Reactivity Study of Pyridyl-Substituted 1-Metalla-2,5-diaza-cyclopenta-2,4-dienes of Group 4 Metallocenes

Article abstract of DOI:10.1002/chem.201601337

In this work the reactivity of 1-metalla-2,5-diaza-cyclopenta-2,4-dienes of group 4 metallocenes, especially of the pyridyl-substituted examples, towards small molecules is investigated. The addition of H₂, CO₂, Ph?C≡N, 2-py?C≡N, 1,3-dicyanobenzene or 2,6-dicyanopyridine results in exchange reactions, which are accompanied by the elimination of a nitrile. For CO₂, a coordination to the five-membered cycle occurs in case of Cp*₂Zr(N=C(2-py)?C(2-py)=N). A 1,4-diaza-buta-1,3-diene complex is formed by H-transfer in the conversion of the analogous titanocene compound with CH₃?C≡N, PhCH₂?C≡N or acetone. For CH₃?C≡N a coupling product of three acetonitrile molecules is established additionally. In order to split off the metallocene from the coupled nitriles, we examined reactions with HCl, PhPCl₂, PhPSCl₂and SOCl₂. In the last case, the respective thiadiazole oxides and the metallocene dichlorides were obtained. A subsequent reaction produced thiadiazoles.

Full text of DOI:10.1002/chem.201601337

DOI: 10.1002/chem.201601337
Full Paper
&
Coordination Chemistry |Hot Paper|
Reactivity Study of Pyridyl-Substituted 1-Metalla-2,5-diaza-
cyclopenta-2,4-dienes of Group 4 Metallocenes
Lisanne Becker, Fabian Reiß, Kai Altenburger, Anke Spannenberg, Perdita Arndt, Haijun Jiao,
and Uwe Rosenthal*[a]
Dedicated to Prof. Dr. Angelika Brꢀckner on the occasion of her 60th birthday
Chem. Eur. J. 2016, 22, 1 – 14
1
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
Abstract: In this work the reactivity of 1-metalla-2,5-diaza-
cyclopenta-2,4-dienes of group 4 metallocenes, especially of
the pyridyl-substituted examples, towards small molecules is
investigated. The addition of H2, CO2, PhÀCꢀN, 2-pyÀCꢀN,
1,3-dicyanobenzene or 2,6-dicyanopyridine results in ex-
change reactions, which are accompanied by the elimination
of a nitrile. For CO2, a coordination to the five-membered
cycle occurs in case of Cp*2Zr(N=C(2-py)ÀC(2-py)=N). A 1,4-
diaza-buta-1,3-diene complex is formed by H-transfer in the
conversion of the analogous titanocene compound with
CH3ÀCꢀN, PhCH2ÀCꢀN or acetone. For CH3ÀCꢀN a coupling
product of three acetonitrile molecules is established addi-
tionally. In order to split off the metallocene from the cou-
pled nitriles, we examined reactions with HCl, PhPCl2,
PhPSCl2 and SOCl2. In the last case, the respective thiadiazole
oxides and the metallocene dichlorides were obtained. A
subsequent reaction produced thiadiazoles.
verse compounds and afforded, for example, amines,[8] azain-
doles,[7c] ketones,[8a,9] pyrroles[10] or other heterocycles.[11] The
Introduction
In many catalytic reactions all-C as well as heterometallacycles
are predicted as important intermediates. Therefore, their syn-
thesis, structure and especially their reactivity are of considera-
ble interest.[1]
second one is the exchange of [Cp’2M] (M=Ti, Zr) by S2Cl2,[9c,12]
[13a]
PCl3
or other electrophiles.[13b] For the 1-metalla-2-aza-cyclo-
penta-2,4-dienes of group 4 metallocenes (D), the products of
an alkyne–nitrile CÀC coupling, similar exchanges by
NiCl2(PPh3)2 and a subsequent reaction with alkynes to pyri-
dines were intensively investigated.[14]
In recent years, we gained access to nitrogen-containing
heterometallacycles of group 4 metallocenes by reactions of
Cp’2M(h2-Me3SiC2SiMe3) (Cp’=Cp=h5-cyclopentadienyl, Cp’2 =
rac-ebthi=rac-1,2-ethylene-1,1’-bis(h5-tetrahydroindenyl), Cp’=
Cp*=h5-pentamethylcyclopentadienyl; M=Ti, Zr) with ni-
triles.[2] We studied the synthesis and structure of the obtained
complexes and were now interested in their reactivity. In the
In our research, we focused especially on the 1-metalla-2,5-
diaza-cyclopenta-2,4-dienes (E). These were established by CÀC
couplings of aryl nitriles at [Cp*2M] (M=Ti, Zr) and isolated in
high yields.[15] This method could also be extended to dicyano-
benzenes, thus forming tri- and tetranuclear compounds.[16] Ex-
literature several conversions of group 4 metallocenes with RÀ ploring the reactivity of Cp*2M(N=C(Ph)ÀC(Ph)=N) (M=Ti, Zr)
CꢀN were already reported, but we noticed that only for some
of these products the reactivity was investigated. These con-
versions resulted in new group 4 metallocene compounds or
in a split-off of [Cp’2M].
An example for the first case is the formation of the diimine
complex Cp*2Zr[N=C(H,pTol)]2 upon addition of H2 to
Cp*2Zr(N=C(pTol)ÀC(pTol)=N).[3] Other products of reactions of
group 4 metallocenes with nitriles are known to dimerise to di-
nuclear compounds (Scheme 1, A)[4] or to couple with acetone,
anilines,[5] dimethyl acetylenedicarboxylate[6] or further nitriles[7]
to heterometallacycles (B, C).
towards H2, CO2 or CH3ÀCꢀN showed that the CÀC bond of
the five-membered metallacycles was easily cleaved.[15a] This re-
sulted in the elimination of one PhÀCꢀN and its replacement
by the substrate. Unfortunately, experiments to hydrolyse the
complexes were unsuccessful. In case of the macrocyclic prod-
ucts, their rapid decomposition under inert conditions imped-
ed the investigation of their reactivity.[16]
Recently, we isolated pyridyl-substituted 1-metalla-2,5-diaza-
cyclopenta-2,4-dienes in the reactions of Cp*2M(h2-Me3SiC2-
SiMe3) (M=Ti, Zr) with cyanopyridines.[17] In contrast to their
aryl-substituted analogues these compounds are easier to
access and more stable. This was rationalised by DFT calcula-
tions and proven by testing the catalytic activity of Cp*2Ti(N=
C(R)ÀC(R)=N) (R=Ph, 2-py, 4-py) in the ring-opening polymeri-
sation of e-caprolactone.[17] Moreover, the tetranuclear
[Cp*2Ti(N=C(m-C6H3N)ÀC=N)]4, obtained in the reaction of
Cp*2Ti(h2-Me3SiC2SiMe3) with 2,6-dicyanopyridine, did not de-
compose rapidly under inert conditions. These differences
might influence the reactivity of the 1-metalla-2,5-diaza-cyclo-
penta-2,4-dienes significantly. The higher stability of the pyrid-
yl-substituted compounds should enable different reactions or
a higher selectivity towards new metallocene complexes. Fur-
thermore, the split-off of [Cp*2M] is more probable to result in
stable organic substances. Besides, the compounds contain
vacant coordination sites with the N atoms of the pyridine
ring, which might lead to fascinating new complexes.[18] Owing
to this, the investigation of the reactivity of the pyridyl-substi-
tuted compounds is of considerable interest.
For the second case, the split-off of [Cp’2M] (M=Ti, Zr) from
the nitrile complexes, two different reaction types were report-
ed. The first is the hydrolysis, which was investigated for di-
Scheme 1. Selected examples of group 4 metallocene complexes, estab-
lished in reactions with nitriles.
[a] Dr. L. Becker, Dr. F. Reiß, Dr. K. Altenburger, Dr. A. Spannenberg, Dr. P. Arndt,
Dr. H. Jiao, Prof. Dr. U. Rosenthal
Leibniz-Institut fꢀr Katalyse e.V. an der Universitꢁt Rostock
Albert-Einstein-Straße 29a, 18059 Rostock (Germany)
E-mail: uwe.rosenthal@catalysis.de
Homepage: www.catalysis.de
Supporting information and ORCIDs from the authors for this article can be
found under http://dx.doi.org/10.1002/chem.201601337.
&
&
Chem. Eur. J. 2016, 22, 1 – 14
www.chemeurj.org
2
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
À12.3 kcalmolÀ1). The values for the first step are less exergon-
Results and Discussion
ic than those for the analogous phenyl-substituted compounds
Firstly, we investigated reactions of Cp*2M(N=C(2-py)ÀC(2-py)= (Ti: DG=À7.3 kcalmolÀ1, Zr: DG=À5.8 and À8.8 kcalmolÀ1).
N) (1a: M=Ti, 2: M=Zr)[17] with H2, CO2 and CH3ÀCꢀN. By
comparing the results to those of the phenyl-substituted ana-
logues we wanted to estimate the influence of the substitu-
ents at the 1-metalla-2,5-diaza-cyclopenta-2,4-dienes on their
reactivity. Secondly, we tested the possibility of separating the
metallocene and the coupled nitriles to gain access to unsatu-
rated organic compounds.
This explains the necessity of higher temperatures for the con-
versions of the pyridyl-substituted complexes 1a and 2.
Reactions with CO2
As in case of H2, at ambient temperature no reaction of 1a
with CO2 was observed. Thus, the solution was saturated with
CO2 and warmed to 508C for 4 d. This resulted in free 2-pyÀCꢀ
N and in 6 (Scheme 3) by the formal CÀC coupling of the re-
maining coordinated 2-cyanopyridine and CO2. Complex 6 was
isolated as a brown solid in high yield. In the 13C NMR and the
IR spectra the typical signals of the C=O and C=N moieties of
the five-membered metallacycle were found at 165.7 (C=O),
154.9 ppm (or 157.5) (C=N), and 1647 (C=O), 1578 cmÀ1 (C=N),
respectively.
Reactions with H2
The conversions of 1a and 2 with H2 were monitored by NMR
spectroscopy and the spectra are given in the Supporting In-
formation. Upon warming to 508C for one day both com-
pounds started to react with H2 to free 2-cyanopyridine and
the hydride complexes 3 and 4 (Scheme 2). Therefore, the
cleavage of the CÀC bond of the five-membered cycle was
necessary. The typical resonances of the imine N=CH groups
1
are found at 9.22 (3) and 9.78 ppm (4) in the H NMR and at
155.4 (3) and 160.1 ppm (4) in the 13C NMR spectra. For the ti-
tanium compound a completion of this conversion could not
be achieved by further warming. Instead, 1a and 3 seemed to
decompose. Thus, 3 was not isolable and was solely character-
ised by NMR spectroscopy.
Scheme 3. Reaction of 1a with CO2.
A similar exchange reaction with CO2 has been observed for
the analogous complex Cp*2Ti(N=C(Ph)ÀC(Ph)=N) (1b). In con-
trast to the conversion of 1a this proceeded at ambient tem-
perature within minutes.[15a] The reaction of 1b with CO2 pro-
vides more Gibbs free energy (DG=À10.0 kcalmolÀ1) than that
of 1a with CO2 (DG=À0.7 kcalmolÀ1), which is consistent with
the experiment.
Scheme 2. Reaction of 1a and 2 with H2.
Coupling of CO2 and a CꢀN bond at an organometallic com-
pound was rarely observed. To the best of our knowledge only
two further examples are known in which CO2 inserted into
the MÀC bond of h2-iminoacyl complexes.[20] Aside from that,
coupling of PhÀCꢀN and CO2 to a four-membered bridge (N=
CÀCÀO) in a bimetallic complex was described.[21] These con-
versions are of considerable interest, in regard of the CO2-fixa-
tion and -activation as well as of potential further reactions.
Considering this, we also investigated the reaction of the zir-
conium complex 2 with CO2. For the analogous compound
Cp*2Zr(N=C(Ph)ÀC(Ph)=N) the addition of CO2 at ambient tem-
perature afforded a mixture.[15a] The higher stability of the pyr-
idyl-substituted cycle 2 is supposed to induce a higher selec-
tivity. We observed an immediate colour change from yellow
to brown upon bubbling CO2 through a solution of 2 at ambi-
ent temperature. The subsequent work-up resulted in the iso-
lation of brown crystals of 7 in high yield (Scheme 4).
In case of the zirconium compound the diimine complex 5
was formed additionally upon further warming to 508C. There-
fore, the free 2-pyÀCꢀN inserted into the MÀH bond of 4
(Scheme 2). Similar insertions were already described for di-
1
verse zirconocene hydride species.[13b,19] In the H and 13C NMR
spectra the typical resonances of the imine N=CH moieties
were with 9.96 and 160.4 ppm, respectively, in the same range
as for 3 and 4. A complete conversion of 2 to 4 and further to
5 could not be achieved. Thus, 4 and 5 were, like 3, not isolat-
ed and are solely described by NMR spectroscopy.
The comparison of these results to those of reactions of the
aryl-substituted analogues with H2 shows that the obtained
products are similar.[3,15a] The difference is that, in those cases
the conversions could be completed within 3 days of warming,
without hints for follow-up products. Moreover, the analogous
zirconium compounds, with two Ph[15a] or pTol[3] groups instead
of 2-py, reacted already at ambient temperature with H2.
To understand this difference, we computed the Gibbs free
energies for the formation of 3 and 4 from 1a and 2 (DG=
À2.2 and 0.8 kcalmolÀ1, respectively) and of 5 from 4 (DG=
The molecular structure of 7 is depicted in Figure 1. It shows
that by the reaction of 2 with CO2 a new CÀN bond was
formed (C1ÀN1) and that the CO2 was coordinated via one of
its O atoms to the Zr centre. As a consequence of this, the O2-
Chem. Eur. J. 2016, 22, 1 – 14
www.chemeurj.org
3
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
two coupled 2-cyanopyridine units. The coordination of CO2 re-
sults in a doubling of these resonances for 7.
The formation of 7 is in many aspects quite remarkable.
Firstly, in contrast to all other reactions observed so far for 1-
metalla-2,5-diaza-cyclopenta-2,4-dienes of group 4 metals this
is the first time the CÀC bond of the five-membered metallacy-
cle remained intact. Secondly, related group 4 metallocene
complexes were isolated only once.[22] In those cases HBEt2 or
ClAlMe2 were coordinated to 1-metalla-2-aza-cyclopenta-2,4-
dienes (Scheme 5, A and B). Regarding these compounds, 7
might be described as Lewis acid-base adduct (Scheme 5) in
which the lone pair of the N1 atom stabilises the formal cat-
ionic C1 atom, which was formed by the coordination of CO2
to the Zr centre. Moreover, complexes, which are similar to 7,
were assumed as intermediates in the metallacycle transfer re-
actions to main-group elements like sulfur.[12b] However, heat-
ing 7 to 508C resulted in the elimination of CO2 and the forma-
tion of 2 and not in the elimination of [Cp*2Zr]. This is a good
indicator for a Lewis acid-base adduct.
Scheme 4. Reaction of 2 with CO2.
Figure 1. Molecular structure of 7 in the solid state. H atoms are omitted for
clarity. The thermal ellipsoids correspond to 30% probability. Selected bond
lengths [ꢁ] and angles [8]: O1-Zr1 2.276(1), N1-Zr1 2.222(2), N2-Zr1 2.201(2),
C1-O1 1.286(2), C1-N1 1.450(2), C2-N1 1.286(2), C3-N2 1.270(2), C2-C3
1.518(3), O2-C1-O1 130.0(2), C2-N1-C1 136.7(2), N1-C2-C3 108.6(2), N2-C3-C2
117.4(2), C2-N1-Zr1 123.8(1), C3-N2-Zr1 120.2(1).
C1-O1 angle is reduced to 130.0(2)8 and the C1=O1 double
bond is elongated to 1.286(2) ꢁ. The new four-membered met-
allacycle is planar with a mean deviation from the best plane
of 0.002 ꢁ.
Scheme 5. Adduct complexes comparable to 7.
A third aspect, which makes the formation of 7 so remark-
able, is that DFT calculations predicted an exchange reaction
to 7a (Scheme 4) to be thermodynamically more probable
(DG=À0.9 kcalmolÀ1, for 7: DG=0.1 kcalmolÀ1). Thus, the se-
lective reaction to 7 must be kinetically favoured. In case of
the analogous titanium complex 1a the Gibbs free energy for
a similar coordination is highly endergonic (DG=11.1 kcal
molÀ1). Therefore, it is in comparison to the exchange thermo-
dynamically not favourable.
Furthermore, the characterisation of 7 by X-ray crystallogra-
phy offers us the opportunity to determine the impact of the
coordination of CO2 on the five-membered metallacycle in 2.
The C2ÀN1 distance (1.286(2) ꢁ) in 7 is slightly longer than the
C3ÀN2 bond (1.270(2) ꢁ), which is the same as in 2 (1.274(2),
1.273(2) ꢁ).[17] The CÀC bond in 7 (C2ÀC3 1.518(3) ꢁ) is in com-
parison to 2 (1.543(2) ꢁ) slightly shortened, but is still in the
range of an elongated Csp2ÀCsp2 single bond. The NÀZr distan-
ces (N1ÀZr1 2.222(2), N2ÀZr1 2.201(2) ꢁ) are longer than in 2
(2.124(1), 2.134(1) ꢁ), which is more pronounced in case of
N1ÀZr1. The angles in the five-membered cycle are likewise in-
fluenced by the coordination. The N1-C2-C3 angle (108.6(2)8)
in 7 is smaller than the N2-C3-C2 angle (117.4(2)8), which is in
the same range as in 2 (117.0(1), 116.3(1)8). The C-N-Zr angles
were both widened in comparison to 2, which is again more
distinct in case of C2-N1-Zr1 (7: C2-N1-Zr1 123.8(1), C3-N2-Zr1
120.2(1)8; 2: 113.80(9), 114.09(8)8).
Reactions with CH3ÀCꢀN
Regarding the fact that the reactivity of the aryl- and pyridyl-
substituted 1-metalla-2,5-diaza-cyclopenta-2,4-dienes towards
H2 and CO2 was similar, we expected an exchange of one 2-cy-
anopyridine upon addition of CH3ÀCꢀN to 1a or 2. Thus, an
equilibrium between 1a/2, Cp*2M(N=C(2-py)ÀC(CH3)=N) and
Cp*2M(N=C(CH3)ÀC(CH3)=N) (M=Ti, Zr) would be established
(Scheme 6).
The five-membered cycle remained nearly planar (mean de-
viation from the best plane: 0.03 ꢁ). The angle between this
plane and that of the four-membered cycle (Zr1-O1-C1-N1) is
6.76(9)8. In summary, the coordination of CO2 has only a slight
impact on the bond lengths and angles in the five-membered
ring in 7. Besides, the resonances of the C=N bonds in the IR
spectra are in the same range as for 2 (7: 1665, 1581 cmÀ1; 2:
1654, 1583 cmÀ1).[17] But, the coordination of CO2 influences
1
the NMR spectra. In case of 2 four multiplets in the H NMR as
well as six signals in the 13C NMR spectra were assigned to the
Scheme 6. Possible exchange reaction at 1a and 2 with CH3ÀCꢀN.
&
&
Chem. Eur. J. 2016, 22, 1 – 14
www.chemeurj.org
4
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
1
Dissolving 1a and 2 separately in CH3ÀCꢀN resulted for,
both compounds, in free 2-pyÀCꢀN, but the expected prod-
ucts were not identified. For a better understanding of this, we
computed the formation free energies of the stepwise ex-
change of both 2-pyÀCꢀN moieties in 1a and 2 by CH3ÀCꢀN.
All four values are endergonic by 1.1 (Ti) and 1.9 kcalmolÀ1 (Zr)
for the first step as well as 3.9 (Ti) and 4.6 kcalmolÀ1 (Zr) for
the second step, and therefore in contrast to the analogous
conversions of the phenyl-substituted complexes (Ti: À1.5 and
À2.2 kcalmolÀ1; Zr: À1.4 and À0.9 kcalmolÀ1).[15a] Regarding
this, the exchange of one 2-pyÀCꢀN by CH3ÀCꢀN might be
possible, although the resulting equilibrium favours 1a (ratio
87:13) or 2, respectively (97:3). The second exchange is unlikely
to proceed. This difference in the reactivity of the phenyl- and
pyridyl-substituted 1-metalla-2,5-diaza-cyclopenta-2,4-dienes is
quite fascinating, but even more interesting is the outcome of
the reactions that occurred instead.
The H NMR spectrum of the reaction mixture was identical to
that of the conversion of 1a with CH3ÀCꢀN, except for the ab-
sence of the signals of the NÀH moieties of 8 and 9 and of the
Me groups of 9. This confirmed our assumption of acetonitrile
being the source of the additional H atoms in 8 (spectra are
given in the Supporting Information).
The characterisation of compound 8 by X-ray crystallography
(Figure 2) offered us the opportunity to directly compare this
1-metalla-2,5-diaza-cyclopent-3-ene to the 1-metalla-2,5-diaza-
cyclopenta-2,4-diene 1a. The CÀN bonds are with 1.363(2) and
1.365(1) ꢁ in the range of Csp2ÀNsp2 single bonds and distinctly
longer than in 1a: 1.274(2), 1.271(2) ꢁ.[15a] In contrast to this,
the CÀC bond is shortened from 1.531(2) ꢁ in 1a to 1.402(2) ꢁ
in 8. Furthermore, complex 8 is folded along the NÀN axis
(angle between the planes defined by Ti1, N1, N2 and N1, C1,
C2, N2: 24.53(5)8), thus showing an envelope conformation,
while 1a was nearly planar (mean deviation from the best
plane 0.02 ꢁ).
In case of the zirconium complex 2, we obtained a mixture
of diverse compounds. Unfortunately, separation attempts
were not successful and no definite zirconocene species were
identified. The titanium complex 1a was converted to two
new compounds. The solubility of these products is very simi-
lar, though slightly different in n-hexane. Thus, it was possible
to isolate a small amount of both substances in high purity as
purple (8) and yellow solid (9; Scheme 7).
Figure 2. Molecular structure of 8 in the solid state. H atoms, except for H1
and H2, are omitted for clarity. The thermal ellipsoids correspond to 30%
probability. Selected bond lengths [ꢁ] and angles [8]: C1-C2 1.402(2), C1-N1
1.363(2), C2-N2 1.365(1), N1-Ti1 2.024(1), N2-Ti1 2.026(1), N1-C1-C2 113.6(1),
N2-C2-C1 114.0(1), C1-N1-Ti1 114.35(8), C2-N2-Ti1 113.23(8).
Scheme 7. Reaction of 1a with excess CH3ÀCꢀN.
The purple solid was identified as the 1,4-diaza-buta-1,3-
diene complex 8. Formally, it was established by the addition
of two H atoms to the five-membered metallacycle in 1a.
Comparable complexes of group 4 metals were already de-
scribed, but mostly, they were formed by a CÀC coupling of
isonitriles[23] or by coordination of 1,4-diaza-buta-1,3-dienes.[24]
To the best of our knowledge, solely three such group 4 metal-
locene compounds were obtained by a CÀC coupling of two
nitriles and an H-transfer. The first one was established by
warming Cp*2Ti(N=C(Fc)ÀC(Fc)=N) [Fc=-(h5-C5H4)Fe(h5-C5H5)] to
808C,[15b] the second, a trinuclear compound, was obtained
upon addition of adiponitrile to Cp*2Ti(h2-Me3SiC2SiMe3)[16] and
the third was achieved in the reaction of Cp*2Zr(h2-
Me3SiC2SiMe3) with PhCH2ÀCꢀN.[2c] In case of the zirconocene
complex the source of the additional H atoms is supposed to
be the deprotonation of further PhCH2ÀCꢀN. For the other
two compounds their origin is unknown. The additional H
atoms in 8 seem to be provided by the CH3ÀCꢀN. A subse-
quent coupling of the remaining fragment (CH2ÀCN) at [Cp*2Ti]
might have formed the six-membered metallacycle in 9. In
order to prove that CH3ÀCꢀN is the source of the additional H
atoms in 8, we investigated the reaction of 1a with CD3ÀCꢀN.
The conversion of the cyclopentadiene in 1a to a cyclopen-
tene in 8 by H-transfer has only a slight impact on the position
of the Cp* and pyridyl group resonances in the 1H and
13C NMR spectra. In comparison to 1a, they are only slightly
shifted for 8. The signals of the NÀH moieties were identified
1
at 7.99 ppm in the H NMR and at 3345 cmÀ1 in the IR spec-
trum.
The second product of the reaction of 1a with CH3ÀCꢀN is
9 (Scheme 7). It was established by a trimerisation of CH3ÀCꢀN
at [Cp*2Ti], accompanied by an H-transfer. In consequence to
this, 9 contains a CꢀN triple, a C=N double as well as a CÀN
single bond and one of the former three methyl groups lost all
its H atoms. The six-membered cycle in 9 is therefore highly
unsymmetrical. Owing to this, two resonances were identified
1
in the H NMR (2.01 and 2.19 ppm) and in the 13C NMR spec-
trum (24.2 and 27.8 ppm) for the CH3 groups. The signal of the
1
NÀH group in the H NMR spectrum was found at 4.27 ppm.
Resonances at 124.6, 158.8 and 165.1 ppm in the 13C NMR
spectrum were assigned to the CꢀN triple, the C=N double
and the CÀN single bond respectively. In the IR spectrum of 9,
Chem. Eur. J. 2016, 22, 1 – 14
www.chemeurj.org
5
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
resonances for the NÀH (3278 cmÀ1), the CꢀN (2156 cmÀ1) and
the C=N unit (1603 and 1546 cmÀ1) were identified.
both reactions to be similar. In the beginning one nitrile
should be eliminated and CH3ÀCꢀN coordinated to the metal.
In case of the phenyl-substituted complex a subsequent CÀC
coupling leads to a five-membered metallacycle with a Ph and
a CH3 substituent. Such an exchange might be possible for 1a
too (DG=1.1 kcalmolÀ1), but it should result in an equilibrium
in favour of 1a. The excess of acetonitrile might influence this
and promote the generation of a complex with 2-pyÀCꢀN and
CH3ÀCꢀN coordinated to [Cp*2Ti]. Instead of the formation of
the five-membered cycle, the next step could be an H-transfer
to the coordinated 2-cyanopyridine. Further coupling and rear-
rangement steps might afford 8 and 9.
The structure of 9 was confirmed by X-ray crystallography
(Figure 3). It shows the nearly planar six-membered metallacy-
cle (mean deviation from the best plane: 0.02 ꢁ) and the three
unequal CÀN bonds. While C5ÀN3 (1.153(3) ꢁ) corresponds to
a CꢀN triple bond, C3ÀN2 (1.278(3) ꢁ) is a C=N double bond
and C1ÀN1 (1.313(3) ꢁ) a shortened CÀN single bond. The re-
maining bond lengths of the six-membered cycle indicate an
unusual bonding situation. The NÀTi bond lengths (N1ÀTi1
2.078(2), N2ÀTi1 1.887(2) ꢁ) differ significantly from each other.
2
2
The C1ÀC2 distance (1.418(3) ꢁ) is shorter than a Csp ÀCsp
single and longer than a Csp2=Csp2 double bond. Whereas, the
C2ÀC3 distance (1.442(4) ꢁ) is in the range of a Csp2ÀCsp2 single
bond. This bonding situation was in case of a tetrameric tung-
sten complex, with similar six-membered cycles, explained by
a partial M=N double bond character, which influences the
other bonds in the metallacycle as well.[25]
Reactions with other nitriles
To confirm our assumed formation path, especially the nitrile-
exchange step, we investigated the reactivity of 1a towards
other nitriles. At first we studied the conversion with PhCH2À
CꢀN, which also contains acidic a-protons. This resulted in free
2-cyanopyridine and in 8, but not in an exchange of the ni-
triles. The source of the additional H atoms might have been
PhCH2ÀCꢀN. Unfortunately, we were not able to identify prod-
ucts, which contain the remainders of the nitrile. A similar reac-
tion to 8 and free 2-pyÀCꢀN was observed upon addition of
acetone to 1a.
In addition, we examined the reactivity of 1a towards PhÀ
CꢀN, which contains no H atoms neighbouring the CꢀN
moiety and thus might hinder the formation of 8. Besides, DFT
calculations predicted that a stepwise exchange of both 2-pyÀ
CꢀN units by two PhÀCꢀN molecules is thermodynamically un-
likely (1st step: DG=4.1 kcalmolÀ1, 2nd step: DG=4.6 kcal
molÀ1). As for PhCH2ÀCꢀN, no reaction was observed at ambi-
ent temperature. But, warming a 1:1 mixture of 1a and PhÀCꢀ
N partly resulted in the formation of the mixed species 1c
(Scheme 8). The symmetric complex 1b was not established.
Upon warming a solution of 1a with an excess of PhÀCꢀN
both compounds 1c and 1b formed, but 1c only in minor
concentration and 1b in traces. This demonstrates that the ex-
change of 2-pyÀCꢀN in 1a by another nitrile is possible, al-
though it is thermodynamically not probable, and that a high
concentration or temperature is necessary to achieve it. The re-
verse reaction of 1b with 2-pyÀCꢀN to 1a, 1c and PhÀCꢀN al-
ready proceeded at ambient temperature within 2 d. In the re-
sulting mixture the main component was 1a.
Figure 3. Molecular structure of 9 in the solid state. H atoms, except for H1
are omitted for clarity. The thermal ellipsoids correspond to 30% probability.
Selected bond lengths [ꢁ] and angles [8]: C1-N1 1.313(3), C1-C2 1.418(3), C1-
C4 1.502(3), C2-C5 1.416(3), C2-C3 1.442(4), C3-N2 1.278(3), C3-C6 1.508(3),
C5-N3 1.153(3), N1-Ti1 2.078(2), N2-Ti1 1.887(2), N1-C1-C2 122.8(2), N2-C3-C2
120.2(2), C1-N1-Ti1 131.0(2), C3-N2-Ti1 140.8(2), N3-C5-C2 178.4(3).
The formation of 9 by a trimerisation of CH3ÀCꢀN is remark-
able. Coupling reactions of titanocene fragments with nitriles
to comparable six-membered titanacycles (Scheme 1, C) were
already described, but in those cases only two nitriles were in-
volved. Thus, they contained an H atom or a CH2 unit as sub-
stituent instead of the CꢀN group.[7a,26] Moreover, a trimerisa-
tion of CH3ÀCꢀN to a six-membered cycle as in 9 was solely
observed for a tungsten complex.[25] In comparable com-
pounds, formed in reactions of diverse metal species with
CH3ÀCꢀN, both N atoms of the established metallacycles were
protonated.[27]
Moreover, we were interested to determine if a nitrile ex-
change is also possible for the multinuclear analogues of 1a
and 1b. This might probably lead to mixed di-, tri- or tetranu-
clear species. As recently published, the reaction of Cp*2Ti(h2-
Me3SiC2SiMe3) with 1,3-dicyanobenzene or 2,6-dicyanopyridine
Though complexes, which are similar to 8 and 9 were al-
ready described, their formation by the reaction of 1a with
CH3ÀCꢀN is fascinating. On the one hand, it is interesting that
an excess of acetonitrile is needed for this conversion, al-
though formally a 2:3 ratio of the substrates should be suffi-
cient. On the other hand, for the analogous compound
Cp*2Ti(N=C(Ph)ÀC(Ph)=N) (1b) the addition of acetonitrile re-
sulted simply in an exchange of the nitriles and not in an H-
transfer or a trimerisation.[15a] Still, we suppose the first steps of
Scheme 8. Exchange reaction with PhÀCꢀN.
&
&
Chem. Eur. J. 2016, 22, 1 – 14
www.chemeurj.org
6
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
Scheme 9. Exchange of dicyano compounds in the macrocyclic complexes 10 and 11 under formation of 12.
afforded the tri- and tetranuclear compounds 10[16] and 11[17]
(Scheme 9).
with 1 equiv HClaq. But still, we observed the formation of
a mixture. Furthermore, we tried to increase the selectivity by
applying only 0.5 equiv of ethereal HCl. This resulted in a mix-
ture as well, in which we identified 1a, 13 and 8 among other
compounds. Though, we were not able to isolate the respec-
tive diimine, these experiments showed us, that in principle
the separation of the metallocene and the coupled nitriles
should be possible. In order to obtain a more stable diimine,
we added Me3SiCl to 1a. Unfortunately, no reaction occurred
(Scheme 10), not at ambient temperature and not upon warm-
ing to 608C for 8 d.
The reaction of 10 with 2,6-dicyanopyridine proceeded at
ambient temperature. Thereby, a 1:2 ratio was sufficient for the
complete conversion of 10. The observed products are free
1,3-dicyanobenzene and several titanocene species.
These are suspected to be mono- or dinuclear 1-titana-2,5-
diaza-cyclopenta-2,4-dienes. The addition of an excess of 2,6-
dicyanopyridine to 10 (ratio 5:1), led to only three compounds,
which are assumed to be the possible mononuclear species.
The main product in all cases is 12 (Scheme 9). This complex
was already established by the coupling of two equivalents
2,6-dicyanopyridine at Cp*2Ti(h2-Me3SiC2SiMe3).[17] In case of the
addition of 1,3-dicyanobenzene to 11 (ratio 3:1) no reaction
was observed at ambient temperature. Upon warming to 608C
for 12 h a mixture formed, with 12 as main component. In con-
trast to the reaction of 10 with 2,6-dicyanopyridine no elimina-
tion of the dicyano compound was observed and the conver-
sion was incomplete. This resembles the reactions of the mon-
onuclear species 1a and 1b with PhÀCꢀN and 2-pyÀCꢀN, re-
spectively. Unfortunately, these conversions of 10 and 11 did
not result in the isolation of mixed di-, tri- or tetranuclear com-
pounds.
Scheme 10. Attempts to generate diimines by reactions of 1a with HCl or
Me3SiCl.
The investigated reactions of 1a and 2 with H2, CO2, CH3ÀCꢀ
N and other nitriles showed several differences in comparison
to those of the phenyl-substituted analogues. In some cases
the improved stability of the pyridyl-substituted 1-metalla-2,5-
diaza-cyclopenta-2,4-dienes induced longer reaction times and
higher temperatures, in others it caused different reactions
and resulted in new metallocene compounds. For our attempts
to split off the [Cp*2M] from the coupled nitriles these findings
are very promising.
In our further attempts to separate [Cp*2M] from the cou-
pled nitriles we tried to generate five-membered heterocycles
by the reaction with PhPCl2, PhPSCl2, Me2SiCl2 or SOCl2. For
comparable compounds, like 1-metallacyclopenta-2,4-dienes or
1-metalla-2-aza-cyclopenta-2,4-dienes of group 4 metallocenes,
such conversions were already realised.[9c,12] These produced
phospholes, thiophene oxides, thiacyclopentenes or isothia-
zoles. The addition of PhPCl2, PhPSCl2 or Me2SiCl2 to 1a result-
ed in 13 and in free 2-pyÀCꢀN.[29] Formal phospholes, siloles or
products, which contain the remainders of PhPCl2 or PhPSCl2,
were not identified. The reaction with SOCl2 was tested for the
1-titana-2,5-diaza-cyclopenta-2,4-dienes Cp*2Ti(N=C(R)ÀC(R)=N)
with R=2-py (1a), R=6-CN-2-py (12) and R=2-Br-5-py (14).[30]
The outcome of these conversions was for all compounds simi-
lar. Therefore, we describe in the following mainly the reaction
of 1a with SOCl2.
Attempts to separate [Cp*2M] from the coupled nitriles
At first, we tried to generate diimines by the treatment of 1a
with HCl or Me3SiCl. Complex 1a reacted vigorously with ethe-
real HCl, resulting in Cp*2TiCl2 (13) and numerous other com-
pounds (Scheme 10). The diimine was not isolated, probably
due to decomposition as in case of its phenyl-substituted ana-
logue (in reaction of 1b with HCl).[15a,28] For directing this fur-
ther conversion to the respective diketone, we performed it
The addition of SOCl2 to 1a at À788C followed by warming
of the mixture to ambient temperature afforded the dichloride
Chem. Eur. J. 2016, 22, 1 – 14
www.chemeurj.org
7
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
13 and the thiadiazole oxide 15a (Scheme 11). The separation
of these two products was difficult. Their solubility is similar
and thus the isolated crystal fractions consisted of yellow crys-
tals of 15a and red ones of 13, which were separated manual-
ly. Due to this method, the yield of pure 15a was rather low
(40%).
oxygen in this complex would explain the necessity of an
excess of Cp*2TiCl2 for the complete conversion of 15a to 16a.
As mentioned above, the reactions of the other pyridyl-sub-
stituted 1-titana-2,5-diaza-cyclopenta-2,4-dienes with SOCl2 are
similar to that of 1a. The difference is that, the subsequent re-
actions were in both cases faster than for 15a. Thus, we were
not able to isolate 15b or 15c in pure form. But, 16c was ob-
tained as crystals (Figure 4). The values of the bond lengths of
the five-membered cycle (N1ÀS1 1.625(2), C1ÀN1 1.331(3), C1À
C1A 1.436(5) ꢁ) and the fact that it is nearly planar (mean devi-
ation from the best plane 0.007 ꢁ) indicate some electron de-
localisation, as already described for other thiadiazoles.[35] The
pyridyl substituent and the five-membered cycle are tilted by
42.77(9)8 (angle between the planes defined by S1, N1, C1,
C1A, N1A and C2, C3, N2, C4, C5, C6).
Scheme 11. Reaction with SOCl2.
Interestingly, if 13 and 15a were not separated, a subsequent
reaction started. This resulted in the thiadiazole 16a, in the
elimination of the Cp* ligand and in a formal Ti species [Cp*Ti]
with only one Cp* ligand (Scheme 12). The eliminated Cp*
ligand was identified as decamethylfulvalene (Cp*2),[31] tetrame-
thylfulvene[32] and pentamethylcyclopentadiene (Cp*H),[33] by
comparison of the typical NMR signals to those published in
the literature (below all three species are mentioned as [Cp*]
for a better readability). In order to complete this conversion
the mixture was warmed to 608C for 38 h. Afterwards, we
identified [Cp*], the formal [Cp*Ti] species, the thiadiazole 16a
and, unfortunately, still 15a. Thus, Cp*2TiCl2 (13) seems to be
necessary for the conversion of 15a to 16a. To test this as-
sumption, we dissolved pure 15a and warmed it to 858C. This
did not result in any kind of reaction. Then, we added an
excess of 13 and again warmed the mixture to 858C. Here, the
conversion of 15a to 16a started.
Figure 4. Molecular structure of 16c in the solid state. H atoms are omitted
for clarity. The thermal ellipsoids correspond to 30% probability. Selected
bond lengths [ꢁ] and angles [8]: N1-S1 1.625(2), C1-N1 1.331(3), C1-C1A
1.436(5), C1-C2 1.481(4), N1-C1-C2 119.3(2), N1-C1-C1A 112.9(1), C1-N1-S1
107.6(2), N1-S1-N1A 99.0(2).
An explanation for this might be the formation of the formal
[Cp*Ti] species. We suppose that by the reaction of 15a with
13, the Cp* ligand is eliminated and the O atom transferred to
the titanium. Subsequent to this, a complex could be formed,
which might resemble the tetrameric [(Cp*Ti)4O6].[34] Unfortu-
nately, we were not able to isolate and characterise this new
compound in detail. But, we suppose that the difference to
the literature known example could be, that it contains Cl and
O atoms as bridging ligands between the [Cp*Ti] moieties
(Scheme 12). A higher amount of chlorine in comparison to
In order to obtain 16a in pure form, we studied the reac-
tions of SOCl2 with Cp*2Ti(N(H)ÀC(2-py)=C(2-py)ÀN(H)) (8) and
Cp*2Zr(N=C(2-py)ÀC(2-py)=N) (2). In case of 8, we expected at
first the formation of a dihydrothiadiazole oxide, which could
be converted to 16a by the elimination of water. The dichlor-
ide 13 should remain intact and might be separated more
easily. But the reaction afforded an inseparable mixture of 16a,
13, [Cp*] and the formal [Cp*Ti] species. The addition of SOCl2
to the zirconocene complex 2 resulted in Cp*2ZrCl2 and in 15a.
Scheme 12. Subsequent reaction of 13 and 15 to 16, [Cp*] and the formal [Cp*Ti] species.
Chem. Eur. J. 2016, 22, 1 – 14 www.chemeurj.org
&
&
8
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
In contrast to the conversions of the titanium compound 1a,
no subsequent reaction to 16a occurred although the mixture
was warmed to 808C for 3 d.
In order to compare the reactivity towards SOCl2 between
the pyridyl- and the phenyl-substituted complexes, we investi-
gated the reaction of 1b with SOCl2 as well. This produced the
thiadiazole 16d and 13 (16d was identified by comparison of
the typical NMR data to those published in the literature[36]).
The formation of the thiadiazole oxide 15d, of the formal
[Cp*Ti] species and of [Cp*] was not observed.
of SOCl2 to different 1-titana-2,5-diazacyclopenta-2,4-dienes af-
forded at first the thiadiazole oxides 15 and the dichloride 13.
In a subsequent reaction these were converted to the thiadia-
zoles 16, [Cp*] and to a formal [Cp*Ti] species.
Experimental Section
All operations were carried out under argon with standard Schlenk
techniques or in a glovebox. The solvents were purified with the
Grubbs-type column system “Pure Solv MD-5” and dispensed into
thick-walled glass Schlenk bombs equipped with Young-type
Teflon valve stopcocks. Cp*2M(N=C(2-py)ÀC(2-py)=N) (1a: M=Ti, 2:
M=Zr),[17] Cp*2Ti(N=C(Ph)ÀC(Ph)=N) (1b),[15a] [Cp*2Ti(N=C(m-
Conclusions
C6H4)ÀC=N)]3
(10),[16]
[Cp*2Ti(N=C(m-C6H3N)ÀC=N)]4
(11),[17]
Herein, we presented a detailed study on the reactivity of 1-
metalla-2,5-diaza-cyclopenta-2,4-dienes of group 4 metallo-
cenes towards small molecules. At first, we investigated the re-
actions of the pyridyl-substituted complexes Cp*2M(N=C(2-
py)ÀC(2-py)=N) (1a: M=Ti, 2: M=Zr) with H2, CO2, CH3ÀCꢀN
and other nitriles. We observed several differences by compar-
ing the results to those of the analogous aryl-substituted com-
pounds.
Cp*2Ti(h2-Me3SiC2SiMe3)[37] and Cp*2Ti(N=C(2-Br-5-py)ÀC(2-Br-5-py)=
N) (14)[17] were synthesised according to the published procedures.
2-Cyanopyridine, PhCH2ÀCꢀN, benzonitrile and Me2SiCl2 (all Sigma–
Aldrich) were purified by distillation and stored under argon. 2,6-
Dicyanopyridine (Sigma–Aldrich) was purified by sublimation. The
solid 1,3-dicyanobenzene (Sigma–Aldrich) was dried in vacuum
prior to use. Ethereal HCl (HCl in diethyl ether, 2.0m, Sigma–Al-
drich), Me3SiCl (Sigma–Aldrich), PhPCl2 (Acros) and PhPSCl2 (TCI)
were used as received. SOCl2 was purified according to the pub-
lished procedure.[38] The following instruments were used: NMR
spectra: Bruker AV300 and AV400. 1H and 13C chemical shifts are
given in ppm and were referenced to the solvent signals:
[D6]benzene (dH =7.16 ppm, dC =128.0 ppm), CDCl3 (dH =7.26 ppm,
dC =77.2 ppm), [D6]DMSO (dH =3.33 ppm, dC =39.5 ppm), [D8]THF
(dH =3.58 ppm, dC =67.6 ppm); IR: Bruker Alpha FT-IR spectrome-
ter; MS: Finnigan MAT 95-XP from Thermo-Electron; elemental
analysis: Leco Tru Spec elemental analyzer; melting points: Met-
tler-Toledo MP70. Melting points are uncorrected and were mea-
sured in sealed capillaries.
In some cases the improved stability of the pyridyl-substitut-
ed 1-metalla-2,5-diaza-cyclopenta-2,4-dienes effected longer re-
action times and higher temperatures. This was observed for
the conversion of 1a with H2 and CO2 as well as of 2 with H2,
in which the obtained products were similar to those of the
aryl-substituted analogues. In other cases different reactions
were caused by the higher stability and resulted in new metal-
locene species. The reaction of 2 with CO2 afforded a fascinat-
ing coordination product, while the conversion of the phenyl-
substituted analogue resulted in a mixture. The special feature
of this reaction of 2 is that the five-membered metallacycle
stayed intact. The higher stability of the pyridyl-substituted
cycle also induced a different reaction with CH3ÀCꢀN. In case
of the phenyl-substituted complexes the addition of acetoni-
trile resulted in an exchange of the nitriles. In the reaction with
1a, the 1,4-diaza-buta-1,3-diene complex 8 was established by
an H-transfer and a six-membered cycle in 9 by a remarkable
coupling of three former CH3ÀCꢀN moieties. Similar H-transfers
were observed for PhCH2ÀCꢀN and acetone, which generated
8 as well.
The exchange of the nitriles was also accomplished for 1a,
but only upon the addition of PhÀCꢀN, which lacks acidic pro-
tons neighbouring the CꢀN moiety. Besides, an excess of PhÀ
CꢀN and elevated temperatures were necessary for this reac-
tion. The same was observed upon addition of 1,3-dicyanoben-
zene or 2,6-dicyanopyridine to the trinuclear compound 10 or
the tetranuclear complex 11, respectively. An explanation for
this seems again to be the higher stability of the pyridyl-substi-
tuted 1-metalla-2,5-diaza-cyclopenta-2,4-dienes in comparison
to their phenyl-substituted analogues.
Diffraction data for 7, 8 and 16c were collected on a Bruker Kappa
APEX II Duo diffractometer using MoKa radiation for 7 and 8 and
CuKa radiation for 16c. For 9 the diffraction data were collected on
a STOE-IPDS II diffractometer using MoKa radiation. The structures
were solved by direct methods and refined by full-matrix least-
squares procedures on F2 with the SHELXTL software package.[39]
Diamond was used for graphical representations.[40]
Crystal data for 7: C33H38N4O2Zr, M=613.89, monoclinic, space
group C2/c, a=28.5767(7), b=12.9297(3), c=16.4976(4) ꢁ, b=
109.2273(7)8, V=5755.6(2) ꢁ3, T=150(2) K, Z=8, 60815 reflections
collected, 6952 independent reflections (Rint =0.0370), final
R
values (I>2s(I)): R1 =0.0297, wR2 =0.0710, final R values (all data):
R1 =0.0395, wR2 =0.0774, 371 parameters.
Crystal data for 8: C32H40N4Ti, Mr =528.58, monoclinic, space
group C2/c, a=19.3954(9), b=16.1976(7), c=18.8007(9) ꢁ, b=
110.995(1)8, V=5514.3(4) ꢁ3, T=150(2) K, Z=8, 76393 reflections
collected, 6672 independent reflections (Rint =0.0236), final
R
values (I>2s(I)): R1 =0.0321, wR2 =0.0865, final R values (all data):
R1 =0.0363, wR2 =0.0911, 352 parameters.
Crystal data for 9: C26H37N3Ti, Mr =439.48, orthorhombic, space
group P212121, a=9.9475(3), b=13.8264(3), c=17.3814(4) ꢁ, V=
2390.6(1) ꢁ3, T=150(2) K, Z=4, 42072 reflections collected, 5762
independent reflections (Rint =0.0547), final R values (I>2s(I)): R1 =
0.0299, wR2 =0.0662, final R values (all data): R1 =0.0365, wR2 =
0.0675, 287 parameters.
In light of these results, we examined in the second part of
this work reactions to separate the [Cp*2M] from the coupled
nitriles. Our first experiments, using HCl, Me3SiCl, PhPCl2,
PhPSCl2 and Me2SiCl2 and the titanium complex 1a were un-
successful. If a reaction occurred, it resulted in a mixture or in
the formation of Cp*2TiCl2 and the free nitrile. But, the addition
Crystal data for 16c: C12H6Br2N4S, Mr =398.09, monoclinic, space
group C2/c, a=14.8696(5), b=7.7513(3), c=13.3462(4) ꢁ, b=
122.7486(14)8, V=1293.76(8) ꢁ3, T=150(2) K, Z=4, 4364 reflections
collected, 1142 independent reflections (Rint =0.0197), final R values
Chem. Eur. J. 2016, 22, 1 – 14
www.chemeurj.org
9
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
(I>2s(I)): R1 =0.0261, wR2 =0.0674, final R values (all data): R1 =
0.0270, wR2 =0.0681, 87 parameters.
Reactions with CO2
Synthesis of 6: Cp*2Ti(N=C(2-py)ÀC(2-py)=N) (1a; 0.159 g,
0.3 mmol) was given into a three-necked-flask equipped with
a pressure release valve and dissolved in toluene under stirring.
CO2 (using solid CO2) was passed through a P4O10 filled column
and then bubbled through the solution of 1a for at least 30 min
to saturate the solution and the atmosphere. Subsequently, the
flask was closed and warmed to 508C for 4 d. All volatiles were re-
moved in vacuum. The remaining brown solid was dissolved in n-
hexane. Upon cooling to À788C a brown precipitate formed,
which was separated by decanting and dried in vacuum to give
pure 6. Yield: 0.124 g (0.26 mmol, 88%); m.p.: 1868C (dec. under
CCDC 1468162 (7), 1468163 (8), 1468164 (9) and 1468165 (16c)
contain the supplementary crystallographic data for this paper.
These data are provided free of charge by The Cambridge Crystal-
lographic Data Centre.
Computational details
Structure optimisations were carried out at the BP86[41] density
functional level of theory with the TZVP basis set[42] for C, H, N O
as well as LANL2DZ[43] for Ti and Zr by using Gaussian 09 program
package.[44] The optimised geometries are characterised as energy
minimums at the potential energy surface from frequency calcula-
tions at the same level of theory (BP86/TZVP), that is, the energy
minimum structure has only real frequencies, while the transition-
state structure has only one imaginary frequency. Furthermore, the
thermal corrections to enthalpy and Gibbs free energy at 298 K
from the frequency analysis are added to the total electronic
energy. Therefore, we either used the corrected enthalpy (DH) or
the Gibbs free energy (DG) at 298 K for our energetic comparisons.
The energetic data and the Cartesian coordinates are given in the
Supporting Information.
1
Ar); H NMR (300 MHz, 297 K, C6D6): d=1.73 (s, 30H, C5Me5), 6.64
(m, 1H, py), 7.26 (m, 1H, py), 8.59 (m, 1H, py), 8.71 ppm (m, 1H,
py); 13C NMR (75 MHz, 297 K, C6D6): d=11.6 (C5Me5), 124.1 (C5Me5),
123.0, 124.1, 135.8, 149.9 (Cpy), 154.9, 157.5 (ipsoCpy or C=N),
165.7 ppm (C=O); IR (ATR): n˜ =2905 (w), 1647 (vs, C=O), 1578 (m,
C=N), 1377 (s), 743 cmÀ1 (s); MS: m/z (CI): 467 (100) [M+H]+, 318
(9) [Cp*2Ti]+, 137 (49) [Cp*+2H]+, 105 (68) [(2-pyÀCꢀN)+H]+; ele-
mental analysis calcd (%) for C27H34N2O2Ti (466.45 gmolÀ1): C 69.52,
H 7.35, N 6.01; found: C 69.43, H 7.27, N 6.33.
Synthesis of 7: A toluene solution of Cp*2Zr(N=C(2-py)ÀC(2-py)=N)
(2; 0.285 g, 0.5 mmol) was transferred into a three-necked-flask
equipped with a pressure release valve. CO2 (using solid CO2) was
passed through a P4O10 filled column and then bubbled through
the solution. The colour changed immediately from yellow to
brown. All volatiles were removed in vacuum and the brown resi-
due was dissolved in a toluene/n-hexane mixture (3:2). The solu-
tion was allowed to stand at À788C and a brown solid formed. It
was isolated by decanting and dried in vacuum to give pure 7.
Crystals suitable for an X-ray crystal structure analysis were ob-
tained by cooling a warm toluene/n-hexane mixture (3:2, 508C) of
7 slowly to ambient temperature. Yield: 0.245 g (0.40 mmol, 80%);
m.p.: 215–2188C (dec. under Ar); 1H NMR (300 MHz, 297 K, C6D6):
d=1.81 (s, 30H, C5Me5), 6.51 (m, 1H, py), 6.54 (m, 1H, py), 7.19 (m,
1H, py), 7.22 (m, 1H, py), 7.87 (m, 1H, py), 8.00 (m, 1H, py), 8.12
(m, 1H, py), 8.45 ppm (m, 1H, py); 13C NMR (75 MHz, 297 K, C6D6):
d=10.9 (C5Me5), 118.9 (C5Me5), 122.0, 122.1, 124.4, 127.1, 135.5,
135.9, 147.2, 148.1 (Cpy), 152.1, 152.4 (ipsoCpy), 157.5, 161.2,
165.4 ppm (C=N or C=O); IR (ATR): n˜ =2905 (w), 1704 (vs, C=O),
1665 (m, C=N), 1581 (m, C=N), 1222 (s, C-O), 737 cmÀ1 (s); MS: m/z
(CI): 596 (2) [MÀO]+, 136 (100) [Cp*+H]+, 105 (39) [(2-pyÀCꢀ
N)+H]+; elemental analysis calcd (%) for C33H38N4O2Zr
(613.92 gmolÀ1): C 64.56, H 6.24, N 9.13; found: C 64.32, H 6.15, N
9.57.
Reactions with H2
Preparation of 3: Cp*2Ti(N=C(2-py)ÀC(2-py)=N) (1a) was dissolved
in C6D6 and the red solution was transferred into a sealable J-
Young NMR tube. The sample was degassed using a freeze-pump-
thaw method and filled with 1 atm of H2 at À1968C. Subsequently,
it was first warmed to ambient temperature, then heated to 508C
1
for 1 d. H and 13C NMR spectra were recorded before and after the
addition of H2 and after the heating. Compound 3 could not be
isolated and characterised in detail because of its incomplete for-
1
mation and its subsequent reaction upon further warming. H NMR
(400 MHz, 297 K, C6D6): d=1.80 (s, 30H, C5Me5), 6.64 (m, 1H, py),
7.12 (m, 1H, py), 8.24 (m, 1H, py), 8.55 (m, 1H, py), 9.22 ppm (s,
1H, N=CH). The signal of the TiÀH moiety could not be identified;
13C NMR (100 MHz, 297 K, C6D6): d=12.2 (C5Me5), 114.0 (C5Me5),
117.6, 119.3, 123.4, 135.6 (Cpy), 149.8 (ipsoCpy), 155.4 ppm (N=CH).
Preparation of 4 and 5: Cp*2Zr(N=C(2-py)ÀC(2-py)=N) (2) was dis-
solved in C6D6 and the yellow solution was transferred into a seala-
ble J-Young NMR tube. The sample was degassed using a freeze-
pump-thaw method and filled with 1 atm of H2 at À1968C. Subse-
quently, it was first warmed to ambient temperature, then heated
to 508C. After 1 d of warming, 4 started to form and after 4 d com-
pound 5 was established. Both complexes could not be isolated
and characterised in detail because of their incomplete formation.
Analytical data of 4: 1H NMR (400 MHz, 297 K, C6D6): d=1.87 (s,
30H, C5Me5), 4.80 (s, 1H, Zr-H), 6.54 (m, 1H, py), 7.08 (m, 1H, py),
7.66 (m, 1H, py), 8.72 (m, 1H, py), 9.78 ppm (s, 1H, N=CH);
13C NMR (100 MHz, 297 K, C6D6): d=11.8 (C5Me5), 115.1 (C5Me5),
119.2, 123.1, 125.9, 132.5 (Cpy), 155.7 (ipsoCpy), 160.1 ppm (N=CH).
Analytical data of 5: 1H NMR (300 MHz, 297 K, C6D6): d=1.72 (s,
30H, C5Me5), 6.84 (m, 2H, py), 7.30 (m, 2H, py), 7.94 (m, 2H, py),
8.64 (m, 2H, py), 9.96 ppm (s, 2H, N=CH); 13C NMR (75 MHz, 297 K,
C6D6): d=11.4 (C5Me5), 117.6 (C5Me5), 116.7, 123.5, 137.3 (Cpy), 149.8
(ipsoCpy), 160.4 ppm (N=CH). The forth Cpy signal could not be iden-
tified.
Reactions with CH3ÀCꢀN
Synthesis of 8 and 9: Cp*2Ti(N=C(2-py)ÀC(2-py)=N) (1a; 0.284 g,
0.5 mmol) was dissolved in CH3ÀCꢀN (10 mL) and the solution was
stirred for 24 h. The colour changed from red to brown. All vola-
tiles were removed in vacuum and the remaining brown solid
washed with n-hexane to give at first a dark-violet and then
a brown fraction. The remaining yellow residue was dried in
vacuum to give 9 in high purity. Cooling the dark-violet solution to
À788C resulted in a brown solid, which was separated by decant-
ing. The dark-violet filtrate was evaporated to dryness to give pure
8. For a higher yield of both products the brown solid and the
brown solution were worked up further. Crystals of 8 suitable for
X-ray analysis were obtained by cooling a saturated n-hexane solu-
tion of 8 to À788C. Crystals of 9 suitable for X-ray analysis formed
at ambient temperature by slow evaporation of a C6D6 solution.
&
&
Chem. Eur. J. 2016, 22, 1 – 14
www.chemeurj.org
10
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
1
Analytical data of 8: Yield: 0.082 g (0.16 mmol, 29%); m.p.: 1938C
(dec. under Ar); 1H NMR (300 MHz, 297 K, C6D6): d=1.79 (s, 30H,
C5Me5), 6.47 (m, 2H, py), 6.85 (m, 2H, py), 7.36 (m, 2H, py), 7.99 (s,
2H, N-H), 8.57 ppm (m, 2H, py); 13C NMR (75 MHz, 297 K, C6D6): d=
12.0 (C5Me5), 117.8 (C5Me5), 119.7, 123.8, 134.4, 149.5 (Cpy),
144.3 ppm (ipsoCpy or C=NH); IR (ATR): n˜ =3345 (vw, N-H), 2894 (w),
1578 (s, C=C), 1442 (s, C-N), 780 cmÀ1 (vs); MS: m/z (CI): 528 (100)
[M]+, 318 (4) [Cp*2Ti]+, 137 (96) [Cp*+2H]+, 105 [(2-pyÀCꢀN)+H]+;
elemental analysis calcd (%) for C32H40N4Ti (528.57 gmolÀ1): C 72.72,
H 7.63, N 10.60; found: C 72.56, H 7.66, N 10.63.
Analytical data of the formal [Cp*Ti] species: H NMR (300 MHz,
297 K, C6D6): d=2.08 (C5Me5); 13C NMR (75 MHz, 297 K, C6D6): d=
13.3 (C5Me5), 134.0 ppm (C5Me5).
Preparation of 15b and 16b: Cp*2Ti(N=C(6-CN-2-py)ÀC(6-CN-2-
py)=N) (12) was synthesised in situ from Cp*2Ti(h2-Me3SiC2SiMe3)
(0.488 g, 1.0 mmol) and 2,6-dicyanopyridine (0.285 g, 2.2 mmol) as
described in the published procedure.[17] It was dissolved in tolu-
ene and the solution was cooled to À788C. A 0.2m toluene solu-
tion of SOCl2 (5.0 mL, 1.0 mmol) was added. This resulted in the
formation of a red solid. After warming to ambient temperature
the red solution was filtered off the now brownish solid. The resi-
due was washed twice with cold toluene and dried in vacuum.
NMR spectra revealed that it consisted of 13 and 15b. (Attempts
to separate this mixture were unsuccessful due to similar solubility
and the subsequent reaction.) The brownish solid was dissolved in
toluene and the brown solution was warmed to 608C for 7 d. NMR
spectra were measured before and after the heating. They revealed
the conversion of 13 and 15b to 16b, decamethylfulvalene (Cp*2),
tetramethylfulvene and the formal [Cp*Ti] species.
Analytical data of 9: Yield: approximately 0.069 g (0.016 mmol,
1
29%; contaminated by unidentified impurities). H NMR (400 MHz,
4
297 K, C6D6): d=1.57 (s, 30H, C5Me5), 2.01 (d, J(1H-1H)=0.9 Hz, 3H,
4-Me), 2.19 (s, 3H, 6-Me), 4.27 ppm (s, 1H, N-H); 13C NMR (100 MHz,
297 K, C6D6): d=11.5 (C5Me5), 24.2 (6-Me), 27.8 (4-Me), 78.0 (C2),
120.3 (C5Me5), 124.6 (C5), 158.8 (C3), 165.1 ppm (C1) (Numbering is
consistent with molecular structure Figure 3); IR (ATR): n˜ =3278
(vw, N-H), 2905 (w), 2156 (m, CꢀN), 1603 (m, C=N), 1546 (w, C=N),
417 cmÀ1 (vs); MS: m/z (CI): 440 (83) [M+H]+, 318 (2) [Cp*2Ti]+, 137
(100) [Cp*+2H]+, 123 (3) [(N=C(CH3)ÀC(CN)=C(CH3)ÀN(H)) +2H]+.
Analytical data of 15b: 1H NMR (300 MHz, 297 K, C6D6): d=6.47
(m, 2H, py), 6.58 (m, 2H, py), 7.22 ppm (m, 2H, py); 13C NMR
(75 MHz, 297 K, C6D6): d=116.3 (CꢀN), 127.2, 129.4, 137.8 (Cpy),
132.3 (ipsoCpyÀCꢀN), 150.4 (ipsoCpy), 164.9 ppm (C=N); MS: m/z (CI):
307 (49) [M+H]+.
Reactions with SOCl2
Analytical data of 16b: 1H NMR (300 MHz, 297 K, C6D6): d=6.53
(m, 2H, py), 6.65 (m, 2H, py), 7.40 ppm (m, 2H, py); 13C NMR
(75 MHz, 297 K, C6D6): d=117.0 (CꢀN), 126.5, 129.3, 137.3 (Cpy),
132.8 (ipsoCpyÀCꢀN), 152.9, 158.7 ppm (ipsoCpy or C=N). MS: m/z (CI):
291 (100) [M+H]+.
Preparation of 15a: A toluene solution of Cp*2Ti(N=C(2-py)ÀC(2-
py)=N) (1a; 0.263 g, 0.5 mmol) was cooled to À788C and then
a 0.2m toluene solution of SOCl2 (2.5 mL, 0.5 mmol) was added.
Subsequently, the mixture was warmed to ambient temperature to
complete the conversion to 13 and 15a. By cooling the mixture to
À788C a red precipitate of 13 formed, which was separated by de-
canting. The filtrate was concentrated. After the addition of n-
hexane to it, the solution was cooled to À788C. Yellow and red
crystals formed, which were separated manually. The yellow crys-
tals were identified as 15a, the red ones as 13.
Preparation of 15c and 16c: A toluene solution of Cp*2Ti(N=C(2-
Br-5-py)ÀC(2-Br-5-py)=N) (14; 0.360 g, 0.5 mmol) was cooled to
À788C and then a 0.2m toluene solution of SOCl2 (2.5 mL,
0.5 mmol) was added. After warming to ambient temperature the
mixture consisted of 13, 15c and 16c. (Attempts to separate 15c
from these compounds were unsuccessful due to similar solubility.)
After 2 weeks at ambient temperature the mixture contained 13,
15c (minor concentration), 16c, the formal [Cp*Ti] species and
[Cp*]. Crystals of 16c formed by cooling a THF solution of this mix-
ture to À788C.
Analytical data of 15a: Yield: 0.050 g (0.20 mmol, 40%); m.p.:
1
1058C (dec. under Ar); H NMR (300 MHz, 297 K, C6D6): d=6.50 (m,
2H, py), 6.93 (m, 2H, py), 7.53 (m, 2H, py), 7.96 ppm (s, 2H, py);
13C NMR (75 MHz, 297 K, C6D6): d=124.7, 125.1, 136.4, 148.4 (Cpy),
150.8, 167.6 ppm (ipsoCpy or C=N); IR (ATR): n˜ =1581 (m, C=N),
1432 (w, C=N), 1123 (s), 540 cmÀ1 (vs); MS: m/z (CI): 257 (100)
[M+H]+, 241 (87) [MÀO]+, 105 (14) [(2-pyÀCꢀN)+H]+.
Conversion of 15a to 16a: A mixture of 15a and 13 was dissolved
in C6D6 and given into a sealable J-Young NMR tube and allowed
to stand at ambient temperature for 3 d. Subsequently, it was
warmed to 608C for 38 h. NMR spectra were measured after 3 d at
ambient temperature, after 2, 14, 26 and 38 h at 608C. They re-
vealed the conversion of 15a and 13 to 16a, decamethylfulvalene
(Cp*2), tetramethylfulvene, pentamethylcyclopentadiene (Cp*H)
and the formal [Cp*Ti] species.
Analytical data for 15c: 1H NMR (300 MHz, 297 K, C6D6): d=6.53
(m, 2H, py), 6.61 (m, 2H, py), 7.90 ppm (m, 2H, py); 1H NMR
(400 MHz, 300 K, [D8]THF): d=7.70 (m, 2H, py), 7.82 (m, 2H, py),
8.60 ppm (m, 2H, py); 13C NMR (75 MHz, 297 K, C6D6): d=124.1,
139.3, 150.4 (Cpy), 137.3, 155.7 (ipsoCpy), 164.4 ppm (C=N); 13C NMR
(100 MHz, 300 K, [D8]THF): d=128.7, 140.4, 151.7 ppm (Cpy). The
signals of the quaternary C atoms could not be detected. MS: m/z
(ESI-TOF): 436.9 [M+Na]+.
Analytical data for 16c: 1H NMR (300 MHz, 297 K, C6D6): d=6.75
(m, 4H, py), 8.24 (m, 2H, py); 1H NMR (300 MHz, 300 K, [D8]THF):
d=7.63 (m, 2H, py), 7.77 (m, 2H, py), 8.54 ppm (m, 2H, py);
13C NMR (75 MHz, 297 K, C6D6): d=127.9, 137.8, 150.0 (Cpy), 149.7,
155.8 ppm (ipsoCpy or C=N); 13C NMR (100 MHz, 300 K, [D8]THF): d=
128.8, 139.3, 150.7 ppm (Cpy). The remaining signals could not be
detected in neither solvent. MS: m/z (ESI-TOF): 398.9 [M+H]+; ele-
mental analysis calcd (%) for C12H6N4Br2S (398.08 gmolÀ1): C 36.21,
H 1.52, N 14.07, S 8.05; found: C 36.24, H 1.40, N 13.94, S 8.23.
Analytical data of 16a: 1H NMR (300 MHz, 297 K, C6D6): d=6.59
(m, 2H, py), 7.09 (m, 2H, py), 7.70 (m, 2H, py), 8.25 ppm (m, 2H,
py); 13C NMR (75 MHz, 297 K, C6D6): d=123.3, 124.1, 136.1, 148.9
(Cpy), 153.3, 161.0 ppm (ipsoCpy or C=N).
1
Analytical data of Cp*2: H NMR (300 MHz, 297 K, C6D6): d=1.15 (s,
6H, 1/1’-Me), 1.67 (s, 12H, Me), 1.76 ppm (s, 12H, Me); 13C NMR
(75 MHz, 297 K, C6D6): d=11.1 (Me), 19.7 (Me), 60.2 ppm (C1/C1’).
The remaining signals were not unambiguously identified.
Preparation of 16d: Cp*2Ti(N=C(Ph)ÀC(Ph)=N) (1b) was synthes-
ised in situ from Cp*2Ti(h2-Me3SiC2SiMe3) (0.244 g, 0.5 mmol) and
PhÀCꢀN (0.10 mL, 1.0 mmol) as described in the published proce-
dure.[15a] Then, the toluene solution was cooled to À788C and
a 0.2m toluene solution of SOCl2 (2.5 mL, 0.5 mmol) was added.
The mixture was warmed to ambient temperature and all volatiles
Analytical data of tetramethylfulvene: 1H NMR (300 MHz, 297 K,
C6D6): d=1.69 (s, 6H, Me), 1.85 (s, 6H, Me), 5.33 ppm (s, 2H, CH2).
1
Analytical data of Cp*H: H NMR (300 MHz, 297 K, C6D6): d=0.99
(d, 3J=7.8 Hz, 3H, 1-Me), 1.74 (m, 6H, Me), 1.80 (s, 6H, Me),
2.42 ppm (m, 1H, Cp-H).
Chem. Eur. J. 2016, 22, 1 – 14
www.chemeurj.org
11
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
11611; b) L. Becker, F. Strehler, M. Korb, P. Arndt, A. Spannenberg, W.
Baumann, H. Lang, U. Rosenthal, Chem. Eur. J. 2014, 20, 3061–3068.
[16] L. Becker, P. Arndt, A. Spannenberg, H. Jiao, U. Rosenthal, Angew. Chem.
Int. Ed. 2015, 54, 5523–5526; Angew. Chem. 2015, 127, 5614–5617.
[17] G. Tomaschun, K. Altenburger, F. Reiß, L. Becker, A. Spannenberg, P.
Arndt, H. Jiao, U. Rosenthal, Eur. J. Inorg. Chem. 2016, 272–280.
[18] F. Reiß, K. Altenburger, L. Becker, K. Schubert, H. Jiao, A. Spannenberg,
D. Hollmann, P. Arndt, U. Rosenthal, Chem. Eur. J. 2016, 22, 3361–3369.
[19] Selected examples and citations therein: a) K. A. Kreutzer, R. A. Fisher,
W. M. Davis, E. Spaltenstein, S. L. Buchwald, Organometallics 1991, 10,
4031–4035; b) G. A. Luinstra, U. Rief, M. H. Prosenc, Organometallics
1995, 14, 1551–1552; c) U. Blaschke, F. Menges, G. Erker, R. Frçhlich,
Eur. J. Inorg. Chem. 1999, 621–632.
were removed in vacuum. The remaining solid, consisting of 13
and 16d was separated by column chromatography (cyclohexane/
ethyl acetate 8:1). Yield: 0.049 g (0.197 mmol, 39%); 1H NMR
(300 MHz, 297 K, C6D6): d=6.82 (m, 4H, Ph), 6.97 (m, 2H, Ph),
7.11 ppm (m, 4H, Ph); 13C NMR (75 MHz, 297 K, C6D6): d=128.3,
130.3, 132.0 (CPh), 130.9 (ipsoCPh), 168.3 ppm (C=N). (For the com-
parison of the NMR data with the literature known the 1H NMR
spectrum was also measured in CDCl3
and the 13C NMR in
[36a]
[D6]DMSO[36b]). MS: m/z (CI): 239 (100) [M+H]+.
Acknowledgements
[20] V: M. Vivanco, J. Ruiz, C. Floriani, A. Chiesi-Villa, C. Guastini, Organome-
tallics 1990, 9, 2185–2187; Fe: A. Klose, E. Solari, C. Floriani, A. Chiesi-
Villa, C. Rizzoli, N. Re, J. Am. Chem. Soc. 1994, 116, 9123–9135.
[21] A. Mendiratta, C. C. Cummins, Inorg. Chem. 2005, 44, 7319–7321.
[22] P. Binger, F. Sandmeyer, C. Krueger, Organometallics 1995, 14, 2969–
2976.
We thank our technical and analytical staff for assistance. Fi-
nancial support by the DFG (RO 1269/9-1) is gratefully ac-
knowledged.
[23] Selected publications and citations therein: a) L. R. Chamberlain, L. D.
Durfee, P. E. Fanwick, L. M. Kobriger, S. L. Latesky, A. K. McMullen, B. D.
Steffey, I. P. Rothwell, K. Foltin, J. C. Huffman, J. Am. Chem. Soc. 1987,
109, 6068–6076; b) M. G. Thorn, P. E. Fanwick, I. P. Rothwell, Organome-
tallics 1999, 18, 4442–4447; c) C. Ramos, P. Royo, M. Lanfranchi, M. A.
Pellinghelli, A. Tiripicchio, Eur. J. Inorg. Chem. 2005, 3962–3970; d) R.
Fernꢃndez-Galꢃn, A. AntiÇolo, F. Carrillo-Hermosilla, I. Lꢄpez-Solera, A.
Otero, A. Serrano-Laguna, E. VillaseÇor, Organometallics 2012, 31, 8360–
8369.
[24] Selected publications and citations therein: a) T. Zippel, P. Arndt, A.
Ohff, A. Spannenberg, R. Kempe, U. Rosenthal, Organometallics 1998,
17, 4429–4437; b) J. Scholz, G. A. Hadi, K.-H. Thiele, H. Gçrls, R. Wei-
mann, H. Schumann, J. Sieler, J. Organomet. Chem. 2001, 626, 243–259;
c) A. C. Benjamin, A. S. P. Frey, M. G. Gardiner, C. L. Raston, B. W. Skelton,
A. H. White, J. Organomet. Chem. 2008, 693, 776–780; d) S. Anga, K.
Naktode, H. Adimulam, T. K. Panda, Dalton Trans. 2014, 43, 14876–
14888.
Keywords: exchange reaction · metallacycles · metallocenes ·
reactivity · small molecules
[1] a) T. Beweries, M. Haehnel, U. Rosenthal, Catal. Sci. Technol. 2013, 3, 18–
28; b) T. Beweries, U. Rosenthal, Nat. Chem. 2013, 5, 649–650.
[2] a) L. Becker, V. V. Burlakov, P. Arndt, A. Spannenberg, W. Baumann, H.
Jiao, U. Rosenthal, Chem. Eur. J. 2013, 19, 4230–4237; b) L. Becker, P.
Arndt, A. Spannenberg, U. Rosenthal, Chem. Eur. J. 2014, 20, 12595–
12600; c) L. Becker, M. Haehnel, A. Spannenberg, P. Arndt, U. Rosenthal,
Chem. Eur. J. 2015, 21, 3242–3248.
[3] a) P. T. Wolczanski, Ph.D. Thesis, California Institute of Technology, 1981;
b) J. E. Bercaw, D. L. Davies, P. T. Wolczanski, Organometallics 1986, 5,
443–450.
[4] Selected examples: a) E. J. M. de Boer, J. H. Teuben, J. Organomet. Chem.
1977, 140, 41–45; b) J. Streuff, M. Feurer, G. Frey, A. Steffani, S. Kacpr-
zak, J. Weweler, L. H. Leijendekker, D. Kratzert, D. A. Plattner, J. Am.
Chem. Soc. 2015, 137, 14396–14405.
[5] K. M. Doxsee, J. B. Farahi, H. Hope, J. Am. Chem. Soc. 1991, 113, 8889–
8898.
[6] S. Zhou, D. Liu, Y. Liu, Organometallics 2004, 23, 5900–5902.
[7] Selected examples: a) R. Beckhaus, I. Strauß, T. Wagner, Angew. Chem.
Int. Ed. Engl. 1995, 34, 688–690; Angew. Chem. 1995, 107, 738–740;
b) R. Choukroun, C. Lorber, L. Vendier, Organometallics 2007, 26, 3784–
3790; c) J. Zhao, S. Zhang, W.-X. Zhang, Z. Xi, Coord. Chem. Rev. 2014,
270–271, 2–13.
[8] a) G. Erker, D. Kowalski, R. Noe, C. Krꢂger, M. Nolte, Tetrahedron Lett.
1994, 35, 6665–6668; b) D. Wingbermꢂhle, C. Sarter, G. Erker, Inorg.
Chim. Acta 1994, 222, 193–198.
[9] a) Y. Yamamoto, D. Matsumi, R. Hattori, K. Itoh, J. Org. Chem. 1999, 64,
3224–3229; b) N. Suzuki, T. Shimura, Y. Sakaguchi, Y. Masuyama, Pure
Appl. Chem. 2011, 83, 1781–1788; c) S. L. Buchwald, B. T. Watson, R. T.
Lum, W. A. Nugent, J. Am. Chem. Soc. 1987, 109, 7137–7141; d) T. Taka-
hashi, C. Xi, Z. Xi, M. Kageyama, R. Fischer, K. Nakajima, E.-i. Negishi, J.
Org. Chem. 1998, 63, 6802–6806.
[25] T. M. Cameron, A. S. Gamble, K. A. Abboud, J. M. Boncella, Chem.
Commun. 2002, 1148–1149.
[26] a) K. M. Doxsee, J. B. Farahi, J. Am. Chem. Soc. 1988, 110, 7239–7240;
b) N. A. Petasis, D. K. Fu, Organometallics 1993, 12, 3776–3780.
[27] a) Al: M. R. Kopp, B. Neumꢂller, Z. Anorg. Allg. Chem. 1999, 625, 739–
745; b) Ga: M. R. Kopp, T. Krꢅuter, A. Dashti-Mommertz, B. Neumꢂller,
Organometallics 1998, 17, 4226–4231; c) In: M. R. Kopp, B. Neumꢂller,
Z. Anorg. Allg. Chem. 1999, 625, 1246–1248; d) Co: M. L. Durꢃn, J. A.
Garcꢆa-Vꢃzquez, C. Gꢄmez, A. Sousa-Pedrares, J. Romero, A. Sousa, Eur.
J. Inorg. Chem. 2002, 2348–2354; e) Ru: N. A. Foley, T. B. Gunnoe, T. R.
Cundari, P. D. Boyle, J. L. Petersen, Angew. Chem. Int. Ed. 2008, 47, 726–
730; Angew. Chem. 2008, 120, 738–742; f) Ir: J. Yuan, R. P. Hughes, A. L.
Rheingold, Organometallics 2009, 28, 4646–4648; g) Fe: J. Nishijo, N.
Nishi, Eur. J. Inorg. Chem. 2006, 3022–3027; h) U: M. A. Antunes, M.
Dias, B. Monteiro, A. Domingos, I. C. Santos, N. Marques, Dalton Trans.
2006, 3368–3374; i) Zn: C. Redshaw, S. Jana, C. Shang, M. R. J. Elsegood,
X. Lu, Z. X. Guo, Chem. Commun. 2010, 46, 9055–9057.
[28] P. Kang, C. S. Foote, J. Am. Chem. Soc. 2002, 124, 9629–9638.
[29] The decomposition of 1a into the dichloride 13, the free nitrile and
other unidentified compounds also occurred upon dissolving it in
CDCl3. This explains why chlorinated solvents were not used in this re-
activity study.
[10] S. Yu, M. Xiong, X. Xie, Y. Liu, Angew. Chem. Int. Ed. 2014, 53, 11596–
11599; Angew. Chem. 2014, 126, 11780–11783.
[30] We also studied the conversion of the tetranuclear complex 11 with
SOCl2, but owing to the poor solubility of the obtained products no
precise analytical data were obtained.
[31] J. P. Falkenhagen, C. Limberg, S. Demeshko, S. Horn, M. Haumann, B.
Braun, S. Mebs, Dalton Trans. 2014, 43, 806–816.
[11] S. Yu, R. Sun, H. Chen, X. Xie, Y. Liu, Chem. Eur. J. 2015, 21, 1420–1424.
[12] a) P. J. Fagan, W. A. Nugent, J. Am. Chem. Soc. 1988, 110, 2310–2312;
b) P. J. Fagan, W. A. Nugent, J. C. Calabrese, J. Am. Chem. Soc. 1994, 116,
1880–1889.
[13] a) N. Avarvari, P. Le Floch, F. Mathey, J. Am. Chem. Soc. 1996, 118,
11978–11979; b) A. Maraval, A. Igau, B. Donnadieu, J.-P. Majoral, Eur. J.
Org. Chem. 2003, 385–394.
[32] N. A. Pushkarevsky, S. N. Konchenko, A. V. Virovets, M. Scheer, Organo-
metallics 2013, 32, 770–779.
ˇ
[14] Selected examples: a) T. Takahashi, F.-Y. Tsai, M. Kotora, J. Am. Chem. Soc.
2000, 122, 4994–4995; b) T. Takahashi, F.-Y. Tsai, Y. Li, H. Wang, Y. Kondo,
M. Yamanaka, K. Nakajima, M. Kotora, J. Am. Chem. Soc. 2002, 124,
5059–5067.
[15] a) L. Becker, P. Arndt, H. Jiao, A. Spannenberg, U. Rosenthal, Angew.
Chem. Int. Ed. 2013, 52, 11396–11400; Angew. Chem. 2013, 125, 11607–
[33] A. A. Grishina, S. M. Polyakova, R. A. Kunetskiy, I. Cꢆsarovꢃ, I. M. Lyapkalo,
Chem. Eur. J. 2011, 17, 96–100.
[34] M. Kessler, S. Schꢂler, D. Hollmann, M. Klahn, T. Beweries, A. Spannen-
berg, A. Brꢂckner, U. Rosenthal, Angew. Chem. Int. Ed. 2012, 51, 6272–
6275; Angew. Chem. 2012, 124, 6377–6380.
[35] I. Rozas, J. Phys. Org. Chem. 1992, 5, 74–82.
&
&
Chem. Eur. J. 2016, 22, 1 – 14
www.chemeurj.org
12
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
[36] a) 1H NMR data: X.-G. Duan, X.-L. Duan, C. W. Rees, T.-Y. Yue, J. Chem.
Soc. Perkin Trans. 1 1997, 2597–2602; b) 13C NMR data: B. Kamienski, W.
Schilf, J. Sitkowski, L. Stefaniak, G. A. Webb, J. Crystallogr. Spectrosc. Res.
1989, 19, 1003–1008.
[37] V. Burlakov, A. V. Polyakov, A. I. Yanovsky, Y. T. Struchkov, V. B. Shur, M. E.
Vol’pin, U. Rosenthal, H. Gçrls, J. Organomet. Chem. 1994, 476, 197–
206.
[38] L. Friedman, W. P. Wetter, J. Chem. Soc. A 1967, 36–37.
[39] G. M. Sheldrick, Acta Crystallogr. A 2008, 64, 112–122.
[40] K. Brandenburg, DIAMOND, Crystal Impact Gbr, 2007, Bonn, Germany.
[41] J. P. Perdew, Phys. Rev. B 1986, 33, 8822–8824.
A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara,
K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O.
Kitao, H. Nakai, T. Vreven, J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro,
M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, T. Keith,
R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S.
Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox,
J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Strat-
mann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L.
Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dan-
nenberg, S. Dapprich, A. D. Daniels, O. Farkas, J. B. Foresman, J. V. Ortiz,
J. Cioslowski, D. J. Fox, Gaussian, Inc., Wallingford CT, 2010.
[42] A. Schꢅfer, C. Huber, R. Ahlrichs, J. Chem. Phys. 1994, 100, 5829–5835.
[43] P. J. Hay, W. R. Wadt, J. Chem. Phys. 1985, 82, 299–310.
[44] Gaussian 09, Revision C.01, M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E.
Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Men-
nucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian,
Received: March 21, 2016
Published online on && &&, 0000
Chem. Eur. J. 2016, 22, 1 – 14
www.chemeurj.org
13
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
FULL PAPER
How nitrile cycles react: The reactions
of pyridyl-substituted 1-metalla-2,5-
diaza-cyclopenta-2,4-dienes of group 4
metallocenes towards small molecules
were investigated. In case of the addi-
tion of H2, CO2 and different nitriles ex-
change products were obtained. A CÀC
coupling and an H-transfer were ob-
served in the reaction with CH3ÀCꢀN.
The addition of SOCl2 resulted in the
formation of Cp*2MCl2 and in thiadia-
zole oxides.
&
Coordination Chemistry
L. Becker, F. Reiß, K. Altenburger,
A. Spannenberg, P. Arndt, H. Jiao,
U. Rosenthal*
&& – &&
Reactivity Study of Pyridyl-Substituted
1-Metalla-2,5-diaza-cyclopenta-2,4-
dienes of Group 4 Metallocenes
Group 4 Metallocenes
The picture highlights the central role of the symmetrically
pyridyl substituted 1-metalla-2,5-diaza-cyclopenta-2,4-
dienes Cp*2M(N=C(2-py)ÀC(2-py)=N), the youngest
members of the group 4 metallocene (M = Ti, Zr) family.
The mystic tree and its reflection on the water surface
connects these substrates with the unsymmetrically and
fascinating molecules gained from their reactions with small
molecules like H2, CO2, CH3ÀCꢀN, and SOCl2. For example,
for the zirconium complex a reversible coordination of CO2
to the five-membered cycle was observed. In case of the
titanium compound and CH3ÀCꢀN a coupling product of
three acetonitrile molecules was obtained. For more details,
see the Full Paper by U. Rosenthal et al. on page &&ff.
&
&
Chem. Eur. J. 2016, 22, 1 – 14
www.chemeurj.org
14
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!