochemically generated M(CO)5 (M ) Cr, Mo, W)
fragments in alkane and alkane/rare gas matrices.13 George et
al. also observed transient species by FTIR assigned as alkane
complexes upon irradiation of Cp*Re(CO)3 in hydrocarbon
solvents.9 Recently, low-temperature observation of the metal
fragment [CpRe(CO)2] in cyclopentane by Ball and Geftakis
provided direct spectroscopic evidence for the formation of an
alkane complex,10 although the data could not distinguish
between a fluxional η2-C,H and an η2-H,H geometry for the
complex.
Earlier studies of the [Tp′Rh(L)] fragment show it to be
capable of activating a wide variety of alkane and arene C-H
bonds (eq 2). In this report we will elucidate several of the key
determine the observed product ratios. Kinetic simulations based
on the results of isotopic labeling studies with several Tp′Rh-
(L)(R)D complexes has allowed for the determination of the
relative rates of each of these processes. Of particular relevance
to these studies is the preliminary report by Flood on similar
rearrangements of [CnRh(L)RD]+ complexes (Cn ) 1,4,7-
triazacyclononane), which will be compared with the present
results.14
mechanistic features of the interaction of the [Tp′Rh(L)] metal
fragment with several alkanes (L ) CNneopentyl, Tp′ ) tris-
(3,5-dimethylpyrazolyl)borate). Reductive elimination from
linear alkyl hydride complexes is examined. Additionally,
secondary alkyl hydride complexes are observed to rearrange
intramolecularly to give the more stable linear n-alkyl hydride
isomers. Kinetic isotope effects on both reductive elimination
Results and Discussion
Preparation of Tp′Rh(L)(R)H Complexes. Synthesis of a
series of Tp′Rh(L)(R)Cl complexes (R ) Me (1-Cl), Et (2-Cl),
n-Pr (3-Cl), n-Bu (4-Cl), n-pentyl (5-Cl), n-hexyl (6-Cl),
isopropyl (3′-Cl), and sec-butyl (4′-Cl); L ) neopentylisocya-
nide)15 is accomplished in reasonable yield (71%, 1-Cl to 34%,
6-Cl) by reaction of Tp′Rh(L)Cl2 with the appropriate Grignard
reagent.16 Yields of the secondary alkyl chloride complexes are
much lower than those obtained for linear alkyl complexes, 31%
for 3′-Cl and 13% for 4′-Cl. The sec-butyl derivative 4′-Cl is
formed as a mixture of diastereomers in a 1:1 ratio and can be
separated by fractional crystallization from methanol. All of
(7) (a) Buchanan, J. M.; Stryker, J. M.; Bergman, R. G. J. Am. Chem.
Soc. 1986, 108, 1537. (b) Periana, R. A.; Bergman, R. G. J. Am. Chem.
Soc. 1986, 108, 7332. (c) Bullock, R. M.; Headford, C. E. L.; Hennessy,
K. M.; Kegley, S. E.; Norton, J. R. J. Am. Chem. Soc. 1989, 111, 3897. (d)
Parkin, G.; Bercaw, J. E. Organometallics 1989, 8, 1172. (e) Gould, G. L.;
Heinekey, D. M. J. Am. Chem. Soc. 1989, 111, 5502. (f) Stahl, S. S.;
Labinger, J. A.; Bercaw, J. E. J. Am. Chem. Soc. 1996, 118, 5961. (g) Wang,
C.; Ziller, J. W.; Flood, T. C. J. Am. Chem. Soc. 1995, 117, 1647. (h)
Mobley, T. A.; Schade, C.; Bergman, R. G. J. Am. Chem. Soc. 1995, 117,
7822. (i) Chernaga, A.; Cook, J.; Green, M. L. H.; Labella, L.; Simpson, S.
J.; Souter, J.; Stephens, A. H. H. J. Chem. Soc., Dalton Trans. 1997, 3225.
(8) Examples of inverse kH/kD: (a) Cp*Ir(PMe3)(H)(C6H11) in C6D6 at
130 °C, 0.7(1) (ref 7a). (b) Cp*Rh(PMe3)(H)(C2H5) in toluene-d8 at -30
°C, 0.5(1) (ref 7b). (c) Cp2W(H)(CH3) in CD3CN at 72.6 °C, 0.75(4) (ref
7c). (d) Cp*2W(H)(CH3) in C6D6 at 100 °C, 0.77(7) (ref 7d). (e) [Cp2Re-
(H)(CH3)][Cl] in CD2Cl2 at 9 °C, 0.8(1) (ref 7e). (f) [CnRh(PMe3)(H)-
(CH3)][BAr4] in C6D6 at 75 °C, 0.74(2) (ref 7g). (g) (tmeda)Pt(CH3)(H)(Cl)
in CH3OH at -27 °C, 0.29 (ref 7f).
1
these chloro derivatives have been fully characterized by H
NMR, 11B NMR, 13C NMR, and IR spectroscopies, as well as
by elemental analysis, and are assigned as octahedral RhIII
complexes containing η3 Tp′ ligands.17 X-ray diffraction studies
have been conducted on complexes 2-Cl and 6-Cl, confirming
the structural conclusions made on the basis of spectroscopic
data (see Supporting Information, Figures S-1 and S-2). Metrical
parameters for the two structures show little difference in the
(9) By TRIR: Sun, X.-Z.; Grills, D. C.; Nikiforov, S. M.; Poliakoff,
M.; George, M. W. J. Am. Chem. Soc. 1997, 119, 7521.
(14) Flood, T. C.; Janak, K. E.; Iimura, M.; Zhen, H. J. Am. Chem. Soc.
2000, 122, 6783.
(15) In this manuscript, compound numbers refer to the length of the
alkyl chain; primes denote 2-alkyl (iso- or sec-) isomers; greek letters (R,
â, etc.) denote the location of deuterium relative along the chain relative to
the metal.
(16) (a) Synthesis of 1-Cl, 3-Cl, and 6-Cl: Wick, D. D.; Jones, W. D.
Inorg. Chem. 1997, 36, 2723. (b) Synthesis of 5-Cl: see ref 1m.
(17) Northcutt, T. O.; Lachicotte, R. J.; Jones, W. D. Organometallics
1998, 14, 5148.
(10) By NMR: Geftakis, S.; Ball, E. J. Am. Chem. Soc. 1998, 120, 9953.
(11) (a) Brown, C. E.; Ishikawa, Y.; Hackett, P. A.; Rayner, D. M. J.
Am. Chem. Soc. 1990, 112, 2530. (b) Klassen, J. K.; Selke, M.; Sorensen,
A. A.; Yang, G. K. J. Am. Chem. Soc. 1990, 112, 1267. (c) Morse, J. M.,
Jr.; Parker, G. H.; Burkey, T. J. Organometallics 1989, 8, 2471.
(12) Gross, C. L.; Girolami, G. S. J. Am. Chem. Soc. 1998, 120, 6605.
(13) (a) Graham, M. A.; Perutz, R. N.; Poliakoff, M.; Turner, J. J. J.
Organomet. Chem. 1972, 34, C34. (b) Perutz, R. N.; Turner, J. J. J. Am.
Chem. Soc. 1975, 97, 4791.

Tp′Rh(CNneopentyl)(R)H Complexes)
J. Am. Chem. Soc., Vol. 123, No. 30, 2001 7259
Table 1. Summary of Kinetic Data for Reductive Elimination of
RH from Tp′Rh(L)(R)H in C6D6 at 26 °Ca
complex
kobs, s-1
Tp′Rh(L)(Me)H (1)
Tp′Rh(L)(Et)H (2)
Tp′Rh(L)(n-Pr)H (3)
Tp′Rh(L)(n-Bu)H (4)
Tp′Rh(L)(n-pentyl)H (5)
Tp′Rh(L)(n-hexyl)H (6)
4.51(3) × 10-5
1.82(7) × 10-4
2.63(7) × 10-4
2.77(14) × 10-4
2.70(8) × 10-4
2.76(6) × 10-4
a Errors are indicated in parentheses at the 95% confidence level
(2σ).
coordination geometry at rhodium, including the Rh-CR-Câ
angle of the alkyl group. Full structural details are included in
the Supporting Information.
Figure 1. Distribution of species for the reaction of Tp′Rh(L)(i-Pr)-
Cl, 3′-Cl, with [Cp2Zr(H)2]2 in C6D6 at 26 °C.
Complete conversion to the corresponding alkyl hydride
complexes is effected by reaction of the alkyl chloride com-
plexes with [Cp2ZrH2]2 in benzene.18 Conversion is generally
rapid (e.g. τ1/2 ≈ 200 s for conversion of 3′-Cl to Tp′Rh(L)-
(i-Pr)H, 3′). The hydride resonance for the primary alkyl hydride
complexes 1-6 appears at ca. δ -14.9 (JRh-H ) 24.8 Hz) while
the hydride resonance for the two secondary alkyl hydride
complexes, 3′ and 4′, appears at δ -15.29 (JRh-H ) 25.2 Hz)
in C6D6 solvent.
Scheme 1
Reductive Elimination of Primary Alkyl Hydrides in
Benzene. Reductive elimination of alkane from the rhodium
primary alkyl hydride complexes 1-6 in C6D6 was monitored
by 1H NMR spectroscopy at 26 °C. The reaction was found to
be first order in metal complex, producing alkane and Tp′Rh-
(L)(C6D5)D as the only products. No intermediates or isomers
were detected. The rate of reductive elimination was determined
for each complex by least-squares fit of the data to an
exponential decay. The results are summarized in Table 1. Loss
of methane from 1 is about 4× slower than loss of ethane from
2, which is slightly slower (1.5×) than loss of alkane from the
longer alkyl chain derivatives. The rate of alkane loss is identical
from alkyl hydride complexes with a chain of three or more
carbon atoms. An earlier study of 1 showed evidence for a minor
bimolecular component to the displacement of methane by
benzene.18 Consequently, all experiments reported here were
carried out in neat benzene solvent, so that the rate of formation
of Tp′Rh(L)(C6D5)D would appear first order.
Scheme 2
Isomerization of Secondary Alkyl Hydride Complexes. In
benzene-d6 solution at ambient temperature Tp′Rh(L)(i-Pr)H,
3′, prepared from 3′-Cl and [Cp2ZrH2]2, is observed to rearrange
over the course of an hour to give the n-propyl isomer, 3. The
secondary alkyl hydride complex is easily distinguished from
the primary complex by virtue of the difference in chemical
shift and coupling constant of the hydride ligand.19 The
isomerization of 3′ to 3 proceeds with a half-life of ap-
proximately 25 min. Competitive with isomerization is the
reductive elimination of propane from 3′ to give Tp′Rh(L)-
(C6D5)D. Loss of propane from 3 produced in situ must also
be occurring at the same rate as that measured from the
independently prepared n-propyl hydride, contributing to the
appearance of Tp′Rh(L)(C6D5)D. The distribution of species is
shown in Figure 1. An empirical fit of the kinetic model shown
in Scheme 1 to the observed data required the presence of a
direct pathway from 3′ to Tp′Rh(L)(C6D5)D as Tp′Rh(L)-
(C6D5)D is formed more rapidly than 3. Importantly, the
rearrangement of 3′ to 3 must be occurring intramolecularly
since C6D6 solvent would scavenge any [Tp′Rh(L)] produced
by reductive elimination followed by dissociation of alkane.
The sec-butyl derivative shows similar behavior. Reaction
of a 1:1 mixture of diastereomers of 4′-Cl with [Cp2Zr(H)2]2
affords a 1:1 mixture of diastereomers of 4′. Both diastereomers
are observed to decay at the same rate to afford the rearranged
n-butyl hydride, 4. Loss of butane from 4′ to give Tp′Rh(L)-
(C6D5)D is competitive with the rearrangement process (Scheme
2). Identical rates are observed when the reaction is performed
with a single diastereomer. Reaction of one diasteriomer of 4′-
Cl (isolated by crystallization) with [Cp2Zr(H)2]2 gives only one
diastereomer of 4′, implying that the Cl/H exchange is ste-
reospecific and that reversible reductive bond formation and
oxidative bond cleavage of the â-C-H bonds is not occurring.
To further examine the nature of the rearrangement, the
isotopically labeled isopropyl deuteride complex 3′-d1 was
prepared by use of [Cp2ZrD2]2 as the hydride source in the
(18) Wick, D. W.; Reynolds, K. A.; Jones, W. D. J. Am. Chem. Soc.
1999, 121, 3974.
(19) This phenomenon has been noted by others: Janowicz, A. H.;
Bergman, R. G. J. Am. Chem. Soc. 1983, 105, 3929.
reaction with 3′-Cl. A doublet of 1:1:1 triplets (δ 0.832, JHH
7.1 Hz, JHD ) 1.2 Hz) is observed for the methyl groups in
)
free propane produced during the reductive elimination of

7260 J. Am. Chem. Soc., Vol. 123, No. 30, 2001
Northcutt et al.
Scheme 3. Reductive Bond Formation from 3′-d1 Is
Irreversible
Measurement of Kinetic Isotope Effects. Use of [Cp2Zr-
(D)2]2 in place of [Cp2Zr(H)2]2 allows for the formation of the
deuteride isotopomers, 1-4′-d1. As mentioned above, for
complexes 3′-d1 and 4′-d1 the deuterium label is observed only
in the secondary position of the alkane upon reductive elimina-
tion. No scrambling of the label into primary sites is observed.
No secondary hydrides were observed by 1H NMR spectroscopy
during the rearrangement and elimination of alkane-d1 from 3′-
d1 and 4′-d1, consistent with an irreversible reductive bond
formation process. These latter observations allow for the
measurement of a kinetic isotope effect for the fundamental step
involving reductive bond formation to generate an alkane
σ-complex (eq 7), which corresponds to the ratio of the observed
propane from 3′-d1 (δ 0.848, JHH ) 7.2 Hz). The methylene
resonance, a septet in fully protonated propane (δ 1.25), appears
as a broad multiplet of low intensity. These results are consistent
with the formation of CH3CHDCH3 in loss of propane from
3′-d1.
The isopropyl hydride isotopomer, Tp′Rh(L)(CD(CH3)2)H,
was not observed, as might be expected for a reversible reductive
bond formation/oxidative bond cleavage process at the â-carbon
(Scheme 3). Furthermore, the deuterium label was observed only
in the secondary positions in both 3â-d1 and free propane-d1 in
rate constants (kbc + kbd)H/(kbc + kbd)D in Schemes 1 and 2,
since re-addition to a secondary C-H bond is slow relative to
migration and insertion into a terminal C-H bond or dissocia-
tion. For 3′-d1 the measured isotope effect (kH/kD) was 2.1, a
normal isotope effect. For 4′-d1 a similar value was obtained
(kH/kD ) 2.7).
2
the H NMR spectrum of the reaction at intermediate times.
The lack of isotopic scrambling into the R-position of 3′-d1
implies that rearrangement and/or dissociation are faVored oVer
actiVation of the secondary C-H bond, i.e., reductiVe bond
formation in isopropyl hydride 3′ is irreVersible.
The isotope effect on the reverse reaction, oxidative bond
cleavage of a C-H bond in an alkane complex, has also
been determined. Photolysis of Tp′Rh(L)(η2-PhNdCdNCH2-
C(CH3)3), 7, at 5 °C in C6F6 in the presence of CH2D2 leads to
the formation of a mixture of C-H and C-D activated products,
Similar observations are made with the sec-butyl deuteride
derivative. Deuterium substitution in 4′-d1 resulted in the
appearance of label in both secondary positions in 4-d1 produced
in situ (eq 6). The concentration of the label is slightly higher
in the â position (17% 4â-d1) than in the γ position (14% 4γ-
1
presumably via a methane σ-complex (eq 8). H{2H} NMR
2
d1) as observed by H NMR spectroscopy. (See Supporting
Information.)
analysis at 9 °C at early reaction times reveals a 4.3:1 mixture
of Tp′Rh(L)(CHD2)H, 1a-d2, and Tp′Rh(L)(CH2D)D, 1b-d2.
Assuming that the σ-CH2D2 complex has equal access to the
C-H and C-D bonds (vide infra),20 this ratio reflects the kinetic
isotope effect on oxidative bond cleavage in an alkane complex,
which is again observed to be a normal kinetic isotope effect.
Upon standing at room temperature, this mixture equilibrates
to yield a 1.7:1 ratio of 1a-d2 and 1b-d2. A more thorough,
quantitative treatment of H/D scrambling will be given in a
subsequent section.
The selectivity of the [Tp′Rh(L)] fragment for activation of
primary C-H bonds over secondary C-H bonds in normal
alkanes has been previously reported.1m Conclusions regarding
the kinetic vs thermodynamic origins of this selectivity remained
uncertain in this earlier work. Since only primary alkyl hydride
products were observed, it could be possible that secondary
C-H activation occurred initially but that these kinetically
formed species rearranged rapidly under the reaction conditions
to the thermodynamically preferred primary derivatives. The
present experiments show, however, that the lifetime of the
isopropyl hydride is sufficiently long that detection of 3′ under
the conditions employed in the low-temperature photochemical
activation of propane by Tp′Rh(L)(η2-PhNdCdNCH2C(CH3)3),
7, would have been possible were it present within the reaction.
Given the lack of secondary C-H bond activation in the earlier
alkane activation experiments and the observed rearrangement
of 3′ to 3, it can be concluded that the actiVation of the primary
C-H bonds by [Tp′Rh(L)] is both kinetic and thermodynamic
in nature.
Alkane σ-complexes are increasingly implicated as significant
intermediates in C-H bond activation, and the observation of
(20) In the σ-cyclopentane complex described by Ball, the metal was
believed to rapidly migrate from one C-H to the other within the same
methylene unit, although an η2-H,H binding geometry could not be ruled
out. In studies of an agostic methyl group by Shapley in Os3(CH2D)(CO)11,
migration between C-H and C-D was fast on the NMR time scale and a
preference for binding C-H vs C-D was only 1.25 (see ref 27). In the
present system, a similar slight preference for binding C-H vs C-D bonds
may be present, although studies of [Cp*Rh(CO)] in liquid krypton show
a distinct preference for binding C-D bonds over C-H bonds (see ref 3).

Tp′Rh(CNneopentyl)(R)H Complexes)
J. Am. Chem. Soc., Vol. 123, No. 30, 2001 7261
an inverse kinetic isotope effect for reductive elimination from
metal alkyl hydride complexes is commonly cited as providing
evidence for these species.8 These inverse isotope effects are
often explained by the presence of a preequilibrium involving
an alkane σ-complex followed by a rate-determining alkane
dissociation step with the observed isotope effect being the
product of the equilibrium isotope effect and the kinetic isotope
effect, K‚k (eq 9). Inverse isotope effects, however, can also be
accommodated by single-step mechanisms provided the transi-
tion state is sufficiently late (i.e., a very product-like transition
state).21
2
Figure 2. Sample H{1H} NMR spectrum of the rearrangement of
Tp′Rh(L)(CH2CH2CH3)D (3-d1, b) to Tp′Rh(L)(CDHCH2CH3)H (3R-
d1, 9, 2 diastereomers) and Tp′Rh(L)(CH2CH2CDH2)H (3γ-d1, [) in
C6H6. Free propane-d1 (O) appears at δ 0.67.
loss: (1) scrambling of deuterium onto the R-carbon and (2)
scrambling of deuterium into the terminal methyl group of the
n-alkyl chain. The label was observed in both the R- and
â-positions with ethyl complex 2-d1. The R-scrambled product,
2R-d1, grew to a maximum of 23% of all observable Tp′Rh(L)
species, while the â-migratory product, 2â-d1, grew to a
maximum of ∼16% over 450 min (eq 10). Similar reactivity
was observed by Bergman in the doubly labeled complex
Cp*Rh(PMe3)(*CH2CH3)D although relative amounts were not
quantified.7b
Normal kinetic isotope effects were observed for the reductive
elimination of alkane from 3′ and 4′. Additionally, no H/D
exchange at the secondary position was observed over the course
of the reaction. These observations are consistent with the trend
noted by Bercaw and Labinger, in which only complexes
exhibiting H/D scrambling showed inverse isotope effects.7f The
lack of scrambling in secondary alkyl deuteride complexes
allows the isotope effect on reductive bond formation from a
secondary alkyl hydride to be unambiguously assigned.
The C-H/D bond activation of CH2D2 by 7 is, to the best of
our knowledge, the first direct measurement of an internally
competitive kinetic isotope effect on the oxidative bond cleavage
of a σ-complexed alkane by a transition metal complex. The
use of an inert solvent, C6F6, precludes side reactions with
solvent. Photolysis and spectroscopy at low temperature place
the reaction under kinetic control, i.e., no rearrangement of the
initially formed bond activation products occurs. Statistically,
C-H and C-D bond activation are equally likely in CH2D2,
making the observed product ratio a direct reflection of the
isotope effect for oxidative bond cleavage of the σ-complexed
methane. The value obtained, 4.3, is normal and larger than
the isotope effect observed on reductive bond formation from
3′ and 4′. Any H/D scrambling occurring before acquisition of
the first 1H{2H} NMR spectrum would serve only to decrease
the observed product ratio, making 4.3 a minimum value for
the isotope effect on oxidative bond cleavage. If we assume
that the isotope effect on reductive bond formation to generate
an alkane complex is similar for different alkanes then these
two kinetic isotope effects can be combined to give an
equilibrium isotope effect of ∼0.5 for the preequilibrium (K)
prior to alkane dissociation in eq 9. This inverse value is similar
to that seen for the overall loss of alkane from Tp′Rh(L)(CD3)D
vs Tp′Rh(L)(CH3)(H), where kH/kD ) 0.62.18
Complexes 3-d1 to 6-d1 exhibit similar behavior as monitored
2
by H{1H} NMR spectroscopy (Figure 2), with R-scrambled
products growing to a maximum of 18-23% of the observable
species. As mentioned above, incorporation of the deuterium
label into the terminal methyl group of the alkyl ligand was
also observed. The amount of the terminally labeled species
decreased as the chain length of the alkyl ligand increased, with
∼8% of the label observed at the terminal methyl group in 3γ-
d1 to a minimum of ∼4% observed in 5ꢀ-d1. The methyl
resonance of free hexane-d1 obscured the small amount of label
present in the methyl group of 6φ-d1, so that the amount of
scrambled product could not be quantified. No other intermedi-
ates were observed during the course of the reaction.
Detailed Mechanism of Isomerization and Scrambling.
The isomerization and scrambling processes observed in the
Tp′Rh alkyl hydride system can be accommodated by several
different reaction mechanisms. However, the intramolecular
nature of the reactions as well as the selectivity of the activation
and isomerization process allow several of these possibilities
to be discounted. Previous work by Bergman7a and Flood14 has
shown that the intramolecular rearrangement of alkyl hydride
complexes does not take place via a reversible â-hydride
elimination/insertion pathway. The observation of retention of
deuterium-labeling with the appropriate carbon in the present
studies also argues against this pathway. A mechanism in which
the terminal methyl group of a straight-chained alkyl ligand is
activated to give a metallocyclic complex that then eliminates
to give an isomerized n-alkyl complex can be ruled out for
several reasons. With increasing chain length, which might be
expected to make the metallocycle more stable and thus lead
to increased appearance of the label at the terminus, migration
of the label monotonically decreases. Also, the RhV species
Also of relevance is the direct measurement of isotope effects
on C-H vs C-D oxidative bond cleavage of cyclohexane and
neopentane by [Cp*Rh(CO)]. In these fast IR studies, direct
kinetic evidence was obtained for normal isotope effects on
oxidative bond cleavage (kH/kD ) 4.4 at 25 °C),3 almost exactly
what is seen in the present case using CH2D2.
Scrambling and Migration in Primary Alkyl Deuteride
Complexes. The isotopic scrambling of 1-d1 to 6-d1 were
monitored by 2H{1H} NMR spectroscopy at 26 °C. As described
in a previous paper, Tp′Rh(L)(CH3)D, 1-d1, scrambles with
Tp′Rh(L)(CH2D)H, 1R-d1, at a rate competitive with loss of
CH3D (see below).18 For the other alkyl derivatives, two
processes were generally observed to be competitive with alkane
(21) Cheng, T.-Y.; Bullock, R. M. J. Am. Chem. Soc. 1999, 121, 3150-
3155.

7262 J. Am. Chem. Soc., Vol. 123, No. 30, 2001
Northcutt et al.
Scheme 4
Scheme 6
Scheme 5
opportunities for dissociation of alkane as the metal center
migrates down the alkyl chain. The amount of scrambling of
the deuterium label into the R position, however, is nearly the
same for all linear alkyl hydride complexes with alkyl groups
with 2 or more carbons (∼20%). A similar primary σ-alkane
complex is involved in the R-scrambling process in each of
these, and consequently, the extent of R-scrambling is nearly
identical in all of the alkyl deuteride complexes.
Appearance of the label in the terminal methyl group of the
longer chain alkanes (i.e. butyl or pentyl) requires migration
from secondary-to-secondary C-H bonds in σ-alkane com-
plexes. Secondary-to-secondary migration is also observed in
the rearrangement of sec-butyl complex 4′-d1. The deuterium
label appears in both secondary positions (â and γ) of the n-butyl
complex that is formed (Scheme 6), indicating that migration
occurs to each end of the chain. Appearance of the label in the
δ-position during the scrambling of n-butyl deuteride complex
4-d1 also necessitates migration from secondary-to-secondary
C-H bonds.
One additional feature present in these systems needs to be
addressed, namely, the hapticity of the Tp′ ligand during these
reactions. Earlier studies by Bergman and Jones suggest that
the RhI species commonly contain an η2-Tp′ ligand whereas
the RhIII species have an η3-Tp′ ligand. Ultrafast studies of
Tp′Rh(CO)2 by Bergman show direct evidence for such an η3-
to η2-Tp′ conversion, and the rate is extremely fast (τ ) 200
ps).5a Such interconversions, if they occur in the present system,
are also likely to be rapid and reversible and therefore do not
have to be specifically included in a kinetic model of the
mechanism. It should be kept in mind, however, that any
hapticity changing equilibria are included in the rate constants
that are obtained.
Kinetic Simulation of Alkane Migration: Methane. The
detailed mechanism for the scrambling of the deuterium label
into the methyl group of 1-d1 is presented in Scheme 7. Five
intrinsic rate constants govern the process, with the relative rates
controlling the observed product ratios. With only three observ-
able species in the present study (1-d1, 1R-d1, and methane-d1),
an exact solution for all of the rate constants is not possible.
Several assumptions may be made, however, which simplify
the scheme and render the problem solvable: (1) all C-H bonds
at a given carbon are accessed equally20 and that exchange
between η2-C,H coordination at a given carbon of a σ-alkane
complex is sufficiently rapid that it is kinetically unimportant,
generated as an intermediate in such a mechanism would
necessitate an η3 or η1 Tp′ ligand.22
Most consistent with the observed data is the presence of
alkane complexes as intermediates in the isomerization and
migration processes. The earlier flash studies by Bergman and
Harris with Tp′Rh(CO)2 provided direct evidence for the
intermediacy of such a species.5a Once such an alkane complex
is formed, the relative rates of migration, insertion into a C-H
bond, and dissociation will control the observed product ratios.
This detailed mechanistic sequence can be applied to the present
studies.
Reductive bond formation from a secondary alkyl hydride
complex such as isopropyl hydride 3′ generates a secondary
σ-alkane complex. Migration to a primary C-H bond followed
by oxidative bond cleavage of the primary C-H bond, and
dissociation of the alkane to generate the 16e- fragment which
activates C6D6 solvent, are competitive processes, as observed
by growth of the n-propyl hydride 3 and Tp′Rh(L)(C6D5)D
(Scheme 4). Lack of insertion into a secondary C-H bond
implies that the barrier for a secondary C-H bond activation
must be higher than that for alkane dissociation and/or migration
from a secondary position to a primary position in an alkane
σ-complex.
σ-Alkane complexes also account for the scrambling of
deuterium label into the R- and terminal-methyl positions in
the normal alkyl deuteride complexes such as 3-d1 (Scheme 5).
The intermediacy of alkane complexes also explains the decrease
in label migration to the terminal end of linear alkyl hydride
complexes with increasing chain length. With each additional
methylene unit the percent migration decreases as there are more
(2) the isotope effect on reductive bond formation to generate
(22) η1 Tp ligands are extremely rare. For an example with Rh see:
Paneque, M.; Sirol, S.; Trujillo, M.; Gutierrez-Puebla, E.; Monge, A.;
Carmona, E. Angew. Chem., Int. Ed. 2000, 39, 218.
RE
a σ-alkane complex is the same for all alkyl hydrides (kH
/
RE
kD ) 2.1), and (3) the rate constant for oxidative bond

Tp′Rh(CNneopentyl)(R)H Complexes)
J. Am. Chem. Soc., Vol. 123, No. 30, 2001 7263
Figure 4. Free energy diagram for the activation of CH3D by [Tp′Rh-
(CNR)].
Figure 3. Distribution of species for the rearrangement of 1-d1 in C6D6
at 26 °C. Symbols represent observed data. Solid lines represent
simulated data.
Scheme 7
Table 2. Summary of Simulated Rate Constants for Isotopic
Scrambling of 1-d1 in C6D6 at 26 °Ca
rate constant
k, s-1
H
kRE
kRE
4.18 (9) × 10-4
1.99 (4) × 10-4
2.15 × 106
8.0 (2) × 105
D
H
kOA
D
kOA
4.3 (4) × 105
kd
a Rate constants that were allowed to vary independently are indicated
in bold. Errors are indicated in parentheses at the 95% confidence level
(2σ).
Scheme 8
cleavage to a C-H bond in an alkane complex is similar to
that observed by Bergman and Harris in their study of oxidative
addition by the [Tp′Rh(CO)] fragment (kOA ) 2.15 × 106
H
s-1).5a It will be shown that this latter assumption is necessary
to numerically simulate the observed data, but that the actual
Value assumed for this rate constant is irrelevant to the
selectiVities available to the σ-alkane complex.
Simulations were carried out using Frieden’s KINSIM/
FITSIM program, which incorporates Gear’s method for the
solution of “stiff” differential equations and permits optimization
of rate constants for a given mechanism.23 Figure 3 shows the
agreement between the observed and simulated results for 1-d1.
Summarized in Table 2 are the results of the kinetic simulation
performed using the assumptions presented above. While the
absolute values of the fast rate constants are dependent upon
the value used in assumption 3 above, the relatiVe values remain
in good agreement with previous estimates of the strength of
the alkane-metal interaction (∼10 kcal/mol), established in
studies of photochemically generated metal fragments in alkane/
rare gas matrices.11c The value for Keq obtained for the observed
equilibration of 1-d1 and 1R-d1 based on the results of the
simulation is 7.1, in good agreement with the observed value
of 6.3 near the end of the experiment.18 The value for the kinetic
isotope effect on oxidative bond cleavage (kOAH/kOAD) obtained
from the simulated values is 5.0, only slightly higher than the
isotope effect measured in the photochemical oxidative addition
of CH2D2 by 7 of 4.3. The good agreement of the simulated
values with the directly measured values provides support for
the validity of the assumptions concerning isotope effects made
in the simulation model.
Kinetic Simulation of Alkane Migration: Ethane. Similar
treatment of scrambling and migration data for ethyl deuteride
complex 2-d1 yields rates of oxidative bond cleavage, reductive
bond formation, and dissociation as well as the rate of primary-
to-primary migration (Scheme 8). The value for the rate constant
for oxidative bond cleavage of C-D bonds (kOAD) obtained in
the simulation of scrambling in 1-d1 was used for simulation
of 2-d1 and held constant (i.e., the isotope effect for oxidative
bond cleavage was maintained at kOAH/kOAD ) 4.97). With 2-d1,
H
unchanged when kOA was varied over 6 orders of magnitude.
The free energy diagram for the oxidative addition of CH3D
by [Tp′RhL] is presented in Figure 4. In this diagram, x
represents the assumed barrier for oxidative bond cleavage of
a C-H bond in an alkane complex, kOAH, and cannot be
determined from the current experiments. In essence, the free
energy of the (n2-CH3D) σ-complex cannot be determined. The
H
value for kOA measured by Bergman and Harris, 2.15 × 106
s-1, corresponds to a value for x of 8.2 kcal/mol. The simulation
provides the relative rates for the processes available to the
methane σ-complex. The relative rates of C-H oxidative bond
cleavage to dissociation to C-D oxidative bond cleavage are
15:2:1, and include the statistical factors appropriate for CH3D.
For CH4 itself, the rate of oxidative bond cleavage to dissociation
can be predicted to be 11:1 (4kOAH/kd).
The barrier to dissociation of methane to generate [Tp′Rh-
(L)] from the methane σ-complex of 9.4 kcal/mol (1.2 + x) is
(23) Prof. Carl Frieden, Department of Biological Chemistry, Box 8231,
Washington University School of Medicine, St. Louis, MO 63110. Available
on the web; search KINSIM.

7264 J. Am. Chem. Soc., Vol. 123, No. 30, 2001
Northcutt et al.
Scheme 9
Figure 5. Distribution of species for the rearrangement of 2-d1 in C6H6
at 26 °C. Symbols represent observed data. Solid lines represent
simulated data.
Scheme 10
Figure 6. Free energy diagram for the activation of CH3CH2D by
[Tp′Rh(CNR)].
Table 3. Summary of Simulated Rate Constants for Isotopic
Scrambling of 2-d1 in C6D6 at 26 °Ca
rate constant
k, s-1
H
kRE
kRE
6.58 (9) × 10-4
3.13 (9) × 10-4
2.15 × 106
D
H
present in the rearrangement of the ethyl deuteride 2-d1 (Figure
5) than was observed in the reaction of 1-d1 (Figure 3) due to
the increased relative rate of dissociation and the presence of
another pathway, namely migration to the “â” position of ethane.
Similar migratory behavior was observed for Cp*Rh(13CH2-
CH3)D by Bergman and Periana.7b
kOA
D
kOA
4.33 × 105
km11
kd
3.0 (5) × 106
1.55 (9) × 106
a Rate constants that were allowed to vary independently are indicated
in bold. Errors are indicated in parentheses at the 95% confidence level
(2σ).
Kinetic Simulation of Alkane Migration: Propane. Kinetic
simulation of deuterium scrambling and migration in propyl
deuteride complex 3-d1 proved to be too complex to solve
uniquely. Several new rate constants are introducedsprimary-
to-secondary migration (km12), secondary-to-primary migration
(km21), and dissociation from a secondary center (kd2)swith no
additional observed data compared to 2-d1. The proposed
mechanism for rearrangement is presented in Scheme 9. The
irreversible nature of the reductive elimination of propane from
isopropyl deuteride complex 3′-d1 yields a mechanism for
rearrangement identical to that for 3-d1 (Scheme 10), apart from
statistical factors, beginning at the isopropyl isomer 3′-d1 rather
than the n-propyl isomer. While this additional data might appear
to provide the necessary observations to permit a simulation of
this kinetic system, the problem still arises that the free energy
of the secondary σ-propane complex cannot be determined from
these experiments. Consequently, the energy of the secondary
σ-propane complex was assumed to be the same as that of the
an additional rate constant appears for migration of the metal
from the primary C-H bond on one methyl group to the other
(km11). The comparison of simulated and observed data is shown
in Figure 5, and rate constants are summarized in Table 3 and
presented as a free energy diagram in Figure 6.
Relative rates of oxidative bond cleavage, migration, and
H
dissociation of C2H5D are as follows: 2kOA > km11 > kd >
D
kOA (10:7:4:1). Oxidative bond cleavage of a C-D bond is
the slowest process involved, as was the case for scrambling in
1-d1. Like methane, oxidative bond cleavage of a methyl C-H
bond in C2H6 is faster than ethane dissociation but only by a
factor of 4 instead of 11 (∆∆Gq is only 0.9 kcal/mol).
Consequently only ∼23% R-scrambling is seen prior to ethane
loss compared with ∼50% in methane. Migration from primary
C-H bond to primary C-H bond occurs at approximately twice
the rate of dissociation. The effects of the relative rates on
product ratios are easily identifiable in the distribution of species
plots, with much less of the R-scrambled product (∼23%)

Tp′Rh(CNneopentyl)(R)H Complexes)
J. Am. Chem. Soc., Vol. 123, No. 30, 2001 7265
Table 4. Summary of Simulated Rate Constants for Isotopic
Scrambling of 3-d1 and 3′-d1 in C6D6 at 26 °Ca
rate constant
k, s-1
H
kRE
kRE
1.31 (2) × 10-3
6.2 (2) × 10-4
2.15 × 106
D
H
kOA
kOA
4.33 × 105
D
km12
km21
kd1
4.3 (9) × 106
4.3 × 106
1.26 (47) × 106
3.9 (14) × 106
8.1 (5) × 10-4
1.79 (14) × 10-3
kd2
kRE2
D
kCl
a Rate constants that were allowed to vary independently are indicated
in bold. Errors are indicated in parentheses at the 95% confidence level
(2σ).
Figure 7. Distribution of species for the rearrangement of 3-d1 in C6H6
at 26 °C. Symbols represent observed data. Solid lines represent
simulated data.
Figure 9. Free energy diagram for the activation of CH3CH2CH2D by
[Tp′Rh(CNR)]. Barriers to oxidative addition/reductive elimination from
the secondary position are omitted for clarity.
σ-alkane complex, oxidative bond cleavage and migration to
the interior of the chain occur at about the same rate; dissociation
is ∼4-5 times slower, as in ethane.
Figure 8. Distribution of species for the rearrangement of 3′-d1 in
C6H6 at 26 °C. Symbols represent observed data. Solid lines represent
simulated data.
These results are consistent with the observations made in
photolytic activation of propane by 7, which showed only
formation of primary activation product 3.1m When bound to
the middle of the propane chain, dissociation and migration to
the end are much faster than activation. Hence, no secondary
activation is observed, leading to the observed kinetic selectivity
for the activation of primary C-H bonds.
primary σ-propane complex.24 While few quantitative data are
in the literature, this assumption is consistent with the photo-
acoustic data measured by Rayner et al., in which Cr(CO)5-
(hexane) vs Cr(CO)5(cyclohexane) were found to have similar
stabilities.25
A simultaneous simulation using kinetic data for both
reactions with the above constraints allows for determination
of all rate constants (5 variables, 8 observables). The comparison
of observed and fitted data is shown in Figures 7 and 8. The
rate constants obtained from the simultaneous kinetic simulation
are presented in Table 4 and presented as a free energy diagram
in Figure 9. The simulation again provides information regarding
the relative rates of the processes determining the fate of the
alkane complex. For the propyl deuteride complex, the relative
The equilibrium constant separating a primary σ-alkane
complex and a secondary σ-alkane complex cannot be estab-
lished from the rate constant ratio km12/km21, since the assumption
that the primary and secondary σ-complexes are at the same
energy amounts to an assumption that these rate constants are
equal. These experiments only tell about the relative barrier
heights seen by each σ-alkane complex separately, as indicated
in the free energy diagram in Figure 9. One cannot learn about
the absolute free energies of the alkane complexes from these
studies. Two studies that address the relative stabilities of various
alkane complexes are those using TRIR measurements of alkane
binding to [W(CO)5]11a and fast IR studies looking at competi-
tion between Kr and alkane with [Cp*Rh(CO)].26 These studies
generally showed that alkane binding was favored for longer
linear alkanes over ethane by 1-2 kcal/mol and that cyclic
alkanes bound more tightly than linear alkanes (by ∼2 kcal/
mol). The nature of the binding of the linear alkanes could not
be determined (i.e., primary vs secondary binding).
H
rate is of the following order: km21 ≈ kd2; 2kOA ≈ km12 > kd1
D
> kOA ≡ 1.1:1:10:10:3:1.
The following key points can be drawn from the simulated
data. First, when propane is bound in the middle of the alkane
as a secondary σ-alkane complex, dissociation and migration
to the end of the chain occur at nearly the same rate. Second,
when propane is bound at the end of the alkane as a primary
(24) Attempts to simultaneously simulate both the n-propyl and isopropyl
experiments showed a strong correlation coefficient relating kd2 and km21
(0.99999), consistent with the concept that the energy of the secondary
alkane complex cannot be determined from these experiments.
(25) Brown, C. E.; Ishikawa, Y.; Hackett, P. A.; Rayner, D. M. J. Am.
Chem. Soc. 1990, 112, 2530. The error limits on the alkane binding strengths
make the data of little value for determining the relative binding energies
in the present system.
Kinetic Simulation of Alkane Migration: Butane. A similar
kinetic simulation of the rearrangement of n-butyl deuteride 4-d1
(Scheme 11) and sec-butyl deuteride 4′-d1 (Scheme 12) has been
possible by varying simultaneously the same five rate constants

7266 J. Am. Chem. Soc., Vol. 123, No. 30, 2001
Northcutt et al.
Table 5. Summary of Simulated Rate Constants for Isotopic
Scheme 11
Scrambling of 4-d1 and 4′-d1 in C6H6 at 26 °Ca
rate constant
k, s-1
H
kRE
kRE
1.01 (3) × 10-3
4.82 (15) × 10-4
2.15 × 106
3.5 (8) × 106
3.5 (8) × 106
1.2 (6) × 107
1.08 (35) × 106
1.71 (49) × 106
7.67 (21) × 10-4
D
H
kOA
kOA
4.33 × 105
D
km12
km21
km22
kd1
kd2
kRE2
a Rate constants that were allowed to vary independently are indicated
in bold. Errors are indicated in parentheses at the 95% confidence level
(2σ).
Scheme 12
Figure 10. Distribution of species for the rearrangement of 4-d1 in
C6H6 at 26 °C. Symbols represent observed data. Solid lines represent
simulated data.
Figure 11. Distribution of species for the rearrangement of 4′-d1 in
C6H6 at 26 °C. Symbols represent observed data. Solid lines represent
simulated data.
2°-1° migration by ∼4:1. Dissociation of the 2° σ-alkane
complex occurs about 1.5× faster than the 1° σ-alkane complex.
Kinetic Simulation of Alkane Migration: Pentane and
Hexane. The similar isomerizations and alkane loss from pentyl
deuteride complex 5-d1 could be observed, but the small quantity
of end-to-end migration made simulation difficult. In the case
of hexyl deuteride complex 6-d1, the migration to the opposite
end of the chain could not be quantified, both because of the
small quantity of the complex formed and because of the
similarity of the chemical shift of the deuterium in 6φ-d1 and
hexane-d1. In both cases, however, R-deuterium scrambling was
seen to the same extent as in the n-propyl and n-butyl deuteride
species (Figures 12 and 13). The distribution of species in these
systems can be simulated using the rate constants obtained from
as for 3-d1 and 3′-d1. In addition, the new rate constant for
migration from a secondary alkane σ-complex to a secondary
alkane σ-complex km22 was also varied, thereby giving the
appropriate ratio of the two isomers 4â-d1 and 4γ-d1. The
resulting rate constants are listed in Table 5, and the fits are
shown in Figures 10 and 11. For the primary alkane σ-complex
of butane, oxidative bond cleavage is favored over dissociation
by ∼6:1 and over 1°-2° migration by ∼2:1. For the secondary
alkane σ-complex of butane, 2°-2° migration is favored over
(26) McNamara, B. K.; Yeston, J. S.; Bergman, R. G.; Moore, C. B. J.
Am. Chem. Soc. 1999, 121, 6437.

Tp′Rh(CNneopentyl)(R)H Complexes)
J. Am. Chem. Soc., Vol. 123, No. 30, 2001 7267
Third, to solve the ethyl deuteride scrambling simulation it
is also assumed that the isotope effect on oxidative bond
cleavage is the same in the formation of an n-alkyl hydride
complex as it is in the formation of the methyl hydride complex
D
(kOAH/kOA ) 5.0). This is a reasonable assumption in that in
either case the factors that make reaction of one isotope favored
over another (zero point energies, bending energies) are similar
in this series of compounds. The independent measurement of
kOAH/kOAD with CH2D2 confirms that this assumption is reason-
able.
Finally, we assumed in our treatment that equilibration
between the various C-H and C-D bonds on a given carbon
center in an alkane σ-complex is rapid, and that these species
are equally accessible. In the σ-cyclopentane complex described
by Ball, the metal was believed to rapidly migrate from one
C-H to the other within the same methylene unit, although an
η2-H,H binding geometry could not be ruled out.10 In studies
of an agostic methyl group by Shapley in Os3(CH2D)(CO)11,
migration between C-H and C-D was fast on the NMR time
scale and a preference for binding C-H vs C-D was only
1.25.27 More to the point, in a recent study by Etienne, an
intramolecular agostic equilibrium isotope effect was directly
measured to be 1.2 in a niobium ethyl derivative.28 In the present
system, a similar slight preference for binding C-H vs C-D
bonds is also likely to be present. This equilibrium effect
between η2-C,H and η2-C,D alkane σ-complexes is of small
(but probably real) importance in the kinetic treatment used here.
One alternate way to view this effect is to say that all of the
Figure 12. Distribution of species for the rearrangement of 5-d1 in
C6H6 at 26 °C. Symbols represent observed data. Solid lines represent
simulated data.
D
kOA values obtained here include in them an equilibrium
contribution to the rate constant, estimated to contribute ∼25%
of the kinetic isotope effect. In the absence of relevant data
here, and considering the small magnitude of this correction,
the present treatment using the assumption of equal access to
C-H and C-D bonds was adopted. One contrary report
suggests that intermolecular equilibrium isotope effects favor
binding of C-D bonds, but in contrast to the present system
the alkane in this example was weakly bound and freely
exchanged with other alkanes in the krypton solvent prior to
oxidative bond cleavage.3
Figure 13. Distribution of species for the rearrangement of 6-d1 in
C6H6 at 26 °C. Symbols represent observed data. Solid lines represent
simulated data.
D
the isomerization of butyl deuterides and varying only kRE
,
since no new fundamental rate constants are now needed. The
agreement shown in the figures is quite good, again indicating
the validity of the kinetic model.
It is also worthwhile to compare the quantitative results
obtained here with the qualitative results obtained by Flood on
the deuterium scrambling in [CnRh(L)RD]+ complexes (L )
P(OMe)3, Cn ) 1,4,7-triazacyclononane).14 In his case, butyl,
hexyl, and even decyl groups were observed to undergo
complete equilibration of the deuterium into the R-position at
5 °C. The observed (statistically corrected) equilibrium constant
was 1.7, favoring deuterium on carbon, and is exactly what was
observed in the present system with CH2D2 as the substrate (eq
8). Furthermore, upon heating to 40 °C Flood observed
subtantial quantities of deuterium migration into the terminal
methyl group of even the n-decyl ligand, indicating that the
metal is capable of migrating down a 10-carbon chain without
dissociating. Clearly, the [CnRh(L)]+ fragment binds alkanes
more tightly than [Tp′Rh(L)], perhaps due to its positive charge.
Several assumptions have been made in these detailed kinetic
simulations, and further discussion is warranted. First, since the
σ-alkane complexes are not directly observed as intermediates
an assumption must be made regarding the stability or free
energy of these species. The direct observation by Bergman and
Harris of the conversion of an alkane σ-complex to an alkyl
hydride in the very similar fragment Tp′Rh(CO) offers an
excellent estimate for this value. Also, as mentioned earlier,
this value ties to the absolute values of the “fast” processes
(kOAH, kOAD, km, kd), but does not affect their relatiVe values.
As the latter are what are needed to determine selectivities, the
conclusions drawn here are independent of the value used for
H
kOA
!
Second, in dealing with the problems presented by the isotope
effects introduced as a result of using deuterium to monitor the
rearrangements of the complexes, we used the reductive
elimination rates for the isopropyl hydride vs the isopropyl
Conclusions
D
deuteride complex to obtain kREH/kRE ) 2.1 for the isotope
The isotopic scrambling and rearrangement of complexes of
the type Tp′Rh(CNR)(R)(H/D) has been shown to proceed
intramolecularly with alkane σ-complexes as intermediates.
Rearrangement of secondary alkyl hydrides to their more stable
primary alkyl hydride isomers was observed. Scrambling of
effect on alkane reductive bond formation to give an alkane
σ-complex. While the absolute rate constants for reductive bond
formation from primary alkyl hydride and alkyl deuteride
complexes are likely to be different, it is reasonable to assume
that the ratio of these rates (i.e. the isotope effect) is of similar
size. It is this assumption that allows one to solve the methyl
deuteride simulation.
(27) Calvert, R. B.; Shapley, J. R. J. Am. Chem. Soc. 1978, 100, 7726.
(28) Jaffart, J.; Mathieu, R.; Etienne, M.; McGrady, J. E.; Eisenstein,
O.; Maseras, F. Chem. Commun. 1998, 2011.

7268 J. Am. Chem. Soc., Vol. 123, No. 30, 2001
Northcutt et al.
stirring over H2SO4, washing with aqueous KMnO4 and water, and
distilling from dark purple solutions of tetraglyme/benzophenone ketyl.
Tetrahydrofuran-d8, benzene-d6, and methane-d2 were purchased from
Cambridge Isotope Lab. The liquids were distilled under vacuum from
dark purple solutions of benzophenone ketyl and stored in ampules
with Teflon sealed vacuum line adaptors. The preparations of KTp′,29
Tp′Rh(L)Cl2,16a Tp′Rh(L)(Me)Cl,16a Tp′Rh(L)(n-Pr)Cl,16a Tp′Rh(L)-
(i-Pr)Cl,16a Tp′Rh(L)(n-Pr)H1m, Tp′Rh(L)(n-pentyl)Cl1m, Tp′Rh(L)-
(n-pentyl)H,1m Tp′Rh(L)(η2-PhNdCdNCH2C(CH3)3),30 and neopen-
tylisocyanide (L)31 have been previously reported. We have additionally
noted that in the preparation of Tp′Rh(CNneopentyl)Cl2, the fac-isomer
of RhCl3(CH3CN)3 must be used. Caution must be taken during the
preparation of this intermediate from RhCl3‚3H2O (do not heat) or the
unreactive mer-isomer can be formed. The zirconium reagents [Cp2-
ZrH2]2 and [Cp2ZrD2]2 were prepared according to literature methods.32
All Grignards with the exception of n-pentylmagnesium chloride and
n-hexylmagnesium chloride were purchased from Aldrich Chemical
Co. Hexafluorobenzene (C6F6) was purchased from the Aldrich
Chemical Co, dried over 4 Å Linde molecular sieves for 24 h, degassed
using three freeze-pump-thaw cycles, and stored in the dri-lab.
All NMR spectra were recorded on a Bruker AMX400 (1H, 2H, 11B,
13C). All 1H and 2H chemical shifts are reported in ppm (δ) relative to
tetramethylsilane and referenced using chemical shifts of residual
solvent resonances (benzene-d6, δ 7.15). 11B NMR spectra were
referenced to external BF3‚OEt2 (δ 0.0). All temperatures for kinetic
experiments were calibrated relative to the chemical shift differences
in the NMR spectra of neat ethylene glycol. Accurate temperature
control ((0.1 °C) was maintained using a Bruker BVT-3000 VT control
unit. Resealable NMR tubes with Teflon valves were purchased from
the Brunfeldt Co. Infrared spectra were recorded on a Mattson
Instruments 6020 Galaxy Series FTIR and processed with First:Acquire
v1.52 software. Analyses were obtained from Desert Analytics. A
Siemens SMART system with a CCD area detector was used for X-ray
structure determination.
Preparation of Tp′Rh(L)(Et)Cl, 2-Cl. Tp′Rh(L)Cl2 (155 mg, 0.27
mmol) was placed in a 50 mL round-bottom flask with a Teflon coated
stir bar. Addition of 25 mL of THF resulted in a yellow orange solution.
Ethylmagnesium chloride (0.20 mL, 0.40 mmol, 1.5 equiv) was added
dropwise to the stirring solution over 5 min and the resulting brown
solution stirred for 15 min. Excess Grignard reagent was quenched with
a saturated solution of NH4Cl(aq) until all had reacted to give a clear
solution. The pale yellow solution was filtered through glass wool and
solvent removed under vacuum. Purification was by preparative TLC
using hexanes-THF (4:1, V/V) as the mobile phase. Recrystallization
at -20 °C from a CH2Cl2 solution layered with hexanes yielded 97
mg (0.17 mmol, 64%) of bright yellow X-ray quality crystals. NMR
data for 2-Cl: 1H NMR (C6D6) δ 5.712 (s, 1 H, pzH), 5.604 (s, 1 H,
pzH), 5.566 (s, 1H, pzH), 3.613 (m, CHHCH3, 1 H), 3.385 (m,
CHHCH3, 1 H), 2.981 (s, 3H, pzCH3), 2.775 (s, 3H, pzCH3), 2.656
(ABq, 2H, NCH2), 2.343 (s, 3H, pzCH3), 2.230 (s, 3H, pzCH3), 2.154
(s, 3H, pzCH3), 2.095 (s, 3H, pzCH3), 1.355 (t, JHH ) 7.6 Hz, 3H,
CH2CH3), 0.716 (s, 9H, CH2C(CH3)3). 13C{1H} NMR (C6D6) δ 155.70,
153.76, 153.35, 146.93, 145.41, 145.17 (s, pzCq), 111.15, 110.33, 109.26
(s, pzCH), 58.79 (s, NCH2), 34.54 (s, C(CH3)3), 29.23 (s, C(CH3)3),
21.48 (s, RhCH2CH3), 17.25, 15.63, 15.39, 14.92 (s, pzCH3), 13.98 (d,
JRhH ) 18.7 Hz, RhCH2CH3). IR (THF, cm-1) 2521 (w, B-H), 2203
(s, CNR). Anal. Calcd for C23H38BClN7Rh: C, 49.18; H, 6.82; N, 17.45.
Found: C, 48.54; H, 6.80; N, 17.46.
Figure 14. Bar graph showing relative rates of reactions of σ-alkane
complexes. For propane and butane, the processes for the primary and
secondary σ-alkane complexes are separated.
deuterium label into both the R-position and terminal methyl
group was observed in linear alkyl deuteride complexes 2-d1 to
6-d1. Relative rates of dissociation, migration, and bond activa-
tion from alkane complexes have been elucidated through kinetic
simulation, allowing several general conclusions to be drawn.
For methane, C-H bond activation is strongly favored over
dissociation of methane from the metal-methane σ-complex.
For ethane-hexane, activation of the terminal C-H bond of
the alkane is favored over both dissociation and migration to
the â position of the σ-alkane complex.
Propane and the higher alkanes offer an opportunity to learn
about a number of competitive processes available to alkane
complexes, as alluded to in the discussion of eqs 3-5. The
relative rates of these processes available to alkane complexes
of methane, ethane, propane, and butane are shown schemati-
cally in Figure 14. Due to the ability to obtain only relative
rates for a given alkane complex, these rates cannot be compared
from one substrate to another, or even from one alkane
σ-complex to another.
The following general conclusions can be made about these
processes for the fragment [Tp′Rh(CNR)] interacting with a
linear alkane: (1) C-H bond activation is favored over
dissociation of alkane in a primary alkane σ-complex by a factor
of ∼five, (2) primary-to-secondary and secondary-to-primary
migrations are faster than dissociation from their respective
alkane σ-complexes by a factor of 2-3 while secondary-to-
secondary migration is about 7× faster than dissociation, (3)
C-H activation is the fastest process in primary alkane
complexes, but the slowest in secondary alkane complexes (it
is not even observed in the latter case, unless no other type of
bond is available (e.g. cycloalkanes)), and (4) in a secondary
alkane σ-complex, secondary-to-secondary migration is about
3× faster than secondary-to-primary migration, giving rise to
the effect of rapid migration down a linear chain.
Preparation of Tp′Rh(L)(n-Bu)Cl, 4-Cl. Synthesis of 4-Cl was
identical to that of 2-Cl except that 190 mg of Tp′Rh(L)Cl2 (0.36 mmol)
and 0.27 mL of 2.0 M n-butylmagnesium chloride (0.54 mmol, 1.5
equiv) were used. Workup and recrystallization of 4-Cl were identical
to that of 2-Cl. Yield: 86 mg (0.146 mmol, 41%). NMR data for 4:
1H NMR (C6D6) δ 5.699 (s, 1 H, pzH), 5.608 (s, 1 H, pzH), 5.556 (s,
1H, pzH), 3.442 (m, CHHCH2CH2CH3, 1 H), 3.290 (m, CHHCH2CH2-
CH3, 1 H), 2.969 (s, 3H, pzCH3), 2.776 (s, 3H, pzCH3), 2.642 (ABq,
Inverse isotope effects on alkane loss from an alkyl hydride
vs an alkyl deuteride complex have been shown to be the result
of an inverse preequilibrium isotope effect, produced by two
different but normal isotope effects in the reductive bond
formation/oxidative bond cleavage preequilibrium steps.
Experimental Section
General Considerations. All manipulations were performed under
an N2 atmosphere, either on a high-vacuum line using modified Schlenk
techniques or in a Vacuum Atmospheres Corporation Glovebox.
Tetrahydrofuran and toluene were distilled from dark purple solutions
of benzophenone ketyl. Alkane solvents were made olefin-free by
(29) Trofimenko, S. Inorg. Synth. 1970, 12, 99.
(30) Hessel, E. T.; Jones, W. D. J. Am. Chem. Soc. 1992, 114, 6087.
(31) Schuster, R. E. Organic Synthesis; Wiley: New York, 1973; Collect.
Vol. V, p 772.
(32) Shriver, D. F. Inorg. Synth. 1979, 19, 223.

Tp′Rh(CNneopentyl)(R)H Complexes)
J. Am. Chem. Soc., Vol. 123, No. 30, 2001 7269
2H, NCH2), 2.360 (s, 3H, pzCH3), 2.220 (s, 3H, pzCH3), 2.148 (s, 3H,
pzCH3), 2.086 (s, 3H, pzCH3), 1.987 (m, CH2CHHCH2CH3, 1 H), 1.762
(m, CH2CHHCH2CH3, 1 H), 1.690 (m, CH2CH2CH2CH3, 2 H), 1.041
(t, JHH ) 7.2 Hz, 3H, RhCH2CH2CH2CH3), 0.709 (s, 9H, CH2C(CH3)3).
13C{1H} NMR (C6D6) δ 153.13, 151.17, 150.83, 144.38, 142.84, 142.59
(s, pzCq), 108.58, 107.81, 106.70 (s, pzCH), 56.27 (s, NCH2), 36.61
(s, RhCH2CH2CH2CH3), 32.01 (s, C(CH3)3), 26.68 (s, C(CH3)3), 26.18
(s, RhCH2CH2CH2CH3), 18.37 (d, JRhH ) 19.0 Hz, RhCH2CH2CH2-
CH3), 14.72, (s, pzCH3), 14.69 (s, RhCH2CH2CH2CH3), 14.52, 13.05,
12.83, 12.36 (s, pzCH3). IR (THF, cm-1) 2524 (w, B-H), 2205 (s,
CNR). Anal. Calcd for C25H42BClN7‚CH2Cl2: C, 46.3; H, 6.57; N,
14.53. Found: C, 45.21; H, 6.37; N, 13.62.
Preparation of Tp′Rh(L)(n-pentyl)Cl, 5-Cl. Synthesis of 5-Cl was
identical to that of 2-Cl except that 238 mg of Tp′Rh(L)Cl2 (0.41 mmol)
and 1.50 mL of 0.30 M n-hexylmagnesium chloride (0.450 mmol, 1.1
equiv) were used. Workup and recrystallization of 5-Cl were identical
to that of 2-Cl. Yield: 64.7 mg (0.107 mmol, 26.1%).
Preparation of Tp′Rh(L)(n-hexyl)Cl, 6-Cl. Synthesis of 6-Cl was
identical to that of 2-Cl except that 274 mg of Tp′Rh(L)Cl2 (0.48 mmol)
and 1.50 mL of 0.37 M n-hexylmagnesium chloride (0.555 mmol, 1.2
equiv) were used. Workup and recrystallization of 6-Cl were identical
to that of 2-Cl. Yield: 104 mg (0.168 mmol, 34.7%). NMR data for
6-Cl: 1H NMR (C6D6): δ 5.699 (s, 1 H, pzH), 5.608 (s, 1 H, pzH),
5.556 (s, 1H, pzH), 3.442 (m, CHHCH2CH2CH3, 1 H), 3.290 (m,
CHHCH2CH2CH3, 1 H), 2.969 (s, 3H, pzCH3), 2.776 (s, 3H, pzCH3),
2.642 (ABq, 2H, NCH2), 2.360 (s, 3H, pzCH3), 2.220 (s, 3H, pzCH3),
2.148 (s, 3H, pzCH3), 2.086 (s, 3H, pzCH3), 1.987 (m, CH2CHHCH2-
CH3, 1 H), 1.762 (m, CH2CHHCH2CH3, 1 H), 1.690 (m, CH2CH2CH2-
CH3, 2 H), 1.041 (t, JHH ) 7.2 Hz, 3H, RhCH2CH2CH2CH3), 0.709 (s,
9H, CH2C(CH3)3). 13C{1H} NMR (C6D6): δ 153.2, 151.2, 150.8, 144.4,
142.9, 142.6 (s, pzCq), 108.6, 107.8, 106.7 (s, pzCH), 56.3 (s, NCH2),
34.3, 33.2, 32.5 (s, RhCH2(CH2)4CH3), 32.0 (s, C(CH3)3), 26.7 (s,
C(CH3)3), 23.4 (s, RhCH2(CH2)3CH2CH3), 18.8 (d, JRhH ) 19.0 Hz,
RhCH2(CH2)4CH3), 14.7, 14.4, 13.1, 12.8, 12.3 (s, pzCH3). IR (THF,
cm-1): 2521 (w, B-H), 2205 (s, CNR). Anal. Calcd for C27H46BClN7-
Rh: C, 52.49; H, 7.50; N, 15.87. Found: C, 52.94; H, 7.59; N, 15.47.
Preparation of Tp′Rh(L)(sec-butyl)Cl, 4′-Cl. Synthesis of 4′-Cl
was identical to that of 2-Cl except that 124 mg of Tp′Rh(L)Cl2 (0.23
mmol) and 0.23 mL of 2.0 M sec-butylmagnesium chloride (0.46 mmol,
2 equiv) were used. Workup and recrystallization of 4′-Cl were identical
to that of 2-Cl. Yield: 14.4 mg (0.024 mmol, 10.5%). Recrystallization
from methanol at -20 °C allows for partial crystallization of one
stereoisomer of 4′-Cl. NMR data for 4′-Cl: 1H NMR (C6D6) δ 5.695,
5.686 (s, 1 H, pzH), 5.638 (bs, 2 H, pzH), 5.566, 5.560 (s, 1H, pzH),
4.89 (bm, 1H, RhCH(CH2CH3)CH3), 3.046, 3.031 (s, 3H, pzCH3),
2.775, 2.772 (s, 3H, pzCH3), 2.672, 2.660 (ABq, 2H, NCH2), 2.401,
2.370 (s, 3H, pzCH3), 2.216, 2.212 (s, 3H, pzCH3), 2.184, 2.172 (s,
3H, pzCH3), 2.118, 2.108 (s, 3H, pzCH3), 1.940 (bm, 4H, RhCH(CH2-
CH3)CH3), 1.049 (bt, JHH ) 8.4 Hz, 6H, RhCH(CH2CH3)CH3), 0.707,
0.694 (s, 9H, CH2C(CH3)3). 13C{1H} NMR (C6D6) δ 153.2, 151.2,
150.8, 144.4, 142.9, 142.6 (s, pzCq), 108.6, 107.8, 106.7 (s, pzCH),
56.3 (s, NCH2), 34.3, 33.2, 32.5 (s, RhCH2(CH2)4CH3), 32.0 (s,
C(CH3)3), 26.7 (s, C(CH3)3), 23.4 (s, RhCH2(CH2)3CH2CH3), 18.8 (d,
JRhH ) 19.0 Hz, RhCH2(CH2)4CH3), 14.7, 14.4, 13.1, 12.8, 12.3 (s,
pzCH3). IR (THF, cm-1) 2521 (w, B-H), 2205 (s, CNR).
(m, 2H, RhCH2CH2CH2CH3), 1.669 (m, 2H, RhCH2CH2CH2CH3), 1.107
(t, JHH ) 7.6 Hz, 3H, RhCH2CH2CH2CH3), 0.658 (s, 9H, CH2C(CH3)3),
-14.993 (d, JRhH ) 24.8 Hz, 1H RhH). For 6: 1H NMR (C6D6) δ
5.808 (s, 1 H, pzH), 5.637 (s, 1 H, pzH), 5.620 (s, 1H, pzH), 2.652
(ABq, 2H, NCH2), 2.576 (s, 3H, pzCH3), 2.564 (s, 3H, pzCH3), 2.360
(s, 3H, pzCH3), 2.274 (s, 3H, pzCH3), 2.194 (s, 3H, pzCH3), 2.172 (s,
3H, pzCH3), 1.927 (m, 2H, RhCH2CH2CH2CH2CH2CH3), 1.636 (m,
2H, RhCH2CH2CH2CH2CH2CH3), 1.466 (m, 2H, RhCH2CH2CH2CH2-
CH2CH3), 1.384 (m, 2H, RhCH2CH2CH2CH2CH2CH3), 0.904 (t, JHH
) 7.6 Hz, 3H, RhCH2CH2CH2CH2CH2CH3), 0.648 (s, 9H, CH2C-
(CH3)3), -14.928 (d, JRhH ) 24.8 Hz, 1H RhH). For 4′: 1H NMR (C6D6)
δ 5.818 (s, 2 H, pzH), 5.636 (s, 2 H, pzH), 5.564 (s, 1H, pzH), 5.558
(s, 1H, pzH), 2.694 (ABq, 4H, NCH2), 2.603 (s, 3H, pzCH3), 2.597 (s,
3H, pzCH3), 2.438 (s, 6H, pzCH3), 2.363 (s, 3H, pzCH3), 2.320 (s,
3H, pzCH3), 2.302 (s, 6H, pzCH3), 2.248 (s, 6H, pzCH3), 2.194 (s,
6H, pzCH3), 1.216 (t, JHH ) 6.8 Hz, 3H, CH(CH3)CH2CH3), 1.190 (t,
JHH ) 7.2 Hz, 3H, CH(CH3)CH2CH3), 0.658 (s, 9H, CH2C(CH3)3),
-15.294 (d, JRhH ) 24.8 Hz, 1H RhH), -15.302 (d, JRhH ) 25.2 Hz,
1H RhH).
Kinetics of Elimination of Alkane from Tp′Rh(L)(R)H. Genera-
tion of rhodium alkyl hydride complexes was accomplished by the
general method outlined above, with 1 µL of dimethoxyethane (DME)
added as an internal standard. The reductive elimination of alkane from
1
1-6 was monitored by H NMR spectroscopy in C6D6 at 26 °C. For
kinetic analysis neopentyl isocyanide resonances for both Tp′Rh(L)-
(C6D5)D and alkyl hydride were integrated relative to DME at regular
intervals. Data analysis was carried out using Microsoft Excel (see
Supporting Information).
Preparation of Tp′Rh(L)(R)D. To a resealable NMR tube was
added 8.0 mg of ethyl derivative 2-Cl (0.014 mmol) and 22.3 mg (0.049
mmol) of [Cp2ZrD2]2. Benzene (0.55 mL) and benzene-d6 (1 µL) were
added via syringe. The reaction mixture was shaken vigorously for 1
min and placed into a pre-shimmed probe. The reaction is general and
was used to form the isotopomers 1-d1 to 4′-d1. For 2-d1: {1H}2H NMR
(C6H6) δ 2.528 (RhCHDCH3), 2.180 (RhCH2CH2D), -14.825. 3-d1:
{1H}2H NMR (C6H6) δ 2.339, 2.048 (RhCHDCH2CH3), 1.313 (RhCH2-
CH2CH2D), -14.791 (RhD). 4-d1: {1H}2H NMR (C6H6) δ 2.438, 2.082
(RhCHD(CH2)2CH3), 1.937 (RhCH2CHDCH2CH3), 1.656 (Rh(CH2)2-
CHDCH3), 1.126 (RhCH2(CH2)2CH2D), -14.782 (RhD). 5-d1: {1H}2H
NMR (C6H6) δ 2.366, 2.056 (RhCHD(CH2)3CH3), 1.008 (RhCH2(CH2)3-
CH2D), -14.810 (RhD). 6-d1: {1H}2H NMR (C6H6) δ 2.332, 2.029
(RhCHD(CH2)4CH3), 14.829 (RhD). 3′-d1: {1H}2H NMR (C6H6) δ
-15.122 (RhD). 4′-d1: {1H}2H NMR (C6H6) δ -15.218 (RhD).
Kinetics of Isomerization of Tp′Rh(L)(R)D. Generation of rhodium
alkyl deuteride complexes was accomplished by the general method
outlined above, with 1 µL of benzene-d6 added as an internal standard.
Isomerization and migration of the deuterium label in 1-4′ was
monitored by {1H}2H NMR spectroscopy in C6H6 at 26 °C. For kinetic
analysis all signals were integrated relative to C6D6 at regular intervals
(see Figure 2). The relative amounts of each species were measured
by calculating the ratio of the integral value of one species to the sum
of the integral value for all species (all species contain one deuterium).
Kinetic simulation was carried out using KINSIM/FITSIM (see
Supporting Information). Agreement is generally good, with the largest
deviations being observed for the lighter hydrocarbons (methane, ethane,
propane) at longer reaction times due to their volatility.
Preparation of Tp′Rh(L)(R)H. To a resealable NMR tube was
added 7.8 mg of ethyl derivative 2-Cl (0.014 mmol) and 12.3 mg (0.027
mmol) of [Cp2ZrH2]2. Benzene-d6 (0.55 mL) was added via syringe.
The reaction mixture was shaken vigorously for 1 min, with complete
Determination of Isotope Effects on Oxidative Addition. Tp′Rh-
(L)(η2-PhNdCdNCH2C(CH3)3) (6 mg, 0.009 mmol) was placed in a
resealable NMR tube. C6F6 (0.6 mL) was added via syringe. The
resulting bright yellow solution was degassed three times. The vacuum
manifold was filled with 0.105 atm of CH2D2 (2.0 mmol). CH2D2 was
condensed into the NMR tube and carefully thawed. (CAUTION:
Extremely high pressure!) The sample was photolyzed for 6 min at 6
°C until the bright yellow solution became completely bleached. The
sample was placed in a -10 °C methanol/water bath and the pressure
released. The NMR probe (at 9 °C) was shimmed with a C6D6 sample
of Tp′Rh(C6D5)D of similar concentration and solvent height. {2H}1H
NMR spectra were acquired unlocked. The methyl signals of the two
C-X bond activation products were integrated relative to the single
neopentyl isocyanide resonance for 1-d2. This ratio was used to
1
conversion to 2 by H NMR spectroscopy. The unreacted [Cp2ZrH2]2
and Cp2ZrCl2 precipitate to the bottom of the sample. The reaction is
general and was used to form hydrides 1-4′. Data for 3, 5, and 3′
were reported earlier. For 2: 1H NMR (C6D6) δ 5.825 (s, 1 H, pzH),
5.660 (s, 1 H, pzH), 5.642 (s, 1H, pzH), 2.628 (ABq, 2H, NCH2), 2.570
(s, 3H, pzCH3), 2.558 (s, 3H, pzCH3), 2.354 (s, 3H, pzCH3), 2.269 (s,
3H, pzCH3), 2.194 (s, 3H, pzCH3), 2.170 (s, 3H, pzCH3), 1.619 (t, JHH
) 7.6 Hz, 3H, CH2CH3), 0.658 (s, 9H, CH2C(CH3)3), -14.936 (d, JRhH
) 24.8 Hz, 1H RhH). For 4: 1H NMR (C6D6) δ 5.816 (s, 1 H, pzH),
5.646 (s, 1 H, pzH), 5.627 (s, 1H, pzH), 2.667 (ABq, 2H, NCH2), 2.572
(s, 3H, pzCH3), 2.561 (s, 3H, pzCH3), 2.368 (s, 3H, pzCH3), 2.267
(s, 3H, pzCH3), 2.206 (s, 3H, pzCH3), 2.183 (s, 3H, pzCH3), 1.931

7270 J. Am. Chem. Soc., Vol. 123, No. 30, 2001
Northcutt et al.
determine the kinetic isotope effect. For 1a-d2: {2H}1H NMR (C6F6)
δ 0.091 downfield from CH2D2. For 1b-d2: {2H}1H NMR (C6F6) δ
0.195 downfield from CH2D2.
Supporting Information Available: Tables of kinetic fits
and experimental details of the X-ray structural determinations
of 2-Cl and 6-Cl (PDF). This material is available free of charge
via the Internet at http://pubs.acs.org.
Acknowledgment is made to the U.S. Department of Energy,
grant FG02-86ER13569, for their support of this work.
JA003944X