Reversible beta-hydrogen elimination of three-coordinate iron(II) alkyl complexes: Mechanistic and thermodynamic studies

Article abstract of DOI:10.1021/om049415+

High-spin organometallic complexes have not received extensive mechanistic study, despite their potential importance as unsaturated intermediates in catalytic transformations. We have found that, with a suitably bulky bidentate ligand, three-coordinate, high-spin alkyl complexes of iron(II) are stable. They undergo isomerization and exchange reactions of the alkyl group through β-hydride elimination and reinsertion, and the β-hydride elimination step is rate-limiting. The alkyl complexes transfer a β-hydrogen atom to C=C, C=N, and C=O double bonds and undergo deprotonation by Bronsted acids. The reversible β-hydride elimination reactions can be used to explore relative M-C bond energies. Competition experiments and density functional calculations demonstrate an enthalpic preference for alkyl isomers with iron bound to the terminal carbon of the alkyl fragment. This preference arises from steric and electronic effects. The steric preference could be overcome with a phenyl substituent, which steers iron to the benzylic position. A Hammett correlation and density functional calculations suggest that the substituent effect is attributable to resonance stabilization of partial negative charge on the alkyl ligand.

Full text of DOI:10.1021/om049415+

5226
Organometallics 2004, 23, 5226-5239
Rever sible Beta -Hyd r ogen Elim in a tion of
Th r ee-Coor d in a te Ir on (II) Alk yl Com p lexes: Mech a n istic
a n d Th er m od yn a m ic Stu d ies
J avier Vela,† Sridhar Vaddadi,‡ Thomas R. Cundari,*,‡ J eremy M. Smith,†,§
Elizabeth A. Gregory,† Rene J . Lachicotte,† Christine J . Flaschenriem,† and
Patrick L. Holland*,†
Department of Chemistry, University of Rochester, Rochester, New York 12627, and
Department of Chemistry, University of North Texas, Denton, Texas 76203
Received J uly 28, 2004
High-spin organometallic complexes have not received extensive mechanistic study, despite
their potential importance as unsaturated intermediates in catalytic transformations. We
have found that, with a suitably bulky bidentate ligand, three-coordinate, high-spin alkyl
complexes of iron(II) are stable. They undergo isomerization and exchange reactions of the
alkyl group through â-hydride elimination and reinsertion, and the â-hydride elimination
step is rate-limiting. The alkyl complexes transfer a â-hydrogen atom to CdC, CdN, and
CdO double bonds and undergo deprotonation by Brønsted acids. The reversible â-hydride
elimination reactions can be used to explore relative M-C bond energies. Competition
experiments and density functional calculations demonstrate an enthalpic preference for
alkyl isomers with iron bound to the terminal carbon of the alkyl fragment. This preference
arises from steric and electronic effects. The steric preference could be overcome with a phenyl
substituent, which steers iron to the benzylic position. A Hammett correlation and density
functional calculations suggest that the substituent effect is attributable to resonance
stabilization of partial negative charge on the alkyl ligand.
In tr od u ction
â-Hydrogen elimination (â-HE) is a characteristic
transformation of these unsaturated alkyl complexes.
While isolable, unsaturated early metal alkyl complexes
that contain â-hydrogen atoms are common, electroni-
cally unsaturated alkyl complexes of the late metals are
often unstable toward alkene loss through irreversible
â-HE. In contrast, we recently reported isolable unsat-
urated alkyliron(II) complexes that cleanly isomerize the
Transition metal complexes with a full complement
of 18 valence electrons at the metal are the most
prevalent in organometallic chemistry. However, it is
electronically unsaturated complexes that undergo many
of the characteristic and varied reactions of organome-
tallic catalysis. At the forefront of transition metal
chemistry, synthetically useful processes including R-ole-
fin polymerization, hydrogenation, hydroformylation,
and hydroamination have soluble hydrocarbyl com-
plexes of unsaturated metals as key intermediates.
Understanding the properties of unsaturated organo-
metallic complexes is therefore crucial in understanding
and improving those processes, as well as in developing
potential applications for novel transformations such as
C-H and C-C bond activation.1,2 Hence chemists have
synthesized and studied reactive unsaturated organo-
metallic complexes with alkyl ligands.3-5
(2) Arakawa, H.; Aresta, M.; Armor, J . N.; Barteau, M. A.; Beckman,
E. J .; Bell, A. T.; Bercaw, J . E.; Creutz, C.; Dinjus, E.; Dixon, D. A.;
Domen, K.; DuBois, D. L.; Eckert, J .; Fujita, E.; Gibson, D. H.; Goddard,
W. A.; Goodman, D. W.; Keller, J .; Kubas, G. J .; Kung, H. H.; Lyons,
J . E.; Manzer, L. E.; Marks, T. J .; Morokuma, K.; Nicholas, K. M.;
Periana, R.; Que, L.; Rostrup-Nielson, J .; Sachtler, W. M. H.; Schmidt,
L. D.; Sen, A.; Somorjai, G. A.; Stair, P. C.; Stults, B. R.; Tumas, W.
Chem. Rev. 2001, 101, 953-996.
(3) Representative examples: (a) Wilkinson, G. Angew. Chem. 1974,
86, 664-667. (b) Smith, G. M.; Carpenter, J . D.; Marks, T. J . J . Am.
Chem. Soc. 1986, 108, 6805-6807. (c) Hermes, A. R.; Girolami, G. S.
Organometallics 1987, 6, 763-768. (d) Richeson, D. S.; Mitchell, J . F.;
Theopold, K. H., J . Am. Chem. Soc. 1987, 109, 5868-5670. (e) Romeo,
R. Comments Inorg. Chem. 1990, 11, 21-57. (f) Power, P. P.
Chemtracts: Inorg. Chem. 1994, 6, 181-195.
(4) (a) Brintzinger, H. H.; Fischer, D.; Muelhaupt, R.; Rieger, B.;
Waymouth, R. M. Angew. Chem., Int. Ed. Engl. 1995, 34, 1143-1170.
(b) Kersten, J . L.; Kucharczyk, R. R.; Yap, G. P. A.; Rheingold, A. L.;
Theopold, K. H. Chem. Eur. J . 1997, 3, 1668-1674. (c) Stahl, S.;
Labinger, J . A.; Bercaw, J . E. Angew. Chem., Int. Ed. 1998, 37, 2181-
2192. (d) Beswick, C. L.; Marks, T. J . Organometallics 1999, 18, 2410-
2412. (e) Pariya, C.; Theopold, K. H. Curr. Sci. 2000, 78, 1145-1131.
(5) (a) Urtel, H.; Meier, C.; Eisentrager, F.; Rominger, F.; J oschek,
J . P.; Hofmann, P. Angew. Chem., Int. Ed. 2001, 40, 781-784. (b)
J ohansson, L.; Tilset, M. J . Am. Chem. Soc. 2001, 123, 739-740. (c)
MacAdams, L. A.; Kim, W.-K.; Liable-Sands, L. M.; Guzei, I. A.;
Rheingold, A. L.; Theopold, K. H. Organometallics 2002, 21, 952-960.
(d) Stambuli, J . P.; Buehl, M.; Hartwig, J . F. J . Am. Chem. Soc. 2002,
124, 9346-9347. (e) Yamashita, M.; Hartwig, Y. F. J . Am. Chem. Soc.
2004, 126, 5344-5345. (f) Vedernikov, A. N.; Caulton, K. G. Angew.
Chem., Int. Ed. 2002, 41, 4102-4104.
* To whom correspondence should be addressed. E-mail: holland@
chem.rochester.edu.
† University of Rochester.
‡ University of North Texas.
§ Current address: Department of Chemistry & Biochemistry, New
Mexico State University, Las Cruces, NM 88003.
(1) (a) Collman, J . P.; Hegedus, L. S.; Norton, J . R.; Finke, R. G.
Principles and Applications of Organotransition Metal Chemistry;
University Science Books: Sausalito, CA, 1987. (b) Spessard, G. O.;
Miessler, G. L. Organometallic Chemistry; Prentice Hall: New J ersey,
1997. (c) Crabtree, R. H. The Organometallic Chemistry of the Transi-
tion Metals, 2nd ed.; Wiley: New York, 1994; pp 53, 54, and 188-190.
(d) Bullock, R. M. Isotope Effects in Reactions of Transition Metal
Hydrides. In Transition Metal Hydrides; Dedieu, A., Ed.; VCH: New
York, 1991; pp 278-283.
10.1021/om049415+ CCC: $27.50 © 2004 American Chemical Society
Publication on Web 10/01/2004

Reversible Beta-Hydrogen Elimination
Organometallics, Vol. 23, No. 22, 2004 5227
Sch em e 1
F igu r e 1. Diketiminate ligands used in this work.
Ch a r t 1
alkyl complexes of iron(II) (Chart 1). The synthesis of
some of these compounds has been reported previ-
ously.6,8 Alkyl complexes were derived from the addition
of Grignard or alkyllithium reagents to LMeFe(µ-Cl)2Li-
(THF)2 (1a ) or LtBuFeCl (1b).7g For example, the tert-
butyl complex LtBuFetBu (2b ) was prepared from
tBuMgCl and LtBuFeCl (Scheme 1a).6 This complex was
fully characterized and its molecular structure deter-
mined by X-ray crystallography. Upon heating, this
tertiary complex undergoes complete thermal rear-
rangement to its isobutyl isomer LtBuFeiBu (3b).6 When
the analogous complex with the smaller ligand LMe was
used (Figure 1), the isomerization was too rapid to
observe the presumed formation of the tertiary isomer
LMeFetBu by 1H NMR spectroscopy, and the only
isolated product was isobutyl complex LMeFeiBu (3a ).
alkyl group, probably through reversible â-HE.6 These
complexes contain a monoanionic, bulky diketiminate
ligand (Figure 1: L ) 2,4-bis(2,6-diisopropylphenylimi-
no)pent-3-yl, LMe, series a ; or 2,2,6,6-tetramethyl-3,5-
bis(2,6-diisopropylphenylimino)hept-4-yl, LtBu, series b),
which maintains d6 iron(II) in a three-coordinate geom-
etry with a formal valence electron count of only 12. In
this report, we present full data supporting reversible
â-HE as the alkyl isomerization mechanism, explore the
reactions of the transient alkene-hydride intermediate,
and use the reversibility of this olefin deinsertion/
insertion chemistry to evaluate metal-carbon bonding
preferences in the unsaturated alkyl complexes. We
supplement our mechanistic and reactivity studies with
DFT calculations.
(7) (a) Eckert, N. A.; Smith, J . M.; Lachicotte, R. J .; Holland, P. L.
Inorg. Chem. 2004, 43, 3306-3321. (b) Vela, J .; Stoian, S.; Flaschen-
riem, C.; Mu¨nck, E.; Holland, P. L. J . Am. Chem. Soc. 2004, 126, 4522-
4523. (c) Eckert, N. A.; Bones, E. M.; Lachicotte, R. J .; Holland, P. L.
Inorg. Chem. 2003, 42, 1720-1725. (d) Holland, P. L.; Cundari, T. R.;
Perez, L. L.; Eckert, N. A.; Lachicotte, R. J . J . Am. Chem. Soc. 2002,
124, 14416-14424. (e) Andres, H.; Bominaar, E. M.; Smith, J . M.;
Eckert, N. A.; Holland, P. L.; Mu¨nck, E. J . Am. Chem. Soc. 2002, 124,
3012-3025. (f) Smith, J . M.; Lachicotte, R. J .; Pittard, K. A.; Cundari,
T. R.; Lukat-Rodgers, G.; Rodgers, K. R.; Holland, P. L. J . Am. Chem.
Soc. 2001, 123, 9222-9223. (g) Smith, J . M.; Lachicotte, R. J .; Holland,
P. L. Chem Commun. 2001, 1342-1343.
Resu lts a n d Discu ssion
Syn t h esis a n d Isom er iza t ion of Alk yl Com -
p lexes. As part of our investigations aimed at unveiling
the low-coordinate chemistry of the late first-row transi-
tion metals,7 we isolated paramagnetic, three-coordinate
(6) Vela, J .; Smith, J . M.; Lachicotte, R. J .; Holland, P. L. Chem.
Commun. 2002, 2886-2887.
(8) Smith, J . M.; Lachicotte, R. J .; Holland, P. L. Organometallics
2002, 21, 4808-4814.

5228 Organometallics, Vol. 23, No. 22, 2004
Vela et al.
F igu r e 3. Transition state for step iv (mechanism B,
Scheme 1b).
Ta ble 1. Ra te Con sta n ts for Rea ction of LMeF eiBu
(3a ) w ith Alk en es a t 336 K in C6D6
k (10-3 s-1
)
F igu r e 2. Decay of isobutyl complexes 3a (C6D6, 325 K)
in the presence of excess ethylene. [Fe]T ) 45 mM, [C2H4]0
) 380 mM. Exponential fits to the data are shown.
[Fe] (mM)
[alkene]0 (mM)
CH2dCH2
CH2dCH2
CH2dCH2
CH2dCH2
CH2dCHCH3
CH2dCHCH3
CH2dCHCH3
CH2dCHCF3
CH2dCHCF3
45
45
110
45
49
49
49
49
49
380
128
840
760
211
485
823
319
490
1.80 ( 0.09
1.86 ( 0.02
1.83 ( 0.09
1.82 ( 0.09
2.1 ( 0.1
To test for the intermediacy of alkene-bound species
in the isomerization reaction, a solution of isobutyl
compound LMeFeiBu was heated to 60 °C in an NMR
tube in the presence of excess (2-19 equiv) ethylene,
cleanly giving a solution of ethyl complex LMeFeEt (4a ),
isobutylene, and the remaining ethylene (Scheme 1a).
2.3 ( 0.1
2.1 ( 0.1
1.61 ( 0.05
1.50 ( 0.06
should always depend on alkene concentration. The data
are most consistent with mechanism B with iv as the
rate-limiting step for isomerization.
The alkene exchange reaction from LMeFeiBu to LMe
-
FeEt is reminiscent of chain transfer in R-olefin poly-
merizations catalyzed by late transition metals9 includ-
ing iron,10 in which low coordinate active catalysts are
thought to be involved.
An Eyring plot of rate constants for reaction of 3a
with ethylene yielded the activation parameters ∆Hq )
+18.4 ( 0.5 kcal‚mol-1 and ∆Sq ) -17 ( 2 cal‚mol-1‚K-1
.
Mech a n ism of Alk yl Tr a n sfer . Two possible mech-
anisms have been proposed for chain transfer in alkene
polymerization,9 and Scheme 1b shows these mecha-
nisms in the context of the alkyl exchange of the three-
coordinate iron complexes. Mechanism A starts with
alkene binding (i) to form four-coordinate alkene alkyl
intermediate 5, followed by intramolecular hydride
transfer (ii) and release of the new alkene (iii). Mech-
anism B has â-hydride elimination (iv) to form an alkene
hydride intermediate 6. This intermediate could ex-
change alkene (v) and undergo â-hydrogen transfer to
give a new alkyl complex (vi). All steps are potentially
reversible.
Steady-state analysis was used to derive rate laws,
because no intermediates are observed. Mechanism A
predicts that the rate of consumption of LRFeiBu is
dependent on the concentration of ethylene and 2-me-
thylpropene (Scheme 1b). In other words, in mechanism
A added alkene is part of the transition state in each
elementary step. In contrast, the same kind of analysis
applied to mechanism B yields a rate law in which one
of the terms (kiv[3], Scheme 1b) is independent of alkene
concentration.
The negative entropy of activation is consistent with the
transition state for â-HE, because motion of the alkyl
ligand is restricted in the rate-determining transition
state (Figure 3).11,12 Positive entropies of activation have
been observed in cases where the rate-limiting step is
alkene dissociation (v) from an intermediate such as 6.13
To gain more insight into the nature of the transition
state, alkene exchange was carried out with the deu-
terated compound LMeFe(iBu-d9) (d9-3a ). Comparison of
the rate constants for the reaction of ethylene with 3a
and d9-3a gave a normal deuterium kinetic isotope effect
(KIE) of kH/kD ) 2.2 ( 0.2 (325-347 K), Table 2, I.14
Normal kinetic isotope effects of this magnitude are
common for â-H eliminations,1d,15 for example, (Ph3P)2-
(CO)Ir(n-octyl) (kH/kD ) 2.28 ( 0.20),16 (PhMe2P)2(acac)-
Co-Et (kH/kD ) 2.30 ( 0.05),17 Cp*2ScCH2CHDPh (kH/
kD ) 2.0 ( 0.3),18 (PCP)RhR (kH/kD ) 1.4, PCP ) 2,6-
(CH2P-tBu2)2C6H3, R ) Et or nPr).11
The isomerization of tert-butyl complex 2b to isobutyl
complex 3b follows first-order kinetics as well. The
activation parameters of ∆Hq ) +20 ( 1 kcal‚mol-1 and
(11) Van der Boom, M. E.; Higgitt, C. L.; Milstein, D. Organome-
tallics 1999, 18, 2411-2419.
The rate of conversion of isobutyl complex 3a to ethyl
complex 4a is first order in [3a ] (Figure 2) and inde-
pendent of [ethylene]. Exchange of other alkenes having
large electronic differences, such as propene or 3,3,3-
trifluoropropene instead of ethylene, occurred with rate
constants similar to that for ethylene (Table 1). Assum-
ing the steady-state approximation above, these obser-
vations are inconsistent with mechanism A, whose rate
(12) Ozawa, F.; Ito, T.; Yamamoto, A. J . Am. Chem. Soc. 1980, 102,
6457-6463.
(13) Romeo, R.; Alibrandi, G.; Scolaro, L. M. Inorg. Chem. 1993, 32,
4688-4694.
(14) An Eyring plot for the temperature dependence of the rate
constants of the ethylene exchange data for 3a D gives ∆HDq ) +18.5
( 1.2 kcal/mol and ∆SDq ) -18 ( 2 cal/mol‚K (see Supporting
Information). Unfortunately, the error bars for the activation param-
eters are large enough to prevent any useful distinction from the
unlabeled compound 4a .
(15) A normal KIE of kH/kD ) 4.9 ( 0.5 has also been reported for
â-H elimination of ethyl groups on an Fe(100) surface. Burke, M. L.;
Madix, R. J . J . Am. Chem. Soc. 1991, 111, 3675-3684.
(16) Evans, J .; Schwartz, J .; Urquhart, P. W. J . Organomet. Chem.
1974, 81, C37-C39.
(9) (a) Ittel, S. D.; J ohnson, L. K.; Brookhart, M. Chem. Rev. 2000,
100, 1169-1203. (b) Britovsek, G. J . P.; Gibson, V. C.; Wass, D. F.
Angew. Chem., Int. Ed. 1999, 38, 428-447.
(10) (a) Britovsek, G. J . P.; Gibson, V. C.; Kimberley, B. S.; Maddox,
P. J .; McTavish, S. J .; Solan, G. A.; White, A. J . P.; Williams, D. J .
Chem. Commun. 1998, 849-850. (b) Small, B. L.; Brookhart, M.;
Bennett, A. M. A. J . Am. Chem. Soc. 1998, 120, 4049-4050. (c) Small,
B. L.; Brookhart, M. J . Am. Chem. Soc. 1998, 120, 7143-7144.
(17) Ikariya, T.; Yamamoto, A. J . Organomet. Chem. 1976, 120,
257-284.
(18) Burger, B. J .; Thompson, M. E.; Cotter, W. D.; Bercaw, J . E. J .
Am. Chem. Soc. 1990, 112, 1366-1377.

Reversible Beta-Hydrogen Elimination
Organometallics, Vol. 23, No. 22, 2004 5229
Ta ble 2. Kin etic Isotop e Effect of â-HE Rea ction s
of 7a , 7b, 3b, and 3a . Because some compounds with
and without â-H react to scramble alkyl groups, there
is some pathway for exchange that does not require
â-HE. As a result, it is not possible from the crossover
experiments to definitively distinguish between associa-
tive and dissociative mechanistic possibilities for alkene
exchange. This conclusion, based on additional experi-
ments, differs from that proposed in the preliminary
communication.6,19
of Bu tyl Com p lexes
I. Ethylene Exchange into 3a , [Fe] ) 45 mM in C6D6
T (K)
kH (10-3 s-1
)
kD (10-3 s-1
)
kH/kD
293
303
312
325
336
347
358
0.031 ( 0.007
0.083 ( 0.001
0.248 ( 0.004
0.82 ( 0.01
1.89 ( 0.01
4.5 ( 0.1
0.38 ( 0.01
0.82 ( 0.02
2.1 ( 0.2
2.16 ( 0.06
2.3 ( 0.1
2.1 ( 0.2
5.7 ( 0.7
In the absence of experimental data to distinguish A
from D olefin exchange, calculations were carried out
to probe the nature of the olefin exchange step. The
crucial issue in comparing the associative and dissocia-
tive pathways involves the energetic accessibility of the
putative five-coordinate [LFe(H)(olefin)2] and three-
coordinate [LFeH] intermediates. The energy of three-
coordinate [L′FeH] (where L′ is the truncated diketim-
II. Isomerization of 2b to 3b, [Fe] ) 40 mM in C6D6
T (K)
kH (10-3 s-1
)
kD (10-3 s-1
)
kH/kD
325
336
347
358
369
380
0.129 ( 0.003
0.54 ( 0.01
1.50 ( 0.04
2.9 ( 0.3
0.56 ( 0.04
1.16 ( 0.03
3.0 ( 0.2
2.7 ( 0.2
2.5 ( 0.3
5.1 ( 0.8
inate N2C3H5 )
- 20 was minimized at the B3LYP/6-31G(d)
∆Sq ) -11 ( 4 cal‚mol-1‚K-1 are similar to those for
alkene exchange. The reaction rate is independent of
the concentration of added isobutylene. Monitoring the
rate of isomerization of LtBuFeC(CD3)3, d9-2b, to d9-3b
gave a normal kinetic isotope effect of kH/kD ) 2.6 (
0.2 (347-358 K, Table 2, II). The similar kinetic isotope
effects, along with the lack of dependence on alkene
concentration and similar activation parameters, strongly
indicate that the isomerization and alkyl transfer reac-
tions share a common rate-determining step, in which
a C-H bond of the alkyl ligand is cleaved. The most
reasonable such step is â-hydride elimination to give
an alkene hydride intermediate, because this intramo-
lecular reaction does not consume or produce free
alkene. The slower rates with the bulkier ligand LtBu
compared to LMe are consistent with an increase in
coordination number at iron in the transition state of
â-HE (Figure 3). Compounds 7a (R ) methyl) and 8a
(R ) neo-pentyl) do not react with excess ethylene or
1-hexene for days at 120 °C, due to their lack of
â-hydrogen atoms.
Na tu r e of th e Alk en e Exch a n ge Step . Because
hydride intermediate 6 lies after the rate-determining
step, kinetics studies do not distinguish whether alkene
exchange (v in Scheme 1b) is associative (A) or dissocia-
tive (D). To distinguish these possibilities, we performed
a crossover experiment using two different diketiminate
ligands (LMe and LtBu) and two different alkyl ligands
(isobutyl and ethyl). Assuming that â-H elimination is
the only pathway for crossover, we hypothesized that
scrambling of labels would occur only if alkene dissocia-
tion occurred in at least one of the two compounds (D
pathway of alkene exchange). A lack of scrambling
would indicate that a free alkene never forms in solution
(i.e., both complexes follow an A pathway). Heating (60
°C overnight) a solution of LMeFeiBu (3a ) and LtBuFeEt
(4b), or a solution of LMeFeEt (4a ) and LtBuFeiBu (3b),6
produced a mixture at equilibrium containing all four
compounds. This suggests that alkene exchange occurs
through a dissociative mechanism.6
level of theory, giving an Fe-H bond distance of 1.66
Å. Adding ethylene gave a binding enthalpy of only 7.8
kcal/mol for the first ethylene. Because the activation
barrier for a D mechanism is primarily the loss of olefin
from 6, this weak binding energy suggests that such a
mechanism could take place rapidly. Despite investiga-
tion of several conformations, no stable bis(olefin)
complex [L′Fe(H)(C2H4)2] was found as an energy mini-
mum, and the second ethylene dissociated upon geom-
etry optimization. The intermediacy of a bis(olefin)-
hydride complex is even less plausible for the more
sterically congested experimental ligand systems. There-
fore, the calculations argue against an associative
pathway and favor a dissociative pathway for the olefin
exchange step. We have synthesized a diketiminate-
supported iron hydride,21 and preliminary experiments
confirm that it inserts alkenes as postulated in the
dissociative mechanism for alkene exchange.
Gen er a lity of th e Hyd r id e Tr a n sfer Rea ction . We
were interested that LMeFeCH2CH(CH3)2 (3a ), a 12-
electron late-metal alkyl containing â-hydrogens, un-
derwent alkyl transfer to ethylene to form the ethyl
complex LMeFeCH2CH3 (4a ) (complete by NMR, 85%
isolated yield).6 To investigate the synthetic utility of
this reaction, 3a was heated to 60 °C with a stoichio-
metric amount of different alkenes in toluene overnight.
Using 3,3,3-trifluoropropene instead of ethylene leads
to formation of LMeFeCH2CH2CF3 (9a ), whereas the
reaction of 3a with styrene yields LMeFeCH(CH3)Ph
(10a ). Both 9a and 10a were isolated in good yields (up
to 80%), and their connectivity was confirmed by X-ray
crystallography (Figure 4). Interestingly, the R carbon
in complex LMeFeCH(CH3)Ph (10a ) is stereogenic, mak-
ing the two halves of the diketiminate ligand chemically
and magnetically inequivalent. This is evident in the
1H NMR spectrum: each of the isopropyl CH and CH3
(19) The mechanism through which direct alkyl exchange can occur
is unclear at this time, but may proceed through alkyl-for-hydride
exchange with unobserved hydride complexes formed from â-hydrogen-
containing alkyl complexes. In support of this idea, thermolysis of LMe
-
FeCH3 (7a ) with [LtBuFeH]2 gave a mixture containing 7b.
21
However, further control experiments with com-
pounds that do not contain â-hydrogens question the
validity of conclusions from crossover experiments.
Scrambling does not occur between neo-pentyl (LR-
FeCH2tBu) and methyl (LR′FeMe) compounds of differ-
ent diketiminate ligands, but complex LMeFeMe (7a )
does react with isobutyl complex 3b to form a mixture
(20) We have used models of this type successfully to understand
the electronic structure and geometries of the methyl complexes LtBu
-
-
MCH3 (M ) Fe, Co) and bridging dinitrogen complexes [LtBu
7b
FeNNFeLtBu
]
(n ) 0, 2-).7f
n
(21) Spectroscopic and kinetic evidence support the presence of the
three-coordinate iron(II) hydride complex LtBuFeH: Smith, J . M.;
Lachicotte, R. J .; Holland, P. L. J . Am. Chem. Soc. 2003, 125, 13752-
13753.

5230 Organometallics, Vol. 23, No. 22, 2004
Vela et al.
Sch em e 2
Sch em e 3
F igu r e 4. Molecular structures of (a) LMeFeCH2CH2CF3
(9a ), (b) LMeFeCH(CH3)Ph (10a ), (c) LMeFeOCHPh2 (11),
(d) LMeFeN(CH3)(CH2Ph) (12), (e) [LMeFe(µ-CH2CN)]2 (13),
and (f) LMeFeCy (16a ). Thermal ellipsoids are shown at 50%
probability. Hydrogen atoms (and aryl groups in 13) have
been omitted for clarity.
signals from the ligand splits into two different signals
(Scheme 2, see further discussion below).
We then explored the extension of the alkyl transfer
reaction to heteroatom-substituted analogues of an
olefin.22 Reacting 3a with 1 molar equiv of benzophe-
none in hot toluene gives LMeFeOCHPh2 (11), the
product of hydride transfer to the ketone (Scheme 3a).
Similarly, the stoichiometric reaction of 3a with (N-
methyl)benzophenone imine, PhCHNMe, gave the ami-
do complex LMeFeN(CH3)(CH2Ph) (12, Scheme 3b).
Complexes 11 and 12 were fully characterized and their
structures confirmed by X-ray diffraction (Figure 4).
2-Methylpropene is the only non-iron byproduct detected
reactions. For instance, compound 8a reacts quantita-
tively with 1 molar equiv of 2,6-diisopropylaniline, and
the only iron-containing product is LMeFeNH(2,6-diiso-
propylphenyl) (12);7a C5H12 was detected by GC/MS.
Similarly, all alkyl complexes LMeFeR react with aceto-
nitrile to form compound 13 (Scheme 3c). X-ray crystal-
lography showed that 13 is a dimer in which two [LMeFe]
fragments are linked by two deprotonated acetonitrile
ligands (Figure 4). The byproduct was alkane when R
did not have â-hydrogens. Alkene was detected (GC/MS)
when the original alkyl complex had â-hydrogens,
suggesting that the hydride may be a more kinetically
accessible base.
1
by H NMR in the reactions between the hydrocarbyls
such as 3a with olefins, a ketone, or an imine. Therefore,
the alkyl complexes can react as their unobserved
alkene hydride isomer, to add Fe-H across a C-C,
C-N, or C-O double bond.
Lin ea r ver su s Br a n ch ed Th r ee-Coor d in a te Ir on -
(II) Alk yls. With understanding of the general mecha-
nistic features and the scope of the reactions of the
iron(II) alkyls, we set about learning thermodynamics
from the relative stabilities of the different alkyl com-
plexes. We mentioned that the tert-butyl complexesLR-
FetBu (2a , and isolable 2b6) quantitatively rearrange
to their corresponding primary isomers LRFeiBu (3a and
3b). Similarly, the reactions of LMeFe(µ-Cl)2Li(THF)2
Alk yl Com p lexes Ca n Act a s Br øn sted Ba ses. The
three-coordinate iron(II) alkyls also engage in acid-base
(22) (a) Abdur-Rashid, K.; Clapham, S. E.; Hadzovic, A.; Harvey, J .
N.; Lough, A. J .; Morris, R. H. J . Am. Chem. Soc. 2002, 124, 13104-
13118. (b) Emslie, D. J . H.; Piers, W. E.; Parvez, M.; McDonald, R.
Organometallics 2002, 21, 4226-4240. (c) Liang, F.; J acobsen, H.;
Schmalle, H. W.; Fox, T.; Berke, H. Organometallics 2000, 19, 1950-
1962. (d) Lin, Z.; Marks, T. J . J . Am. Chem. Soc. 1990, 112, 5513-
5525. (e) J ordan, R. F.; Bajgur, C. S.; Dasher, W. E.; Rheingold, A. L.
Organometallics 1987, 6, 1041-1051.

Reversible Beta-Hydrogen Elimination
Organometallics, Vol. 23, No. 22, 2004 5231
with isopropylmagnesium chloride (through direct me-
tathesis) and of LMeFeiBu (3a ) with 3,3,3-trifluoropro-
pene (through alkyl transfer) yield the linear products
LMeFeCH2CH2CH3 (14a )23 and LMeFeCH2CH2CF3 (9a ),
respectively. In an attempt to increase the ratio of
secondary to primary positions, we also prepared the
compound LMeFenBu (15a ). Neither 9a , 14a , nor 15a
undergoes rearrangement to its branched isomer upon
bond (∼100.5 kcal/mol).26 However, metal-carbon bonds
are known to become disproportionately strong as the
correspondinghydrogen-carbon bondstrength increases,27-29
which makes the preference of iron for benzylic positions
surprising. To investigate this feature further, we
generated a series of phenethyl/R-methylbenzyl com-
plexes bearing electron-donating groups in the para
position, by treating LMeFeCy (16a ) with a stoichiomet-
ric amount of the substituted styrene. Each reaction
proceeded smoothly to give a mixture of primary and
benzylic products. In all cases, 16a was completely
1
prolonged heating (110 °C, 5 days), as observed by H
NMR. Therefore, among C4 and C3 hydrocarbyl com-
plexes, the thermally stable isomers are the compounds
with the primary carbon bound to iron (formally, the
1,2-insertion products). The isomerization equilibria in
these reactions therefore lie toward the linear isomer
with an equilibrium constant of at least Keq > 100 (n/
iso or iso/tert ratio),24 implying that for these transfor-
mations the free energy, ∆G°, is more negative than
-2.7 kcal/mol.
1
consumed, as evidenced by H NMR spectroscopy.
The equilibrium constants (Keq) for the p-Me-, OMe-,
and -NMe2-substituted benzyl/phenethyl systems were
1
measured at different temperatures by H NMR spec-
troscopy. The ratio of benzylic to primary isomer de-
creases with greater electron-donating ability of the
substituent (see Supporting Information).30 At room
temperature (298 K), the population of primary isomer
changes in the order 6% (H), 11% (p-Me), 18% (p-OMe),
20% (p-NMe2). There is a relationship between the
logarithm of the equilibrium ratio of isomers and the
Hammett constants σp,31 σ+,31 and σp(ω)32 (the linear
correlation coefficients at 333 K were 0.929, 0.939, and
0.971, respectively). A plot of ln Keq vs σp(ω) is shown
in Figure 5.33 The positive slope of 1.2-1.3 indicates that
there is a slightly greater preference for the benzylic
isomer when less electron-donating groups are present
on the aryl ring.
The branched complex LMeFeCH(CH3)Ph (10a ), for-
mally a 2,1-insertion product formed from the alkyl
transfer reaction between LMeFeiBu (3a ) and styrene,
stands in contrast with the observed preference for
linear chains in all the other alkyl complexes. Formation
of 10a occurs regardless of the synthetic route employed,
as reacting LMeFe(µ-Cl)2Li(THF)2 with PhCH2CH2MgCl
in ether at room temperature overnight gave the same
1
branched product. Nonetheless, a closer look at the H
NMR spectrum of 10a revealed one minor set of signals
that remained after several recrystallizations. To in-
1
vestigate this phenomenon, we used H NMR spectros-
copy to monitor the synthesis of 10a from [LMeFeCl]2
and (PhCH2CH2)2Mg in toluene-d8. The mysterious set
of signals appears immediately upon mixing the orga-
nomagnesium and iron chloride precursors, before pro-
ceeding to the equilibrium mixture. The initial pattern
consists of 10 signals, typical of a C2v symmetry molecule
(Scheme 2).25 Therefore, we assign these resonances to
the primary complex LMeFeCH2CH2Ph (p r im ), the
kinetic product of the metathesis reaction.
(26) (a) Blanksby, S. J .; Ellison, G. B. Acc. Chem. Res. 2003, 36,
255-263. (b) Brocks, J . J .; Beckhaus, H.-D.; Beckwith, A. L. J .;
Ru¨chardt, C. J . Org. Chem. 1998, 63, 1935-1943. (c) Tsang, W. In
Energetics of Organic Free Radicals; Martinho Simo˜es, J . A., Green-
berg, A., Liebman, F., Eds.; Blackie Academic and Professional:
London, 1996; pp 22-58. (d) McMillen, D. F.; Golden, D. M. Annu.
Rev. Phys. Chem. 1982, 33, 493. (c) Kerr, J . A. In CRC Handbook of
Chemistry and Physics, 71st ed.; Lide, R. L., Ed.; CRC Press: Boca
Raton, FL, 1990; pp 9_86-9_98.
(27) (a) Halpern, J . Acc. Chem. Res. 1982, 13, 238-244. (b) Bruno,
J . W.; Marks, T. J .; Morss, L. R. J . Am. Chem. Soc. 1983, 105, 6824-
6832. (c) Stoutland, P. O.; Bergman, R. G.; Nolan, S. P.; Hoff, C. D.
Polyhedron 1988, 7, 1429-1440. (d) Schock, L. E.; Marks, T. J . J . Am.
Chem. Soc. 1988, 110, 7701-7713. (e) Marks, T. J .; Gagne, M. R.;
Nolan, S. P.; Schock, L. E.; Seyam, A. M.; Stern, D. Pure Appl. Chem.
1989, 61, 1665-1672. (f) Nolan, S. P.; Stern, D.; Marks, T. J . J . Am.
Chem. Soc. 1989, 111, 7844-7853.
1
Integration of peaks from H NMR spectra shows that
the equilibrium ratio of p r im to 10a is about 1:9 at 25
°C, but this ratio decreases with temperature. The
relation between equilibrium constants and tempera-
ture was fit to a van’t Hoff plot (see Supporting
Information), giving ∆H° ) -2.8 ( 0.1 kcal‚mol-1 and
∆S° ) -4.9 ( 0.3 cal‚mol-1‚K-1. The negative value of
∆S° shows that the reaction is entropically unfavorable,
most likely because the branched benzylic product (10a )
is more vibrationally restricted from steric effects. The
entropic effect is overcome by the larger and favorable
enthalpic contribution.
(28) (a) Halpern, J . ACS Symp. Ser. 1990, 428, 100-112. (b)
Martinho Simo˜es, J . A.; Beauchamp, J . L. Chem. Rev. 1990, 90, 629-
688. (c) Drago, R. S.; Wong, N. M.; Ferris, D. C. J . Am. Chem. Soc.
1992, 114, 91-98. (d) Hoff, C. D. Prog. Inorg. Chem. 1992, 40, 503-
561. (e) Diogo, H. P.; Simoni, J . A.; Minas da Piedade, M. E.; Dias, A.
R.; Martinho Simo˜es, J . A. J . Am. Chem. Soc. 1993, 113, 2764-2774.
(f) Freiser, B. S. Acc. Chem. Res. 1994, 27, 353-360. (g) Martinho
Simo˜es, J . A.; Minas da Piedade, M. E. In Energetics of Organic Free
Radicals; Martinho Simo˜es, J . A., Greenberg, A., Liebman, F., Eds.;
Blackie Academic and Professional: London, 1996; pp 169-195.
(29) (a) Clot, E.; Besora, M.; Maseras, F.; Me´gret, C.; Eisenstein,
O.; Oelckers, B.; Perutz, R. N. Chem. Commun. 2003, 490-491. (b)
Li, G.; Zhang, F. F.; Pi, N.; Chen, H. L.; Zhang, S. Y.; Chan, K. S.
Chem. Lett. 2001, 30, 284-285. (c) Wick, D. D.; J ones, W. D.
Organometallics 1999, 18, 495-505. (d) Bennett, J . L.; Wolczanski, P.
T. J . Am. Chem. Soc. 1997, 119, 10696-10719. (e) Holland, P. L.;
Andersen, R. A.; Bergman, R. G.; Huang, J .; Nolan, S. P. J . Am. Chem.
Soc. 1997, 119, 12800-12814.
It is tempting to ascribe this enthalpic change to the
fact that the benzylic C-H bond in ethylbenzene (∼85
kcal/mol) is substantially weaker than a primary C-H
(23) We have reported that using the larger diketiminate LtBu gives
isolable LtBuFetBu6 and LtBuFeiPr.8 These complexes rearrange to their
linear isomers on extended heating.
(30) Due to the small change in equilibrium ratios with varying
substitution, the ∆H° values derived from van’t Hoff plots are too
similar to allow comparison on this basis.
(24) Calculation based on 1H NMR detection limit of 1%. Ebsworth,
E. A. V.; Rankin, D. W. H.; Cradock, S. Structural Methods in Inorganic
Chemistry, 2nd ed.; CRC Press: Boca Raton, FL, 1991; pp 1-114.
(25) Two different isopropyl-CH3 signals are seen for C2v symmetric
iron(II) diketiminate complexes because steric bulk prevents free
rotation of the C(aryl)-N bond. See: Panda, A.; Stender, M.; Wright,
R. J .; Olmstead, M. M.; Klavins, P.; Power, P. P. Inorg. Chem. 2002,
41, 3909-3916. The R and â protons in the phenethyl group are not
visible by 1H NMR, as is typical for paramagnetic iron(II) alkyl
complexes.
(31) (a) Hansch, C.; Leo, A.; Taft, R. W. Chem. Rev. 1991, 91, 165-
195. (b) Smith, M. B.; March, J . Advanced Organic Chemistry, 5th ed.;
Wiley: New York, 2001; pp 368-375.
(32) Recent calculations provide Hammett constants corrected for
solvent effects: Domingo, L. R.; Perez, P.; Contreras, R. J . Org. Chem.
2003, 68, 6060-6062.
(33) A plot using σ- showed a poor correlation. Phenethyl groups
with more electron-withdrawing substituents had an undetectably
small amount of the terminal isomer.

5232 Organometallics, Vol. 23, No. 22, 2004
Vela et al.
Ta ble 3. Ca lcu la ted (B3LYP /6-31G(d ))
Th er m od yn a m ic P a r a m eter s for L′F eR; En er gies
in k ca l/m ol
D(HR)a
∆Hb
D(Fe-R) - D(Fe-H)
∆Hâ-HE
c
R
H
Me
Et
nPr
iPr
iBu
tBu
PhCH2CH2
PhCHMe
104.2
105.0
101.1
101.0
98.6
101.0
96.5
0.0
7.6
11.3
9.5
11.6
11.5
15.2
12.0
4.9
0.0
-6.8
-14.4
-12.7
-17.2
-14.7
-22.9
-15.7
-24.4
21.8
20.4
17.8
16.1
13.9
16.0
23.0
100.5
84.7
a
HR bond energies from NIST, except R ) phenethyl, which
b
was calculated (see text). Calculated enthalpy for reaction in eq
1. c Calculated enthalpy for â-hydride elimination.
F igu r e 5. Hammett plot for the isomerization equilibria
between para-substituted phenethyl and R-(methyl)benzyl
complexes in C6D6, 333-373 K, [Fe]T ) 63 mM.
Com p u ta tion s on th e Sta bility of Th r ee-Coor d i-
n a te Ir on (II) Alk yls. To obtain more information on
the relative stability of the diketiminate iron(II) alkyl
complexes, density functional calculations were per-
formed at the B3LYP level, using either the truncated
diketiminate C3N2H5- (L′)20 or the full ligand LMe. The
calculated energies were converted to enthalpies by
applying zero-point energy and finite temperature cor-
rections calculated from B3LYP/6-31G(d) vibrational
frequencies.34
F igu r e 6. Plot of calculated D(Fe-R) vs D(H-R) for L′FeR
compounds (L′ ) N2C3H5-).
The calculated energies of iron alkyl compounds give
relative bond energies through the conceptual reaction
L′FeH + HR f L′FeR + H
(1)
Assuming that there is minimal influence on the other
bonds in the complex, the iron-ligand bond enthalpies
(relative to iron-hydride) become
B3LYP/6-31G(d) optimization of the lowest energy
conformers obtained from a molecular mechanics-based
conformational searching of LMeFeiPr and LMeFenPr
shows that H(nPr) < H(iPr) by 3.2 kcal/mol. This agrees
with a lower limit of 2.7 kcal/mol based on the detection
D(Fe-R) - D(Fe-H) ) D(H-R) - ∆H - D(H-H)
(2)
1
limit of H NMR spectroscopy (see above).24 Using a
where D represents the bond dissociation enthalpy, and
the bond enthalpies of HR and H2 are determined from
the literature enthalpies of formations for HR, R, and
H.35 Relative homolytic energies, D(L′Fe-R) - D(L′Fe-
H) estimated from eq 2 are shown in Table 3. A plot of
relative Fe-R energies against the corresponding H-R
energies appears in Figure 6. This plot shows that the
Fe-R bond strength increases with the corresponding
H-R BDE. Without considering the hydride and benzyl
ligands, the “selectivity slope”29 for the saturated alkyl
ligands in this ∆D(L′Fe-R)/D(H-R) plot is 1.8. A much
poorer correlation is seen for heterolytic bond energies,
so it appears appropriate to describe the Fe-C bonds
as being primarily covalent.
similar strategy, LMeFeCH2CH2Ph was calculated to be
3.1 kcal/mol higher in energy than LMeFeCH(CH3)Ph,
in agreement with the observed reaction enthalpy of 2.8
( 0.1 kcal/mol. The high accuracy of the QM/MM results
suggest that the DFT methodology is fundamentally
sound for these systems.
Calculations using the truncated L′ ligand gave poorer
agreement (∆H ) 2.1 and 7.1 kcal/mol for the two
reactions, respectively). In each case, the secondary
isomer is calculated to be unreasonably stable, strongly
suggesting that steric effects are important in determin-
ing the energies of the experimental compounds. More
extensive calculations were carried out on the truncated
L′ models for two reasons: computational expediency,
and because we were more interested in the inherent
(not influenced by the steric clash of the diketiminate
and the alkyl ligands) enthalpic contributions to the
iron-carbon bonds.
Figure 6 verifies that the calculated benzylic Fe-C
bond strength is much greater than what would be
predicted from the correlation between the other Fe-C
and H-C bond energies.36 It also shows that the Fe-H
bond is disproportionately strong relative to the Fe-C
bond, as found by other workers.27
X-r a y Str u ctu r es. The molecular structures of LtBu
FeiBu (3b), LMeFeMe (7a ), LMeFeCH2CH2CF3 (9a ), LMe
-
-
(34) Hehre, W. J .; Radom, L.; Schleyer, P. v. R.; Pople, J . A. Ab Initio
Molecular Orbital Theory; Wiley: New York, 1986.

Reversible Beta-Hydrogen Elimination
Organometallics, Vol. 23, No. 22, 2004 5233
[LMeFe(µ-CH2CN)]2 (13)
Ta ble 4. Selected Bon d Dista n ces (Å) a n d An gles (d eg) for Com p ou n d s 11-13
LMeFeOCHPh2 (11)
LMeFeN(CH3)(CH2Ph) (12)
Fe-O
1.8076(16)
142.57(8), 122.84(7)
94.48(7)
140.42(16)
Fe-Namide
1.8715(18)
133.57(7), 133.74(7)
92.66(7)
122.69(14), 123.54(15)
121.08(16), 121.23(17)
Fe-Nµ
2.0378(15)
2.1409(18)
93.35(5)
110.40(6), 109.70(6)
120.57(14), 121.49(14)
N-Fe-O
N-Fe-Na
Fe-O-C
C-N-CAr
N-Fe-Namide
N-Fe-Na
Fe-N-C
Fe-Cµ
N-Fe-Na
N-Fe-Nµ
C-N-CAr
119.72(17), 120.69(15)
C-N-CAr
a
Diketiminate bite angle.
Ta ble 5. Selected Str u ctu r a l Da ta for
Th r ee-Coor d in a te Alk yl Ir on (II) Dik etim in a tes
Fe-C
C-N-CAr
R
LMe
LtBu
LMe
LtBu
Me
Et
2.003(2)
2.009(3)a
120.99(12)
120.76(12)
120.3(2)a
120.7(2)
128.56(11)a
2.033(3)a
2.019(6)a
iBu
2.055(3)
2.027(4)a
119.9(2)a
120.6(2)
126.3(2)
128.1(2)
126.3(2)a
127.9(2)
CH2tBu
(CH2)2CF3
iPr
2.027(6)
120.6(4)
122.7(4)
2.048(3)a
2.079(4)a
125.5(2)a
127.6(2)
Cy
CHMePh
2.036(2)
2.060(2)
120.96(12)
119.6(2)
120.8(2)
F igu r e 7. Molecular structures of LRFeiBu with LtBu (3b,
in red) and LMe (3a ), showing the ligand tert-butyl groups
pushing the aryl groups. ORTEP diagrams at 50% prob-
ability are viewed along an axis perpendicular to the
diketiminate plane. Hydrogen atoms and isopropyl groups
are omitted for clarity.
tBu
127.0(2)a
125.3(2)
a
Refs 6, 7, 8.
FeCHMePh (10a),LMeFeOCHPh2 (11),LMeFeN(CH3)(CH2-
Ph) (12), [LMeFe(µ-CH2CN)]2 (13), and LMeFeCy (16a )
were determined by single-crystal X-ray diffraction.
Selected data for compounds 11, 12, and 13 appear in
Table 4. Complete structural parameters for these and
all new compounds, along with experimental data used
in crystal structure determination, are included in the
Supporting Information.
Selected structural data for the three-coordinate iron-
(II) alkyl complexes appear in Table 5. The iron-carbon
bond length increases with the substitution at CR in both
series LMeFe-R and LtBuFe-R with R ) 1° < 2° < 3°.
Regardless of steric hindrance from the diketiminate
ligand, there is free and fast Fe-C bond rotation in
solution at room temperature in all the alkyl complexes.
In one especially clear example, complex 16a has
distinct equatorial and axial protons in the solid-state
structure (Figure 4f), but fast equilibration of conform-
ers on the 1H NMR time scale yields a single set of
cyclohexyl signals at room temperature.
backbone of the larger ligand LtBu, which pushes the
aryl groups toward the open coordination sites on iron.
The C-N-CAr angle is a quantitative measure of this
steric effect: 119(2)° for 1a and 128.4(2)° for 1b.
However, there is some ambiguity in the interpretation
of the C-N-CAr angle because of the difference in
coordination number at iron in the two iron chloride
complexes. This situation is remedied by the isolation
and crystallographic characterization of complexes 3a
and 3b, LRFe-iBu, and 7a and 7b, LRFe-Me. Because
iron is three-coordinate in each complex, the C-N-CAr
angle in these compounds is a direct measurement of
the steric effect of the backbone substituent. In com-
plexes of LtBu this angle is consistently pushed 6-8°
toward the open site on iron by the tert-butyl groups,
in complexes with the same coordination number (Table
5, bold numbers). This aryl pushing effect is illustrated
in Figure 7 for the isobutyl complexes.
We have reported that addition of LRLi to FeCl2-
(THF)1.5 gives products with different coordination
numbers: a four-coordinate iron chloride complex with
LMe, LMeFe(µ-Cl)2Li(THF)2 (1a ), and a three-coordinate
iron chloride complex with LtBu, LtBuFeCl (1b).7g The
difference in iron coordination number was proposed to
arise from the steric bulk of the tert-butyl groups in the
Discu ssion
Three-coordinate diketiminate iron(II) alkyl com-
plexes undergo a variety of transformations depending
on the steric constraints of the ancillary ligand used,
the presence of â-hydrogens, and the degree of substitu-
tion at carbon. These iron(II) alkyl complexes are
potentially useful synthetic precursors to low-coordinate
inorganic compounds, as exemplified by the syntheses
of alkoxide 11, amide 12, and bimetallic 13. This finding
is significant because it opens a new synthetic route to
alkoxide and amido complexes of iron(II) supported by
diketiminato ligands. The alkyl complexes reported here
and elsewhere6,8,37 represent the first series of hetero-
leptic three-coordinate iron(II) hydrocarbyl complexes.
(35) Bartmess, J . E. Negative Ion Energetics Data. In NIST
Chemistry WebBook, NIST Standard Reference Database Number 69;
Linstrom P. J ., Mallard, W. G., Eds.; National Institute of Standards
and Technology: Gaithersburg, MD, 2003 (http://www.nist.gov). The
enthalpy values for (PhCH2CH2•) and (PhHMeC•) were not found in
the literature, so they were calculated at the B3LYP/6-31G(d) level of
theory with pure (5d and 7f) d and f functions.
(36) We have not attempted to draw correlations to the homolytic
bond energies of substituted benzyl groups, because the bond energies
are affected by unusual radical stabilization energies. Pratt, D. A.;
Dilabio, G. A.; Mulder, P.; Ingold, A. K. U. Acc. Chem. Res. 2004, 37,
334-340.
(37) Sciarone, J . J .; Meetsma, A.; Hessen, B.; Teuben, J . H. Chem.
Commun. 2002, 1580-1581.

5234 Organometallics, Vol. 23, No. 22, 2004
Vela et al.
Compound 11 is a novel alkoxide complex based on the
lowed by rapid isomerization of the alkyl group.45,46
These mechanistic studies of alkyl isomerization were
commonly hindered by the necessity of creating an open
coordination site prior to the isomerization reaction.
Brookhart and co-workers have studied chain isomer-
ization and transfer in cationic alkyl complexes [(di-
imine)MIIR]+ of palladium and nickel, in studies rel-
evant to chain running and chain transfer in R-olefin
polymerizations.47 The cationic complexes have â-agostic
interactions that stabilize branched over linear alkyl
complexes, and addition of an exogenous ligand (ether,
CH3CN, DMSO, or ethylene) gives square planar com-
plexes that do not show â-agostic interactions and in
which linear alkyl complexes are more stable. Isomer-
ization occurs only from the agostic species,47a,c and
interestingly when R ) Et and L′ is labeled ethylene,
there is no exchange, ruling out a concerted â-hydrogen
transfer pathway as well as a five-coordinate bis(olefin)
transition state for the isomerization.47b In contrast,
alkene exchange from [(diimine)MR(ethylene)+] (R ) C2
alkyl or higher * Me) is associative and happens
through a five-coordinate transition state.47b
Complete alkene exchange was recently observed in
bis(arylimino)pyridine (BIP) complexes of cobalt(I).48
Treatment of (BIP)CoR (R ) nPr, nBu) complexes with
ethylene yields (BIP)CoEt and free alkene (propene or
1-butene). The reaction rate is first order in [Co] and
zero order in [ethylene], and it is faster for the n-propyl
than for the n-butyl derivative. Moreover, exchange with
propene or ethylene occurs at similar rates. These
observations suggested a mechanism involving an alk-
ene hydride intermediate, analogous to mechanism B
in Scheme 1.
easily accessible diketiminate ligand LMe 38
Alkoxide
.
complexes with the bulkier ligand LtBu were recently
found to be active initiators for the living polymerization
of lactides.38 The insertion of imines into metal-hydride
bonds giving amides39 is of interest because of the role
that imine insertion plays in catalytic hydrogenation
and transfer hydrogenation processes.40
We have shown here that the three-coordinate iron
alkyl complexes with â-hydrogen are capable of reacting
as an alkyl complex (protonation by amine) or as a hy-
dride complex (protonation by acetonitrile, reduction of
ketone to alkoxide, and reduction of imine to amide). It
is likely that the different reactivity of the substrates
is controlled by the rate of reaction with the alkyl com-
plex. It is well-known that N-H protons transfer more
quickly than C-H protons.41 Therefore, it is likely that
the iron alkyl is protonated directly in the reaction with
amine, but the relatively slow rate of alkyl group pro-
tonation by acetonitrile allows â-hydrogen-containing
alkyl complexes time to rearrange. Because the hydride
complex is less hindered than the alkyl complex, it
preferentially acts as the base in the reaction leading
to 13.
We have reported other reactions that proceed directly
from the alkyl isomer. For example, CO inserts into
three-coordinate LtBuFeR (R ) Me, CH2tBu, iPr).8 Even
when R contained â-hydrogen atoms, these reactions
yielded acyl rather than formyl products. Because of the
very rapid rate of the carbonylation reactions, we have
not been able to distinguish whether the reason is the
greater reactivity of carbon ligands toward CO insertion
or because reaction with CO is faster than â-HE.
Alk yl Isom er iza tion P r oceed s th r ou gh â-Hy-
d r id e E lim in a t ion /In ser t ion . There have been few
mechanistic investigations on alkyl rearrangements at
late transition metals, despite their relevance to useful
synthetic processes. Tamaki and Kochi found that the
alkyl groups in trans-(PPh3)Me2AuR complexes (L )
PPh3, R ) hydrocarbyl) isomerize through â-hydride
elimination.42 Reger and co-workers found an analogous
mechanism for isomerization of CpFe(CO)(PPh3)(alkyl)43
and (R2NCS2)M(PEt3)(alkyl) (M ) Pd, Pt).44 Bennett and
Maitlis have also implicated â-hydride elimination/
insertion in studies of alkyl isomerization within coor-
dinated acyl groups, and the available evidence supports
the idea that carbonyl deinsertion is rate-liming, fol-
Our results indicate that alkyl group isomerization
and transfer in high-spin iron(II) complexes occur
through â-HE. The highest energy barrier for these
reactions is â-HE from the starting hydrocarbyl com-
plex, as shown from the lack of rate dependence on
alkene and by the H/D KIE. This mechanism is consis-
tent with the slower isomerization rates for LtBu com-
plexes: as a result of the significantly higher steric
constraints, a transition state like that shown in Figure
3, with a higher coordination number at iron, is more
difficult to achieve with LtBu than for LMe
.
It is likely that the intermediate alkene hydride
complex can reversibly dissociate alkene to form a three-
coordinate iron hydride. We have independently syn-
thesized a relevant hydride compound, [LtBuFeH]2,
(38) Gibson, V. C.; Marshall, E. L.; Navarro-Llobet, D.; White, A. J .
P.; Williams, D. J . J . Chem. Soc., Dalton Trans. 2002, 4321-4322.
(39) (a) Fryzuk, M. D.; Piers, W. E. Organometallics 1990, 9, 986-
998. (b) Debad, J . D.; Legzdins, P.; Lumb, S. A.; Batchelor, R. J .;
Einstein, F. W. B. Organometallics 1995, 14, 2543-2555. (c) Liang,
F.; Schmalle, H. W.; Berke, H. Inorg. Chem. 2004, 43, 993-999.
(40) (a) Abdur, R.; Clapham, S. E.; Hadzovic, A.; Harvey, J . N.;
Lough, J . A.; Morris, R. H. J . Am Chem. Soc. 2002, 124, 15104-15118.
(b) Zhao, J .; Hesslink, H.; Hartwig, J . F. J . Am Chem. Soc. 2001, 123,
7220-7227. (c) Casey, C. P.; Singer, S. W.; Powell, D. R.; Hayashi, R.
K.; Kavana, M. J . Am Chem. Soc. 2001, 123, 1090-1100. (d) Hartwig,
J . F. J . Am Chem. Soc. 1996, 118, 7010-7011. Blum, O.; Milstein, D.
J . Am. Chem. Soc. 1995, 117, 4582-4594. (e) Lorkovic, I. M.; Duff, R.
R., J r.; Wrighton, M. S. J . Am. Chem. Soc. 1995, 117, 3617-3618. (f)
Willoughby, C. A.; Buchwald, S. L. J . Am. Chem. Soc. 1994, 116,
11703-11714.
(44) (a) Reger, D. L.; Garza, D. G.; Baxter, J . C. Organometallics
1990, 9, 873-874. (b) Reger, D. L.; Garza, D. G.; Lebioda, L. Organo-
metallics 1992, 11, 4285-4292. (c) Reger, D. L.; Ding, Y.; Garza, D.
G.; Lebioda, L. J . Organomet. Chem. 1993, 452, 263-270.
(45) (a) Ellis, P. R.; Pearson, J . M.; Haynes, A.; Adams, H.; Bailey,
N. A.; Maitlis, P. M. Organometallics 1994, 11, 3213-3226. (b) Haynes,
A.; Mann, B. E.; Morris, G. E.; Maitlis, P. M. J . Am. Chem. Soc. 1993,
113, 4093-4100.
(46) (a) Bennett, S. L.; Charles, R. J . Am. Chem. Soc. 1972, 94, 666-
667. (b) Bennett, M. A.; Charles, R.; Mitchell, T. R. B. J . Am. Chem.
Soc. 1978, 100, 2737-2743. (c) Bennett, M. A.; J effery, J . C.; Robertson,
G. B. Inorg. Chem. 1980, 19, 3763-3767. (d) Bennett, M. A.; Crisp, G.
T. Organometallics 1986, 5, 1792-1800.
(47) (a) Tempel, D. J .; J ohnson, L. K.; Huff, R. L.; White, P. S.;
Brookhart, M. J . Am. Chem. Soc. 2000, 122, 6686-6700. (b) Shultz,
L. H.; Tempel, D. J .; Brookhart, M. J . Am. Chem. Soc. 2001, 123,
11339-11355. (c) Leatherman, M. D.; Svejda, S. A.; J ohnson, L. K.;
Brookhart, M. J . Am. Chem. Soc. 2003, 125, 3068-3081. (d) J enkins,
J . C.; Brookhart, M. J . Am. Chem. Soc. 2004, 126, 5827-5842.
(48) Gibson, V. C.; Tellmann, K. P.; Humphries, M. J .; Wass, D. F.
Chem. Commun. 2002, 2316-2317.
(41) Eigen, M. Angew. Chem., Int. Ed. Engl. 1964, 3, 1-19.
(42) Tamaki, A.; Magennis, S. A.; Kochi, J . K. J . Am. Chem. Soc.
1974, 96, 6140-6148.
(43) (a) Reger, D. L.; Culbertson, E. C. J . Am. Chem. Soc. 1976, 98,
2789-2794. (b) Reger, D. L.; Culbertson, E. C. Inorg. Chem. 1977, 16,
3104-3107. (c) Reger, D. L.; McElligot, P. J . J . Organomet. Chem. 1981,
216, C12-C14.

Reversible Beta-Hydrogen Elimination
Organometallics, Vol. 23, No. 22, 2004 5235
which dissociates to a monomer at high dilution or high
diamagnetic and have a formal electron count of 16 at
the metal as well. Kinetics data for these rhodium
systems are most consistent with an alkene hydride
intermediate (like the one in mechanism B, Scheme 1b).
Diketiminate complexes of other metals have different
trends toward â-hydride elimination. For example, it is
possible to isolate the four-coordinate olefin hydride
platinum complex L1PtH(η2-CH2dCHtBu) (L1 ) 2,4-bis-
(2,6-dimethyl-4-tertbutylphenylimino)pent-3-yl).54 Mean-
while, Warren has reported a series of four-coordinate
nickel alkyls with general formula L2NiR(2,4-lutidine)
(R ) Me, Et, Pr, L2 ) 2,4-bis(2,4,6-trimethylphenylimi-
no)pent-3-yl).55 Despite the isolation of a three-coordi-
nate, formally 11-electron phenyl complex of manganese-
(II), LMeMnPh,56 the analogous n-butyl compound
decomposed, probably through â-H elimination. There-
fore, the three-coordinate geometry imposed by the
presence of the diketiminate ligand does not suffice to
stabilize the metal-alkyl complexes, neither does it
solely explain the reversible olefin insertion chemistry
of the iron(II) alkyls described here.
Rela tive Sta bilities of Alk yl Com p lexes a n d
F e-C Bon d En er gies. Linear saturated alkyl com-
plexes of three-coordinate iron(II) are clearly favored
over their branched counterparts. We chose to investi-
gate this trend through DFT calculations on a less
sterically hindered model in which the bulky aryl groups
are removed. Although larger models gave good quan-
titative agreement with the observed enthalpy changes,
we anticipated that the smaller models would generate
purely electronic trends in bond energies. This is
confirmed by a slope of 1.8 in the plot of relative Fe-C
bond energies versus the corresponding H-C energies
for the saturated alkyl ligands (Figure 6). A similar
preference for linear alkyl ligands has been observed
in analogous plots for compounds of titanium (slope )
1.36),29d rhodium (slope ) 1.22),29c and iridium (slope
) 2).27c
Because terminal alkyl-H bonds are calculated to
have higher bond homolytic or dissociation energies
(BDEs) than the internal alkyl-H bonds, the large
calculated selectivity slopes indicate that the metal
prefers the terminal position, even with minimized
steric effects. Consistent with this trend, the iron
cyclohexyl complexes (containing only secondary posi-
tions) are exceptionally good precursors to other alkyl
complexes, through loss of cyclohexene. However, ben-
zylic isomers are greatly favored, even when they are
secondary and have an accessible primary isomer.
Unusual favoritism for benzylic M-C bonds has been
seen in the literature for M-C trends established using
C-H activation reactions.29c,d The potential reasons for
heightened stability of benzyl complexes are (1) the
presence of C-H‚‚‚M agostic interactions,47,57 (2) η3-
coordination, commonly encountered in transition metal
benzyl and allyl chemistry,58 (3) inductive stabilization
temperature.21 Control experiments show that [LtBu
FeH]2 reacts with alkenes to form the expected alkyl
complexes. Density functional computations (see above)
indicate that three-coordinate L′FeH binds ethylene
with a modest enthalpy of 7.8 kcal/mol. These data
combined support the idea that hydride complexes are
the most likely intermediates in alkyl group exchanges
between metals.9,49
-
â-Hyd r ogen Elim in a tion Is Rever sible. Alkene
insertion into metal-hydrogen bonds is common among
early metals, and this feature is the basis for the useful
hydrozirconation reaction.50 However, alkyl complexes
(particularly of the late transition metals) with â-H often
decompose in the absence of stabilizing interactions
such as geometric constraints.1 Hence, â-hydrogen
elimination (â-HE) is often regarded as a foe in synthetic
chemistry,1,51 and generally accepted guidelines for
making stable metal alkyls include avoiding having â-H,
preventing the M-C-C-H unit from adopting the
coplanar conformation of the transition state for â-HE
(cf. Figure 3), or avoiding vacant sites cis to the alkyl
ligand.1c None of these conditions are met in the three-
coordinate iron(II) alkyl complexes described here, yet
they are stable toward irreversible alkene loss. This
stability does not arise from kinetic trapping, because
rapid â-HE clearly occurs at room temperature.
The results of the DFT calculations can be used to
calculate the enthalpy change for putative â-HE. Using
a thermodynamic cycle, we find that the enthalpies for
converson of L′FeR to alkene and L′FeH range between
+13.9 kcal/mol (R ) tBu) and +23.0 kcal/mol (R )
PhMeCH), as shown in Table 3. There is a favorable
entropy gain with â-HE, but for reactions in which 1
particle goes to 2, T∆S is usually only about 10 kcal/
mol,52 not enough to balance the unfavorable ∆H
contribution. Therefore, theory confirms that enthalpic
contributions are responsible for the endothermicity of
â-HE from the three-coordinate iron(II) alkyl complexes.
Other examples of stable late-metal alkyls with â-H
exist in the literature, and some experimental evidence
is consistent with reversible olefin insertion occurring
in these systems. Among these are Gibson’s bis(imino)-
pyridine cobalt(I) alkyls (see above), which show com-
plete alkyl transfer by a â-H elimination pathway.48
Although these 16-electron diamagnetic cobalt com-
plexes reportedly decompose in the absence of an
incoming olefin, this was attributed to possible ligand
cyclometalation rather than to irreversible olefin loss
after â-HE. Chan has reported rhodium(III) porphyrin
complexes that also undergo clean alkyl rearrangement,
favoring isomers with electron-withdrawing substitu-
ents at the R-position.53 These rhodium(III) alkyls are
(49) Deng, L.; Margl, P.; Ziegler, T. J . Am. Chem. Soc. 1999, 121,
3998-4004.
(50) (a) Hart, D. W.; Schwartz, J . J . Am. Chem. Soc. 1974, 96, 8113-
8116. (b) Negishi, E.; Takahashi, T. Synthesis 1988, 1-19. (c) Wipf,
P.; Xu, W.; Takahashi, H.; J ahn, H.; Coish, P. D. G. Pure Appl. Chem.
1997, 69, 639-644. (d) Chirik, P. J .; Day, M. W.; Labinger, J . A.;
Bercaw, J . E. J . Am. Chem. Soc. 1999, 121, 10308-10317.
(51) (a) Zhou, J .; Fu, G. C. J . Am. Chem. Soc. 2004, 126, 1340-
1341. (b) Ca´rdenas, D. J . Angew. Chem., Int. Ed. 2003, 42, 384-387.
(c) Ca´rdenas, D. J . Angew. Chem., Int. Ed. 1999, 38, 3018-3020. (d)
Giovannini, R.; Stu¨demann, T.; Dussin, G.; Knochel, P. Angew. Chem.,
Int. Ed. 1998, 37, 2387-2390.
(53) Mak, K. W.; Xue, F.; Mak, T. C. W.; Chan, K. S. J . Chem. Soc.,
Dalton Trans. 1999, 3333-3334. Mak, K. W.; Chan, K. S. J . Am. Chem.
Soc. 1998, 120, 9686-9687.
(54) Fekl, U.; Kaminsky, W.; Goldberg, K. I. J . Am. Chem. Soc. 2003,
125, 15286-15287.
(55) (a) Wiencko, H. L.; Kogut, E.; Warren, T. H. Inorg. Chim. Acta
2003, 345, 199-208. (b) Puiu, S. C.; Warren, T. H. Organometallics
2003, 22, 3974-3976.
(52) Minas da Piedade, M. E.; Martinho Simo˜es, J . A. J . Organomet.
Chem. 1996, 518, 167-180.
(56) Chai, J .; Zhu, H.; Roesky, H. W.; Magull, J . Organometallics
2004, 23, 1177-1179.

5236 Organometallics, Vol. 23, No. 22, 2004
Vela et al.
of partial negative charge on a carbon atom bearing an
electron-withdrawing substituent, or (4) resonance sta-
bilization of the partial negative charge through charge
delocalization into the aromatic group.
We recently reported that reaction of LtBuFeNHAr
with PhCCH (pKa ) 23) or tBuOH (pKa ) 17) leads to
complete protonation of amide, giving free aniline (pKa
) 30) and forming LtBuFeCCPh or LtBuFeOtBu, respect-
ively.7a,63 This can be understood as a favoritism for a
complex with the more stable anionic ligand (the ligand
with the lowest pKa). Also, LMeFeN(CH3)(CH2Ph) (12)
did not show any sign of rearrangement (1H NMR) to
In 10a and the other three-coordinate alkyls we can
rule out option (1). None of the alkyl compounds have
agostic C-H‚‚‚Fe interactions in the solid state (H‚‚‚Fe
> 2.50 Å).59 We can similarly rule out η3-coordination
(2) because the crystal structure of 10a shows simple
the carbon-bound isomers LMeFeCH2NHCH2Ph or LMe
-
1
σ-coordination of the benzylic carbon, and the H NMR
FeCHPh(NHCH3) in solution, even after prolonged
heating. Therefore R2N-, a harder ligand and a weaker
base, forms a more stable complex than R3C- (pKa ca.
48). Similar observations were made about the impor-
tance of ionic contributions to metal-ligand bonding in
nickel29e and rhenium29a complexes. Because of reso-
nance stabilization, a benzylic C-H bond has a sub-
stantially lower pKa (ca. 41) than that of an alkane (pKa
) 48),63 and 10a is more stable than the linear phen-
ethyl complex p r im (substantially more than would be
predicted from the Fe-R/H-R correlation alone). There-
fore, one can sum up our thermodynamic findings on
alkyl, amido, and alkoxo complexes by stating that the
Fe-X bonds in three-coordinate iron(II) complexes have
substantial polarity and can benefit energetically from
stabilization of the partial negative charge on the ligand.
In support of this idea, recent theoretical studies by
Harvey showed that electrostatic effects can explain the
bonding energetics in a series of transition metal-alkyl
complexes.64 In our three-coordinate complexes, this
polarity becomes particularly important when hard
nitrogen and oxygen ligands replace an alkyl ligand (a
stronger base and a less stable anion).
shows no evidence that a species other than p r im
coexists in solution. In addition, the DFT-optimized
structures of sterically nonhindered L′FeR models showed
no signs of geometric distortion or short distances that
would indicate the presence of agostic interactions or
aryl π-coordination. If inductive electron withdrawal (3)
were important for the stability of 10a , then the complex
LMeFeCH2CH2CF3 (9a ) would be expected to rearrange
(at least partially) to its branched isomer LMeFeCH-
(CH3)(CF3). However, after heating a solution of 9a , no
new species was seen by 1H NMR (see above). Inductive
effects operate in other alkyl complexes; for example
Reger observed 50% isomerization of (Me2NCS2)Pd-
(PEt3)CH2CH2CF3 to (Me2NCS2)Pd(PEt3)CH(CH3)-
(CF3).44b In Reger’s system, replacing the CF3 group
with a cyano group led to complete formation of the
branched isomer.44b,43c
The remaining explanation for the higher stability of
10a is stabilization through resonance stabilization of
the partial negative charge (4).60 This model readily
explains the dependence of the equilibrium ratio of
phenethyl isomers on the para-substituent (Figure 5)61
and the stability of 10a compared to the other alkyl
complexes in Figure 6. Interestingly, this resonance
stabilization takes place without any geometric distor-
tion of the iron-bound carbon atom toward sp2 hybrid-
ization. We note in passing that repulsive π-interactions
between filled metal d orbitals and filled ligand p
orbitals62 are unlikely to be important in alkyl com-
plexes, which lack nonbonding electrons on the ligand.
Con clu sion s a n d Im p lica tion s
Three-coordinate, paramagnetic iron(II) diketiminate
alkyl complexes undergo reversible â-hydrogen elimina-
tion. Our three-coordinate alkyliron(II) complexes add
to a growing number of transition metal alkyls contain-
ing â-hydrogen that are stable in the absence of geo-
metric constraints. A common feature in these systems
is their ability to undergo reversible â-HE (olefin
deinsertion/reinsertion) reactions. Therefore it appears
appropriate to distinguish between a stable metal alkyl
and a â-HE stable metal alkyl. Learning from systems
similar to the three-coordinate iron(II) diketiminates
and the factors responsible for their unusual stability
could lead to strategies for the stabilization of other late-
metal alkyls and their incorporation into catalytic
pathways.51
As a result of the omnipresent, rapid equilibration of
the â-H-containing alkyls with a small amount of an
iron hydride, these alkyl complexes can be considered
as synthetic equivalents of a three-coordinate iron(II)
hydride complex. We have used this feature to synthe-
size a variety of low-coordinate iron complexes contain-
ing different alkyl, alkoxo, and amido ligands. Experi-
mental and theoretical data show that the driving force
for alkyl rearrangement and transfer is enthalpic.
Linear alkyls are preferred over their branched isomers,
unless resonance stabilization is possible. Resonance
(57) Agostic C-H‚‚‚M bonding affects the reactivity of many early-
and late-metal hydrocarbyl complexes: (a) Schmidt, G. F.; Brookhart,
M. J . Am. Chem. Soc. 1985, 107, 1443-1444. (b) Guo, Z.; Swenson, D.
C. Organometallics 1994, 13, 1424. (c) Grubbs, R. H.; Coates, G. W.
Acc. Chem. Res. 1996, 29, 85-93. (d) Calhorda, M. J . Chem. Commun.
2000, 801-809. (e) Shultz, L. H.; Brookhart, M. Organometallics 2001,
20, 3975-3982. (f) J ensen, V. R.; Thiel, W. Organometallics 2001, 20,
4852-4862. (g) Yu, Z.-X.; Houk, K. N. Angew. Chem., Int. Ed. 2003,
42, 808-811.
(58) (a) Mechria, A.; Bavoux, C.; Bouachir, F. J . Organomet. Chem.
2003, 677, 53-56. (b) Lee, B. Y.; Bazan, G. C.; Vela, J .; Komon, Z. J .
A.; Bu, X. J . Am. Chem. Soc. 2001, 123, 5352-5353. (c) Lesueur, W.;
Solari, E.; Floriani, C.; Chiesi-Villa, A.; Rizzoli, C. Inorg. Chem. 1997,
36, 3354-3362. (d) Bhandari, G.; Rheingold, A. L.; Theopold, K. H.
Chem. Eur. J . 1995, 1, 199-203.
(59) In diimine iron species, agostic interactions are proposed to
stabilize the alkyl species that are intermediates in R-olefin polymer-
ization:10,49 (a) Griffiths, E. A. H.; Britovsek, G. J . P.; Gibson, V. C.;
Gould, I. R. Chem. Commun. 1999, 1333-1334. (b) Khoroshun D. V.;
Musaev, D. G.: Vreven, T.; Morokuma, K. Organometallics 2001, 20,
2007-2026.
(60) We cannot rule out the possibility that LMeFeCHMePh (10a ) is
stabilized somewhat by π stacking between the phenyl rings on the
hydrocarbyl and diketiminate ligands. In the X-ray structure of 10a ,
the CAr-CAr′ contacts are between 3.61 and 4.98 Å. Two distances,
C32-C44 (3.751 Å) and C42-C54 (3.615 Å), lie within the 3.3-3.8 Å
range typical of π stacking interactions. See: J aniak, C. J . Chem. Soc.,
Dalton Trans. 2000, 3885-3896.
(61) Other aromatic groups such as pyrrole and pyridine also direct
metal binding to the “benzylic” carbon atom: (a) Settambolo, R.; Pucci,
S.; Bertozzi, S.; Lazzaroni, R. J . Organomet. Chem. 1995, 489, C50-
C51. (b) Settambolo, R.; Caiazzo, A.; Lazzaroni, R. J . Organomet. Chem.
1996, 506, 337-338.
(62) (a) Mayer, J . M. Comments Inorg. Chem. 1988, 8, 125-135. (b)
Caulton, K. G. New J . Chem. 1994, 18, 25-41.
(63) Values of pKa were obtained from ref 31b.
(64) Harvey, J . N. Organometallics 2001, 20, 4887-4895.

Reversible Beta-Hydrogen Elimination
Organometallics, Vol. 23, No. 22, 2004 5237
magnesium chloride (2 M in diethyl ether), phenethylmagne-
sium chloride (1 M in THF), tert-butylmagnesium chloride (2
M in diethyl ether), cyclohexylmagnesium chloride (2 M in
diethyl ether), ethylene (99.5+%), propylene (99+%), and 3,3,3-
trifluoropropene (99%). 2-Methylpropene (isobutylene, 99%)
was further purified by vacuum transfer into a Schlenk bomb
containing activated molecular sieves and stored overnight
prior to use. p-Methyl- and p-methoxystyrene were purchased
from Aldrich and were degassed and passed through activated
alumina prior to use. p-Dimethylaminostyrene was prepared
from p-(dimethylamino)benzaldehyde and triphenylmeth-
ylphosphonium bromide as described in the literature.68
Compounds LMeFe(µ-Cl)2Li(THF)2 (1a ),7g [LMeFeCl]2,7g LtBuFeCl
stabilization of the partial negative charge on the iron-
bound carbon apparently can overpower even severe
steric effects.
These metal-carbon bonding preferences are of rel-
evance to the degree of branching in polyolefins made
with late-metal catalysts. During the development of
homogeneous, late-metal R-olefin polymerization, it was
noted that the iron and cobalt catalysts gave linear
polymers,10 whereas nickel and palladium catalysts
were able to produce polymer with a high degree of
branching.9,47 Brookhart’s studies on group 10 metals
have shown that “chain walking” (through reversible
â-HE) occurs faster than insertion,47 and branching is
generally a consequence of the rate of chain walking
relative to the rate of olefin insertion.9 The linear nature
of the iron- and cobalt-generated polymers would then
imply that little chain walking takes place. The studies
reported here show that three-coordinate alkyliron(II)
complexes show a strong thermodynamic preference for
linear alkyl isomers, and this feature may be shared
by unsaturated intermediates in the polymerization
process. The fast rate of reversible â-HE in these low-
coordinate iron(II) alkyls suggests that catalytically
capable iron complexes could also undergo substantial
chain walking, yet the population of growing chains may
be skewed so far toward the thermodynamically favored
n-alkyl isomers that essentially all chain growth occurs
from them. Of course, the magnitude of this preference
is likely to depend on the metal and coligands, as these
intimately affect M-C bond energies and insertion
barriers.
(1b),7g LtBuFetBu (2b),6 LMeFeiBu (3a ),6 LMeFeEt (4a ),6 LtBu
-
FeCH2tBu (8b),8 and LtBuFeMe (7b),7d,e were prepared by
known procedures.
LMeF eCH2tBu (8a ). 8a was prepared by a similar procedure
to that reported for LMeFe-iBu6 and LtBuFeCH2tBu.8 A solution
of neo-pentyllithium (34.0 mg, 431 µmol) in diethyl ether (4
mL) was added to a solution of LMeFe(µ-Cl)2Li(THF)2 (300 mg,
431 µmol) in diethyl ether (4 mL). After 4 h stirring at room
temperature, the mixture was filtered through Celite and the
solvent pumped down. The bright yellow solid was extracted
with pentane (2 mL) and cooled to -38 °C to give crystals (188
mg, 63% yield). Anal. Found(calcd): C, 74.57(74.98); H, 8.63-
(9.62); N, 5.57(5.14). µeff (C6D6) ) 5.8(3) µB. 1H NMR (C6D6, 21
°C): 130 (1, R-CH), 116 (9, n-pentyl-(CH3)3), 72 (6, (CH3)2-L),
-12 (4, m-CH), -14 (12, iPr-CH3), -72 (2, p-CH), -103 (12,
iPr-CH3), -145 (4, iPr-CH). Vis (toluene): 467 nm (720
M-1cm-1), 492 nm (930 M-1cm-1). The other alkyl complexes
were synthesized through an analogous route; the alkyl donor
is listed.
LMeF eCHMeP h (10a ). 10a was prepared from phenethyl-
magnesium chloride; yield 99%. Anal. Found(calcd): C, 76.90-
1
(76.80); H, 8.47(8.74); N, 4.89(4.84). µeff (C6D6) ) 6.0(3) µB. H
NMR (C6D6, 21 °C): 132 (1, R-CH), 60 (6, (CH3)2-L), 45 (2, Ph-
o- or m-CH), 31 (2, Ph-o- or m-CH), -11 (2, L(Ar)-m-CH), -11
(2, L(Ar)-m-CH), -19 (6, iPr-CH3), -19 (6, iPr-CH3), -57 (1,
Ph-p-CH), -77 (2, L(Ar)-p-CH), -98 (2, iPr-CH), -119 (6, iPr-
CH3), -132 (6, iPr-CH3), -155 (2, iPr-CH).69 Vis (toluene): 470
nm (2040 M-1 cm-1).
LMeF eMe (7a ). 7a was prepared from methylmagnesium
chloride; yield 90%. Anal. Found(calcd): C, 73.55(73.76); H,
9.21(9.08); N, 5.78(5.73). µeff (C6D6) ) 6.4(4) µB. 1H NMR (C6D6,
21 °C): 130 (1, R-CH), 72 (6, (CH3)2-L), -11 (4, m-CH), -20
(12, iPr-CH3), -74 (2, p-CH), -119 (16, iPr-CH3, iPr-CH). Vis
(pentane): 466 nm (740 M-1cm-1), 491 nm (1080 M-1cm-1).
LMeF en P r (14a ). 14a was prepared from isopropylmagne-
sium chloride; yield 81%. Anal. Found(calcd): C, 74.11(74.40);
H, 9.59(9.37); N, 5.15(5.42). µeff (C6D6) ) 5.8(3) µB. 1H NMR
(C6D6, 21 °C): 130 (1, R-CH), 96 (3, nPr-CH3), 69 (6, (CH3)2-
L), -12 (4, m-CH), -19 (12, iPr-CH3), -75 (2, p-CH), -120
(16, iPr-CH3, iPr-CH). Vis (toluene): 463 nm (760 M-1 cm-1),
490 nm (930 M-1 cm-1).
LMeF en Bu (15a ). 15a was prepared from n-butyllithium,
yield 97%. Anal. Found(calcd): C, 74.66(74.70); H, 9.74(9.50);
N, 5.29(5.28). µeff (C6D6) ) 5.9(3) µB. 1H NMR (C6D6, 21 °C):
130 (1, R-CH), 91 (2, nBu-γ-CH2), 69 (6, (CH3)2-L), 65 (3, nBu-
ω-CH3), -12 (4, m-CH), -19 (12, iPr-CH3), -75 (2, p-CH), -119
(16, iPr-CH3, iPr-CH). Vis (toluene): 464 nm (820 M-1 cm-1),
491 nm (700 M-1 cm-1).
LMeF eCy (16a ). 16a was prepared from cyclohexylmagne-
sium chloride; yield 84%. Anal. Found(calcd): C, 75.05(75.52);
H, 9.43(9.42); N, 5.03(5.03). µeff (C6D6) ) 5.7(3) µB. 1H NMR
(C6D6, 21 °C): 134 (1, R-CH), 115 (4, Cy-γ-CH2), 64 (6, (CH3)2-
L), -13 (4, m-CH), -18 (12, iPr-CH3), -77 (2, p-CH), -120
Exp er im en ta l Section
Gen er a l Con sid er a tion s. Manipulations were performed
under a nitrogen atmosphere by standard Schlenk techniques
or in an M. Braun Unilab N2-filled glovebox maintained at or
below 1 ppm of O2 and H2O. Glassware was dried at 130 °C
overnight. H and 19F NMR data were recorded on a Bruker
1
Avance 400 MHz spectrometer at the specified temperature.
1H shifts are reported in ppm relative to residual protiated
solvent in C6D6 (7.13 ppm) or THF-d8 (3.58 ppm). 19F shifts
are referenced to R,R,R-trifluorotoluene (δ -63.73 ppm). In the
kinetic and equilibrium studies, the NMR probe temperature
was calibrated using the ethylene glycol method.65 Solution
magnetic susceptibilities were determined at 294 K by the
Evans method.66 Microanalyses were performed at Desert
Analytics (Tucson, AZ). Pentane, diethyl ether, tetrahydrofu-
ran (THF), toluene, and acetonitrile were purified by passage
through activated alumina and “deoxygenizer” columns from
Glass Contour Co. (Laguna Beach, CA). Deuterated benzene
and THF were dried over CaH2, then over Na, and then
vacuum distilled into a storage container or directly into the
NMR tube. tert-Butyl chloride-d9 was purchased from Cam-
bridge Isotope Labs and used after filtration through Celite.
tert-Butyl-d9-magnesium chloride, neo-pentyllithium, and bis-
(phenethyl)magnesium were prepared by known procedures.67
The following compounds were purchased from Aldrich and
used as received: Ethylmagnesium chloride (2 M in THF),
isopropylmagnesium chloride (2 M in diethyl ether), n-butyl-
(65) (a) Ammann, C.; Meier, P.; Merbach, A. E. J . Magn. Reson.
1982, 46, 319-321. (b) Kaplan, M. L.; Bovey, F. A.; Cheng, H. N. Anal.
Chem. 1975, 47, 1703-1705.
(66) (a) Schubert, E. M. J . Chem. Educ. 1992, 69, 62. (b) Evans, D.
F. J . Chem. Soc. 1959, 2003-2005.
(68) Iida, T.; Itaya, T. Tetrahedron 1993, 49, 10511-10530.
(69) Although the stereogenic CR makes all proton groups chemically
and magnetically inequivalent, some signals are unresolved because
of peak broadness caused by the paramagnetic iron center.
(67) (a) Wakefield, B. J . Organomagnesium Methods in Organic
Synthesis; Academic Press: London, 1995. (b) Wakefield, B. J . Orga-
nolithium Methods; Academic Press: London, 1988.

5238 Organometallics, Vol. 23, No. 22, 2004
Vela et al.
(12, iPr-CH3), -130 (4, iPr-CH). Vis (toluene): 461 nm (1280
M-1 cm-1), 486 nm (1080 M-1 cm-1).
80 °C overnight. The solvent was pumped down, and the
remaining bright orange solid was extracted with pentane (3
mL), filtered through Celite, and cooled to -38 °C to give
crystals (246 mg, 81%). Anal. Found(calcd): C, 76.63(76.19);
H, 10.20(10.17); N, 4.65(4.56). µeff (C6D6) ) 5.7(3) µB. 1H NMR
(C6D6, 21 °C): 110 (1, R-CH), 121 (18, C(CH3)3-L), -5 (4,
m-CH), -28 (12, iPr-CH3), -107 (6, iPr-CH, p-CH), -118 (12,
iPr-CH3). Vis (toluene): 521 nm (660 M-1 cm-1).
d 9-LMeF eiBu (d 9-3a ). d9-tert-Butylmagnesium chloride (735
µmol, prepared in situ) was filtered into a diethyl ether
solution containing 1a (460 mg, 662 µmol). After 2 h stirring
at room temperature, the white precipitate was allowed to
settle and the solution was filtered through Celite. Concentra-
tion under vacuum afforded a solid that was extracted with
pentane (4 mL), filtered, concentrated (to 2 mL), and cooled
Syn th eses of Alk yl, Alk oxid e, a n d Am id o Com p lexes
via Alk yl Tr a n sfer Rea ction s: LMeF eCH2CH2CF 3 (9a ). A
Schlenk flask was loaded with LMeFe-iBu (200 mg, 377 µmol),
toluene (4 mL), and 3,3,3-trifluoropropene (condensed from a
calibrated volume bulb: 380 µmol), then heated to 80 °C
overnight. After pumping down the solvent, the remaining
solid was extracted with pentane (3 mL), filtered through
Celite, and cooled to -38 °C to give yellow crystals (174 mg,
80% yield). Anal. Found(calcd): C, 66.89(67.36); H, 8.36(7.95);
1
to -38 °C, affording crystals: 211 mg (59%). H NMR (C6D6,
21 °C): 130 (1, R-CH), 70 (6, (CH3)2-L), -12 (4, m-CH), -18
(12, iPr-CH3), -74 (2, p-CH), -114 (12, iPr-CH3), -132 (4, iPr-
CH).
d 9-LtBu F etBu (d 9-2b). d9-2b was made in a similar way to
d9-LMeFeiBu from d9-tert-butylmagnesium chloride (368 µmol,
prepared in situ) and 1b (175 mg, 294 µmol): 137 mg (77%).
1H NMR (C6D6, 21 °C): 130 (1, R-CH), 46 (18, C(CH3)3-L), -7
(4, m-CH), -26 (12, iPr-CH3), -102 (6, iPr-CH, p-CH), -150
(12, iPr-CH3).
1
N, 5.11(4.91). µeff ) 5.6(3) µB. H NMR (C6D6, 21 °C): 125 (1,
R-CH), 66 (6, (CH3)2-L), -9 (4, m-CH), -20 (12, iPr-CH3), -79
(2, p-CH), -121 (16, iPr-CH3, iPr-CH). 19F NMR (C6D6, 21
°C): 128 (CF3). Vis (toluene): 460 nm (710 M-1 cm-1), 490 nm
(921 M-1 cm-1). Heating a solution of 9a at 80 °C for extended
periods of time did not result in formation of new compounds.
Formation of the secondary isomer LMeFeCH(CH3)(CF3) would
1H NMR Obser va tion of LMeF eCH2CH2P h (p r im ). Bis-
(phenethyl)magnesium (22 mg, 99 µmol) and [LMeFeCl]2 (92
mg, 90 µmol) were mixed in C6D6 (2 mL) inside a vial. The
vial was stirred at room temperature for 10 min, and then the
mixture was allowed to settle for 5 min. The soluble fraction
was decanted and filtered and the resulting yellow solution
monitored by 1H NMR at this point. The primary complex was
observed initially, and upon heating to 60 °C, complete
conversion to the equilibrium mixture was achieved within 2
h. Spectral assignment for LMeFeCH2CH2Ph (p r im ): 1H NMR
(C6D6, 21 °C): 128 (1, R-CH), 92 (2, (Ph)-o- or m-CH), 67 (6,
(CH3)2-L), 40 (2, (Ph)-o- or m-CH), 31 (1, (Ph)-p-CH), -11 (4,
L(Ar)-m-CH), -20 (12, iPr-CH3), -76 (2, L(Ar)-p-CH), -120
(16, iPr-CH and iPr-CH).
1
have resulted in a loss of symmetry of the HNMR spectrum
as seen for LMeFeCH(CH3)(Ph) (10a ) (see above).
LMeF eOCHP h 2 (11). 11 was synthesized similarly, from
LMeFeiBu (238 mg, 445 µmol) and dry benzophenone (81.2 mg,
445 µmol); green crystals (262 mg, 89% yield). Anal. Found-
(calcd): C, 76.58(76.81); H, 8.00(7.98); N, 4.19(4.27). µeff ) 5.1-
1
(3) µB. H NMR (C6D6, 21 °C): 110 (1, R-CH), 110 (4, (OR)-o-
CH), 52 (6, (CH3)2-L), 29 (4, (OR)-m-CH), 21 (2, (OR)-p-CH),
-16 (4, L(Ar)-m-CH), -23 (12, iPr-CH3), -75 (2, L(Ar)-p-CH),
-119 (16, iPr-CH3, iPr-CH). Vis (toluene): 492 nm (540 M-1
cm-1).
Su bstitu ted P h en eth yl Com p lexes. Syn th esis. A re-
sealable J . Young NMR tube was loaded with LMeFeCy (12 mg,
25 µmol), the desired styrene (1.02 equiv), and C6D6 (0.4 mL).
The tube was agitated for a day at room temperature or
warmed to 50 °C for 3 h, achieving in both cases clean
conversion to a mixture of benzylic/primary complexes.
LMeF eN(CH3)CH2P h (12). 12 was synthesized from LMe
-
FeiBu (400 mg, 754 µmol) and N-benzylidenemethylamine (98
µL, 754 µmol); orange crystals (317 mg, 72% yield). Anal.
Found(calcd): C, 74.42(74.98); H, 8.80(8.50); N, 6.80(7.09). µeff
) 5.6(3) µB. 1H NMR (C6D6, 21 °C): 116 (1, R-CH), 87 (2, (NR)-
o-CH), 35 (2, (NR)-m-CH), 27 (1, (NR)-p-CH), 7 (6, (CH3)2-L),
-13 (4, m-CH), -20 (12, iPr-CH3), -81 (2, p-CH), -119 (16,
iPr-CH3, iPr-CH). Vis (toluene): 373 nm (6200 M-1 cm-1), 422
nm (3400 M-1 cm-1), 491 nm (760 M-1 cm-1).
1
LMeF eCHMe(p-Me-C6H5). H NMR (C6D6, 21 °C): 116 (1,
R-CH), 86 (3, (Ph)-p-CH3), 62 (6, (CH3)2-L), 45 (2, (Ph)-o- or
m-CH), 30 (2, (Ph)-o- or m-CH), -12 (doublet, 2 × 2, L(Ar)-
m-CH), -20 (6H, iPr-CH3), -20 (6H, iPr-CH3), -78 (2, L(Ar)-
p-CH), -102 (2, iPr-CH), -122 (6, iPr-CH3), -111 (6, iPr-CH3),
-160 (2, iPr-CH).69
[LMeF e(µ-CH2CN)]2 (13). A Schlenk flask was loaded with
LMeFeiBu (31a ) (200 mg, 387 µmol), acetonitrile (16.0 mg, 387
µmol), and pentane (4 mL) and heated to 60 °C overnight with
constant stirring. The solution was filtered through Celite,
concentrated (2 mL), and cooled to -38 °C to give yellow
crystals (125 mg, 73% yield). Anal. Found(calcd): C, 71.96-
(72.50); H, 8.66(8.48); N, 8.88(8.44). µeff (C6D6, per dimer) )
LMeF eCHMe(p-MeO-C6H5). 1H NMR (C6D6, 21 °C): 116
(1, R-CH), 62 (6, (CH3)2-L), 44 (2, (Ph)-o- or m-CH), 25 (2, (Ph)-
o- or m-CH), 17 (3, (Ph)-p-OCH3), -11 (doublet, 2 × 2, L(Ar)-
m-CH), -19 (doublet, 2 × 6H, iPr-CH3), -78 (2, L(Ar)-p-CH),
-100 (2, iPr-CH), -122 (6, iPr-CH3), -112 (6, iPr-CH3), -135
(2, iPr-CH).69
1
LMeF eCHMe(p-NMe2-C6H5). H NMR (C6D6, 21 °C): 116
1
6.5(3) µB. H NMR (C6D6, 21 °C): 23, 20, 0.1, -6.0, -29, -41,
(1, R-CH), 59 (6, (CH3)2-L), 43 (2, (Ph)-o- or m-CH), 20 (2, (Ph)-
o- or m-CH), 16 (6, (Ph)-p-N(CH3)2), -11 (2, L(Ar)-m-CH), -11
(2, L(Ar)-m-CH), -18 (6, iPr-CH3), -18 (6, iPr-CH3), -76 (2,
L(Ar)-p-CH), -105 (2, iPr-CH), -119 (6, iPr-CH3), -129 (6,
iPr-CH3), -160 (2, iPr-CH).69
-73; spectrum is complicated and difficult to assign, possibly
indicating impurities or a slow dynamic process in solution.
Vis (toluene): 435 nm (4100 M-1 cm-1).
Kin etics Stu d ies. A resealable NMR tube was loaded with
a solution of 3a , d9-3a , 2b, or d9-2b (ca. 0.04 M in C6D6). The
NMR spectrometer probe was equilibrated at the given tem-
perature,65 and 1H NMR spectra were collected periodically.
A capillary with a solution of complex LtBuFeCl in C6D6 was
used as an internal standard.70 No significant variation in the
total iron concentration versus internal standard was observed
over time, and no intermediates were detected. A plot of
starting material concentration (y) against reaction time (t)
was analyzed with KaleidaGraph v. 3.51. This curve was used
to find the best fit to the general first-order integrated kinetic
equation y ) M1 + M2[exp(-kt)], where M1 and M2 are
constants and k is the first-order rate constant.71 This fitting
Equ ilibr iu m Stu d ies. Internal-to-terminal ratios (Keq)
were obtained by 1H NMR at different temperatures65 for each
substituted pair. Total iron concentration was 63 mM in each
case. Integrations were compared against an internal standard
as done for the kinetics experiments. No systematic variation
of normalized total iron concentration ((2%) was observed
during the course of the experiments. Error bars for integra-
tions obtained from NMR spectra were calculated assuming a
confidence level of 1%.24 Error bars for thermodynamic pa-
rameters were calculated on the basis of the maximum and
miminum deviations of the van’t Hoff plot through the error
limits of the individual data points.
LtBu F eiBu (3b). 3b was prepared from thermal isomeriza-
tion of LtBuFetBu (2b) (300 mg, 488 µmol) in toluene (4 mL) at
(70) LtBuFeCl was used as an integration standard because its
relaxation time is similar to the compounds studied here.

Reversible Beta-Hydrogen Elimination
Organometallics, Vol. 23, No. 22, 2004 5239
procedure also yielded error bars as calculated from the
standard deviations from the individual data points. In the
experiments with volatile alkenes, the solution was frozen in
a liquid nitrogen bath, its headspace was evacuated, and then
a known amount of gaseous alkene was condensed in from a
calibrated volume bulb at 294 K. Alkene was identified in the
Monte Carlo75 technique to determine the lowest energy
structure. After an initial equilibration of 1000 time steps, the
molecular dynamics (MD) calculations were performed at 300
K for 200 000 iteration steps. The MD time step was 0.005 ps.
The lowest energy conformer was used as a starting point for
B3LYP/6-31G(d) geometry optimization of LMeFeR.
1
volatile materials extracted from the reaction mixture by H
NMR and GC/MS.
X-r a y Str u ctu r es. Crystalline samples were grown in the
glovebox from pentane or ether solutions at -38 °C. Each
sample was rapidly mounted under Paratone-8277 onto a glass
fiber and immediately placed in a cold nitrogen stream at -80
°C on the X-ray diffractometer. X-ray intensity data were
collected on a standard Bruker-axs SMART CCD Area Detec-
tor System equipped with a normal focus molybdenum-target
X-ray tube operated at 2.0 kW (50 kV, 40 mA). A total of 1121
frames of data (2424 for LMeFeCH2CH2CF3 and LMeN(Me)(CH2-
Ph)) were collected using a narrow frame method with scan
widths of 0.3° in ω. Frames were integrated to a maximum 2θ
angle of 56.6° with SAINT. The final unit cell parameters (at
-80 °C) were determined from the least-squares refinement
of three-dimensional centroids of >4000 reflections for each
crystal. Data were corrected for absorption with the SADABS77
program.
The space groups were assigned using XPREP, and the
structures were solved by direct methods and refined employ-
ing full-matrix least-squares on F2 (Bruker-axs, SHELXTL-
NT,78 version 5.10). All non-H atoms were refined with
anisotropic thermal parameters. Hydrogen atoms were in-
cluded in idealized positions. The structures refined to good-
ness-of-fit values and final residuals found in the Supporting
Information.
Alk yl Cr ossover Exp er im en ts. A solution of two com-
plexes (ca. 40 mM each) in 0.40 mL of C6D6 was heated to 60,
80, and 100 °C in an oil bath for overnight periods and
monitored by 1H NMR spectroscopy. Concentrations were
calculated from integration of resonances for alkyl-â-CH
protons (120 to 80 ppm), diketiminate backbone methyl (80 to
50 ppm), and tBu (30 to 50 ppm) groups and p-CH (-60 to
-75 ppm) from aryl rings in LMe
.
Com p u ta tion a l Meth od s. Calculations on L′FeR model
systems (L′ ) C3N2H5-) employed the Gaussian98 quantum
chemistry program package.72 The systems were modeled with
the B3LYP hybrid density functional73 and the 6-31G(d) all-
electron basis set, which includes an f function on iron. For
simple L′FeR complexes all plausible conformations were
manually generated and evaluated at the B3LYP/6-31G(d)
level of theory, and the lowest energy conformations were used.
Geometry optimizations were performed on all atoms without
constraint of symmetry. Frequencies were calculated to con-
firm all species as minima (zero imaginary frequencies) or
transition states (one imaginary frequency). All calculations
were performed on the quintet state of L′FeR, which was
indicated in previous experiments and calculations to be the
ground state.7b The unrestricted Kohn-Sham formalism was
employed. Calculated energetics include enthalpic and zero-
point energies determined at 1 atm and 298.13 K using
B3LYP/6-31G(d) analytical frequencies.
Ack n ow led gm en t. This work was sponsored by the
National Science Foundation (CHE-0112658 to P.L.H.
and CHE-0309811 to T.R.C.) and the Department of
Energy (FG02-97ER14811 to T.R.C.). We also thank the
A. P. Sloan Foundation (Research Fellowship to P.L.H.)
and the University of Rochester (Hooker Fellowship to
J .V.). T.R.C. acknowledges the Chemical Computing
Group for supplying the MOE program. We are indebted
to Prof. Bill J ones and Prof. J oe Dinnocenzo for insight-
ful comments, and Sandip Sur for assistance with
variable-temperature NMR.
The Titan74 program was used to build LMeFeR complexes.
The molecular structures were minimized using the MOE75
package and the MMFF76 force field. The minimized structure
was subjected to conformational analysis using the Hybrid
(71) Espenson, J . H. Chemical Kinetics and Reaction Mechanisms,
2nd ed.; McGraw-Hill: New York, 1995.
(72) Frisch, M. J .; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J . R.; Zakrzewski, V. G.; Montgomery, J . A.;
Stratmann, R. E.; Burant, J . C.; Dapprich, S.; Millam, J . M.; Daniels,
A. D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J .; Barone, V.;
Cossi, M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford,
S.; Ochterski, J .; Petersson, G. A.; Ayala, P. Y.; Cui, Q.; Morokuma,
K.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J . B.;
Cioslowski, J .; Ortiz, J . V.; Stefanov, B. B.; Liu, G.; Liashenko, A.;
Piskorz, P.; Komaromi, I.; Gomperts, R.; Martin, R. L.; Fox, D. J .; Keith,
T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Gonzalez, C.;
Challacombe, M.; Gill, P. M. W.; J ohnson B. G.; Chen, W.; Wong, M.
W.; Andres, J . L.; Head-Gordon, M.; Replogle, E. S.; Pople, J . A.
Gaussian 98 (Revision A.1); Gaussian Inc.: Pittsburgh, PA, 2001.
(73) Becke, A. D. J . Chem. Phys. 1993, 98, 5648-5652.
Su p p or tin g In for m a tion Ava ila ble: Additional kinetics
and thermodynamics data, calculated geometries, and CIF
files. This material is available free of charge via the Internet
at http://pubs.acs.org.
OM049415+
(76) Halgren, T. A. J . Comput. Chem. 1995, 17, 490-519.
(77) The SADABS program is based on the method of Blessing;
see: Blessing, R. H. Acta Crystallogr. A 1995, 51, 33-38.
(78) SHELXTL NT: Structure Analysis Program, version 5.10;
Bruker-AXS: Madison, WI, 1995.
(74) Titan; Wavefunction, Inc., Schro¨dinger, Inc., 1999.
(75) MOE 2002.03; Chemical Computing Group Inc., http://ww-
w.chemcomp.com/.