Dynamic behavior of heterobimetallic derivatives of cycloheptatriene. Chemically induced switch of the vibration mode in molecular oscillators

Article abstract of DOI:10.1021/om050039i

Two new heterobimetallic complexes, (7-exo-triphenylstannyl)- η⁴-cycloheptatrienylirontricarbonyl (3) and (7-exo- triphenylstannyl)-η⁴-cycloheptatrienylrutheniumtricarbonyl (5), have been synthesized via the ligand exchange reactions from triphenyl(cycloheptatrienyl)-tin (1). Both 3 and 5 are fluxional. Facile [1,3]-Fe and [1,3]-Ru haptotropic shifts are the fastest dynamic processes observed in 3 and 5, respectively. Slower diatropic rearrangements involving simultaneous [1,7]-Sn and [1,2]-M (or [1,4]-M) migrations (M = Fe, Ru) were observed in each case. Activation parameters of the fast rearrangements in 3 and 5 were determined via 2D EXSY NMR (¹H. and ¹³C). The mode of intramolecular oscillations was switched in solution by reacting 3 with bromodimethylborane, which yielded cleanly (7-exophenyl(methyl)boryl)- η⁴-cycloheptatrienylirontricarbonyl (14), which exhibited a single type of intramolecular dynamics, viz., [1,7]-B + [1,2]-Fe diatropic rearrangement.

Full text of DOI:10.1021/om050039i

Organometallics 2005, 24, 4519-4527
4519
Dynamic Behavior of Heterobimetallic Derivatives of
Cycloheptatriene. Chemically Induced Switch of the
Vibration Mode in Molecular Oscillators
Ilya D. Gridnev* and Mia Karenina C. del Rosario
COE Laboratory, Department of Chemistry, Graduate School of Science, Tohoku University,
Aoba, Sendai, 980-8578, Japan
Received January 18, 2005
Two new heterobimetallic complexes, (7-exo-triphenylstannyl)-η4-cycloheptatrienyliron-
tricarbonyl (3) and (7-exo-triphenylstannyl)-η4-cycloheptatrienylrutheniumtricarbonyl (5),
have been synthesized via the ligand exchange reactions from triphenyl(cycloheptatrienyl)-
tin (1). Both 3 and 5 are fluxional. Facile [1,3]-Fe and [1,3]-Ru haptotropic shifts are the
fastest dynamic processes observed in 3 and 5, respectively. Slower diatropic rearrangements
involving simultaneous [1,7]-Sn and [1,2]-M (or [1,4]-M) migrations (M ) Fe, Ru) were
observed in each case. Activation parameters of the fast rearrangements in 3 and 5 were
determined via 2D EXSY NMR (1H and 13C). The mode of intramolecular oscillations was
switched in solution by reacting 3 with bromodimethylborane, which yielded cleanly (7-exo-
phenyl(methyl)boryl)-η4-cycloheptatrienylirontricarbonyl (14), which exhibited a single type
of intramolecular dynamics, viz., [1,7]-B + [1,2]-Fe diatropic rearrangement.
Introduction
design of substituents around the rotating bond6 or by
coordination of a certain part of the molecule with
metallic templates7 has been widely used for the con-
struction of molecular devices. Another approach is
based on the photochemically assisted rotation around
double bonds, whereas the direction of rotation can be
controlled by proper design of substituents.8 Supramo-
lecular devices usually apply reversible coordination of
the sliding parts of the molecule to the metal ions via
donor atoms.9 However, surprisingly, metallotropic
rearrangements have not been so far analyzed as the
possible source of a controlled intramolecular motion.
Various types of intramolecular motion are now under
intent attention of chemists in connection with the
efforts aimed at the creation of molecular devices
imitating the natural ATP-1 or DNA-based2 molecular
machines. This goal provides tremendous challenges for
the molecular design and synthesis of appropriate
molecules, and as it has been recently pointed out, the
realization of a truly operative molecular device (imply-
ing the necessity to overcome the stochastic behavior
of molecular assemblies) is probably quite a remote
prospect.3
Although the fastest metallotropic migrations are
comparable in their rates with conformational equilib-
ria, unlike the latter they are strictly regulated elec-
tronically.10 The simple MO theory correctly predicts the
modes of migration for different elements in the com-
pounds with either σ- or π-bonded organometallic
groups.10c The mode and rate of rearrangement can be
Nevertheless, the concept itself is useful in stimulat-
ing chemists to undertake brave and beautiful synthetic
projects. This situation is similar to the total synthesis
of huge natural products, however is much more sus-
ceptible to frustration, since quite often the achieved
synthetic goal does not possess the expected properties
necessary for being a prototype of a molecular device.4
On the other hand, the research in the area of the
potential molecular devices creates a more general and
elaborate view on the molecular structure and, in
particular, on the controlled intramolecular motion that
is a driving force of any molecular machine.5 Thus,
conformational rotation controlled either by clever
(6) (a) Kelly, T. R.; Tellitu, I.; Sestelo, J. P. Angew. Chem., Int. Ed.
Engl. 1997, 36, 1866. (b) Kelly, T. R.; Sestelo, J. P.; Tellitu, I. J. Org.
Chem. 1998, 63, 3655. (c) Kelly, T. R.; De Silva, H.; Silva, R. A. Nature
1999, 401, 150. (d) Kelly, T. R.; Silva, R. A.; De Silva, H.; Jasmin, S.;
Zhao, Y. A. J. Am. Chem. Soc. 2000, 122, 6935.
(7) Kelly, T. R.; Bowyer, M. C.; Bhaskar, K. V.; Bebbington, D.;
Garcia, A.; Lang, F.; Kim, M. H.; Jette, M. P. J. Am. Chem. Soc. 1994,
116, 3657.
(1) (a) Abrahams, J. P.; Leslie, A. G. W.; Lutter, R.; Walker, J. E.
Nature 1994, 370, 621. (b) Noji, H.; Yasuda, R.; Yoshida, M.; Kinoshita,
K., Jr. Nature 1997, 386, 299. (c) Rayment, I.; Holden, H. M.;
Whittacker, M.; Yohn, C.; Lorenz, M.; Holmes, K. C.; Milligan, R. A.
Science 1993, 261, 50.
(8) (a) Feringa, B. L. Acc. Chem. Res. 2001, 34, 504. (b) Koumura,
N.; Geertsema, E. M.; van Gelder, M. B.; Meetsma, A.; Feringa, B. L.
J. Am. Chem. Soc. 2002, 124, 5037.
(9) Ashton, P. R.; Ballardini, R.; Balzani, V.; Credi, A.; Dress, K.
R.; Ishow, E.; Kleverlaan, C. J.; Koichan, O.; Preece, J. A.; Spencer,
N.; Stoddart, J. F.; Venturi, M.; Wenger, S. Chem. Eur. J. 2000, 6,
3558.
(10) (a) Woodward, R. B.; Hoffmann, R. The Conservation of Orbital
Symmetry; Verlag Chemie: Weinheim, 1970. (b) Larrabee, R. B. J. Am.
Chem. Soc. 1971, 93, 1510. (c) Gridnev, I. D.; Schreiner, P. R.; Tok, O.
L.; Bubnov, Yu. N. Chem. Eur. J. 1999, 5, 2042. (d) Gridnev, I. D.;
Tok, O. L. In Fluxional Organometallic and Coordination Compounds;
Gielen, M., Willem, R., Wrackmeyer, B., Eds.; Wiley: The Atrium,
2004; pp 41-83.
(2) (a) Mao, C.; Sun, W.; Shen, Z.; Seeman, N. C. Nature 1999, 397,
144. (b) Yurke, B.; Turberfield, A. J.; Mills, A. P.; Simmel, F. C.;
Neumann, J. L. Nature 2000, 406, 605. (c) Feng, L.; Park, S. H.; Reif,
J. H.; Yan, H. Angew. Chem., Int. Ed. 2003, 42, 4342. (d) Chen, Y.;
Mao, C. J. Am. Chem. Soc. 2004, 126, 8626.
(3) Easton, C. J.; Lincoln, S. E.; Barr, L.; Onagi, H. Chem. Eur. J.
2004, 10, 3120.
(4) Kelly, T. R. Acc. Chem. Res. 2001, 34, 514.
(5) Balzani, V.; Credi, A.; Venturi, M. Chem. Eur. J. 2002, 8, 5525.
10.1021/om050039i CCC: $30.25 © 2005 American Chemical Society
Publication on Web 08/13/2005

4520 Organometallics, Vol. 24, No. 19, 2005
Gridnev and del Rosario
Scheme 1
tuned by appropriate choice of the organometallic group.
These features make the metallotropic rearrangements
promising candidates for the prototypes of molecular
motors. Cotton and Reich were first to demonstrate that
both metals in heterobimetallic complexes can be in-
volved in the synchronous intramolecular motion by the
analysis of the temperature-dependent 1H NMR spectra
of the bimetallic µ-η3:η4-Cp(CO)2Mo(C7H7)Fe(CO)3 com-
plex.11 They concluded that the most probable mecha-
nism of the observed fluxionality is the simultaneous
[1,2] shift of both metals with the probable admixture
of the [1,3] shift mechanism.11 Although quite a few
dynamic heterobimetallic complexes were characterized
afterward,12 the topology of the observed rearrange-
ments has been rarely probed, and the regulations
governing the selectivity of such migrations are un-
known so far. Recently, we have reported the coopera-
tive effects on the character of the intramolecular
dynamics in heterobimetallic complexes: introduction
of the second organometallic group changes the mode
of migration and can either facilitate13 or lock electroni-
cally14 the rearrangements occurring in an organome-
tallic molecule. In this paper we report the synthesis of
new heterobimetalic complexes and the experimental
analysis of their dynamic behavior and demonstrate the
possibility to switch chemically the mode of vibrations
in two-centered molecular oscillators.
Figure 1. ORTEP diagram of complex 3 (50% thermal
ellipsoids). Selected bond lengths (Å): Sn(1)-C(1) 2.217-
(3), C(1)-C(7) 1.472(4), C(1)-C(2) 1.498(4), C(2)-C(3)
1.438(4), C(3)-C(4) 1.398(4), C(4)-C(5) 1.428(5), C(5)-C(6)
1.474(4), C(6)-C(7) 1.335(4), Fe(1)-C(2) 2.172(3), Fe(1)-
C(3) 2.059(3), Fe(1)-C(4) 2.045(3), Fe(1)-C(5) 2.150(3).
Scheme 2
Results
Synthesis and Thermal Rearrangements of Iron-
tricarbonyl and Rutheniumtricarbonyl Complexes
of Triphenyl(cycloheptatrienyl)tin. Iron complex 3
was synthesized by the exchange reaction of triphenyl-
(cycloheptatrienyl)tin (1) with a 2-fold excess of tricar-
bonyl(cyclooctatetraene)iron (2) (Scheme 1). The target
complex 3 is unstable in the air and yields quantita-
tively tricarbonyl(η4-cycloheptatrienyl)iron when passed
through the column. Therefore, to get pure 3, it was
necessary to use a 2-fold excess of 2 to shift the
equilibrium toward the formation of 3 and remove the
excess of 2 by sublimation after the reaction was com-
plete. The structure of 3 has been established by a com-
bination of spectral methods and confirmed by micro-
analysis and single-crystal X-ray structure (Figure 1).
Reaction of tricarbonyl(cyclooctatetraene)ruthenium
(4) with 1 afforded the ruthenium complex 5 (Scheme
2), but the reaction required harder conditions that led
to thermal rearrangements of 5 (vide supra). We have
found that a more convenient access to 5 is the reaction
of 1 with a 1,5-cyclooctadiene(tricarbonyl)ruthenium, 6
(Scheme 2). This exchange reaction proceeds in very
mild conditions (0.5 h at 50 °C). The resulting complex
5 can be chromatographed in the air to give spectrally
pure samples. However, it decomposes slowly in solu-
tion, which prevented its recrystallization and mi-
croanalysis.
Both complexes undergo further rearrangements
upon heating over 80 °C. The iron complex 3 gives
cleanly (η5-C7H7)Fe(CO)2SnPh3 (7)15 (Scheme 3). A
similar transformation yielding (η5-C7H7)Ru(CO)2SnPh3
(8)15 occurs to some extent with the ruthenium complex
5; however the main product of the thermal rearrange-
ment of 5 is a trimetallic complex 9 (Scheme 3). The
(11) Cotton, F.; Reich, C. R. J. Am. Chem. Soc. 1969, 91, 847.
(12) (a) Bennett, M. J.; Pratt, J. L.; Simpson, K. A.; LiShingMan,
L. K. K.; Takats, J. J. Am. Chem. Soc. 1976, 98, 4810. (b) Salzer, A.;
Egolf, T.; von Philipsborn, W. Helv. Chim. Acta 1982, 65, 1145. (c) Ball,
R. G.; Edelmann, F.; Kiel, G.-Y.; Takats, J.; Drews, R. Organometallics
1986, 5, 829. (d) Edelmann, F.; Takats, J. J. Organomet. Chem. 1988,
344, 351. (e) Astley, S. T.; Takats, J. J. Organomet. Chem. 1989, 363,
167. (f) Fu, W.; McDonald, R.; Takats, J.; Bond, A. H.; Rogers, R. D.
Inorg. Chim. Acta 1995, 229, 307. (g) Airoldi, M.; Beringhelli, T.;
Deganello, G.; Gennaro, G.; Moret, M.; Saiano F.; Sironi, A. Inorg.
Chim. Acta 1995, 229, 461. (h) Airoldi, M.; Deganello, G.; Gennaro,
G.; Moret, M.; Sironi, A. J. Chem. Soc., Chem. Commun. 1992, 850. (i)
Airoldi, M.; Deganello, G.; Gennaro, G.; Moret, M.; Sironi, A. Orga-
nometallics 1993, 12, 3964. (j) Wadepohl, H.; Galm, W.; Pritzkow, H.
Angew. Chem. 1990, 102, 701. (k) Wadepohl, H.; Galm W.; Wolf, A. J.
Organomet. Chem. 1993, 452, 193. (l) Wadepohl, H.; Galm, W.;
Pritzkow, H. Organometallics 1996, 15, 570.
(13) Gridnev, I. D.; Tok, O. L.; Gurskii M. E.; Bubnov, Yu. N. Chem.
Eur. J. 1996, 2, 1483.
(14) Gridnev, I. D.; Tok, O. L. J. Am. Chem. Soc. 2003, 125, 14700.

Heterobimetallic Derivatives of Cycloheptatriene
Organometallics, Vol. 24, No. 19, 2005 4521
Figure 2. ORTEP diagram of complex 9 (50% thermal ellipsoids). Selected bond lengths (Å): Sn(1)-Ru(1) 2.6490(6),
Ru(1)-Ru(2) 2.9060(7), Ru(1)-C(1) 2.218(5), Ru(1)-C(2) 2.214(5), Ru(1)-C(3) 2.364(6), Ru(1)-C(7) 2.332(6), Ru(1)-C(11)
1.886(6), Ru(1)-C(12) 1.894(6), Ru(2)-C(4) 2.237(6), Ru(2)-C(5) 2.171(5), Ru(2)-C(6) 2.265(6), Ru(2)-C(9) 1.920(8), Ru-
(2)-C(10) 1.928(7), Ru(2)-C(8) 1.909 (9). Selected bond angles (deg): Sn(1)-Ru(1)-Ru(2) 177.74(2), Ru(1)-Ru(2)-C(11)
174.6(2).
Scheme 3
complex 9 is fluxional, exhibiting only averaged signals
in the 1H and 13C NMR spectra, which are only slightly
broadened at -100 °C. The structure of 9 has been
determined by single-crystal X-ray analysis (Figure 2).
Figure 3. 1H-1H 2D EXSY NMR spectrum (400 MHz,
Silicon and germanium analogues of 9 were previously
toluene-d8, 308 K), of compound 3, mixing time 50 ms.
prepared by the reaction of cycloheptatriene with [{Ru-
Reference 1D spectra taken at 295 K are given for clarity.
(MMe3)(CO)4}2] (M ) Si, Ge);16 in our case complex 9 is
most probably formed by transfer of the irontricarbonyl
5; they can be distinguished experimentally by the
group from 5 to 8 (Scheme 3).
nature of the cross-peaks observed in the EXSY spectra
Topology and Kinetics of the Fastest Intramo-
(see Figures S9, S26).
lecular Rearrangements in Complexes 3 and 5.
Three cross-peaks observed in the phase-sensitive
Totally 13 rearrangements are possible in either 3 or
1H-1H or 13C-13C EXSY spectra of complex 3 (e.g.,
Figure 3) unequivocally attest to the facile [1,3]-Fe
(15) Muhandiram, D. R.; Kiel, G.-Y.; Aarts, G. H. M.; Saez, I. M.;
haptotropic shift constantly occurring in this molecule.
The rate of the rearrangement did not change with the
concentration of the sample (in the 0.01-0.6 mol L-1
range) or with the variation of the inert solvent (toluene,
Reuvers, J. G. A.; Heinekey, D. M.; Graham, W. A. G., Takats, J.;
McClung, R. E. D. Organometallics 2002, 21, 2687.
(16) (a) Howard, J.; Woodward, P. J. Chem. Soc., Dalton Trans.
1975, 59. (b) Brookes, A.; Knox, S. A. R.; Riera, V.; Sosinsky, B. A.;
Stone, F. G. A. J. Chem. Soc., Dalton Trans. 1975, 1641.

4522 Organometallics, Vol. 24, No. 19, 2005
Gridnev and del Rosario
Table 1. Crystallographic Data for 3 and 9
3
9
chemical formula
fw
C
28H22FeO3Sn
C30H22O5Ru2Sn
783.33
581.02
hexane
yellow plate
monoclinic
P 1 21/n 1
6.824(3)
29.68(1)
12.193(5)
90
solvent
hexane
yellow prism
monoclinic
P 1 21/a 1
17.043(7)
14.992(6)
23.022(9)
90
cryst color, habit
cryst syst
space group
a (Å)
b (Å)
c (Å)
R (deg)
â (deg)
103.030(1)
90
2406.2(1)
1.604
109.605(2)
90
γ (deg)
volume (Å3)
5541.3(39)
1.878
D(calcd) (g cm-1
λ (Å)
)
0.7107
Mo KR
173.2
0.7107
radiation
temperature
Z
Mo KR
223.1
4
8
cryst size (cm3)
0.30 × 0.15 × 0.08 0.20 × 0.20 × 0.20
abs coeff (µ) (mm-1
)
1.667
2.009
min./max. transmn
abs corr
0.678/0.875
multiscan
Rigaku/MSC
Mercury CCD
ω
0.467/0.669
multiscan
Rigaku Saturn
difractometer
diffraction geometry
2θ max (deg)
index ranges
ω
55.0
54.9
-6 e h e 6
37 e k e 34
14 e k e 9
9276
-22 e h e 19
-19 e k e 19
-29 e k e 26
39 905
no. reflns collected
no. indep reflns
R(F) (I > 2σ(I))
wR2 (I > 2σ(I))
goodness of fit
3166
8597
0.021
0.045
0.022
0.119
1.201
0.742
benzene, chloroform, methylene chloride, cyclohexane),
thus attesting to the truly intramolecular process.
Activation parameters of this rearrangement (Table
2) show that the [1,3]-Fe haptotropic shift in 3 is
significantly faster than similar rearrangements in
germanium derivatives17 or the parent compound.19 The
rate of the [1,3]-Fe shift in 3 is comparable to that of
the fastest so far known [1,3]-Fe shift in [η4-(C6H6NCO2-
Et)Fe(CO)3].18
Similarly, the rutheniumtricarbonyl complex 5 ex-
hibits [1,3]-Ru haptotropic shift, a rearrangement that
has not been observed previously. The activation bar-
riers of the [1,3]-Fe and [1,3]-Ru haptotropic shifts in 3
and 5 are almost equal (Table 1), and the slightly lower
rates of rearrangement in 5 compared to 3 are due to
the more negative enthropy of activation.
Figure 4. 13C-13C 2D EXSY NMR spectrum of complex
3 (75 MHz, toluene-d8, 308 K), mixing time 80 ms,
relaxation delay 3 s, number of scans 320. Traces of F2
projections (a) and (b) are taken in the positions shown in
the spectrum.
either 3 or 5 occur only in a pendant-like manner; that
is, the metaltricarbonyl group cannot move around the
cycle due to the break of conjugation at C7. Therefore,
any extra cross-peaks observed in the EXSY spectra at
elevated temperatures attest to additional dynamic
processes. From Figure 4 one can see that the carbon
atom C7 of complex 3 (not participating in the fast
exchange) has cross-peaks of comparable intensity with
C1 and C6. Inspecting the sets of the exchange atoms
for all 13 rearrangements possible in 3 (see Scheme S9),
one can conclude that all rearrangements involving â-
or γ-Sn migrations can be excluded, since their occur-
rence must be manifested by observation of the C7-C2,
C7-C5 or C7-C3, C7-C4 exchanges. Therefore, only four
rearrangements bringing the triphenyltin group to the
Analysis of Slower Rearrangements Taking Place
in 3 and 5. Applying higher temperatures (308-328 K)
together with relatively long mixing times (e.g., Figure
4) we detected other rearrangements occurring in 3 and
5. To elucidate their nature, it is important to realize
that the fast haptotropic shifts of the M(CO)3 group in
Table 2. Activation Parameters Determined in This Work
number of
points
compound
reaction
[1,3]-Fe
nucleus
solvent
T, K
EA,a kJ mol-1
ln Aa
3
1H
13C
1H
13C
1H
toluene-d8
C6D12
toluene-d8
C6D12
toluene-d8
268-313
278-308
293-333
293-318
273-313
9
8
8
6
6
71.4 ( 0.8
73.4 ( 5.6
71.0 ( 4.5
70.0 ( 8.5
70.0 ( 11
30.6 ( 0.5
31.4 ( 2.3
28.7 ( 1.8
27.5 ( 2.8
29.7 ( 5.4
5
[1,3]-Ru
14
[1,7]-B+[1,2]-Fe
a The estimated errors were assigned by investigating all potential sources of experimental error (temperature, sample purity, chemical
shift difference, number of independent measurements at each temperature, goodness of fit).20 The parameters capable of causing
uncertainties were manually varied, and the resulting effects of such variations on the activation parameters were summed to give the
estimated errors shown in the table.

Heterobimetallic Derivatives of Cycloheptatriene
Organometallics, Vol. 24, No. 19, 2005 4523
Scheme 4
Scheme 5
in the additional experiment with mixing time 0.01 ms
and the rate matrix derived using the Mestrec EXSY-
Calc software. The rate of the fastest process determined
in this experiment corresponds well to the value derived
from the previously determined activation parameters.
The mean values of the exchange rates for the slower
process show some diversity. Thus, the rates for ex-
changes C7-C1 and C7-C6 are almost equal and notably
higher than any other exchanges. Qualitatively this is
well illustrated by two projection traces shown in the
Figure 4: trace (a) shows that the intensities of the
cross-peaks C7-C1 and C7-C6 are very similar, and
trace (b) demonstrates how the C1-C7 cross-peak com-
pares to the C1-C2 and C1-C.5 We regard these data
as evidence testifying that the diatropic migrations in
3 are not selective, viz., the triphenyltin group migrates
in both directions with comparable speed. Indeed, if only
one diatropic rearrangement preceded (or followed) by
the [1,3]-Fe shift would occur, it should be expected that
only one exchange (either C7-C1 or C7-C6) would be
faster than the others, since this would be the only
exchange repeatedly observed in both rearrangements
(slow and fast+slow).
adjacent position (Scheme 4) can be regarded as can-
didates for the experimentally observed processes.
Qualitatively the set of the observed cross-peaks could
be explained by any of the processes A-D followed (or
preceded) by much faster [1,3]-M shift. On the other
hand, the combination of several slower rearrangements
cannot be excluded based only on the qualitative data.
We attempted the quantitative analysis of the multisite
exchange in the iron complex 3 by the 13C-13C EXSY
experiment carried out at 308 K (Figure 4). The use of
high temperature for the experiment has been restricted
by the simultaneously occurring fast migration and
thermal rearrangement of 3 (vide supra). Hence, the
challenge was to be able to measure the intensities of
quite weak cross-peaks in the 13C-13C EXSY to be
certain that their intensities are not affected by NOEs
and/or relayed NOEs. A sample with the maximal
possible concentration (0.6 mol L-1) has been used for
Thus, we conclude that the available data attest to
two diatropic rearrangements in 3 with a different
direction of tin migration. However we are unable to
distinguish between several possible combinations: A+C,
A+D, B+C, or B+D (Scheme 4).
this experiment. A series of preliminary H-1H EXSY
1
spectra with various mixing times (in the 10-150 ms
interval) has been carried out in order to determine the
maximal mixing time at that temperature not leading
to the saturation of the magnetization transfer due to
the fast process (80 ms). The intensity matrix obtained
by the integration of the experimental phase-sensitive
13C-13C spectrum is shown in the Scheme 5 together
with the row of intensities of diagonal peaks obtained
Looking for the slower rearrangements in the ruthe-
nium complex 5, we have studied its EXSY spectra at
relatively increased temperatures and longer mixing
1
times. We were able to observe in the H-1H EXSY
spectrum taken at 323 K (see Figure S18) low-intensity
cross-peaks; qualitatively the set of those was exactly
the same as in the EXSY spectra of the iron complex 3.
However, the broadness of the signals due to the fast
[1,3]-Ru migrations together with the limited thermal
stability of the samples did not confirm this observation
with the 13C-13C EXSY data.
(17) LiShingMan, L. K. K.; Reuvers, J. G. A.; Takats, J.; Deganello,
G. Organometallics 1983, 2, 28.
(18) Gu¨nther, H.; Wenzl, R. Tetrahedron Lett. 1967, 8, 4155.
(19) Karel, K. J.; Brookhart, M. J. Am. Chem. Soc. 1978, 100, 1619.

4524 Organometallics, Vol. 24, No. 19, 2005
Gridnev and del Rosario
Scheme 6
Reaction of the Irontricarbonyl Complex of
Cycloheptatrienyl(triphenyl)tin 3 with Bromodi-
methylborane; Topology and Kinetics of Intramo-
lecular Rearrangements in the Irontricarbonyl
Complex of Cycloheptatrienyl(phenyl)(methyl)-
borane. Earlier we reported a fast diatropic [1,7]-B +
[1,2]-Fe migration in the irontricabonyl complex of
cycloheptatrienyl(dipropyl)borane (10).13 Complex 3,
differing from 10 by the nature of the σ-bonded orga-
nometallic group, exhibits a different fast rearrange-
ment, viz., the facile [1,3]-Fe shift. We therefore inves-
tigated the possibility to switch the mode of the
intramolecular oscillations in the heterobimetallic com-
plex by transmetalation reaction in solution.
Previously we have tried to prepare the cyclohep-
tatrienylboron derivative by reacting 1 with chlorodipro-
pylborane.21 However, in that case the phenyl groups
exhibited much higher migration ability compared to
the cycloheptatrienyl substituent, and the exchange
reaction resulted in dialkylphenylboron species 11
(Scheme 5). The problem has been overcome by using
trimethyl(cycloheptatrienyl)tin; however the latter com-
pound is synthesized in a very low yield, and the
transmetalation reaction is complicated by further
exchange of substituents.21 Therefore it was tempting
to try 3 in the transmetalation reaction since the
electrophilic irontricarbonyl group is expected to in-
crease the reactivity of the Sn-C(cycloheptatrienyl)
bond in 3; this effect is manifested in its very facile
hydrolysis (vide supra).
Figure 5. Integration matrix of the 2D 13C-13C EXSY
spectrum shown in Figure 4 and rates matrix obtained via
matrix resolution. The auto-signal of C6 overlaps partially
with the tin satellite of the o-phenyl carbon and could not
be accurately integrated. It was assumed, therefore, that
its intensity equals that of C5.
Reaction of 3 with 2 equiv of bromodimethylborane
proceeded smoothly and quantitatively, yielding com-
plex 14 (Scheme 6). NMR monitoring of the reaction
(e.g., Figure 6) revealed the intermediate formation of
the expected complex 12 and bromotriphenyltin, which
apparently react further to give 14 and bromodiphenyl-
(methyl)tin (Scheme 6). The boron complex 14 has been
completely characterized by NMR in solution; attempts
to have it in a pure form were unsuccessful so far.
Figure 6. Monitoring of the transmetalation reaction by
119Sn NMR (112 MHz, toluene-d8, 295 K). Bottom: the
spectrum of the starting complex 3; middle: after 2 h; top:
after 4 h.
1
Figure 7 displays the H-1H EXSY NMR spectrum
diatropic [1,7]-B + [1,2]-Fe migration, was detected in
the whole range of temperatures and mixing times used
for the measurements (up to 318 K and 2 s mixing time).
As could be expected, the topology and the activation
barrier of the fastest rearrangement in 14 (Table 2)
of 14 taken at 288 K. Only one rearrangement, viz., the
(20) Hayes, P. G.; Piers, W. E.; Parvez, M. Organometallics 2005,
24, 1173.
(21) Gridnev, I. D.; Tok, O. L.; Gridneva, N. A.; Bubnov, Yu. N.;
Schreiner, P. R. J. Am. Chem. Soc. 1998, 120, 1034.

Heterobimetallic Derivatives of Cycloheptatriene
Organometallics, Vol. 24, No. 19, 2005 4525
boryl group the character of the intramolecular oscil-
lations becomes completely different: two organome-
tallic groups migrate simultaneously toward each other
due to a simultaneous (diatropic) [1,7]-B + [1,2]-Fe
migration. This switch takes place cleanly and quanti-
tatively at ambient temperature.
Discussion
The present work combined with the earlier results
demonstrates the features of a specific type of intramo-
lecular dynamics, viz., diatropic rearrangements in
binuclear complexes. The ability of two different orga-
nometallic groups to move simultaneously within a
single molecule in a highly organized manner is an
interesting phenomenon both from the theoretical point
of view and with respect to further implications. Al-
though the exact mechanism of the diatropic migrations
requires a separate theoretical study, it is quite clear
that physically the sophisticated intramolecular reor-
ganization occurs as a single event.
The cooperative effects between two organometallic
groups are clearly seen in the rates and modes of
intramolecular dynamics. The involvement of the π-sys-
tem of the cycloheptatrienyl ring in the coordination
with either the η6-bound chromiumtricarbonyl group or
the η4-bound iron- or rutheniumtricarbonyl group ef-
fectively locks fast [1,5]-Sn sigmatropic migrations of
the triphenyltin substituent. Nevertheless, the sigma-
tropic migration of tin is not completely suppressed in
these compounds: it is observed in diatropic migrations
taking place in 2, 3, and 5 with comparable rates and
the topology is determined by the mode of coordination
of the transition metal.
Figure 7. 1H-1H 2D EXSY NMR spectrum (300 MHz,
toluene-d8, 288 K) of compound 14, mixing time 500 ms.
parallel the dynamic behavior of 10; the slightly slower
rate of the [1,7]-B + [1,2]-Fe diatropic migrations can
be attributed to the conjugation of the unoccupied 2p-
AO of boron with the phenyl ring in 14. The same effect
probably explains why we were unable to detect slower
[1,3]-B and [1,3]-Fe shifts observable in 10.
Thus, we have accomplished a chemically induced
switch of the mode of metallotropicic rearrangements
in the fluxional molecule (Scheme 7). In the tin com-
pound 3 fast [1,3]-Fe migrations occur in a pendant-
like manner. When the tin group is exchanged to the
Moreover, from the point of view of haptotropic
migrations, the [1,3]-Fe shift is significantly faster in
complex 3 than in the parent compound19 or Ge- and
Si-substituted derivatives.17 The fast selective diatropic
migration in 10 and 14 clearly demonstrates the coop-
Scheme 7

4526 Organometallics, Vol. 24, No. 19, 2005
Gridnev and del Rosario
Further studies along these lines using different com-
binations of metals and other hydrocarbon frameworks
are underway in our laboratory.
Experimental Section
General Procedures. All experiments were performed
under a dry argon atmosphere. The samples for NMR experi-
ments were prepared in degassed deuterated solvents distilled
prior to use over the appropriate drying agent (Na for toluene
and cyclohexane, CaH2 for CDCl3 and CD2Cl2). 1H, 13C, 11B,
and 119Sn NMR spectra were recorded using JEOL AL-300 and
1
JEOL AL-400 spectrometers. Phase-sensitive H and 13C 2D
EXSY NMR spectra were acquired using a NOESYTP pulse
program, slightly modified in the case of 13C EXSY in order to
allow 1H decoupling during acquisition. The temperature
controller of the spectrometer was calibrated with the standard
samples containing 4% of methanol in methanol-d4 (263-293
K) and 80% of ethylene glycol in DMSO-d6 from Bruker. The
estimated uncertainty in temperature measurements was 0.5
K in the 263-288 and 308-333 K intervals and approximately
1 K at 288-308 K, where both methanol and ethylene glycol
calibration curves become nonlinear and give somewhat
contradictory results. The T1 relaxation times have been
Figure 8. Molecular orbitals for the molecules 3a and 14a
optimized at the MP2/SDD level of theory.
erative effects in the intramolecular dynamics: the [1,7]-
B + [1,2]-Fe rearrangement is faster than either of the
parent rearrangements.
1
measured for all exchanging signals in H and 13C NMR; the
relaxation times varied from 0.8 to 2.5 s. The relaxation delay
was maintained to be longer than 2(T1)av (mixing time aver-
aged for all exchanging signals).22 Mixing times were varied
in the interval 20-1000 ms depending on the exchange rate
at each temperature to optimize the signal-to-noise ratio. The
auto- and cross-peak volumes were determined after phase and
baseline correction of the two dimensions using JEOL Delta
software. The signals arising from tin satellites in all cases
except for the signals of H7 and C7 were integrated together
with the main signals. The tin satellites of H7 and C7 (in the
diagonal as well as in the cross-peaks) were integrated
separately from the main signal, and their volumes were
combined with those of the main peaks.
Of interest is the observed selectivity of the rear-
rangements taking place in the cycloheptatrienyl bime-
tallic complexes. Whereas totally 13 different rearrange-
ments are possible in compound 3, 5, or 14 (see Schemes
S9 and S26), only three rearrangements take place in
3 or 5, and a single rearrangement selectively occurs
in 14. Considering the molecular orbitals in boryl- and
stannyl-substituted iron complexes (Figure 8), one can
conclude that the symmetry of the HOMO must play
an important role in determining the nature of the
allowed migrations.
The HOMO of 3a is almost perfectly symmetrical;
moreover the isolobal orbital density covers the whole
area involved in the [1,3]-Fe haptotropic migration. In
the same area of the molecule 14a the orbital density
changes sign twice; therefore the simple [1,3]-Fe migra-
tion is disfavored in 14a compared to 3a. This conclusion
is in good agreement with the experimental observations
for 3 and 14. Furthermore, the nonsymmetrical distri-
bution of the orbital density with respect to the boron
atom in 14a corresponds to the exact nature of the
observed diatropic rearrangement. Thus, we conclude
that the selectivity of metallotropic rearrangements in
heterobimetallic complexes is regulated by the sym-
metry of frontier orbitals, similarly to simple metallo-
tropic rearrangements.
In complexes such as 3, 5, 11, and 14 two organome-
tallic groups are situated on opposite sides of the
hydrocarbon cycle, and both are involved in the syn-
chronous movement in a regular fashion. Therefore,
physically this molecular unit would be capable of a
rotational motion to a different axis or keeping the
rotation of the two axes synchronized. In addition,
chemically switching the mode of diatropic rearrange-
ment, it is possible to change the direction and speed of
rotation.
Rate constants at each temperature were derived from the
experimental intensities via exchange matrix resolution ap-
proach23 using Mestrec EXSYCalc software. For the determi-
nation of activation parameters of the fast rearrangements six
values obtained from three pairs of equally populated signals
arising from the same exchanging process were averaged. In
the 13C-13C EXSY spectra of complex 3 the auto-signal of C6
overlapped partially with the tin satellite of the o-phenyl
carbon; it was not used in the calculations. The accuracy in
calculation of rate constants is determined by the accuracy of
the measurements and the chemical shift differences. The
signals in our experiments were well resolved; manual varia-
tion of the experimental integral values in the input matrixes
according to the estimated uncertainties in the integration
produced errors not larger than 10% in the averaged rate
constants.
Complex 3. A Schlenk tube was filled with dry argon and
charged with triphenyl(cycloheptatrienyl)tin, 1 (228 mg, 0.52
mmol), and cyclooctatetraenyl(tricarbonyl)iron, 2 (189 mg, 0.78
mmol). Dry toluene (15 mL) was added, and the reaction
mixture was heated with stirring for 10 h at 100 °C. Then it
was cooled, and the solvent was removed in a vacuum. The
excess cyclooctatetraenyl(tricarbonyl)iron was sublimed to the
upper part of the Schlenk tube (100 °C, 1 mmHg), leaving a
brown oily residue at the bottom. It was collected with dry
toluene. After evaporation of toluene, the residue was recrys-
tallized from hexane to give 164 mg (54%) of pure 3 as bright
yellow crystals, mp 91-92 °C. Anal. Found: C, 57.99; H, 3.85.
Calcd for C28H22O3SnFe: C, 57.88; H, 3.79. 1H NMR (400 MHZ,
In conclusion, we have characterized the dynamic
behavior of new heterobimetallic complexes 3 and 5,
revealed some regularities in the selectivity of diatropic
migrations, and demonstrated the possibility to switch
chemically the mode of intramolecular rearrangements.
(22) Frick, A.; Schulz, V.; Huttner, G. Eur. J. Inorg. Chem. 2002,
3129.
(23) Perrin, C. L.; Dwyer, T. G. Chem. Rev. 1990, 90, 935.

Heterobimetallic Derivatives of Cycloheptatriene
Organometallics, Vol. 24, No. 19, 2005 4527
toluene-d8, 295 K): 2.82 (dd, 1H, H7, 3JHH ) 5.0, 5.6 Hz, 2JSnH
) 115, 110 Hz), 2.99 (dd, 1H, H4, 3JHH ) 7.5, 8.9 Hz Hz), 3.93
hexane-toluene, 3:2) to yield 90 mg (60%) of the trimetallic
complex 9 (Rf ) 0.4) as bright yellow crystals together with a
small amount of (η5-C7H7)Ru(CO)2SnPh3 (8).14 Crystals for the
X-ray analysis of 9 were obtained by slow recrystallization
from hexane. 1H NMR (300 MHz, CDCl3, 293 K): 3.96 (s, 7H,
7.29 (m, 9H), 7.42 (m, 6H). 13C NMR (75 MHz, toluene-d8, 295
K): 64.59 (C7), 128.86 (3CHp), 129.00 (6CHm), 137.48 (6CHo),
144.04 (3C-Sn), 196.92 (3CO), 205.05 (2CO). 119Sn NMR (112
MHz, toluene-d8, 295 K): 31.94. FT IR (ν, cm-1): 1942, 1975,
1996, 2073.
Complex 14. A NMR tube was filled with dry argon and
charged with (7-exo-triphenylstannyl)-η4-cycloheptatrienyliron-
tricarbonyl (3) (310 mg, 0.53 mmol). Dry toluene-d8 (1 mL) was
added followed by bromodimethylborane (175 mg, 1.45 mmol).
The reaction progress was controlled by NMR. After 4 h at
ambient temperature the sample contained only bromodiphen-
yl(methyl)tin (δ119Sn ) -72.1) and (7-exo-phenyl(methyl)boryl)-
η4-cycloheptatrienylirontricarbonyl, 14: 1H NMR (300 MHz,
toluene-d8, 295 K): 2.33 (dd, 1H, H7, 3JHH ) 4.2, 4.2 Hz,), 2.87
(dd, 1H, H4, 3JHH ) 7.5, 7.5 Hz), 2.97 (dd, 1H, H1, 3JHH ) 4.2,
3
4
3
(ddd, 1H, H1, JHH ) 5.6, 7.5 Hz, JHH ) 2.0 Hz, JSnH ) 22
3
4
Hz), 4.05 (ddd, 1H, H3, JHH ) 4.5, 7.5 Hz, JHH ) 1.2 Hz),
4.21 (ddd, 1H, H2, JHH ) 4.5, 7.5 Hz, JHH ) 1.2 Hz, JSnH
)
3
4
4
11 Hz), 5.40 (dddd, 1H,H6, JHH ) 5.0, 10.4 Hz, JSnH ) 20
3
3
Hz), 5.27 (dd, 1H, H5, JHH ) 8.9, 10.4 Hz, JSnH ) 35 Hz),
7.17 (m, 9H, 6CHm + 3CHp arom.), 7.46 (m, 6CHo arom, 3JSnH
) 46 Hz). 13C NMR (100 MHz, toluene-d8, 295 K): 36.65 (C7,
1JSnC ) 270, 282 Hz), 60.04 (C4, 4JSnC ) 14 Hz), 68.62 (C1, 2JSnC
) 22 Hz), 83.53 (C2, 3JSnC ) 22 Hz), 90.96 (C3, 4JSnC ) 10 Hz),
126.08 (C5, 3JSnC ) 47 Hz), 128.26 (C6, 2JSnC ) 47 Hz), 128.97
3
4
(2CHo arom., JSnC ) 35 Hz), 129. 53 (CHp arom., JSnC ) 12
Hz), 137.52 (2CHm arom., 3JSnC ) 49 Hz), 138.53 (C-Sn arom.,
1JSnC ) 453, 475 Hz), 212.82 (CO). 119Sn NMR (112 MHz,
toluene-d8, 295K): -135.76. FT IR (ν, cm-1): 1948, 1965, 1981.
Complex 5. A Schlenk tube was filled with dry argon and
charged with triphenyl(cycloheptatrienyl)tin, 1 (250 mg, 0.57
mmol), and 1,5-cyclooctadienyl(tricarbonyl)ruthenium, 4 (146
mg, 0.5 mmol). Dry toluene (15 mL) was added, and the
reaction mixture was heated with stirring for 1 h at 50 °C.
Then it was cooled, and the solvent was removed in a vacuum.
The residue was chromatographed on silica gel (eluent hexane:
toluene 3:1), collecting the fraction with Rf 0.7. 1H NMR (300
2
4
3
4
6.8 Hz), 4.45 (ddd, 1H, H3, JHH ) 5.1, 10.8 Hz, JHH ) 1.0
Hz), 4.49 (dd, 1H, H2, 3JHH ) 4.4, 7.6 Hz, 4JHH ) 1.2 Hz, 4JSnH
) 7 Hz), 5.02 (dd, 1H, H6, 3JHH ) 4.2, 10.4 Hz), 5.61 (ddd, 1H,
H5, JHH ) 10.4, 10.4 Hz, JHH ) 1.0 Hz), 7.10-7.30 (m, 3H,
2CHm + CHp), 7.80 (d, 2CHo, 3JHH ) 6.6 Hz). 13C NMR (75 MHz,
toluene-d8, 223 K): 13.14 (br, Me), 46.62 (br C7), 59.02 (C4),
64.19 (C1), 86.16 (C2), 92.24 (C3), 126.83 (C6), 130.41 (C5),
129.12 (2CH), 132.43 (CHp), 136.46 (2CH), 137.3 (br, C-B
arom.), 212.55 (CO). 11B NMR (MHz, toluene-d8, 295 K): 74.
3
4
MHz, toluene-d8, 295 K): 2.42 (ddd, 1H, H7, JHH ) 5.2, 5.2
3
Hz, JHH ) 1.1 Hz, JSnH ) 106, 111 Hz), 2.75 (dddd, 1H, H4,
4
2
3JHH ) 7.9, 7.9 Hz, JHH ) 1.0, 1.0 Hz), 3.93 (dddd, 1H, H1,
4
3JHH ) 5.3, 7.7 Hz, JHH ) 1.6, 1.6 Hz, JSnH ) 14.7 Hz), 4.06
4
3
(ddd, 1H, H3, JHH ) 4.5, 7.6 Hz, JHH ) 1.5 Hz), 4.19 (ddd,
3
4
1H, H2, 3JHH ) 4.4, 7.6 Hz, 4JHH ) 1.2 Hz, 4JSnH ) 7 Hz), 5.08
(dddd, 1H, H6, 3JHH ) 5.0, 10.4 Hz, 4JHH ) 1.0, 1.5 Hz, 3JSnH
)
Acknowledgment. This work was financially sup-
ported by the COE Program “Giant molecules and
complex systems” of MEXT hosted at Tohoku Univer-
sity, and by Grant-in-Aid for Scientific Research from
MEXT (17034006). The authors thank Professor M. Kira
of Tohoku University for fruitful and stimulating dis-
cussions, and Dr. C. Kabuto for X-ray structure deter-
mination. We are also grateful to the reviewers for
suggesting additional useful experiments.
19 Hz), 5.40 (dd, 1H, H5, JHH ) 7.9, 10.4 Hz, JHH ) 1.2 Hz,
3
4
4JSnH ) 27 Hz), 6.90 (m, 9H, 6CHm + 3CHp arom.), 7.23 (m,
6CHo arom, JSnH ) 43 Hz). 13C NMR (75 MHz, toluene-d8,
3
295 K): 33.98 (C7, 1JSnC ) 293, 306 Hz), 51.47 (C4, 4JSnC ) 14
Hz), 59.00 (C1, 2JSnC ) 20 Hz), 83.62 (C2, 3JSnC ) 22 Hz), 92.11
(C3, JSnC ) 12 Hz), 124.09 (C,5 JSnC ) 50 Hz), 126.57 (C6,
4
3
2JSnC ) 47 Hz), 127.54 (2CHo arom., JSnC ) 48 Hz), 128.02
2
(CHp arom., JSnC ) 11 Hz), 136.23 (2CHm arom., JSnC ) 35
4
3
1
Hz), 137.7 (C-Sn arom., JSnC ) 444, 468 Hz), 197.53 (CO).
119Sn NMR (112 MHz, toluene-d8, 295 K): -140.82. FT IR (ν,
cm-1): 1987 (br), 2054.
Supporting Information Available: Charts of all im-
portant NMR spectra, kinetic curves, analysis and classifica-
tion of possible rearrangements (pdf), and detailed X-ray data
(cif). This material is available free of charge via the Internet
at http://pubs.acs.org.
Complex 9. A solution of (7-exo-triphenylstannyl)-η4-cyclo-
heptatrienylirontricarbonyl, 3 (150 mg, 0.26 mmol), in toluene-
d8 (0.6 mL) was heated under argon for 1 h at 110 °C. Then
the reaction mixture was cooled, the solvent was evaporated,
and the residue was chromatographed on silica gel (eluent
OM050039I