Coordination of 1,4-diazabutadiene ligands to decamethylytterbocene: Additional examples of spin coupling in ytterbocene complexes

Article abstract of DOI:10.1021/om0610142

The paramagnetic 1:1 coordination complexes of (C₅Me ₅)₂Yb with a series of diazabutadiene ligands, RN=C(R′)C(R′)=NR, where R = CMe₃, CHMe₂, adamantyl, p-tolyl, p-anisyl, mesityl when R′ = H and R = p-anisyl when R' = Me, have been prepared. The complexes are paramagnetic, but their magnetic moments are less than expected for the two uncoupled spin carriers, (C ₅Me₅)₂Yb^III(4f¹³) and the diazabutadiene radical anions (S = 1/2), which implies exchange coupling between the spins. The variable-temperature ¹H NMR spectra show that rotation about the R-N bond is hindered and these barriers are estimated. The barriers are largely determined by steric effects, but electronic effects are not unimportant.

Full text of DOI:10.1021/om0610142

2296
Organometallics 2007, 26, 2296-2307
Coordination of 1,4-Diazabutadiene Ligands to
Decamethylytterbocene: Additional Examples of Spin Coupling in
Ytterbocene Complexes
Marc D. Walter, David J. Berg, and Richard A. Andersen*
Department of Chemistry and Chemical Sciences DiVision of Lawrence Berkeley National Laboratory,
UniVersity of California, Berkeley, California 94720
ReceiVed NoVember 4, 2006
The paramagnetic 1:1 coordination complexes of (C5Me5)2Yb with a series of diazabutadiene ligands,
RNdC(R′)C(R′)dNR, where R ) CMe3, CHMe2, adamantyl, p-tolyl, p-anisyl, mesityl when R′ ) H
and R ) p-anisyl when R′ ) Me, have been prepared. The complexes are paramagnetic, but their magnetic
moments are less than expected for the two uncoupled spin carriers, (C5Me5)2YbIII(4f13) and the
diazabutadiene radical anions (S ) 1/2), which implies exchange coupling between the spins. The variable-
1
temperature H NMR spectra show that rotation about the R-N bond is hindered and these barriers are
estimated. The barriers are largely determined by steric effects, but electronic effects are not unimportant.
This paper continues our phenomenological studies of
exchange coupling in decamethylytterbocene complexes by
preparing the 1,4-diazabutadiene complexes (C5Me5)2Yb(RNd
C(R′)C(R′)dNR), where R is an alkyl (Me3C, Me2HC, ada-
mantyl) or aryl group (p-tolyl, p-anisyl, and mesityl) and R′ is
either H or Me, abbreviated Cp*2Yb(dad(R′)-R). Several di-
azabutadiene complexes of lanthanide metallocenes have been
described for Sm,9 Eu,10 and Yb.11-14 However, the reactions
of (indenyl)2Yb(thf)2 and (fluorenyl)2Yb(thf)2 with aryl-
substituted 1,4-diazabutadiene do not yield simple metallocene
adducts, but complexes derived from C-C coupling and C-H
bond activation are observed.15,16
Introduction
Coordination complexes of the type Cp′2Yb(bipy-X), where
Cp′ is a substituted cyclopentadienyl ligand and bipy-X is a
4,4′-disubstituted bipyridine ligand, are molecules in which the
bipyridine ligand is a radical anion.1-3 However, the magnetic
moments in these complexes are lower than expected, which
implies that the spin carriers are coupled: that is, the spin on
the [Cp′2YbIII(4f13)] fragment is correlated with the spin on the
1
bipy-X radical anion (S ) /2) fragment. Although the spin
carriers are antiferromagnetically coupled, the mechanism by
which they couple is not clear; the coupling mechanism is
important, since exchange coupling between f electrons with
paired electrons or with other f electrons is generally weak.4,5
The magnetic moments of the ytterbocene-bipy complexes
Results and Discussion
range from small, where µeff ≈ 0.7 µB, to large, where µeff
≈
Comparison between bipy and dad(R′)-R Ligands. The
frontier molecular orbitals of bipy17,18 and dad(H)-R19 are
isolobal. The σ donor orbitals on each nitrogen atom of bipy
and dad(H)-R transform as a1 + b2 (in C2V symmetry). The
3.3 µB at 300 K, depending on the substituents on the ligands,
which implies that the exchange coupling is strong.1-3 The
terpyridine and substituted-terpyridine complexes of decam-
ethylytterbocene behave similarly.6,7 The implication of strong
antiferromagnetic coupling in these ytterbocene-bipy complexes
provides new ways of thinking about the specific role that
electrons in f orbitals play in bonding in these particular
complexes, where an unpaired electron resides in a ligand
molecular orbital, which is related to the concept of covalence
in f element compounds in general.8
(8) Booth, C. H.; Walter, M. D.; Daniel, M.; Lukens, W. W.; Andersen,
R. A. Phys. ReV. Lett. 2005, 95, 267202.
(9) Recknagel, A.; Noltemeyer, M.; Edelmann, F. T. J. Organomet.
Chem. 1991, 410, 53-61.
(10) Moore, J. A.; Cowley, A. H.; Gordon, J. C. Organometallics 2006,
25, 5207-5209.
(11) Trifonov, A. A.; Kirillov, E. N.; Bochkarev, M. N.; Schumann, H.;
Muehle, S. Russ. Chem. Bull. 1999, 48, 381-384.
(12) Trifonov, A. A.; Kurskii, Y. A.; Bochkarev, M. N.; Muehle, S.;
Dechert, S.; Schumann, H. Russ. Chem. Bull. 2003, 52, 601-606.
(13) Trifonov, A. A.; Fedorova, E. A.; Ikorskii, V. N.; Kurskii, Y. A.;
Dechert, S.; Schumann, H.; Bochkarev, M. N. Eur. J. Inorg. Chem. 2005,
2812-2818.
(14) Trifonov, A. A.; Kirillov, E. N.; Dechert, S.; Schumann, H.;
Bochkarev, M. N. Eur. J. Inorg. Chem. 2001, 2509-2514.
(15) Trifonov, A. A.; Fedorova, E. A.; Fukin, G. K.; Druzhkov, N. O.;
Bochkarev, M. N. Angew. Chem. 2004, 116, 5155-5158.
(16) Trifonov, A. A.; Fedorova, E. A.; Fukin, G. K.; Baranov, E. V.;
Druzhkov, N. O.; Bochkarev, M. N. Chem. Eur. J. 2006, 12, 2752-2757.
(17) McPherson, A. M.; Fieselmann, B. F.; Lichtenberger, D. L.;
McPherson, G. L.; Stucky, G. D. J. Am. Chem. Soc. 1979, 101, 3425-
3430.
* To whom correspondence should be addressed. E-mail: raandersen@
lbl.gov.
(1) Schultz, M.; Boncella, J. M.; Berg, D. J.; Tilley, T. D.; Andersen, R.
A. Organometallics 2002, 21, 460-472.
(2) Walter, M. D.; Schultz, M.; Andersen, R. A. New J. Chem. 2006,
30, 238-246.
(3) Walter, M. D.; Berg, D. J.; Andersen, R. A. Organometallics 2006,
25, 3228-3237.
(4) Berg, D. J.; Boncella, J. M.; Andersen, R. A. Organometallics 2002,
21, 4622-4631.
(5) Rosen, R. K.; Andersen, R. A.; Edelstein, N. M. J. Am. Chem. Soc.
1990, 112, 4588-4590.
(6) Veauthier, J. M.; Schelter, E. J.; Kuehl, C. J.; Clark, A. E.; Scott, B.
L.; Morris, D. E.; Martin, R. L.; Thompson, J. D.; Kiplinger, J. L.; John,
K. D. Inorg. Chem. 2005, 44, 5911-5920.
(18) Kaim, W. J. Am. Chem. Soc. 1982, 104, 3833-3837.
(19) Kaupp, M.; Stoll, H.; Preuss, H.; Kaim, W.; Stahl, T.; van Koten,
G.; Wissing, E.; Smeets, W. J. J.; Spek, A. L. J. Am. Chem. Soc. 1991,
113, 5606-5618.
(7) Kuehl, C. J.; Da Re, R. E.; Scott, B. L.; Morris, D. E.; John, K. D.
Chem. Commun. 2003, 2336-2337.
10.1021/om0610142 CCC: $37.00 © 2007 American Chemical Society
Publication on Web 03/22/2007

Spin Coupling in Ytterbocene Complexes
Organometallics, Vol. 26, No. 9, 2007 2297
Table 1. Reduction Potentials of 1,4-Diazabutadiene
The infrared spectra in the low-energy region show an
absorption that is associated with the (C5Me5)2YbIII fragment
(dad(R′)-R) Ligands20,21
1
(Table 2).4 The solution H NMR chemical shifts at 20 °C are
R′
R
E
1/2(L0/-)a/V
indicative of paramagnetic complexes, as is their temperature
dependence (see Table 5). All of the complexes are paramagnetic
in the solid state, but their µeff values, which lie in the narrow
range between 3.3 and 4.0 µB at 300 K, are lower than expected;
the expected value for the two uncoupled spin carriers, Cp′2-
YbIII(4f13) and a diazabutadiene radical anion (S ) 1/2), is 4.85
µB.23 The lower value of µB implies that the spin carriers are
antiferromagnetically coupled as in the bipy-X complexes.
However, the temperature dependence of the magnetic suscep-
tibility is not simple antiferromagnetic coupling between the
spin carriers, which is also true for the bipy-X complexes; these
details are described below.1-3,8
H
H
H
Me
bipy
CMe3
-2.13
-1.40
-1.47
-1.85
-2.12
p-tolyl
p-anisyl
p-anisyl
a Potentials quoted relative to SCE in 0.1 M [nBu4N][BF4]-DMF at 293
K.
LUMO is of b1 symmetry (in C2V symmetry) in each case as
illustrated in A and B. The energy of the LUMO is measured
Solid-State Crystal Structures. The ORTEP diagrams for
two representative molecules, (C5Me5)2Yb(dad(H)-R), where R
) p-tolyl, p-anisyl, are shown in Figures 1 and 2. The crystal
data and packing diagrams are available as Supporting Informa-
tion. Bond distances and angles for these two complexes are
given in Tables 3 and 4 along with data for related ytterbocene
diazabutadiene complexes.11,12
The two complexes shown in Figures 1 and 2 have ap-
proximate C2V symmetry, and the C5Me5 rings are staggered in
the p-tolyl derivative but eclipsed in the p-anisyl derivative. The
NCCN atoms in the diazabutadiene ligands are planar, since
their torsion angles are 0.5° in each complex. The average value
of the angle formed by the intersection of the planar p-tolyl
ring (NC(ipso)C(ortho)C(ortho)) and the NCCN plane is 21.5°
in (C5Me5)2Yb(dad(H)-p-tolyl); in the p-anisyl complex this
angle is 4.5°. This difference could be due to different
intramolecular steric effects between the N-Ar fragment and
the orientation of the C5Me5 rings in the solid state. That is,
the steric interaction is minimized when the C5Me5 rings are
eclipsed and the NAr fragments are nearly coplanar with the
NCCN fragment, which maximizes the extent of N-C(ipso) π
bonding.
The averaged Yb-C and Yb-N distances in both molecules
are statistically equal, as are the Cp(ring centroid)-Yb-Cp-
(ring centroid) angles. The distances are in the range expected
for a (C5Me5)2YbIII fragment and are similar to those found in
the dad(H)-t-Bu complexes (Table 3).11,12 These values may be
compared to the Yb-C and Yb-N distances in (C5Me5)2Yb-
(py)2 of 2.74 and 2.56 Å, respectively,24 both of which are
significantly longer than the comparable distances for the
complexes given in Table 3. In addition, the Yb-N distance in
(C5H5)2Yb(dad(H)-t-Bu) is significantly shorter than that in (C5-
Me5)2Yb(dad(H)-t-Bu), which is consistent with intramolecular
steric effects.12
by the reduction potential of the neutral ligands. The reduction
potentials of some of the 1,4-diazabutadiene ligands used in
this study are given in Table 1; the value for bipy is listed for
comparison.
The reduction potentials show that the LUMO is lower in
energy for the dad(H)-R ligands when R is a substituted benzene
group relative to bipy, but when R is CMe3, the energy of the
LUMO is essentially the same as that of bipy. Substituent effects
change the reduction potential by 0.73 V,20,21 and these changes
are readily rationalized by inductive and resonance effects when
the substituents on the nitrogen atoms are aromatic rings and
by hyperconjugation when the substituents are alkyl groups.
Since the diazabutadiene ligands are as easy or easier to reduce
than bipy, they are generally thought to be better π-acceptors.22
Synthesis and Physical Properties. All of the dad(R′)-R
complexes are prepared by adding the diazabutadiene ligand to
(C5Me5)2Yb(OEt2) in hydrocarbon solvent at room temperature
(eq 1). The 1:1 complexes are either green or red, and most of
The bond distances in the dad(H)-R fragments support the
view that it is a radical anion. The uncoordinated diazabutadiene
dad(H)-t-Bu, in the trans conformation, has been structurally
characterized.25 The data in Table 4 compare the C-C and C-N
bond distances in the four ytterbocene complexes with those in
the free diazabutadiene. The C-C distances shorten and the
C-N distances lengthen in the complexes relative to the
distances in the free ligand, as expected when the LUMO of
the diazabutadiene illustrated in B is populated. The crystal-
them are difficult to purify by crystallization from the mother
liquor. However, most sublime in the temperature range 160-
220 °C under diffusion pump vacuum, and the sublimed
compounds readily crystallize from pentane or toluene solution.
The pure complexes are high-melting solids, and those that
sublime yield molecular ions in their mass spectra; these and
other physical properties are shown in Table 2.
(23) The uncoupled value of 4.85 µB is obtained by summing the øT
value of each spin carrier; this assumes that all of the crystal field states
2
(20) Coulombeix, J.; Emmenegger, F.-P. HelV. Chim. Acta 1985, 68,
derived from the F7/2 term of Yb(III) are populated.1
248-254.
(24) Tilley, T. D.; Andersen, R. A.; Spencer, B.; Zalkin, A. Inorg. Chem.
1982, 21, 2647-2649.
(25) Huige, C. J. M.; Spek, A. L.; de Boer, J. L. Acta Crystallogr. 1983,
C41, 113-116.
(21) tom Dieck, H.; Renk, I. W. Chem. Ber. 1971, 104, 110-130.
(22) van Koten, G.; Vrieze, K. AdV. Organomet. Chem. 1982, 21, 151-
239.

2298 Organometallics, Vol. 26, No. 9, 2007
Table 2. Solid-State Properties of Ytterbocene 1,4-Diazabutadiene Complexes
Walter et al.
compd
color
mp/°C
IRa/cm-1
µ
eff(300 K)b/µB
(C5Me5)2Yb(dad(H)-p-tolyl)
(C5Me5)2Yb(dad(H)-p-anisyl)
(C5Me5)2Yb(dad(Me)-p-anisyl)
(C5Me5)2Yb(dad(H)-mesityl)
(C5Me5)2Yb(dad(H)-t-Bu)
(C5Me5)2Yb(dad(H)-i-Pr)
(C5Me5)2Yb(dad(H)-adamantyl)
(C5Me4H)2Yb(dad(H)-t-Bu)
(C5H5)2Yb(dad(H)-t-Bu)11
green
red
green-brown
blue-green
red
red
brown-red
red
220 dec
260-262 dec
219-221
313
308
295
3.97
3.72
3.96
230-232 dec
220-222 dec
208-211 dec
238-239
280
295
280
3.63
3.71
3.43
242-245 dec
yellow
3.34
a Ytterbium-C5Me5 ring symmetric tilting frequency.4 b Determined from a plot of øT vs T, where µeff ) 2.828(øT)0.5 and T ) 300 K.
Table 3. Selected Bond Distances (Å) and Angles (deg) of (C5H5)2Yb(dad(H)-t-Bu),11 (C5Me5)2Yb(dad(H)-t-Bu),12
(C5Me5)2Yb(dad(H)-p-tolyl), and (C5Me5)2Yb(dad(H)-p-anisyl)
(C5H5)2Yb-
(C5Me5)2Yb-
(C5Me5)2Yb-
(C5Me5)2Yb-
(dad(H)-t-Bu)a
(dad(H)-t-Bu)b
(dad(H)-p-tolyl)c
(dad(H)-p-anisyl)c
Yb-Cring (mean)
Yb-centroid
centroid-Yb-centroid
Yb-N
2.60
2.33, 2.33
128
2.306(9); 2.306(9)
74.7(3)
117.0
2.69
2.41; 2.40
130
2.385(3); 2.394(3)
75.3(1)
113.3
2.65
2.37; 2.34
137
2.340(5); 2.368(5)
73.4(2)
118.3
2.64
2.36; 2.35
139
2.339(7); 2.337(6)
72.5(2)
119.5
N-Yb-N
R-N-C (av)
a From ref 11. b From ref 12. c This work.
Table 4. Bond Distances (Å) in Diazabutadiene Ligands
from X-ray Crystallography
free
dad(H)-
tBua
(C5Me5)2Yb- (C5H5)2Yb- (C5Me5)2Yb- (C5Me5)2Yb-
(dad(H)-
(dad(H)-
(dad(H)-
(dad(H)-p-
t-Bu)b
t-Bu)c
p-tolyl)d
anisyl)d
A
B
C
1.467(2)
1.267(2)
1.267(2)
1.398(3)
1.339(2)
1.326(5)
1.398(10)
1.299(10)
1.310(10)
1.380(9)
1.342(8)
1.342(8)
1.382(13)
1.335(11)
1.339(11)
Figure 1. ORTEP diagram of (C5Me5)2Yb(dad(H)-p-tolyl) (50%
probability ellipsoids).
a From ref 25. b From ref 12. c From ref 11. d This work.
net magnetic moments in the high-temperature regime are much
lower, in the range of 1-3.9 µB (at 400 K).1-3 This pattern
implies that the reason for the lowered magnetic moments,
relative to the value expected for two uncoupled spin carriers,
is the same: viz., exchange coupling between the unpaired
electron in the (C5Me5)2YbIII fragment with the electron in the
b1 symmetry orbital of the ligand. The extent of coupling and
therefore the coupling constant depends upon several variables,
the principle ones being the overlap between and the relative
energy of the magnetic orbitals.3 Detailed studies similar to those
published in ref 8 are underway to make this qualitative
statement more quantitative. Recent articles claim that the
variation in the magnetic moment as a function of temperature
is due to a tautomeric equilibrium between the diamagnetic [(C5-
Me5)2YbII(4f14)(bipy, S ) 0)] spin isomer and the paramagnetic
[(C5Me5)2YbIII(4f13)(bipy•-, S ) 1/2)] spin isomer.6,26 This
explanation is incorrect for (C5Me5)2Yb(bipy), since the shape
and therefore the population of Yb(II) and Yb(III) signatures
in the Yb LIII-edge XANES do not change from 10 to 400 K.8,27
Figure 2. ORTEP diagram of (C5Me5)2Yb(dad(H)-p-anisyl) (50%
probability ellipsoids).
lographic data and all of the other data clearly support the idea
that the complexes are derived from radical anions and cationic
ytterbocene fragments.
Solid-State Magnetic Measurements. The magnetic sus-
ceptibility measurements as a function of temperature provide
a bulk measure of how the paramagnetic condition changes with
temperature. A plot of ø-1 as a function of temperature for all
the complexes is shown in Figure 3, and a plot of µeff vs T for
some of them is shown in Figure 4. The øT vs T plots are
available as Supporting Information.
The one exception to the generalizations mentioned above is
the ø-1 vs T plot for (C5Me5)2Yb(dad(H)-p-anisyl). Although
the plot is qualitatively similar to that of (C5Me5)2Yb(dad(Me)-
The plots of ø-1 vs T and µeff vs T define a family of curves,
with one exception, in which µeff increases, more or less
smoothly, to 400 K, where the curves appear to saturate with
an effective magnetic moment of 3.9-4.2 µB. These plots are
similar in shape to those for the bipy-X derivatives, though the
(26) Carlson, C. N.; Kuehl, C. J.; Da Re, R. E.; Veauthier, J. M.; Schelter,
E. J.; Milligan, A. E.; Scott, B. L.; Bauer, E. D.; Thompson, J. D.; Morris,
D. E.; John, K. D. J. Am. Chem. Soc. 2006, 128, 7230-7241.
(27) Variable-temperature Yb LIII-edge XANES studies on the com-
pounds reported here and on other bipy-X compounds will be published in
due course (Booth, C. H. Personal communication).

Spin Coupling in Ytterbocene Complexes
Organometallics, Vol. 26, No. 9, 2007 2299
Figure 3. ø-1 vs T plot of (C5Me5)2Yb(dad(R′)-R) compounds.
Figure 5. ø-1 vs T plot of (C5Me5)2Yb(dad(H)-p-anisyl) and (C5-
Me5)2Yb(dad(Me)-p-anisyl).
Figure 6 shows a ø-1 vs T plot for (C5H5)2Yb(dad(H)-t-Bu)
and (C5Me5)2Yb(dad(H)-t-Bu). As the curves are nearly super-
imposable, the substituents on the cyclopentadienyl ligand do
not appreciably change the magnetic moment. This is unlike
the behavior of the bipyridine complexes, where the effective
magnetic moments depend upon the substituents on the cyclo-
pentadienyl ring.1,2
Solution 1H NMR Spectroscopic Studies. (a) General
Considerations. The observed solution 1H NMR chemical shifts
in C6D6 at 20 °C for the complexes described in this paper are
given in Table 5.28 In general, the chemical shifts of the C5Me5
groups are relatively broad and lie in a narrow chemical shift
range from δ 2.8 to 0.15 ppm. The resonances due to the
diazabutadiene ligands vary widely in chemical shift and line
width. Since assignments can be made on the basis of relative
intensities, and in some cases specific substitutions, some
generalizations can be made. The backbone CH’s appear upfield
in the chemical shift range of δ -25 to -146 ppm with the
most shielded resonances associated with the dad(H)-R ligands,
in which R is a substituted benzene ring. When R is an alkyl
group, the resonances appear in a very narrow range, δ -25 to
-28 ppm. The chemical shift of the meta CH groups of the
substituted benzene ring are observed in the range δ +47 to
+62 ppm, but the ortho H’s are observed in only one case. These
chemical shift values clearly show that the complexes are
paramagnetic in solution. In these complexes coordinated
diazabutadiene ligand does not exchange with added free
diazabutadiene ligand over the temperature range of -80 to +90
°C, showing that the dynamic processes, described in detail
below, are due to intramolecular processes.
In this and earlier papers, variable-temperature 1H NMR
spectroscopy is used as a probe of dynamic processes in solution.
Plots of δ vs T-1 indicate if intramolecular exchange is occurring
on the NMR time scale and if the chemical shifts are linear in
T-1. In order to ensure that all intermolecular processes are slow
on the NMR time scale, generally, variable-temperature data
in the presence and absence of added exchangeable ligand have
been obtained.1-3 In the best experiments, coalescence-
Figure 4. µeff vs T plots of (a, top) (C5Me5)2Yb(dad(H)-alkyl)
compounds and (b, bottom) (C5Me5)2Yb(dad(R′)-aryl) compounds.
p-anisyl) (Figure 5), the slope of ø-1 vs T for the former complex
increases much more rapidly with temperature, passing through
a sharp maximum at ∼25 K, then decreases much more rapidly
than in the other complexes. At about 200 K, a sudden increase
occurs over a temperature range of 15 K, before the ø-1 value
approaches that found for the other complexes. The origin of
the difference in the solid-state magnetic behavior of these two
molecules is unknown.
(28) The isotropic chemical shifts (δiso ) δobs - δdia) have not been
determined, since the diamagnetic analogues to the molecules presented in
this paper, e.g. (C5Me5)2Ca(dad(R′)-R), have not been prepared. However,
approximate isotropic shifts (δiso) using the chemical shifts of (C5Me5)2-
Yb(py)2 and the free dad(R′)-R ligands have been calculated (see the
Supporting Information). Errors introduced by nonideal reference com-
1
pounds are expected to be small for H NMR data.

2300 Organometallics, Vol. 26, No. 9, 2007
Walter et al.
decoalescence behavior is observed, from which ∆Gq(Tc) is
obtained and a physical process responsible for the fluxionality
is generally obvious.29,30
When the δ vs T-1 plots are nonlinear, there is insufficient
information in the line shape to postulate a physical process
other than some intramolecular process(es) occurring in solution
is (are) temperature dependent. A physical process that changes
the populations will give rise to the resonances that have
nonlinear dependence on T-1: e.g., geometric changes, intramo-
lecular fluxions, electronic exchange, etc.31-33 In general, our
interest is to obtain ∆Gq(Tc) for chemical exchange processes,
and the issue of nonlinearity is not addressed. However, the
bipy and related N-heterocyclic complexes present a new
challenge, since these ligands in these complexes are radical
anions and the effect of electron transfer, spin delocalization,
and/or exchange coupling on the line shape must be addressed
in a realistic manner. A corollary is the origin of chemical shifts;
we have not dealt explicitly with them until recently,3 since we
have simply viewed them as a number. For lanthanide molecules
with neutral or anionic ligands, that are diamagnetic, the
paramagnetic (isotropic) shifts are mainly due to the pseudo-
contact term, δpc ) -D[G(θ,r)], where D is the magnetic
susceptibility tensor and G(θ,r) is the geometric term which is
related to 3 cos2 θ - 1/r3 , when axial symmetry is present.31
Thus, the isotropic chemical shift depends on the distance r of
the observed nucleus from the paramagnetic center (that is, the
Figure 6. ø-1 vs T plot of (C5Me5)2Yb(dad(H)-t-Bu) and (C5H5)2-
Yb(dad(H)-t-Bu).
spin density is located on the ligand: i.e., when there is
covalence and/or when the ligand is a radical anion, as in the
case of Cp′2Yb(N-heterocyclic bases). Both Fermi contact and
pseudo-contact terms are related to the magnetic susceptibility,
and δ vs T-1 plots are expected to be linear, if Curie behavior
is followed and if the isotopic chemical shifts account for all
of the intramolecular dynamic processes.
distance from the unpaired spin) and D, which is related to T-1
,
(b) Aryl-Substituted Complexes, (C5Me5)2Yb(dad(H)-R)
if the Curie law is followed. This is a good approximation for
the vast number of complexes we have prepared over the years,
since the metal to ligand bonds are not covalent: that is, the
unpaired spin density on the ligands is small, since unpaired f
electrons are localized on the metal center and contributions
from the Fermi contact term or deviations from the Curie law
to the overall isotropic chemical shift are small. We have
recently shown that complexes in which the ligand is a radical
anion, such as Cp′2Yb(bipy-X), the plot of the solid-state
magnetic susceptibility, øT, vs δ is linear for the 6,6′-H chemical
shift at 300 K.3 This is not surprising, since the 6,6′-position
carries a small amount of spin density in the bipy radical anion,18
it is close to the paramagnetic center, and the chemical shift
should depend on the pseudo-contact contribution. For the other
resonances that carry higher spin densities and are further away
from the paramagnetic center, the Fermi contact term, δFc, is
proportional to -ai S av, where ai is the isotropic hyperfine
1
(R ) Aryl). The H NMR data are shown as δ vs T-1 plots,
where δ is the observed chemical shift as in earlier papers.1-3
An alternative method of representing the data is to plot the
reduced chemical shift (ϑ298) (the chemical shift at a given
temperature multiplied by that temperature, δΤ divided by 298
K in order to get convenient numbers) as a function of T.32,36,37
The latter representations are useful, since they show the relation
between solid-state magnetic susceptibility as a function of
temperature when plotted as øT vs T and the reduced chemical
shift of a given resonance as a function of temperature, ϑ298 vs
T.38,39 These plots are available as Supporting Information.
The variable-temperature 1H NMR spectra of the p-tolyl and
p-anisyl complexes are similar, and only δ vs T-1 plots are
shown for the p-tolyl complex in Figure 7; the δ vs T-1 plot
for the p-anisyl complex is available as Supporting Information.
The C5Me5 resonances are essentially independent of tem-
perature, though the p-Me and backbone CH resonances have
a nonlinear dependence on temperature. The resonances at-
tributable to the benzene ring meta and ortho CH’s address the
question of arene ring rotation in solution. At 20 °C, a single
resonance of relative intensity 4H is observed, which broadens,
disappears, and then reappears as two resonances of 2H each
as the temperature is decreased to -70 °C (Figure 7b). The
coalesced resonance sharpens, and the chemical shift is es-
sentially independent of temperature from +20 to +90 °C.
During the temperature study, another resonance that is very
broad at 20 °C emerges from the baseline as the temperature is
decreased as two widely separated resonances due to 2H each.
As the temperature is increased from +20 to +90 °C, this
splitting constant and S is the thermal average of the spin
av
moments along the principal molecular axes, which is related
to the atomic magnetic susceptibility, will play a role.34,35 This
term is an important contributor to the chemical shift when
(29) Maron, L.; Werkema, E. L.; Perrin, L.; Eisenstein, O.; Andersen,
R. A. J. Am. Chem. Soc. 2005, 127, 279-292.
(30) Zi, G.; Blosch, L. L.; Jia, L.; Andersen, R. A. Organometallics 2005,
24, 4602-4612.
(31) (a) La Mar, G. N.; Horrocks, W. D.; Holm, R. H. NMR of
Paramagnetic Molecules; Academic Press: New York, 1973; pp 85-178.
(b) Wicholas, M.; Drago, R. S. J. Am. Chem. Soc. 1968, 90, 6946-6950.
(c) Horrocks, W. D.; Taylor, R. C.; LaMar, G. N. J. Am. Chem. Soc. 1964,
86, 3031-3038. (d) LaLancette, E. A.; Eaton, D. R.; Benson, R. E.; Phillips,
W. D. J. Am. Chem. Soc. 1962, 84, 3968-3970.
(32) Ko¨hler, F. H. Probing Spin Densities by Use of NMR Spectroscopy.
In Magnetism: Molecules to Materials. Models and Experiments; Miller,
J. S., Drillon, M., Eds.; Wiley-VCH: Weinheim, Germany, 2001; pp 379-
430, and references cited therein.
(35) Kurland, R. J.; McGarvey, B. R. J. Magn. Reson. 1970, 2, 286-
301.
(33) Fischer, R. D. NMR-spectroscopy of organo-f-element and pre-
lanthanoid complexes: some current trends. In Fundamental and Techno-
logical Aspects of Organo-f-Element Chemistry; Marks, T. J., Fragala`, I.
L., Eds.; D. Reidel: Dordrecht, The Netherlands, 1985; pp 277-326.
(34) Fischer, R. D. Lanthanide and Actinide Complexes. In NMR of
Paramagnetic Molecules; La Mar, G. N., Horrocks, W. D., Holm, R. H.,
Eds.; Academic Press: New York, 1973; pp 521-553.
(36) Knorr, R.; Polzer, H.; Bischler, E. J. Am. Chem. Soc. 1975, 97,
643-644.
(37) Knorr, R.; Weiss, A.; Polzer, H.; Bischler, E. J. Am. Chem. Soc.
1975, 97, 644-646.
(38) Ko¨hler, F. H.; Schlesinger, B. Inorg. Chem. 1992, 31, 2853-2859.
(39) Heise, H.; Ko¨hler, F. H.; Xie, X. J. Magn. Reson. 2001, 150, 198-
206.

Spin Coupling in Ytterbocene Complexes
Organometallics, Vol. 26, No. 9, 2007 2301
Table 5. 1H NMR Data for Cp′2Yb(dad(R′)-R) Complexesa
Table 6. Barrier of Ortho and Meta CH Site Exchange in
(C5Me5)2Yb(dad(H)-p-tolyl) and (C5Me5)2Yb(dad(H)-p-anisyl)
R
(C5Me5)2Yb-
(dad(H)-p-tolyl)
(C5Me5)2Yb-
(dad(H)-p-anisyl)
compd
Cp′
R′
(C5Me5)2Yb(dad(H)-
p-tolyl)
1.74 (35) -123.7 (320) o-CH
b
ortho H
meta H
ortho H
meta H
m-CH 50.2 (700)
p-Me
58.3 (90)
b
∆νa/Hz
12400
310
27546
10.5
16800
270
37320
10.1
85200
290
189266
10.0
520
220
1155
9.7
(C5Me5)2Yb(dad(H)-
p-anisyl)
1.76 (13) -109.3 (270) o-CH
b
Tc /K
m-CH 46.9 (350)
p-OMe 18.9 (4)
kc/s-1
∆Gqc/kcal mol-1
(C5Me5)2Yb(dad(Me)-
p-anisyl)
1.51 (23) +126.0 (360) o-CH
103.3 (800)
a Signal separation in Hz. b Coalesence temperature (Tc) at 400 MHz
operating frequency. c The free energy of activation, ∆Gq, was determined
by the temperature dependence of the ortho and meta CH proton resonances
in C7D8, as outlined in refs 40-42.
m-CH 50.1 (34)
p-OMe 18.5 (8)
(C5Me5)2Yb(dad(H)-
mesityl)
0.29 (90) -146.2 (240) o-CH3 224.9 (120)
o-CH3 58.2 (64)
m-CH 62.5 (32)
m-CH 50.4 (24)
p-Me
58.2 (64)
(C5Me5)2Yb(dad(H)-
t-Bu)
(C5Me5)2Yb(dad(H)-
i-Pr)
(C5Me5)2Yb(dad(H)-
adamantyl)
0.15 (8)
2.83 (9)
0.14 (8)
-26.6 (200) t-Bu
b
-28.3 (170) CH
108.9 (260)
32.3 (12)
18.9 (650)
16.7 (400)
12.5 (67)
Me
-25.2 (200) CH2
CH2
CH
(C5Me4H)2Yb(dad(H)-
t-Bu)
Me 2.62 (34) -36.0 (180) t-Bu
Me 2.47 (50)
42.7 (480)
H -38.0 (60)
-23.8 (60)
(C5H5)2Yb(dad(H)-
t-Bu)
-42.5 (208) t-Bu
45.8 (46)
a Recorded in benzene-d6 at 20 °C. Observed chemical shifts are given
in ppm. Line widths, full width at half-height (Hz), are given in parentheses.
b These resonances are not observed at 20 °C.
Figure 8. Chemical shift (δ) vs T-1 plot of the 1H NMR resonances
of (C5Me5)2Yb(dad(H)-mesityl) in toluene-d8 at temperatures from
-65 to +80 °C.
to assign them with certainty; however, the broader one at 20
°C is likely to be the ortho CH, since it is closer to the
paramagnetic center. The assignment is not crucial, since the
activation free energy derived from the line shape for both
resonances is identical (Table 6). The temperature behavior of
the resonances is as expected for hindered rotation of the p-tolyl
group about the N-C(ipso) bond. At low temperature the two
p-tolyl rings are oriented as in the solid-state structures (Figure
1 and 2), the molecule has averaged C2V symmetry, and the
ortho and meta CH’s are distal and proximal, respectively,
relative to the ytterbocene fragment. As the temperature is
increased, rotation around the N-p-tolyl bond increases, result-
ing in distal-proximal site exchange with barriers of about 10
kcal mol-1 in each case (Table 6).
1
The H NMR spectrum of the mesityl complex, (C5Me5)2-
Yb(dad(H)-mesityl), is useful, since it provides an assignment
of the meta CH resonances and shows the effect of steric
hindrance on the rotation barrier. The 1H NMR spectrum shows
two inequivalent ortho Me and meta CH resonances at +80 °C
(Figure 8). These inequivalent ortho CMe and meta CH
resonances are attributable to slow distal-proximal site ex-
change that is observed in the p-tolyl and p-anisyl complexes
only at low temperature. Even though distal-proximal site
exchange is slow at 80 °C, the mesityl complex has averaged
C2V symmetry, since the C5Me5 rings are equivalent. As the
temperature is lowered, the C5Me5 rings become inequivalent,
∆Gq(Tc ) 260 K) ) 11 kcal mol-1. Since the distal and
proximal CMe and CH groups are still pairwise equivalent, the
complex contains a time-averaged vertical plane of symmetry.
On further cooling, the CMe group resonances broaden but do
not disappear in the baseline; thus, oscillation of the N-mesityl
Figure 7. Chemical shift (δ) vs T-1 plots of (a, top) the 1H NMR
resonances of (C5Me5)2Yb(dad(H)-p-tolyl) in toluene-d8 at tem-
peratures from -70 to +90 °C and (b, bottom) the ortho and meta
CH 1H NMR resonances of (C5Me5)2Yb(dad(H)-p-tolyl) in toluene-
d8 at temperatures from -70 to +90 °C.
coalesced resonance, due to 4H, is linear in temperature. These
resonances are either the ortho or meta CH’s, and it is difficult

2302 Organometallics, Vol. 26, No. 9, 2007
Walter et al.
Figure 10. Chemical shift (δ) vs T-1 plot of the 1H NMR
resonances of (C5Me5)2Yb(dad(H)-t-Bu) in toluene-d8 at temper-
atures from -70 to +90 °C.
Figure 9. Chemical shift (δ) vs T-1 plot of the 1H NMR backbone
CH/CMe resonances of (C5Me5)2Yb(dad(H)-p-anisyl) and (C5Me5)2-
Yb(dad(Me)-p-anisyl) in toluene-d8 at temperatures from -70 to
+90 °C.
groups is not slow enough to remove the mirror plane,
generating a complex with C1 symmetry.43
1
The variable-temperature H NMR spectra of (C5Me5)2Yb-
(dad(Me)-p-anisyl) are similar to those of (C5Me5)2Yb(dad(H)-
p-tolyl) and (C5Me5)2Yb(dad(H)-p-anisyl); complete δ vs T-1
plots are given in the Supporting Information, and the plots for
the backbone CH and CMe resonances are shown in Figure 9.
The plot shows that the backbone CMe resonance has a
strikingly large temperature dependence; it changes from δ +150
(-70 °C) to -50 ppm (+90 °C). The slope is negative, in
contrast to the backbone CH resonance in (C5Me5)2Yb(dad(H)-
p-tolyl), which is positive, as shown in Figure 9. The different
sign of the slope in the δ vs T-1 plot is expected when a methyl
group replaces a hydrogen atom at a given site in a paramagnetic
molecule when the Fermi contact term dominates the pseudo-
contact term, since the hyperfine coupling constants aH and aMe
have different signs.31 Similarly, the change of sign of the slope
in the δ vs T-1 plot was observed in (C5Me5)2Yb(bipy-X), when
the X substituents in the 4,4′-sites are changed from H to Me.3
(c) Alkyl-Substituted Complexes, (C5Me5)2Yb(dad(H)-R)
(R ) Alkyl). In general, the patterns of chemical shifts in the
alkyl-substituted complexes are similar to those of the aryl
complexes, with one notable exception: viz., when R ) t-Bu.
In this case, the resonances due to the CMe3 groups are not
Figure 11. Chemical shift (δ) vs T-1 plot of the 1H NMR
resonances of (Me4C5H)2Yb(dad(H)-t-Bu) in toluene-d8 at temper-
atures from -70 to +90 °C.
observed at 20 °C in C6D6 or C7D8. However, as shown in the
δ vs T-1 plot in Figure 10, the resonances due to the CMe3
groups are observed at 95 °C (T-1 ) 0.0027 K-1) as a broad
singlet that disappears into the baseline as the temperature is
reduced below 80 °C and then emerge from the baseline as two
widely separated resonances at δ 9.6 and 119 ppm in a 2:1 ratio,
respectively. The chemical shifts of the decoalesced resonances
are then independent of temperature. This temperature behavior
is presumably due to hindered rotation around the N-CMe3
bond that is slow on the NMR time scale below 250 K, resulting
from steric hindrance between the CMe3 groups and the CMe
groups on the C5Me5 rings. Two complexes, (C5H5)2Yb(dad-
(H)-t-Bu)11 and (C5Me4H)2Yb(dad(H)-t-Bu), have been prepared
and characterized as detailed in the Experimental Section and
Table 2 in order to examine the effect of cyclopentadienyl ring
substituents on the rotation barrier.
(40) Lukens, W. W.; Beshouri, S. M.; Stuart, A. L.; Andersen, R. A.
Organometallics 1999, 18, 1247-1252.
(41) Luke, W. D.; Streitwieser, A. J. Am. Chem. Soc. 1981, 103, 3241-
3243.
(42) Sandstro¨m, J. Dynamic NMR Spectroscopy; Academic Press: New
York, 1982.
(43) In all of the diazabutadiene complexes with (C5Me5)2Yb, except
the mesityl derivative just discussed, the chemical shift of the C5Me5
resonances moves downfield, passes through a minimum value at about
-50 to -60 °C, and then moves upfield as the temperature is decreased to
-80 °C. This nonlinear behavior is due to temperature-dependent processes
whose populations and chemical shifts depend on temperature.32 The
averaged chemical shift, δav, is the weighted chemical shift of two nuclei,
δav ) aδA + bδB. If the populations are unequal, a * b, then δav vs T-1
will be the sum of the individual δA vs T-1 and δB vs T-1 plots, which will
be nonlinear if the populations change with temperature. Conversely, if a
) b but δA * δB, δav vs T-1 will be non-linear if δA and/or δB exhibit
nonlinear T-1 dependences. Another contributing factor might be the
movement of the C5Me5 rings relative to the magic angle with temperature,
which is related to the (1 - 3 cos2 θ) term in the pseudo-contact contribution.
The motion of the C5Me5 might be correlated to the motion of the
N-heterocyclic base about the horizontal mirror plane; however, there is
insufficient information in the line shape to delineate the physical process
responsible.
Figure 11 shows the resonances of the C5Me4H derivative as
a δ vs T-1 plot, which is similar to that of (C5Me5)2Yb(dad-
(H)-t-Bu) (Figure 10), with the exception of the temperature at
which the CMe3 resonances disappear into and reappear from
the baseline. In the C5Me4H metallocene, the coalescence
temperature is about 35 °C lower than that for the C5Me5

Spin Coupling in Ytterbocene Complexes
Organometallics, Vol. 26, No. 9, 2007 2303
Table 7. Barrier to CMe3 Site Exchange in
Cp′2Yb(dad(H)-t-Bu)
(C5Me5)2Yb-
(Me4C5H)2Yb-
(dad(H)-t-Bu)
(dad(H)-t-Bu)
∆νa/Hz
∼50000
∼345
13
∼54500
b
Tc /K
∼235
∆GAqc/kcal mol-1
∆GBqc/kcal mol-1
9
8
12
a Distance between the signals in Hz. b Tc ) coalescence temperature
(400 MHz operating frequency). c Tc and ∆ν are determined by an
extrapolation of the temperature dependence of the CMe3 protons in C7D8,
as described in the Experimental Section. The free energy of activation,
∆Gq, is given without specifying the uncertainty in the value.
Section, and the values obtained for the activation energies given
in Table 7 are reasonable, as they are similar to those in
Table 6.
The relative barriers are consistent with steric hindrance
playing a large role in the barrier. However, the X-ray crystal
structures of (C5Me5)2Yb(dad(H)-t-Bu) and (C5H5)2Yb(dad(H)-
t-Bu) do not show short intramolecular distances, though the
Yb-N(av) distance in (C5Me5)2Yb(dad(H)-t-Bu) of 2.390(2) Å
is significantly longer than the equivalent distance in the C5H5
complex of 2.306(9) Å.11,12 The representation of the crystal
structure of the former complex by van der Waals spheres shows
that the CMe3 groups and C5Me5 rings are interlocking (see
the Supporting Information for this representation). Unfortu-
nately, the chemically inequivalent methyl resonances in the
less sterically congested complex, (C5Me4H)2Yb(dad(H)-t-Bu),
just disappear into the baseline and the C5Me4H rotation barrier
cannot be estimated.
Figure 12. Chemical shift (δ) vs T-1 plot of the 1H NMR
resonances of (C5Me5)2Yb(dad(H)-i-Pr) in toluene-d8 at tempera-
tures from -70 to +90 °C.
derivative, consistent with less steric hindrance to rotation. In
addition, the CMe3 resonance is a single resonance down to
-70 °C in (C5H5)2Yb(dad(H)-t-Bu); the δ vs T-1 plot is
available as Supporting Information.
The temperature dependence of the resonances of (C5Me5)2-
Yb(dad(H)-i-Pr) (Figure 12) shows that the CHMe2 resonances
do not decoalesce to -70 °C, which is also consistent with the
steric hindrance to rotation postulate. The variable-temperature
spectra of (C5Me5)2Yb(dad(H)-i-Pr) contain additional informa-
tion, since the slopes of the δ vs T-1 plots for the CH and CMe2
resonances have opposite signs. As mentioned above, this is
the expected behavior, when the Fermi contact term dominates
the pseudo-contact term.31 This behavior shows that the unpaired
spin density is delocalized onto the methyl groups, presumably
by hyperconjugation, consistent with an electronic component
to the N-C rotation barrier. It is noteworthy that the EPR
spectrum of [dad(H)-t-Bu]•- shows unpaired spin density on
the CMe3 groups,44 which was attributed to hyperconju-
gation. Thus, NMR and EPR spectra support the idea of an
electronic contribution to the N-C(R) rotation barrier; how-
ever, steric hindrance to rotation is likely to be the major
contributor.45
(d) Quantitative Evaluation of the N-CMe3 Rotation
Barrier. The rotation barriers in dad(H)-R, where R is aryl,
given in Table 6, are readily obtained since they involve
coalescence of two resonances of equal intensity.40-42 This
graphical method cannot be used, however, for estimating the
free energy barrier in the dad(H)-t-Bu complexes, since the
populations of the exchanging sites are unequal. In a diamagnetic
compound, line-shape analysis is used to solve this problem,
but this method requires that the spin-spin relaxation time (T2)
does not depend on temperature, a condition that is not valid
for paramagnetic compounds, whose line widths often are very
temperature dependent. An analytical method, however, has been
developed for estimating the activation free energies for a 2:1
site exchange.46 This method is outlined in the Experimental
While the work described in this paper was in progress, the
synthesis and crystal structure of (C5Me5)2Yb(dad(H)-t-Bu) were
reported;12 however, the published 1H NMR spectrum is
different from ours. The cell dimensions of our complex agree
with those reported (see the Experimental Section for details),
showing that the solids are identical. Trifonov et al. report that
the CMe3 resonances appear as four unequal area resonances
(δ -6.7, 12.2, 19.2, and 24.2 ppm) at 20 °C in C6D6, whereas
we do not observe these resonances under the same conditions
(Figure 13). We noted, however, that when our material is
allowed to stand at room temperature in an NMR tube that is
sealed with a rubber septum, resonances appear at the chemical
shifts reported by Trifonov et al., suggesting that the reported
resonances are due to decomposition products. In a sealed NMR
tube, (C5Me5)2Yb(dad(H)-t-Bu) is indefinitely stable in C6D6
at 20 °C. Trifonov et al. also report that thf-d8 displaces dad-
(H)-t-Bu, giving (C5Me5)2Yb(thf)x.12 We have reproduced this
observation with neat thf-d8; however, addition of a drop of
thf-d8 to a solution of (C5Me5)2Yb(dad(H)-t-Bu) in C6D6 does
1
not perturb the H NMR chemical shifts at 20 °C. In contrast,
(C5Me5)2Yb(dad(H)-p-tolyl) does not exchange with neat thf-
d8, since the resonances in that solvent are identical with those
obtained in C6D6 at 20 °C. 2,2′-Bipyridine displaces the
diazabutadiene ligand in (C5Me5)2Yb(dad(H)-t-Bu) on mixing,
giving (C5Me5)2Yb(bipy) and free dad(H)-t-Bu; the reverse
reaction does not occur, clearly showing the relative thermo-
dynamic stability of the bipy complex.1 As reported, free bipy
does not exchange with (C5Me5)2Yb(bipy) on the NMR time
scale; however, it does exchange with 4,4′-dimethyl-2,2′-
bipyridine on the chemical time scale.1 These, and other recently
reported exchange studies,47,48 show that the mechanism of
ligand exchange also involves a redox process.
(44) Clopath, P.; v. Zelewsky, A. HelV. Chim. Acta 1972, 55, 52-66.
(45) The low exchange limiting 1H NMR spectrum of (C5Me5)2Yb(dad-
(H)-adamantyl) could not be reached, and the rotation barriers cannot be
determined.
(46) Shanan-Atidi, H.; Bar-Elli, K. H. J. Phys. Chem. 1970, 74, 961-
963.
(47) Wang, J.; Amos, R. I. J.; Frey, A. S. P.; Gardiner, M. G.; Cole, M.
L.; Junk, P. C. Organometallics 2005, 24, 2259-2261.
(48) Evans, W. J.; Davis, B. L. Chem. ReV. 2002, 102, 2119-2136.

2304 Organometallics, Vol. 26, No. 9, 2007
Walter et al.
Figure 13. 1H NMR spectrum of (C5Me5)2Yb(dad(H)-t-Bu) in benzene-d6 at 20 °C.
expressed as øT, vs the reduced observed chemical shift of the
backbone CH/CMe resonance in the solution 1H NMR spectrum,
expressed as ϑ298, for several dad(R′)-R complexes. The reduced
chemical shift (ϑ298) is defined as the observed chemical shift
at a given temperature multiplied by that temperature, δT,
divided by 298 K. These graphs dramatically illustrate the linear
relation between solution- and solid-state magnetic properties
and show that the observed magnetic properties result from the
individual molecule, not from the collection of molecules in
the ensemble. This implies that the phenomena responsible for
the nonlinear behavior in the individual øT vs T and ϑ298 vs T
plots are traceable to the magnetic properties of the individual
molecular complexes and, therefore, to the energy and overlap
of the molecular orbitals.
The origin of the slight curvature at low temperature (-60
to -70 °C) in Figure 14b, but not in Figure 14a, parallels the
nonlinear chemical shift behavior of the backbone H over a
similar temperature range observed for the alkyl derivatives in
Figures 10, 12, S11, and S15 (see Supporting Information),
which is not compensated by changes in øT (in the solid state).
Since all processes are intramolecular in solution, the nonlin-
earity might be due to slowing of some dynamic process, such
as motion of the N-heterocyclic ligand about the horizontal
mirror plane, that does not occur in the solid state in this
temperature regime.
In an earlier article,3 a qualitative symmetry orbital model
was developed, in which the antiferromagnetic exchange
coupling was traced to the interaction between the electron or
hole on the bent-sandwich fragment (C5Me5)2YbIII(f13) which
is of f or d parentage or hybridization therefrom, with the
electron in the radical anion in an orbital of b1 symmetry. This
qualitative coupling scheme depends on the relative orbital
energies and overlap integrals of the individual fragments and
ultimately on the crystal field states of B1 symmetry on the
ytterbocene fragment.
Figure 14. Reduced observed chemical shift (ϑ298) vs øT plots of
1H NMR resonances of backbone-R′ in toluene-d8 at temperatures
from -70 °C to +90 °C of (a, top) (C5Me5)2Yb(dad(R′)-aryl)
systems and (b, bottom) (C5Me5)2Yb(dad(H)-alkyl) systems.
(e) Relation between Solid- and Solution-State Magnetism.
Figure 14 shows plots of the solid-state magnetic susceptibility,

Spin Coupling in Ytterbocene Complexes
Organometallics, Vol. 26, No. 9, 2007 2305
Chart 1
(C5Me5)2Yb(N,N′-di-p-tolyl-1,4-diazabutadiene). N,N′-Di-p-
tolyl-1,4-diazabutadiene49 (0.27 g, 1.1 mmol) and (C5Me5)2Yb-
(OEt2)50 (0.59 g, 1.1 mmol) were weighed into a Schlenk flask
under nitrogen and dissolved in toluene (80 mL), and the dark green-
black solution was stirred at room temperature for 3 h. The solvent
was removed under dynamic vacuum, and the residue was sublimed
under diffusion pump vacuum at 160-180 °C. The sublimed
material was dissolved in ca. 100 mL of pentane, the mixture was
filtered, and the filtrate was concentrated and cooled to -25 °C
overnight to form dark green crystals (0.58 g, 0.83 mmol, 76%).
Mp: 220 °C dec. Anal. Calcd for C36H46N2Yb: C, 64.6; H, 7.02;
1
N, 4.08. Found: C, 64.2; H, 7.02; N, 4.08. H NMR (C6D6, 20
°C): δ 58.3 (6 H, ν1/2 ) 90 Hz, p-Me), 50.2 (4 H, ν1/2 ) 700 Hz,
meta CH), 1.74 (30 H, ν1/2 ) 55 Hz, C5Me5), -123.7 (2H, ν1/2
)
330 Hz, backbone dad-CH). The ortho CH was not observed at 20
°C in C6D6. The EI mass spectrum showed a molecular ion at m/e
680. The parent ion isotopic cluster was simulated (calcd %, obsd
%): 676 (11, 11), 677 (41, 40), 678 (72, 72), 679 (71, 70), 680
(100, 100), 681 (36, 37), 682 (42, 42), 683 (15, 15), 684 (3, 3). IR
(Nujol mull; CsI windows; cm-1): 2720 (w), 1604 (vs), 1553 (m),
1497 (vs), 1328 (s), 1313 (sh w), 1274 (vs), 1181 (m), 1143 (s),
1113 (vw), 1026 (m), 1000 (m), 897 (br s), 830 (vw), 796 (s), 722
(w), 700 (vw), 647 (vw), 616 (vw), 591 (vw), 513 (m), 415 (br w),
384 (br w), 313 (vs), 303 (vs).
(C5Me5)2Yb(N,N′-diisopropyl-1,4-diazabutadiene). N,N′-Di-
isopropyl-1,4-diazabutadiene51 (0.21 g, 1.5 mmol) and (C5Me5)2-
Yb(OEt2) (0.78 g, 1.5 mmol) were weighed into a Schlenk flask
under nitrogen and dissolved in toluene (80 mL), and the bright
red solution was stirred at room temperature for 3 h. The solvent
was removed under dynamic vacuum, and the residue was sublimed
under diffusion pump vacuum at 100-120 °C. The sublimed
material was dissolved in a minimum amount of pentane (ca. 5
mL) and cooled to -80 °C for several days. The compound
crystallized as large deep red blocks (0.48 g, 0.82 mmol, 55%),
and it was very soluble in aromatic and aliphatic hydrocarbons.
Mp: 208-211 °C dec. Anal. Calcd for C28H46N2Yb: C, 57.50; H,
7.93; N, 4.79. Found: C, 57.55; H, 8.04; N, 4.74. 1H NMR (C6D6,
Conclusions
The studies described in this paper, along with those in earlier
papers,1-3 show that the bipy-X and dad(R′)-R ligands, whose
HOMO’s have identical symmetries but different energies, form
complexes with (C5Me5)2Yb that are not diamagnetic. The extent
of paramagnetism in general, as judged by the effective magnetic
moment at 300 K, qualitatively correlates with the reduction
potential of the ligands. However, the magnetic susceptibility
behavior of the individual complexes as a function of temper-
ature is far from simple. This and earlier papers present a
phenomenological description of the systematics of the solid-
state magnetic properties of the ytterbocene complexes with
heterocyclic-nitrogen bases, but a molecular level of understand-
ing of the exchange coupling is not possible from these studies.
However, a qualitative model is postulated that ascribes the
antiferromagnetic coupling to an electron on the Cp′2YbIII
fragment of b1 symmetry with the electron in the b1 molecular
orbital in the ligand radical anion. The tools used by physicists,
such as those described in ref 8, must be employed in order to
achieve a molecular level of understanding; these and related
studies will be reported in due course.
20 °C): δ 108.9 (2 H, ν1/2 ) 260 Hz, CHMe2), 32.3 (12 H, ν1/2
)
12 Hz, CHMe2), 2.83 (30 H, ν1/2 ) 9 Hz, C5Me5), -28.3 (2H, ν1/2
) 170 Hz, backbone dad-CH). The EI mass spectrum showed a
molecular ion at m/e 584. The parent ion isotopic cluster was
simulated (calcd %, obsd %): 580 (11, 10), 581 (43, 42), 582 (72,
75), 583 (68, 66), 584 (100, 100). 585 (29, 30), 586 (41, 36), 587
(13, 12). IR (Nujol mull; CsI windows; cm-1): 2720 (w), 1575
(w), 1548 (w), 1338 (w), 1310 (br w), 1262 (m), 1232 (m), 1160
(m), 1105 (m), 1021 (br m), 800 (br m), 784 (m), 720 (br m), 610
(vw), 475 (vw), 450 (vw), 395 (br vw), 295 (br s).
Experimental Section
General Comments. All reactions, product manipulations, and
physical studies have been carried out as previously described.1-3
The temperatures quoted in the variable-temperature NMR are
obtained through calibration of the probe, in the specific instrument
used, by recording the chemical shift of methanol (low temperature)
and ethylene glycol (high temperature).42 The 1,4-diazabutadiene
ligands were purified by crystallization and/or sublimation prior to
use. The ∆Gq value for an unequal population site exchange system
was determined using eqs 2 and 3.46 This method has been applied
(C5Me5)2Yb(N,N′-di-tert-butyl-1,4-diazabutadiene). N,N′-Di-
tert-butyl-1,4-diazabutadiene52 (0.23 g, 1.37 mmol) and (C5Me5)2-
Yb(OEt2) (0.71 g, 1.37 mmol) were weighed into a Schlenk flask
under nitrogen and dissolved in toluene (80 mL), and the dark red
solution was stirred at room temperature for 3 h. The solvent was
removed under dynamic vacuum, and the residue was sublimed
under diffusion pump vacuum at 180-190 °C. The sublimed
material was dissolved in a minimum amount of pentane and cooled
to -80 °C for several days. The compound crystallized as deep
red blocks (0.45 g, 0.74 mmol, 54%). Mp: 220-222 °C dec. Anal.
Calcd for C30H50N2Yb: C, 58.90; H, 8.24; N, 4.58. Found: C,
Tc
X
∆GAq ) 4.57Tc 10.62 + log
+ ln
+ ln
(2)
(3)
(
)
)
∆V
2π(1 - ∆p)
Tc
X
∆GBq ) 4.57Tc 10.62 + log
(
∆V
2π(1 + ∆p)
to the spin equilibrium in dimethylmanganocene.38 Equa-
tions 2 and 3 give the change in free energy of activation for species
A and B. The terms in these equations are defined by Chart 1,
with Tc, the coalescence temperature, ∆ν, the extrapolated chemical
shift difference in hertz, and the expressions log[X/(2π(1 ( ∆p)],
where ∆p ) -0.33 when the population difference is 1:2, are
evaluated from Figure 2 in ref 46. However, the ∆Gq values
obtained in this way are only valid at the coalescence temperature,
Tc.46
1
58.70; H, 8.14; N, 4.63. H NMR (C6D6, 20 °C): δ 0.15 (30 H,
ν1/2 ) 8 Hz, C5Me5), -26.6 (2H, ν1/2 ) 200 Hz, backbone dad-
CH). The resonance due to the t-Bu protons was not observed at
room temperature. The EI mass spectrum showed a molecular ion
at m/e 612 amu. The parent ion isotopic cluster was simulated (calcd
(50) Tilley, T. D.; Boncella, J. M.; Berg, D. J.; Burns, C. J.; Andersen,
R. A. Inorg. Synth. 1990, 27, 146-149.
(51) tom Dieck, H.; Franz, K.-D.; Majunke, W. Z. Naturforsch. 1975,
30b, 922-925.
(49) Kliegman, J. M.; Barnes, R. K. J. Org. Chem. 1970, 35, 3140-
3143.
(52) Kliegman, J. M.; Barnes, R. K. Tetrahedron 1970, 26, 2555-2560.

2306 Organometallics, Vol. 26, No. 9, 2007
Walter et al.
%, obsd %): 608 (11, 10), 609 (42, 42), 610 (72, 75), 611 (68,
66), 612 (100, 100). 613 (31, 30), 614 (41, 37), 615 (13, 12), 616
(2, 2). IR (Nujol mull; CsI windows; cm-1): 2720 (w), 1542 (w),
1492 (m), 1389 (sh), 1368 (s), 1358 (m), 1263 (s), 1220 (w), 1188
(br vs), 1118 (vw), 1100 (w), 1020 (br m), 990 (m), 895 (m), 800
(m), 782 (m), 760 (w), 721 (m), 612 (br w), 532 (vw), 478 (br m),
383 (br m), 280 (br vs).
Unit Cell Determination of (C5Me5)2Yb(dad(H)-t-Bu). A
crystal measuring 0.12 × 0.20 × 0.22 mm was mounted on a glass
fiber using Paratone N hydrocarbon oil and transferred to a Bruker
SMART 1k CCD diffractometer. Cell constants and an orientation
matrix were obtained of the measured positions of reflections with
I > 10σ to give the following unit cell: a ) 13.7007(29) Å, b )
15.5126(13) Å, c ) 13.3171(18) Å, R ) â ) γ ) 90°. These values
are in agreement with the reported values in the orthorhombic space
group Pna21: a ) 13.7169(1) Å, b ) 15.5660(2) Å, c ) 13.3681-
(2) Å, R ) â ) γ ) 90°.12
(vw), 710 (br w), 652 (w), 628 (w), 530 (w), 480 (vbr w), 390 (vbr
w), 295 (br s), 275 (sh), 245 (m).
(C5Me5)2Yb(N,N′-diadamantyl-1,4-diazabutadiene). N,N′-Di-
adamantyl-1,4-diazabutadiene54 (0.33 g, 1.01 mmol) and (C5Me5)2-
Yb(OEt2) (0.53 g, 1.02 mmol) were weighed into a Schlenk flask
under nitrogen and dissolved in toluene (120 mL), and the dark
red solution was stirred at room temperature for 2 h. The solution
was filtered, concentrated to ca. 50 mL, and cooled to -25 °C
overnight to form brown-red crystals (0.62 g, 0.80 mmol, 79%).
Mp 238-239 °C rev. Anal. Calcd for C42H62N2Yb: C, 65.68; H,
8.14; N, 3.65. Found: C, 65.51; H, 8.31; N, 3.69. 1H NMR (C6D6,
20 °C): δ 18.9 (2H, ν1/2 ≈ 650 Hz, adamantyl CH2), 16.7 (6 H,
ν1/2 ≈ 400 Hz, adamantyl CH2), 12.5 (7H, ν1/2 ) 67 Hz, adamantyl
CH), 0.14 (30 H, ν1/2 ) 8 Hz, C5Me5), -25.2 (2H, ν1/2 ) 200 Hz,
backbone dad CH). IR (Nujol mull; CsI windows; cm-1): 2720
(vw), 2670 (vw), 2650 (vw), 1622 (br w), 1482 (m), 1352 (m),
1341 (m), 1312 (m), 1303 (m), 1272 (m), 1234 (vs), 1185 (m),
1178 (m), 1112 (m), 1185 (w), 1178 (m), 1112 (m), 1088 (vs),
1069 (m), 1012 (br m), 995 (w), 980 (vw), 970 (vw), 954 (w), 938
(w), 912 (m), 812 (br m), 775 (s), 721 (m), 698 (vw), 620 (br w),
470 (br m), 418 (br m), 380 (br w), 355 (br w), 280 (br vs), 245
(w), 221(m).
(C5Me5)2Yb(N,N′-di-p-anisyl-1,4-diazabutadiene). N,N′-Di-p-
anisyl-1,4-diazabutadiene49 (0.33 g, 0.99 mmol) and (C5Me5)2Yb-
(OEt2) (0.51 g, 0.99 mmol) were weighed into a Schlenk flask under
nitrogen and dissolved in toluene (80 mL), and the dark green-
black solution was stirred at room temperature for 1 h. The solvent
was removed under dynamic vacuum, and the residue was sublimed
under diffusion pump vacuum at 200-220°C. The sublimed
material was dissolved in ca. 100 mL of pentane and the mixture
filtered, and the filtrate was concentrated and cooled to -25 °C
overnight to form dark green shiny plates (0.45 g, 0.63 mmol, 64%).
Mp: 260-262 °C dec. Anal. Calcd for C36H46N2O2Yb: C, 60.75;
(C5Me5)2Yb(N,N′-dimesityl-1,4-diazabutadiene). N,N′-Dimesi-
tyl-1,4-diazabutadiene55 (0.44 g, 1.5 mmol) and (C5Me5)2Yb(OEt2)
(0.78 g, 1.5 mmol) were weighed into a Schlenk flask under
nitrogen and dissolved in toluene (80 mL), and the green-blue
solution was stirred at room temperature for 3 h. The solution was
filtered, concentrated, and cooled to -25 °C overnight. The
compound crystallized as dark blue-green crystals (0.52 g, 0.71
mmol, 47%), and it was nearly insoluble in aliphatic hydrocarbons
and moderately soluble in aromatic hydrocarbons. Mp 230-232
°C dec. Anal. Calcd for C40H54N2Yb: C, 65.28; H, 7.40; N, 3.81.
1
H, 6.51; N, 3.94. Found: C, 60.60; H, 6.56; N, 4.21. H NMR
(C6D6, 20 °C): δ 46.9 (4 H, ν1/2 ) 350 Hz, meta CH), 18.9 (6 H,
ν1/2 ) 4 Hz, para OMe), 1.76 (30 H, ν1/2 ) 13 Hz, C5Me5), -109.3
(2H, ν1/2 ) 270 Hz, backbone dad-CH). The ortho CH was not
observed at 20 °C in C6D6. The EI mass spectrum showed a
molecular ion at m/e 712. The parent ion isotopic cluster was
simulated (calcd %, obsd %): 708 (10, 10), 709 (41, 40), 710 (72,
72), 711 (70, 70), 712 (100, 100). 713 (37, 37), 714 (42, 42), 715
(16, 16), 716 (3, 3). IR (Nujol mull; CsI windows; cm-1): 2720
(vw), 1602 (m), 1570 (br w), 1503 (vs), 1300 (m), 1278 (s), 1250
(vs), 1188 (w), 1172 (m), 1112 (sh), 1098 (br m), 1052 (br m),
822 (m), 800 (m), 780 (m), 722 (br w), 673 (br w), 638 (vw), 562
(vw), 532 (w), 510 (vw), 480 (vw), 388 (w), 308 (br s).
1
Found: C, 65.55; H, 7.37; N, 3.98. H NMR (C6D6, 20 °C): δ
224.9 (6 H, ν1/2 ) 120 Hz, mesityl Me), 62.5 (2 H, ν1/2 ) 32 Hz,
mesityl CH), 58.2 (6 H, ν1/2 ) 24 Hz, mesityl Me), 50.4 (2 H, ν1/2
) 24 Hz, mesityl CH), 16.6 (6 H, ν1/2 ) 35 Hz, mesityl Me), 0.29
(30 H, ν1/2 ) 90 Hz, C5Me5), -146.2 (2H, ν1/2 ) 240 Hz, backbone
dad CH).
(C5Me4H)2Yb(N,N′-di-tert-butyl-1,4-diazabutadiene). N,N′-Di-
tert-butyl-1,4-diazabutadiene52 (0.085 g, 0.51 mmol) and (C5-
HMe4)2Yb(OEt2)42 (0.25 g, 0.51 mmol) were weighed into a
Schlenk flask under nitrogen and dissolved in toluene (80 mL),
and the bright red solution was stirred at room temperature for 2
h. The solvent was removed under dynamic vacuum, and the residue
was sublimed under diffusion pump vacuum at 180-190 °C to give
analytically and spectroscopically pure product as red crystalline
material (0.18 g, 0.31 mmol, 60%). Mp 242-245 °C dec. Anal.
Calcd for C28H46N2Yb: C, 57.61; H, 7.94; N, 4.80. Found: C,
[(C5Me5)2Yb(N,N′-di-p-anisyl-2,3-dimethyl-1,4-diazabutadi-
ene)]. N,N′-Di-p-anisyl-2,3-dimethyl-1,4-diazabutadiene53 (0.29 g,
0.98 mmol) and (C5Me5)2Yb(OEt2) (0.52 g, 1.00 mmol) were
weighed into a Schlenk flask under nitrogen and dissolved in toluene
(80 mL), and the dark green-brown solution was stirred at room
temperature for 1 h. The solvent was removed under dynamic
vacuum, and the residue was sublimed under diffusion pump
vacuum at 170-185 °C. The sublimed material was dissolved in
ca. 100 mL of pentane and the mixture filtered, and the filtrate
was concentrated and cooled to -25 °C overnight to form dark
green-brown crystals (0.4 g, 0.54 mmol, 55%). Mp: 219-221 °C
rev. Anal. Calcd for C38H50N2O2Yb: C, 61.69; H, 6.81; N, 3.79.
1
57.31; H, 8.00; N, 4.72. H NMR (C6D6, 20 °C): δ 43.1 (18 H,
ν1/2 ) 480 Hz, dad CMe3), 2.62 (12 H, ν1/2 ) 30 Hz, C5HMe4),
2.47 (12 H, ν1/2 ) 50 Hz, C5HMe4), -36.0 (2H, ν1/2 ) 180 Hz,
backbone dad CH), -38.0 (2H, ν1/2 ) 60 Hz, C5HMe5). The EI
mass spectrum showed a molecular ion at m/e 584. The parent ion
isotopic cluster was simulated (calcd %, obsd %): 580 (12, 11),
581 (43, 42), 582 (73, 74), 584 (100, 100), 585 (29, 29), 586 (41,
41), 587 (13, 14), 588 (2, 2).
1
Found: C, 61.48; H, 6.75; N, 3.66. H NMR (C6D6, 20 °C): δ
126.0 (6H, ν1/2 ) 360 Hz, backbone dad-CMe), 103.3 (4 H, ν1/2
)
800 Hz, ortho CH), 50.1 (4 H, ν1/2 ) 34 Hz, meta CH), 18.5 (6 H,
ν1/2 ) 8 Hz, p-OMe), 1.51 (30 H, ν1/2 ) 23 Hz, C5Me5). The EI
mass spectrum showed a molecular ion at m/e 740. The parent ion
isotopic cluster was simulated (calcd %, obsd %): 736 (11, 11),
737 (47, 46), 738 (72, 72), 739 (71, 70), 740 (100, 100). 741 (38,
40), 742 (42, 42), 743 (16, 16), 744 (3, 3). IR (Nujol mull; CsI
windows; cm-1): 2720 (vw), 1603 (w), 1560 (vw), 1555 (w), 1540
(w), 1500 (vs), 1350 (m), 1300 (m), 1290 (m), 1240 (br vs), 1210
(sh), 1180 (w), 1170 (w), 1105 (br w), 1038 (sh), 1028 (s), 980
(w), 870 (br vw), 838 (m), 820 (w), 800 (br m), 778 (vw), 762
Acknowledgment. This work was supported by the Director,
Office of Energy Research, Office of Basic Energy Sciences,
Chemical Sciences Division, of the U.S. Department of Energy
under Contract No. DE-AC03-76SF00098. We thank Dr. Fred
Hollander and Dr. Allen Oliver (at CHEXRAY, the UC
(54) Walker, I. Ph.D. Thesis, Eberhard-Karls-Universita¨t, Tu¨bingen,
Germany, 2002.
(55) Arduengo, A. J.; Krafczyk, R.; Schmutzler, R.; Craig, H. A.;
Goerlich, J. R.; Marshall, W. J.; Unverzagt, M. Tetrahedron 1999, 55,
14523-14534.
(53) tom Dieck, H.; Svoboda, M.; Greiser, T. Z. Naturforsch. 1981, 36b,
823-832.

Spin Coupling in Ytterbocene Complexes
Organometallics, Vol. 26, No. 9, 2007 2307
representation as well as øT vs T, observed chemical shift δ vs
T-1, and reduced observed chemical shift (ϑ298) and reduced
isotropic chemical shift (ϑ298iso) vs T plots referred to in the text.
This material is available free of charge via the Internet at http://
pubs.acs.org. Structure factor tables are available from the authors.
Crystallographic data were also deposited with the Cambridge
Crystallographic Data Centre. Copies of the data (CCDC Nos.
615276 615277) can be obtained free of charge via http://
www.ccdc.cam.ac.uk/data_request/cif, by e-mailing data_request@
ccdc.cam.ac.uk, or by contacting The Cambridge Crystallographic
Data Centre, 12, Union Road, Cambridge CB 1EZ, U.K. (fax +44
1223 336033).
Berkeley X-ray facility) for assistance with the crystallography,
the German Academic Exchange Service (DAAD) (M.D.W.)
and the NSERC (Canada) (D.J.B.) for fellowships, and Wayne
W. Lukens, Corwin H. Booth, and Prof. Frank H. Ko¨hler (TU
Mu¨nchen) for helpful discussions and Prof. Herbert Schumann
(TU Berlin) for providing the crystal data for (C5Me5)2Yb(dad-
(H)-t-Bu).
Supporting Information Available: CIF files, tables and
figures giving crystallographic data, labeling diagrams, and atomic
positions, anisotropic thermal parameters, bond distances, bond
angles, torsion angles, least-squares planes, and packing diagrams
for (C5Me5)2Yb(dad(H)-p-tolyl) and (C5Me5)2Yb(dad(H)-p-anisyl)
and figures giving (C5Me5)2Yb(dad(H)-t-Bu) as a van der Waals
OM0610142