Low-temperature NMR studies of Zn tautomerism and hindered rotations in solid zincocene derivatives

Article abstract of DOI:10.1021/jp711504g

Using a combination of NMR methods we have detected and studied fluxional motions in the slip-sandwich structure of solid decamethylzincocene (I, [(η-C₅Me₅)Zn(η¹-C₅Me ₅)]). For comparison, we have als

Full text of DOI:10.1021/jp711504g

J. Phys. Chem. A 2008, 112, 3557-3565
3557
Low-Temperature NMR Studies of Zn Tautomerism and Hindered Rotations in Solid
Zincocene Derivatives
Juan Miguel Lopez del Amo,† Gerd Buntkowsky,‡ Hans-Heinrich Limbach,*,† Irene Resa,§
Rafael Ferna´ndez,§ and Ernesto Carmona§
Institut fu¨r Chemie und Biochemie der Freien UniVersita¨t Berlin, Takustrasse 3, D-14195, Berlin, Germany,
Institut fu¨r Physikalische Chemie, Friedrich-Schiller UniVersita¨t, Jena, Helmholtzweg 4, D 07743 Jena,
Germany, and Instituto de InVestigaciones Qu´ımicas, CSIC, AVda. Ame´rico Vespucio 49, 41092 SeVilla, Spain
ReceiVed: December 6, 2007; In Final Form: February 6, 2008
Using a combination of NMR methods we have detected and studied fluxional motions in the slip-sandwich
structure of solid decamethylzincocene (I, [(η5-C5Me5)Zn(η1-C5Me5)]). For comparison, we have also studied
the solid iminoacyl derivative [(η5-C5Me5)Zn(η1-C(NXyl)C5Me5)] (II). The variable temperature 13C CPMAS
NMR spectra of I indicate fast rotations of both Cp* rings in the molecule down to 156 K as well as the
presence of an order-disorder phase transition around 210 K. The disorder is shown to be dynamic arising
from a fast combined Zn tautomerism and η1/η5 reorganization of the Cp* rings between two degenerate
states A and B related by a molecular inversion. In the ordered phase, the degeneracy of A and B is lifted;
that is, the two rings X and Y are inequivalent, where X exhibits a larger fraction of time in the η5 state than
Y. However, the interconversion is still fast and characterized by a reaction enthalpy of ∆H ) 2.4 kJ mol-1
and a reaction entropy of ∆S ) 4.9 J K-1 mol-1. In order to obtain quantitative kinetic information, variable
2
temperature H NMR experiments were performed on static samples of I-d6 and II-d6 between 300 and 100
K, where in each ring one CH3 is replaced by one CD3 group. For II-d6, the 2H NMR line shapes indicate fast
CD3 group rotations and a fast “η5 rotation”, corresponding to 72° rotational jumps of the η5 coordinated Cp*
ring. The latter motion becomes slow around 130 K. By line shape analysis, an activation energy of the η5
2
rotation of about 21 kJ mol-1 was obtained. H NMR line shapes analysis of I-d6 indicates fast CD3 group
rotations at all temperatures. Moreover, between 100 and 150 K, a transition from the slow to the fast exchange
regime is observed for the 5-fold rotational jumps of both Cp* rings, exhibiting an activation energy of 18
kJ mol-1. This value was corroborated by 2H NMR relaxometry from which additionally the activation energies
6.3 kJ mol-1 and 11.2 kJ mol-1 for the CD3 rotation and the molecular inversion process were determined.
Introduction
SCHEME 1: Molecular Structures of Metallocenes
Metallocenes constitute one of the most important families
of organometallic compounds.1 Their general structure corre-
sponds to MCp′2, where M represents a metal and Cp′ represents
any cyclo-pentadienyl-type ligand. Metallocenes are endowed
with a distinguished history, and moreover, they enjoy multiple
applications in different areas of chemistry, such as polymer-
ization catalysis, asymmetric catalysis (e.g., chiral ferrocenyl
ligands), bio-organometallic chemistry, and others.1 The met-
2
allocene era started in 1951 with the preparation of Fe(C5H5)2
and with the recognition by Fischer and Wilkinson et al. of its
correct, now well-known, π-bonded sandwich structure.3 In
1959, Fischer and co-workers reported the preparation of Be-
2+
a structure of type 4 was proposed only for the B(C5Me5)2
cation7 and more recently for Zn(η5-C5Me4SiMe3)(η1-C5Me4-
SiMe3).8 Most beryllocenes9 and zincocenes10 exhibit geometries
of type 3, the so-called slipped-sandwich structure, also referred
to as η5/η1(π). This peculiar structure is characterized by nearly
parallel Cp′ rings, by a relatively weak and long M-C bond to
the η1 ring and by extensive π-electron delocalization within
the latter. The question then arises whether the η1(π) binding
is closer to the η5 or to the η1 binding type. Unfortunately, it is
not easy to study this question experimentally.
4a
(C5H5)2 and Zn(C5H5)2,4b but it was only in the following
decades, with the development of the substituted bulky cyclo-
pentadienyl ligands,1,5 that the similarity between beryllocenes
and zincocenes became apparent. In contrast to the divalent
metallocenes of the transition elements that feature invariably
a sandwich structure, the analogous metallocenes of the main
group elements exhibit a variety of structures that include types
1-5, schematically depicted in Scheme 1.6 To our knowledge
Some of us have reported recently the synthesis and structural
characteristics of the methyl-substituted beryllocenes Be(C5-
Me5)2, Be(C5Me4H)2, and Be(C5Me5)(C5Me4H).9a With the only
exception of Be(C5Me5)2, which in the solid-state exhibits an
* Corresponding author. E-mail: juanmi@chemie.fu-berlin.de.
† Freie Universita¨t Berlin.
‡ Friedrich Schiller Universita¨t Jena.
§ CSIC Sevilla.
10.1021/jp711504g CCC: $40.75 © 2008 American Chemical Society
Published on Web 03/27/2008

3558 J. Phys. Chem. A, Vol. 112, No. 16, 2008
Lopez del Amo et al.
Figure 1. (a) Schematic representation of the decamethylzincocene
molecule (I) in the forms A and B related by inversion symmetry. (b)
Stick representation of the X-ray structure of I. Forms A and B are
shown separately, as well as superimposed (A + B). This last form
corresponds to the actual unit cell observed in the X-ray structure.
almost regular sandwich geometry, possibly as a consequence
of crystal packing effects, they feature η5/η1(π) coordination
of the Cp′ rings. The compounds are fluxional, both in solution
and in the solid-state, as the slipped-sandwich structure is
characterized by a flat potential energy surface. Ab initio
calculations show that the latter contains several minima
exhibiting only small energy differences, associated with
isomeric η5/η1 structures which interconvert via a η5/η5 structure.9a
Indeed, for the parent beryllocene, Be(C5H5)2, two dynamic
processes involving very low activation energies had been
identified by MD calculations at 400 K.11 They consisted of a
1,5-sigmatropic shift of the (η5-C5H5)Be unit around the
periphery of the η1 ring, involving an η5/η2 transition state (Ea
) 5 kJ mol-1), and of a ring-exchange process, or molecular
inversion, that interchanges the roles of the two rings from η5
to η1 and vice versa (Ea ) 8 kJ mol-1). The latter was proposed
to occur through an η3/η3 transition state.11 During the course
of our work, the dynamics of the known beryllocenes have been
Figure 2. Fluxional motions observed in I. (a) The molecular inversion
process or Zn tautomerism: interchange the role of both ring η5 (green)
and ring η1(yellow) by the Zn tautomerism (blue) between the two
equivalent positions. The overall result is equivalent to an inversion
of the molecule. In (b) and (c), the η5 and η1 rotation processes are
shown. (d) Methyl group rotation corresponding to 120° jumps.
versa, due to the partial σ character of the Zn-(η1)Cp* bond. In
other words, the process consists of a Zn displacement coupled
to an interchange of the “hapticity” of both rings. The overall
result of this motion is equivalent to a molecular inversion.
However, as this term corresponds to a symmetry operation and
not to a real molecular rearrangement, we prefer to call the
process “Zn tautomerism”. The process presumably involves
an η3-η3 intermediate.
The second process consists of internal ring rotations, that
is, an interconversion of the carbon atoms and methyl groups
in each ring. The “η5 rotation” (Figure 2b) consists of rotational
72° jumps around the ring axis. This internal rotation leaves
the sequence of bonds unchanged. The “η1 rotation” (Figure
2c) of the η1 bound ring also consists of rotational jumps,
however, coupled to a reorganization of the skeleton. It is also
known as “1,5-sigmatropic shift”.
9
investigated by solid-state 13C and Be NMR spectroscopy.12
Unfortunately, the dynamic range of this method is limited.
As we were interested in studying the solid-state dynamics
of this unusual slipped-sandwich metallocene structure, par-
ticularly the effect of an η5/η1(π) interaction on the activation
energies, we concentrated our efforts on decamethylzincocene
(Zn(C5Me5)2, I) in order to avoid the high toxicity of beryllium
and its compounds. I was prepared in 1986 by the reaction of
NaC5Me5 and ZnCl2 and characterized by gas electron diffrac-
tion as (η5-C5Me5)Zn(η1-C5Me5).10c Subsequent X-ray studies
led also to this geometry for the monomeric molecules of Zn-
(C5Me5)2, with the zinc atom disordered between two equivalent
sites (Figure 1). In the unit cell, two equivalent molecules A
and B are superimposed (A + B). Both molecules together
exhibit an inversion center. As such slip-sandwich geometries
are associated with highly fluxional structures,9,11 we expected
the following dynamic processes.
The first process, the “molecular inversion” or ring-exchange
process, is illustrated in Figure 2a. It does not consist of a real
180° rotation of the whole molecule in the unit cell as one might
infer from this figure. However, it is the result of a displacement
of the central Zn ion between two positions in the unit cell.
This displacement is coupled to the following geometrical
changes of each of the Cp* rings, (i) anticooperative twist jumps
of both rings by approximately 10° and (ii) a simultaneous
reorganization of the skeleton of each ring from a planar
disposition in the η5 form to a nonplanar in the η1 or vice-
Finally, the fastest intramolecular process one can expect is
methyl group rotations which interchange the three equivalent
proton atoms by rotation around the C-C bond as illustrated
in Figure 2d.
Because of the limited dynamic range of solid state 13C NMR,
2
we thought that solid state H NMR line shape analyses and
relaxometry of suitably deuterated decamethylzincocene should
provide much more kinetic information about the above-
mentioned processes. Therefore, we introduced a deuterated
methyl group into each Cp* of I as a molecular probe. For
comparison, we have also prepared the iminoacyl complex
(η5-C5Me5)Zn(η1-C(NXyl)C5Me5) (II) (see Figure 3; Xyl )
2,6-Me2C6H2) containing only one η5 ring as well as the analog

Zn Tautomerism and Hindered Rotations
J. Phys. Chem. A, Vol. 112, No. 16, 2008 3559
Figure 3. (a) Chemical structures of compounds I and II. (b) Left:
structure of I-d6 used for the 2H NMR experiments. Right: deuterated
sample used for the 2H NMR solid state measurements of the iminoacyl
[(η5-C5Me5)Zn(C(NXyl)C5Me5)]. Each ring contains a single CD3 group.
Each methyl group position can accommodate the CD3 group; however,
only a single isotopolog is depicted for each compound.
Figure 4. (a) 13C NMR subspectra and (b) spectra calculated for
different rate constants kAB of the nondegenerate Zn transfer exhibiting
an equilibrium constant of KAB * 1. (c) Calculated spectra for a
degenerate Zn transfer with KAB ) 1. (d) Calculated spectra for a
2
complex containing a CD3 group in each ring for H NMR
measurements. The interest in II arises from the fact that it is
only subject to methyl group rotations and to the “η5 rotation”,
whereas the Zn tautomerism and the η1 rotation are suppressed.
Thus, II was very useful to confirm the analyses of the dynamic
NMR experiments performed on I.
This paper is organized as follows. After a theoretical section
dealing with the relation between NMR spectra and the different
molecular motions illustrated in Figure 2 and an experimental
section, we describe the results of our NMR experiments on I
and II. The latter are then analyzed and discussed.
nondegenerate fast Zn transfer as a function of KAB
.
chemical shift νX(A) while Y exhibits the chemical shift νY-
(A). Similarly, νX(B) and νY(B) represent the intrinsic shifts of
form B. Let us assume a very reasonable assumption for the
13C nuclei of I that
νX(A) ) νY(B)
νX(B) ) νY(A)
i.e. ∆ν ) νX(A) - νX(B) (2)
Theoretical Section
Let us also assume that A dominates, that is, that KAB < 1.
Neglecting any anisotropic nuclear interactions, we obtain for
both rings the usual features of a nondegenerate exchange
process as depicted in Figure 4a. The sum of the two contribu-
tions gives the total line shape depicted in Figure 4b. Thus,
only two lines are observed in the slow exchange regime, which
broaden and coalesce into two new lines. The latter exhibit a
reduced splitting δν ) νX - νY from which the equilibrium
constant can be calculated14
In this section, we discuss how the processes listed in Figure
2 can influence the line shapes of the 13C signals of I measured
under the conditions of magic angle spinning (MAS), cross
1
polarization (CP), and H decoupling.
For a given molecule of I, the Zn tautomerism depicted in
Figure 2a which takes place between two forms A and B is not
necessarily a degenerate process. Therefore, it may be character-
ized by the equilibrium constant
KAB ) (1 - δν/∆ν)/(1 + δν/∆ν)
(3)
KAB ) kAB/kBA ) xB/xA ) exp(∆SAB/R - ∆HAB/RT)
(1)
In Figure 4c, we have considered the degenerate case with
KAB ) 1, which shows the usual features. In Figure 4d, we
have considered the case of very fast exchange, where the value
of the equilibrium constant KAB varies from a very small value
to a value of 1. Here, the two lines continually shift toward
each other and can be obtained from the splitting KAB according
to eq 3 if ∆ν is known from low-temperature spectra.
Here, ∆HAB and ∆SAB represent the reaction enthalpy and
reaction entropy, R is the gas constant, kAB and kBA are the
forward and backward rate constants, and xA and xB are the
corresponding mole fractions.
Generally, exchange broadened NMR line shapes are de-
scribed in terms of the quantum-mechanical density matrix
formalism proposed by Binsch et al.13 Using this theory, various
NMR line shapes of heavy atom nuclei in the presence of
nondegenerate exchange reactions have been calculated.14 As
illustrated at the top of Figure 4, let us denote the upper ring as
“X” and the lower ring as “Y”. Figure 4a shows the calculated
separate line shape contributions of X and of Y to the high
resolution 13C spectrum when the rate constant kAB is increased.
In the slow exchange regime, X exhibits in form A the intrinsic
We note that if the intrinsic line widths are large, it is
sometimes not easy to distinguish the cases of Figure 4c,d.
Experimental Section
Synthesis of I and II and Their d6 Isotopomers. All
preparations and manipulations were carried out under oxygen-
free argon using conventional Schlenk techniques. Solvents were

3560 J. Phys. Chem. A, Vol. 112, No. 16, 2008
Lopez del Amo et al.
rigorously dried and degassed before use. Infrared spectra were
recorded on a Bruker Vector 22 spectrometer. NMR spectra
were recorded on Bruker DRX-500 spectrometers. The 1H, 2H,
and 13C resonances of the solvent were used as the internal
standard, and the chemical shifts are reported relative to TMS.
spectrometer has been given recently.16 Here, only some salient
features are reproduced. All experiments were performed at a
field of 7 T, corresponding to a 2H resonance frequency of 46.3
MHz on a standard Oxford wide bore magnet (89 mm) equipped
2
with a room-temperature shim unit. For the H channel, a 2
kW class AB amplifier from AMT equipped with an RF-
blanking for suppressing the noise during data acquisition was
employed. The RF was fed through a crossed diode duplexer,
connected to the detection preamplifier and through the filters
into the probe. Typical 2H pulse width was 5.0 µs, corresponding
to 50 kHz B1 field in frequency units. To achieve a better
excitation of the echo spectrum, a shortened pulse with of 3.5
µs was employed.
1,2,3,4,5-Pentamethylcyclopentadiene-d3. This ligand was
prepared by the literature procedure reported for C5Me5H.15
2,3,4,5-tetramethylcyclopenten-2-enone (2.22 g, 2.4 mL, 15.29
mmol) was added to a solution of CD3Li‚LiI (36 mL, 18 mmol,
0.5 M Et2O). The remaining methyl lithium was hydrolyzed
with methanol (1 mL) and water (1 mL), and the mixture was
washed with a solution prepared with ammonium chloride (0.24
g), concentrated hydrochloric acid (0.2 mL), and water (0.1 mL).
The aqueous phase was extracted with ether (25 mL), and the
organic layers were stirred for 30 min with 1 mL of HCl, then
washed twice with 50 mL of 5% NaHCO3 solution, and finally
dried over K2CO3 to yield 1.86 g (13.38 mmol, 87%) of the
All experiments were performed using a home-built 5 mm
2H NMR probe. The probe is placed in a dynamic Oxford
CF1200 helium flow cryostat. The sample temperature was
controlled employing an Oxford ITC 503 temperature controller.
During cooling and before and after data acquisition, the sample
temperature was directly controlled via a Cernox sensor placed
in the direct vicinity of the sample. This temperature was used
to calibrate the readings of a second CGR-1-1000 sensor, which
is part of the cryostat. During data acquisition, the first sensor
was disconnected from the ITC 503 and grounded to protect
the ITC from the RF and to avoid distortions of the signal.
1
product. H NMR (500 MHz, C6D6, 20 °C): 0.98 (d, 3H, J )
8 Hz), 1.73 (s, 6H), 1.78 (s, 6H), 2.40 (m, 1H). 2H NMR (500
MHz, C6H6, 20 °C): 0.91 (s, 3D), 1.66 (s, 6D), 1.71 (s, 6D).
13C{1H} NMR (125 MHz, C6D6, 20 °C): 10.9 (2 Meâ), 11.4 (2
MeR), 13.9 (Me), 51.28, 51.44, 51.48 (CH; three isotopomers),
134.1 (Câ), 137.2 (CR).
KC5Me4(CD3). To a suspension of KH (532 mg, 13.26 mmol)
in THF (35 mL) was added a solution of C5Me4(CD3)H (1.86
g, 13.38 mmol) in the same solvent (10 mL). The mixture was
stirred at room temperature during 36 h, and the solvent was
then removed under vacuum. The solid residue was washed
twice with 30 mL of hexane and dried in vacuo to yield 2.14 g
(12.1 mmol, 90%) of the salt.
2
Because of the high sensitivity of the H NMR line shapes of
both I-d6 and II-d6 on the temperature, we carefully monitored
the stabilization and equilibration of the temperature, allowing
stabilization times up to several hours at each temperature before
performing the NMR experiment.
All spectra were recorded in resonance using the solid echo
technique, with an echo spacing of 30 µs and a full 8-step phase
cycle, which removes artifacts from FIDs of the two pulses.
Before Fourier transformation, the echo-signal was phase
corrected, and the imaginary part was zeroed to give fully
symmetric spectra, which are better suited for the line shape
analysis.
2H NMR Data Evaluation. The 2H-echo spectra in the fast
and slow exchange regime were simulated employing a labora-
tory written Matlab program. Instead of numerically performing
the powder integration, the faster analytical expression of the
powder pattern in terms of elliptic integrals17 was used to
calculate the line shape for infinite T2. The 2H NMR echo spectra
of I-d6 in the intermediate exchange regime were simulated by
a self-written Matlab routine using a Liouville formalism of
the exchange process, described recently.18
13C NMR Measurements. All 13C NMR measurements were
performed on a Bruker MSL-300 instrument, operating at 7 T,
equipped with a Chemagnetics-Varian variable-temperature 6
mm pencil CPMAS probe. The magic angle spinning (MAS,
7000 Hz) NMR experiments were performed employing the
{1H}13C CPMAS technique with a cross-polarization contact
time of 5 ms. All 13C chemical shift values were referenced to
solid adamantane.
Zn(C5Me4CD3)2. This compound was prepared in accordance
with the literature procedure10c but by using the deuterated salt.
A mixture of 205 mg (1.5 mmol) of ZnCl2 and 531 mg (3 mmol)
of KC5Me4CD3 was stirred in 30 mL of THF, at room
temperature, for 1 h. The crude product was extracted with
pentane to yield the compound as a white solid that was
1
recrystallized from pentane (320 mg, 0.95 mmol, 65%). H
NMR (500 MHz, C6D6, 25 °C, ppm): 1.87 (Me). 2H NMR (500
MHz, C6H6, 25 °C, ppm): 1.81 (CD3). 13C{1H} NMR (125
MHz, C6D6, 25 °C, ppm): 11.2 (C-Me), 111.8 (C-Me). IR
(Nujol, cm-1): 2063 (νCD3).
(η5-C5Me5)Zn(η1-C(NXyl)C5Me5) (II). Zn(C5Me5)2 (746 mg,
2.2 mmol) and CNXyl (288 mg, 2.2 mmol) were dissolved in
50 mL of hexane, and the resulting solution was stirred at room
temperature for 5 h. 1H NMR monitoring revealed the reaction
was complete. The solvent was then removed under vacuum,
the resulting yellow solid was extracted with 30 mL of hexane
and crystallized from this solvent. Yield: 840 mg, 1.8 mmol,
1
82%. H NMR (500 MHz, C6D6, 25 °C, ppm): 1.70 (s, 15H),
1.46 (s, 3H,), 1.72 (s, 6H), 1.82 (s, 6H), 2.02 (s, 6H), 6.87 (t,
1H, p-CH-Ar), 7.03 (d, 2H, m-CH-Ar). 13C{1H} NMR (125
MHz, C6D6, 25 °C, ppm): 9.8 (η5-CH3), 10.8 (Me), 11.2 (2
Me), 14.2 (2 Me), 18.6 (2 Me), 70.9 (Cq), 105.7 (CH), 107.8
(Cq), 122.5 (C-Me), 128.5 (CH), 137.5 (C-Me), 140.0 (C-
Me) 156.2 (Cq-N), 198.9 (CdN). IR (nujol, cm-1): 1585
Results
(νCdN). Elemental analysis calcd (%) for C29H39NZn:
C
Synthesis and Spectroscopic Properties of I and of II.
Decamethylzincocene can be readily prepared by the reported
procedure10c that involves the reaction of anhydrous ZnCl2 with
NaC5Me5. The potassium salt of the ligand, KC5Me5, may
alternatively be used, and in either case, Zn(C5Me5)2 can be
isolated as an air-sensitive white crystalline solid in moderate
yields (ca. 60%; we have been unable to reproduce the reported
74.62, H 8.35, N 2.99. Found: C 74.6, H 7.7, N 2.9.
Compound II-d6 was prepared as above, starting from I-d6
(320 mg, 0.94 mmol), CNXyl (123 mg, 0.94 mmol), and 30
mL of hexane as the reaction solvent. The yield of the product
from the first crop of crystals was 220 mg, approximately 50%.
The remaining product was discarded.
2
1
Low-Temperature Solid State H NMR Measurements.
96% yield).10c Low-temperature H NMR studies of solutions
A detailed discussion of our home-built three channel NMR
of this complex revealed10b a highly fluxional behavior,

Zn Tautomerism and Hindered Rotations
J. Phys. Chem. A, Vol. 112, No. 16, 2008 3561
TABLE 1: 13C NMR Chemical Shifts and Line Separations
of the Aromatic Signals of I
T/K
δX/ppm
δY/ppm
δX-δY/ppm
276
256
246
236
216
196
186
176
156
110.8
110.8
111.0
111.0
111.0
112.5
113.0
113.0
113.2
0
0
0
0
0
109.2
109.2
109.2
108.9
3.3
3.8
3.9
4.4
Figure 5. Variable temperature 13C CPMAS NMR spectra of I
recorded at 75 MHz (7 T).
at 198.9 ppm. These data are comparable to those reported for
the analogous beryllium iminoacyl (η5-C5Me5)Be(η1-C(NXyl)C5-
Me5).20 The ring carbon nuclei of the (η5-C5Me5)Zn moiety of
II give a 13C{1H} NMR signal at δ 107.8, while those of the
iminoacyl functionality appear at 139.9, 137.5, and 70.9 ppm.
These and other data summarized in the Experimental Section
provide unequivocal support for the structure of II shown in
eq 4, which has been independently confirmed by X-ray
crystallography.21
Variable Temperature 13C CPMAS NMR of Compounds
I and II. The variable temperature 13C CPMAS spectra of
compound I are shown in Figure 5. The spectra recorded at
higher temperatures (276 to 216 K) are characterized by two
lines arising from the aromatic (110.8 ppm) and methyl (10.9
ppm) carbons. The presence of common signals for all of the
aromatic and methyl groups in the spectra is indicative of
degenerate fast ring rotations and fast molecular inversion rates,
whose combination gives rise to a situation in which both rings
and the different positions inside each ring are indistinguishable
by 13C NMR. At lower temperatures (from 216 to 196 K), on
the other hand, the aromatic signal suddenly splits into two lines
with equal intensities within the margin of error. By further
lowering the temperature, the splitting δν ) νX - νY between
these two lines increases gradually. These effects are highlighted
in Figure 6. The sudden splitting of the aromatic line below
200 K is characteristic of a phase transition, in which the
degeneracy of the Zn tautomerism present at the high temper-
atures is lost. In this regime, the spectral features correspond to
those of Figure 4d and not to those of Figure 4c. All chemical
shifts and line separations are summarized in Table 1. In order
to calculate the equilibrium constants KAB below 200 K using
eq 3 and the solid line in Figure 6, we needed to know the
intrinsic chemical shift difference
Figure 6. Experimental and predicted 13C chemical shift difference
δν ) νX - νY of the two aromatic rings of I (Table 1, Figure 5). Above
216 K, a single line is observed corresponding to a disordered solid
with two equivalent rings X and Y, that is, δν ) 0. Below 200 K, both
rings become inequivalent and δν > 0. The solid line was fitted to the
experimental low-temperature data using eq 3.
evinced by the observation of only one methyl signal down to
-100 °C. Similarly, only one 13C resonance was detected at
111.6 ppm for the ring quaternary carbons.10b In this regard, it
is worth mentioning that in half sandwich compounds of the
type (η5-C5Me5)ZnR (R ) Me, Et) and other related complexes
(see below for II), the ring carbon nuclei resonate invariably in
the very narrow range of 107.5-109 ppm.19
Upon adding 1 mol equivalent of CNXyl to solutions of I at
room temperature, a smooth reaction ensues that yields the
iminoacyl II as the only reaction product. IR and NMR data
for II are in agreement with the proposed formulation.
The Zn-η1-C(NXyl)C5Me5 linkage features a distinct IR
absorption at 1590 cm-1 and a low-field 13C NMR resonance
∆ν ) νX(A) - νX(B) ) νY(B) - νY(A)
(5)
between the two rings. Unfortunately, we were not able to reach

3562 J. Phys. Chem. A, Vol. 112, No. 16, 2008
Lopez del Amo et al.
Figure 7. 13C CPMAS NMR spectra of II. The signal obtained for
the aromatic η5 carbons (106.9 ppm) is used in order to extract the
intrinsic chemical shift values of I in the variable temperature spectra
of Figure 6, as explained in the text.
the low temperatures required to obtain this value. Therefore,
we proceeded as follows. We assumed that all chemical shifts
are temperature independent and that the chemical shift of the
η5 bound ring in I is the same as the corresponding ring in II
whose room-temperature 13C CP MAS spectrum is depicted in
Figure 7. The spectrum was not analyzed in detail, but the signal
of the Cp* ring at 106.9 ppm could be easily identified. Thus,
we set
νX(A) = ν(II - η5) ) 106.9 ppm
(6)
Figure 8. 2H NMR spectra of compound II-d6 measured with the solid
echo technique and an echo delay of τ ) 30 µs.
As the average high-temperature chemical shift of the aromatic
carbons of I is given by
constants k5 of the 5-fold rotation of the η5 bonded Cp* ring
can be obtained by comparison of its line shape contribution
with the calculated spectra of Figure S3 of Supporting Informa-
tion. We obtain k5 = 2 × 106 sec-1 at 180 K and k5 = 5 × 103
sec-1 at 130 K. Considering an Arrhenius behavior for this
motion and a preexponential factor log A/s-1 ) 12.6, we obtain
an activation energy for this process of Ea5 = 21 kJ/mol.
Variable Temperature 2H NMR of I-d6. The superimposed
ν
X,Y(298K) ) 1/2 (νX(A) + ν X(B)) ) 110.8 ppm (7)
we obtained a chemical shift value of
ν X(B) )114.9 ppm or
∆ν ) 8 ppm
(8)
Keeping now the value of ∆ν fixed, we could adapt the
calculated solid line in Figure 6 to the experimental data by
varying the reaction enthalpy and the reaction entropy. Thus,
we obtained the values ∆H ) 2.4 kJ/mol and ∆S ) 4.9 J/mol
K. The temperature dependence of the equilibrium constant can
then be expressed as
2
experimental and calculated H NMR spectra of I-d6 recorded
at different temperatures using the solid echo technique, and
echo delays of τ ) 30 µs are shown in Figure 9. At 100 K, a
single Pake doublet is observed for the two CD3 groups of both
rings, exhibiting a quadrupole coupling constant of qcc ) 4qzz/3
) 54 kHz, where qzz corresponds to the separation of the main
peaks. As temperature is increased, the signal is gradually
transformed into a reduced Pake doublet at 150 K, characterized
by a motionally reduced quadrupolar coupling constant value
of qcc ) 23 kHz. To account for the nonplanar methyl group
conformation, the line shapes were simulated by considering
5-fold conical rotations, as explained in the Supporting Informa-
tion (Figure S3).
Variable Temperature 2H NMR Relaxation Times of
Compound I-d6. In this section, we describe the results of a
study of the longitudinal 2H NMR relaxation times of I-d6. They
were measured at 7 T using the saturation recovery technique.
Within the margin of error, single-exponential magnetization
built-up curves were observed at all temperatures. The data
obtained are plotted as a function of the inverse temperature in
Figure 10 and are listed in Table 2.
KAB ) 1.8 exp(-289 / T) for T e 200 K
KAB ) 1 for T g 200 K
(9)
2
Variable Temperature H NMR of II-d6. In Figure 8, the
variable temperature 2H NMR spectra of II-d6 are shown. The
spectrum at 180 K consists clearly of a superposition of two
Pake doublets. We assign the Pake doublet exhibiting the larger
coupling constant of qcc ) 4/3qzz ) 53 kHz to the three fast
rotating CD3 groups of the σ-bonded C(NXyl)Cp* group whose
Cp* ring is unable to rotate in the solid. The narrower Pake
doublet (26 kHz) on the other hand is assigned to the η5 bonded
Cp* ring, subject to fast rotational 72° jumps. As temperature
is decreased, the narrow component gradually disappears as the
rate constants of the η5 rotation become slow in the NMR time
scale, until finally at 130 K almost all of the 2H signal intensity
of the η5 bonded Cp* ring is broadened and superimposed on
the Pake doublet of the nonrotating C(NXyl)C5Me4CD3 unit.
Three T1 minima are observed at 220, 130, and 75 K. Each
minimum corresponds to a different thermally activated kinetic
process in the molecule that modulates the quadrupolar tensor
and serves as a mechanism for the longitudinal relaxation. The
2
An accurate line shape analysis of the H NMR depicted in
Figure 8 is difficult. However, a rough estimate of the rate

Zn Tautomerism and Hindered Rotations
J. Phys. Chem. A, Vol. 112, No. 16, 2008 3563
TABLE 2: 2H Spin-Lattice Relaxation Times at Various
Temperatures
T/K
2H T1/s
T/K
2H T1/s
300
280
270
250
220
190
180
175
165
160
150
1.16
0.33
0.25
0.24
0.16
0.19
0.25
0.80
0.41
0.40
0.16
140
130
120
110
100
95
90
80
70
60
0.15
0.10
0.12
0.15
0.17
0.44
0.35
0.26
0.34
0.46
τ -1 ) kAB + kBA ) A( exp(-EaAB/RT) +
exp(-(EaAB - ∆H)/RT)) (11)
Here, an Arrhenius law is used to describe the dependence of
the forward and the backward rate constants on temperature.
For degenerate processes, eq 10 reduces to
1
T1I
τ
4τ
) C
+
(12)
2
2
2 2
(
)
1 + ωI τ
1 + 4ωI τ
Usually, for intramolecular reactions, the pre-exponential factors
are given by A ) 1012.6 s-1. The solid lines in Figure 10 were
calculated using eqs 10 to 12. The molecular processes giving
rise to these data will be discussed in the next section.
2
Figure 9. Variable temperature H NMR spectra of compound I-d6
recorded at 46.3 MHz with the solid echo sequence.
Discussion
Using a combination of solid state 13C and 2H NMR, we have
studied the molecular dynamics of polycrystalline ZnCp*2 (I).
Variable temperature high-resolution 13C NMR experiments
indicated a fast Zn tautomerism according to Figure 2a, which
is degenerate above 210 K but becomes nondegenerate below
this temperature. The degenerate Zn tautomerism above 210 K
implies an interconversion of the η5 and η1 bound rings. Below
this temperature, the two rings X and Y are inequivalent, where
X exhibits a larger fraction of time in the η5 state than Y. The
2H NMR spectra of I-d6 indicated slow Cp* ring dynamics
around 100 K but fast dynamics at 150 K (Figure 9). By contrast,
the spectra of II-d6 revealed only a single fluxional Cp* ring.
The longitudinal 2H -T1 relaxation times of I-d6 exhibited three
minima (Figure 10).
In this section, we will discuss all observations and develop
a scenario of the fluxional behavior. In particular, we will discuss
the motional model used to simulate the spectra of Figure 9.
Order-Disorder Transition of I and Thermodynamics of
the Associated Zn Tautomerism. The variable-temperature 13
C
CPMAS NMR spectra of I indicated only a single line for all
carbons of the two Cp* rings down to 210 K. Below 210 K,
however, a line splitting was observed which increased when
temperature was further lowered (Figures 5 and 6). Such a
behavior is typical for an order-disorder transition around 210
K, that is, the presence of a fast degenerate process at higher
temperatures whose degeneracy is lifted at lower temperatures,
as observed previously for the case of proton tautomerism in
solids.23 Because of fast 5-fold ring rotations discussed below,
each 13C line represents all carbon atoms of a given Cp* ring;
hence, below 210 K, the two rings of I become inequivalent in
the solid state. One line corresponds to a preferentially η1-bound
ring; the other corresponds to a preferentially η5-bound ring.
This behavior can be explained in terms of the Zn tautomerism
between two forms A and B depicted in Figure 2; in the absence
Figure 10. 2H NMR longitudinal relaxation times of I measured at 7
T (46.3 MHz for H) as a function of the inverse temperature. The
fitting lines were calculated as described in the text.
2
rates of each relaxation mechanism can be described using the
master equation for quadrupolar relaxation in the presence of
exchange between two forms A and B22
4CKAB
(KAB + 1)2 1 + ωI τ
1
T1I
τ
4τ
)
+
(10)
2
2
2 2
(
)
1 + 4ωI τ
where KAB is the equilibrium constant of the interconversion
and τ the corresponding correlation time which can be expressed
as

3564 J. Phys. Chem. A, Vol. 112, No. 16, 2008
Lopez del Amo et al.
of asymmetric isotope substitution, these two forms are related
by a molecular inversion process. The equilibrium constants of
this process could be measured by 13C NMR, giving rise to a
small reaction enthalpy of 2.4 kJ mol-1. It follows that the η1/
η5 mole fractions are different for both rings. However, as
compared to the 13C NMR time scale, the process is still fast
down to 150 K where 13C experiments could be performed.
We note that in previous 13C CPMAS NMR studies on the
related beryllocene,12 a similar line splitting as shown in Figure
5 was observed which increased when the temperature was
lowered. This finding was analyzed in terms of a molecular
inversion process which is degenerate in the whole temperature
regime but which presented a transition from fast to slow
exchange. However, in the corresponding line shape analysis,
temperature-dependent chemical shifts had to be assumed whose
origin remained unexplained. Moreover, this analysis led to an
talline powdered samples of decamethylzincocene ZnCp2* (I).
Very important was the comparison with [(η5-Cp*)Zn(C(NXyl)-
C5Me5)] (II), where only a 5-fold rotation of the η5 bound Cp*
is possible. A second very important key was that both 13C NMR
2
experiments under CPMAS conditions and H NMR of static
powdered samples were measured. Only the combination of both
methods allowed us to analyze the different processes in Figure
2, as 13C NMR is sensitive to the Zn tautomerism and 2H NMR
to the ring rotations.
One important finding was that the Zn transition in solid I is
degenerate above but nondegenerate below 210 K. This corre-
sponds to a novel order-disorder phase transition. The origin
of this ordering can be attributed to the influence of the electric
dipole moment observed in this molecule3a due to its slip-
sandwich disposition. Thus, one can expect a paraelectric
disordered phase at high temperatures and an antiferroelectric
or ferroelectric phase at low temperatures. It would be interesting
in our opinion to further investigate this phenomenon and the
nature of this phase transition.
The second important finding is the observation of similar
activation energies for the 5-fold rotational jumps of the η1 and
the η5 bound Cp* rings in I. This finding is in agreement with
the structure 3 in Scheme 1, where both the η5 and the η1 rings
exhibit π interactions with the metal. In contrast, one would
expect for structure 4 in Scheme 1 a much slower rotation of
the η1(σ) bond ring.
unexpectedly high activation energy value of 36.9 kJ mol-1
,
which was much larger than calculated values11 which were of
the order of 10 kJ mol-1. It would be interesting to perform a
similar analysis of the 13C spectra of beryllocene as proposed
here for decamethylzincocene, which would, however, require
additional studies of model compounds in which the Be
tautomerism is suppressed; there are good chances that beryl-
locene and decamethylzincocene behave in a similar way with
respect to the order-disorder transition associated with the metal
tautomerism.
Longitudinal Relaxation of I-d6. In Figure 10, we assign
the low-temperature minimum to the usual 3-fold methyl group
rotational jumps depicted in Figure 2d. The high-temperature
minimum occurs above the order-disorder transition and
corresponds to the slowest process,that is, to averaged η1/η5
rotations of the Cp* rings (Figure 2b, 2c) which will be
discussed in more detail in the next section. These processes
are degenerate and were calculated using eq 12, using a simple
Arrhenius law.
Acknowledgment. This research has been supported by the
Deutsche Forschungsgemeinschaft, Bonn. We also thank the
Fonds der Chemischen Industrie (Frankfurt), the Junta de
Andaluc´ıa, and the Spanish Ministerio de Educacio´n y Ciencia
(MEC) (to E.C. Project CTQ2004-00409/BQU; FEDER support
to R.F. for a Ramo´n y Cajal fellowship and for I.R. for a
predoctoral fellowship).
Supporting Information Available: Introduction to the
We assign the center minimum to the Zn tautomerism (Figure
2a), which is associated with the interconversion of the η1 and
η5 bound rings. The solid center curve in Figure 10 was
calculated using eqs 10 and 11, and the values of the reaction
enthalpies and entropies obtained by 13C NMR (eq 9). In all
cases, the remaining parameters were then only the constants
C in eqs 10 and 12, which are different for each process, and
the corresponding forward activation energy. Both parameters
were varied until the best agreement of the calculated curves
with the experimental data was obtained. Thus, we obtained
the energies of activation of Ea1,5 ) 18.3 kJ mol-1 for the
averaged η1/η5 rotations, a forward energy of activation of EaAB
) 11.2 kJ mol-1 for the Zn tautomerism, and an energy of
activation of Ea3 ) 6.3 kJ mol-1 for the methyl group rotation.
The corresponding C factor in eq 10 used for the fittings was
2
theory used for the H NMR calculations, as well as model
calculations in which the effects of the predicted tautomerisms
2
of I in static H NMR powder spectra as a function of rates
and equilibrium constants are calculated. This material is
available free of charge via the Internet at http://pubs.acs.org.
References and Notes
(1) (a) Togni, A.; Halterman, R. L.; Metallocenes, Vols. 1 and 2;
Wiley: New York, 1998. (b) Scheirs, J.; Kaminsky, W.; Metallocene-Based
Polyolefins, Vols. 1 and 2; John Wiley & Sons Ltd.: Chichester, 2000. (c)
Togni, A.; Hayashi, T.; Ferrocenes; Verlag Chemie: Weinheim, 1995. (d)
Jaouen, G.; Top, S.; Vessie`res, A.; Alberto, R. J. Organomet. Chem. 2000,
600, 23-36.
(2) (a) Kealy, T. J.; Pauson, P. L. Nature 1951, 168, 1039-1040. (b)
Miller, S. A.; Tebboth, J. A.; Tremaine J. F. J. Chem. Soc. 1952, 632-
635.
(3) (a) Fischer, E. O.; Pfab, W. Z. Naturforsh. 1952, 7b, 377-379. (b)
Wilkinson, G.; Rosenblum, M.; Whiting, M. C.; Woodward, R. B. J. Am.
Chem. Soc. 1952, 74, 2125-2126. (c) Wilkinson, G. J. Am. Chem. Soc.
1952, 74, 6146-6149.
(4) (a) Fischer, E. O.; Hofman, H. P. Chem. Ber. 1959, 92, 482-486.
(b) Fischer, E. O.; Hofmann, H. P.; Treiber, A. Z. Naturforsh. 1959, 14b,
599-608.
(5) (a) Jutzi, P.; Burford, N. Chem. ReV. 1999, 99, 969-990. (b) Janiak,
C.; Schumann, H. AdV. Organomet. Chem. 1991, 23, 291. (c) Hays, M. L.;
Hanusa, T. P. AdV. Organomet. Chem. 1996, 40, 117.
(6) (a) Hanusa, T. P. Organometallics 2002, 21, 2559-2571. (b) Jutzi,
P.; Burford, N. Metallocenes Vol. 1; Togni, A., Haltermann, R. L., Eds.;
Wiley: New York, 1998. (c) Jutzi, P. J. Organomet. Chem. 1990, 400,
1-17.
(7) Voigt, A.; Filipponi, S.; Macdonald, C. L. B.; Gorden, J. D.;
Cowley, A. H. Chem. Commun. 2000. 911-912.
(8) Ferna´ndez, R.; Resa, I.; del R´ıo, D.; Carmona, E. Organometallics
2003, 22, 381-383.
C3 ) 2.4 × 109 s-2
.
The value of the energy of activation of the 3-fold methyl
group rotation corresponds to what is expected for this kind of
motions in the literature.24 The activation energy of the Zn
tautomerism is not much larger; at present, we are not aware of
any reference value. However, we note that this value may be
influenced by the intermolecular interactions in the crystal. The
activation energy for the η1/η5 rotations in I of 18.3 kJ mol-1
is of the order expected for rotations of η5 bound Cp* rings
coordinated to transition metals η5
Conclusions
We have demonstrated the presence of a number of super-
imposed fluxional processes according to Figure 2 in polycrys-

Zn Tautomerism and Hindered Rotations
J. Phys. Chem. A, Vol. 112, No. 16, 2008 3565
(9) (a) Conejo, M. M.; Ferna´ndez, R.; R´ıo, D.; Carmona, E.; Monge,
A.; Ruiz, C.; Ma´rquez, A. M.; Ferna´ndez Sanz, J. Chem. Eur. J. 2003, 9,
4452-4461. (b) Wong, C. J.; Lee, T. Y.; Cho, K. J.; Lee, S. Acta
Crystallogr. Sect. B 1972, 28, 1662-1668. (c) Wong, C. H.; Lee, T. Y.;
Lee, T. J.; Chang, T. W.; Liu, C. S. Inorg. Nucl. Chem. Lett. 1973, 9, 667-
673. (d) Almenningen, A.; Haaland, A.; Lusztyk, J. J. Organomet. Chem.
1979, 170, 271-284. (e) Nugent, K. W.; Beattie, J. K.; Hambley, T. W.;
Snow, M. R. Aust. J. Chem. 1984, 37, 35-45.
(10) (a) Burkey, D. J.; Hanusa, T. P. J. Organomet. Chem. 1996, 512,
165-173. (b) Fischer, B.; Wijkens, P.; Boersma, J.; van Koten, G.; Smeets,
W. J. J.; Spek, L.; Budzelaar, P. H. M. J. Organomet. Chem. 1989, 376,
223-233. (c) Blom, R.; Boersma, J.; Budzelaar, P. H. M.; Fischer, B.;
Haaland, A.; Volden, H. V.; Weidlein, J. Acta Chem. Scand. 1986, A40,
113-120.
(16) (a) Wehrmann, F.; Albrecht, J.; Gedat, E.; Kubas, G. J.; Limbach,
H. H.; Buntkowsky, G. J. Phys. Chem. A 2002, 106, 2855-2861. (b)
Wehrmann, F.; Fong, T.; Morris, R. H.; Limbach, H. H.; Buntkowsky, G.
Phys. Chem. Chem. Phys. 1999, 1, 4033-4041.
(17) Mehring, M. High Resolution NMR Spectroscopy in Solids;
Springer-Verlag: New York, 1983.
(18) (a) Gedat, E.; Schreiber, A.; Albrecht, J.; Shenderovich, I.;
Findenegg, G.; Limbach, H. H.; Buntkowsky, G. J. Phys. Chem. B 2002,
106, 1977. (b) Masierak, W.; Emmler, Th.; Gedat, E.; Schreiber, A.;
Findenegg, G. H.; Buntkowsky, G. J. Phys. Chem. B 2004, 18890-18896.
(19) Resa, I. Ph.D. Thesis, University of Sevilla, 2006.
(20) Conejo, M. M.; Ferna´ndez, R.; Carmona, E.; Andersen, R. A.;
Gutierrez-Puebla, E.; Monge, M. A. Chem. Eur. J. 2003, 9, 4462-4471.
(21) Gutierrez-Puebla, E.; Monge, A., personal communication.
(11) (a) Margl, P.; Schwarz, K. J. Am. Chem. Soc. 1994, 116, 11177-
11178. (b) Margl, P.; Schwarz, K. J. Chem. Phys. 1995, 103, 683-690.
(12) Hung, I.; Macdonald, C. L. B.; Schurko, R. W. Chem. Eur. J. 2004,
10, 5923-5935.
(22) Medycki, W. R.; Latanowicz, E. C. Mol. Phys. 1998, 93, 323-
330.
(13) (a) Alexander, S. J. Chem. Phys. 1962, 37, 971-977. (b) Binsch,
G. J. Am. Chem. Soc. 1969, 91, 1304-1309. (c) Kleier, D. A.; Binsch, G.
J. Magn. Reson. 1970, 3, 146-152.
(23) (a) Limbach, H. H. In Encyclopedia of Nuclear Magnetic Reso-
nance: Supplementary Volume 9, AdVances in NMR; Grant, D. M., Harris,
R. K., Eds.; John Wiley & Sons, Ltd: Chichester, 2002; pp 520-531. (b)
Foces-Foces, M. C.; Echevarr´ıa, A.; Jagerovic, N.; Alkorta, I.; Elguero, J.;
Langer, U.; Klein, O.; Minguet-Bonveh´ı, M.; Limbach, H. H. J. Am. Chem.
Soc. 2001, 123, 7898-7906.
(14) (a) Wehrle, B.; Zimmermann, H.; Limbach, H. H. J. Am. Chem.
Soc. 1988, 110, 7014-7024. (b) Wehrle, B.; Limbach, H. H. Chem. Phys.
1989, 136, 223-247.
(15) Herrmann, W. A.; Salzer, A. In Synthetic Methods of Organome-
tallic and Inorganic Chemistry; Herrmann, W. A., Ed.; Thieme: New York,
1996; Vol. 1, Chap. 3.
(24) Del Rio, A.; Boucekkine, A.; Meinnel, J. J. Comput. Chem. 2003,
24, 2093-2100.