Synthesis of an η2-N2-titanium diazoalkane complex with both imido- metal carbene-like reactivity patterns

Article abstract of DOI:10.1021/ja974303d

The novel η²-N₂-diazoalkane complex Cp*₂Ti(N₂CHSiMe₃) (2) has been prepared by addition of (trimethylsilyl)diazomethane to Cp*₂Ti(C₂H₄) (1). The structure of 2 reveals nearly symmetric Ti-N distances and an N-N distance ~0.1 A? longer than that of the free diazoalkane. Compound 2 loses dinitrogen in solution under mild conditions to give the fulvene complex Cp*FvTiCH2SiMe₃ (4). Thermolysis of 2 in the presence of 1-alkenes yields the trans-α,β-disubstituted titanacyclobutane complexes Cp*₂Ti(CH(SiMe₃)CH(R)CH₂) (R = H (6), R = Ph (7), R = Me (8), R = Et (9)). The regio- and stereochemistry of the titanacyclobutane complexes was determined by a combination of one- and two-dimensional NMR techniques. The NMR assignment was supported in the case of 8 by an X-ray diffraction study. In addition to confirming the regio- and stereochemistry of the metallacycle, the X-ray structure of 8 shows that unlike most titanacyclobutanes, the metallacycle ring is puckered. A kinetic study of the formation of 7 from 2 and styrene revealed that the reaction is first order in 2 and zero order in styrene. The rate constant for this reaction is identical to that measured for the conversion of 2 to 4. The kinetic study supports a mechanism involving initial rate-determining loss of dinitrogen to form a carbene complex intermediate which undergoes hydrogen transfer from a Cp* methyl tO give 4, or in the presence of styrene is trapped to form 7. In addition to its metal carbene-like reactions involving N₂ loss, complex 2 undergoes a variety of transformations in which N₂ is retained in the final product. These include cycloaddition reactions with alkynes and allene, as well as reversible reactions with Lewis bases to form adducts. These transformations, which are similar to reactions of group IV imido complexes, demonstrate the imide-like character of the diazoalkane ligand.

Full text of DOI:10.1021/ja974303d

6316
J. Am. Chem. Soc. 1998, 120, 6316-6328
Synthesis of an η2-N2-Titanium Diazoalkane Complex with Both
Imido- and Metal Carbene-Like Reactivity Patterns
Jennifer L. Polse, Anne W. Kaplan, Richard A. Andersen,* and Robert G. Bergman*
Contribution from the Department of Chemistry, UniVersity of California, Berkeley, California 94720
ReceiVed December 19, 1997
Abstract: The novel η2-N2-diazoalkane complex Cp*2Ti(N2CHSiMe3) (2) has been prepared by addition of
(trimethylsilyl)diazomethane to Cp*2Ti(C2H4) (1). The structure of 2 reveals nearly symmetric Ti-N distances
and an N-N distance ∼0.1 Å longer than that of the free diazoalkane. Compound 2 loses dinitrogen in
solution under mild conditions to give the fulvene complex Cp*FvTiCH2SiMe3 (4). Thermolysis of 2 in the
presence of 1-alkenes yields the trans-R,â-disubstituted titanacyclobutane complexes Cp*2Ti(CH(SiMe3)CH-
(R)CH2) (R ) H (6), R ) Ph (7), R ) Me (8), R ) Et (9)). The regio- and stereochemistry of the
titanacyclobutane complexes was determined by a combination of one- and two-dimensional NMR techniques.
The NMR assignment was supported in the case of 8 by an X-ray diffraction study. In addition to confirming
the regio- and stereochemistry of the metallacycle, the X-ray structure of 8 shows that unlike most
titanacyclobutanes, the metallacycle ring is puckered. A kinetic study of the formation of 7 from 2 and styrene
revealed that the reaction is first order in 2 and zero order in styrene. The rate constant for this reaction is
identical to that measured for the conversion of 2 to 4. The kinetic study supports a mechanism involving
initial rate-determining loss of dinitrogen to form a carbene complex intermediate which undergoes hydrogen
transfer from a Cp* methyl to give 4, or in the presence of styrene is trapped to form 7. In addition to its
metal carbene-like reactions involving N2 loss, complex 2 undergoes a variety of transformations in which N2
is retained in the final product. These include cycloaddition reactions with alkynes and allene, as well as
reversible reactions with Lewis bases to form adducts. These transformations, which are similar to reactions
of group IV imido complexes, demonstrate the imide-like character of the diazoalkane ligand.
Introduction
loss.3,4 As a result, few isolable diazoalkane complexes have
been shown to generate metal carbene complexes.8-12 We
recently reported a novel titanocene diazoalkane complex that
undergoes facile N2 loss.13 This complex also undergoes a
variety of reactions that proceed with retention of N2, including
cycloaddition reactions with alkynes and allene that are strik-
ingly similar to those of group IV imido complexes. This paper
reports the full details of the synthesis of the diazoalkane
complex and of its carbene- and imido-like reactivity patterns.
Diazoalkanes have long been used as synthons for both free
carbenes and metal carbene complexes (eq 1).1-4 In the metal
carbene syntheses, it is generally assumed that an intermediate
diazoalkane complex is involved in the carbene-transfer process.
Diazoalkane complexes have also been implicated as intermedi-
ates in the metal-catalyzed cyclopropanation of olefins by
diazoalkanes (eq 2).2,5-7
Results
Synthesis and Characterization of Cp*2Ti(N2CHSiMe3)
(2). Treatment of a toluene or benzene solution of Cp*2Ti-
(C2H4) (1) with 1 equiv of (trimethylsilyl)diazomethane results
in gas evolution and an immediate color change from lime to
forest green. Use of exactly equimolar amounts of the two
reactants is critical to the clean isolation of the product Cp*2-
Ti(N2CHSiMe3) (2), because it decomposes in the presence of
excess diazoalkane. Removal of the solvent gives a green
powder that may be crystallized from hexanes at -50 °C to
give pure 2 in 60-70% yield (eq 3).
An understanding of the reactivity of diazoalkane complexes
would aid in further development of such catalysts. Unfortu-
nately, while there are numerous examples of transition metal
diazoalkane complexes, they are typically quite resistant to N2
(1) For a very recent example, see: Schwab, P.; Grubbs, R. H.; Ziller,
J. W. J. Am. Chem. Soc. 1996, 118, 100.
(2) Mizobe, Y.; Ishii, Y.; Hidai, M. Coord. Chem. ReV. 1995, 139, 281.
(3) Herrmann, W. A. Angew. Chem., Int. Ed. Engl. 1978, 17, 800.
(4) Sutton, D. Chem. ReV. 1993, 93, 995.
(5) Herrmann, W. A.; Menjon, B.; Herdtweck, E. Organometallics 1991,
10, 2134.
(8) Messerle, L.; Curtis, M. D. J. Am. Chem. Soc. 1980, 102, 7789.
(9) Curtis, M. D.; Messerle, L.; D′Errico, J. J.; Butler, W. M.; Hay, M.
S. Organometallics 1986, 5, 2283.
(10) Nakamura, A.; Yoshida, T.; Cowie, M.; Otsuka, S.; Ibers, J. A. J.
Am. Chem. Soc. 1977, 99, 2108.
(11) Cowie, M.; Loeb, S. J.; McKeer, I. R. Organometallics 1986, 5,
854.
(6) Doyle, M. P.; Peterson, L. S.; Zhou, Q. L.; Nishiyama, H. Chem.
Commun. 1997, 211.
(7) Doyle, M. P.; Peterson, C. S.; Parker, D. L., Jr. Angew. Chem., Int.
Ed. Engl. 1996, 35, 1334.
(12) Clauss, A. D.; Shapley, J. R.; Wilson, S. R. J. Am. Chem. Soc. 1981,
103, 7387.
(13) Polse, J. L.; Andersen, R. A.; Bergman, R. G. J. Am. Chem. Soc.
1996, 118, 8737.
S0002-7863(97)04303-5 CCC: $15.00 © 1998 American Chemical Society
Published on Web 06/12/1998

η2-N2-Titanium Diazoalkane Complex
J. Am. Chem. Soc., Vol. 120, No. 25, 1998 6317
Table 1. Crystal Data for 2, 8, and 18
2
8
18
diffractometer
T (°C)
radiation
Siemens Smart Siemens Smart Siemens Smart
-99.0
Mo KR
-100.0
Mo KR
-109.0
Mo KR
(λ ) 0.71069 Å) (λ ) 0.71069 Å) (λ ) 0.71069 Å)
take-off angle (deg) 6.0
cryst-to-detec
dist (mm)
6.0
60
6.0
60
The modest isolated yield reflects the high solubility of the
complex rather than the formation of any byproducts. When
the reaction is performed in benzene-d6 and followed by H
60
1
scan type
ω(0.3°/frame)
ω(0.3°/frame)
10.0
51.6
12217
4740
2809
ω(0.3°/frame)
10.0
51.9
7493
4601
NMR spectroscopy, 1 equiv of free ethylene is produced in
addition to 2 in >98% yield, which is the only observable
titanium-containing product. Retention of the nitrogen atoms
in 2 was confirmed by mass spectroscopy, elemental analysis,
and X-ray crystallography (vide infra).
scan rate (s/frame) 30.0
2θmax (deg) 46.5
no. of refl. collected 9838
no. of unique refl.
no. of obsd reflns
I > 3.00σ(I)
3629
2904
3941
1
The H NMR spectrum of 2 displays singlets at 0.44 and
no. of variables
414
400
280
refl/parameter ratio 7.01
7.02
0.043
0.050
1.68
0.01
14.07
0.039
0.057
2.15
1.71 ppm for the Me3Si and equivalent Cp* groups, respectively.
In addition, there is a singlet at 4.11 ppm that is assigned to the
methine proton of the diazoalkane ligand. The 13C{1H} NMR
spectrum, in addition to resonances for the Cp* and Me3Si
groups, shows a methine resonance at 99.9 ppm for the
diazoalkane ligand. The IR spectrum does not contain any
strong absorptions in the region typically associated with organic
diazoalkanes (2000-2270 cm-1). The lack of a strong infrared
stretch in this region for the diazoalkane ligand probably results
from reduction of the N-N bond order due to back-bonding
from the metal center (vide infra).14
R
Rw
0.045
0.058
2.23
goodness of fit
p factor
0.01
0.03
Because of the wide variety of possible coordination modes
for diazoalkane ligands, it is often difficult to predict their
structures based on spectroscopic data. We therefore undertook
a single-crystal X-ray diffraction study to determine the
coordination mode of the diazoalkane ligand in 2. The structure
was solved by direct methods and refined using standard least-
squares and Fourier techniques. Crystal data and collection
parameters are given in Table 1. An ORTEP diagram is shown
in Figure 1, and representative bond lengths and angles are given
in Table 2. Other details of the structure determination are
provided in the Experimental Section and as Supporting
Information. The structural study reveals that the diazoalkane
ligand is bound to the titanium in a side-on fashion through the
two nitrogen atoms. Although side-on coordination is less
common than end-on or bridging coordination modes, it is not
unprecedented.2-4,10,15,16 As is usually the case for side-bound
diazoalkane complexes,4 the Ti-N1 (1.979(2) Å) and Ti-N2
(2.012(2) Å) distances for 2 are very similar. The Ti-N2
distance is slightly longer than the Ti-N1 distance, possibly
because N1 is two-coordinate and therefore has a smaller radius
than does N2.17 The Ti-N distances are indicative of single
bonds, and are similar to the Ti-N1 distance (1.980(5) Å) in
the diazoalkane complex Cp2Ti(DEDM) (DEDM ) diethyldia-
zomalonate) (3) (Scheme 1) synthesized by Floriani.15 The other
Ti-N distance in 3 is quite long (2.219(7) Å), possibly because
of a large contribution from resonance structure B (Scheme 1).
The bonding situation in 2 is similar to that in an olefin
adduct. The relatively long N-N bond distance (1.276(3) Å)
compared with that of uncomplexed diazoalkanes (1.12-1.13
Å)18 indicates substantial π-back-bonding from the b2 orbital
Figure 1. ORTEP diagram of Cp*2Ti(N2CHSiMe3) (2).
Table 2. Selected Bond Distances (Å) and Angles (deg) for 2
Ti-N1
1.979(2)
2.0961(4)
1.276(3)
1.840(3)
Ti-N2
Ti-Cp*2
N2-C21
2.012(2)
2.0970(4)
1.342(3)
Ti-Cp*1
N1-N2
Si1-C21
Cp*1-Ti-Cp*2
N1-Ti-Cp*2
N2-Ti-Cp*2
Ti-N2-C21
Si1-C21-N2
140.72(2)
N1-Ti-Cp*1
N2-Ti-Cp*1
N1-Ti-N2
N1-N2-C21
Ti-N1-N2
109.05(6)
108.51(6)
110.03(6)
158.7(2)
122.6(2)
106.68(6)
37.27(9)
131.3(2)
72.8(1)
Scheme 1
(14) The IR spectrum of 2 contains a weak stretch at 1894 cm-1 which
may be due to a reduced N-N bond stretch; however, it has not been
assigned.
(15) Gambarotta, S.; Floriani, C.; Chiesi-Villa, A.; Guastini, C. J. Am.
Chem. Soc. 1982, 104, 1918.
(16) Schramm, K. D.; Ibers, J. A. Inorg. Chem. 1980, 19, 2441.
(17) Pauling, L. The Nature of the Chemical Bond; Cornell University
Press: Ithaca, New York, 1967.
of the titanocene fragment to the diazoalkane LUMO, which is
N-N antibonding.19 The N-N distance is indicative of a partial
N-N double bond, since N-N single bonds are typically ∼1.40
(18) Patai, S. The Chemistry of Diazonium and Diazo Groups, Parts 1
and 2; Wiley: New York, 1978 and references therein.
(19) Jorgensen, W. L.; Salem, L. The Organic Chemist’s Book of
Orbitals; Academic Press: New York, 1973; p 126.

6318 J. Am. Chem. Soc., Vol. 120, No. 25, 1998
Polse et al.
Scheme 2
Figure 2. Isomers of disubstituted metallacyclobutanes.
Scheme 3
Å.20 In this respect, the structure of 2 is similar to that of the
starting olefin complex 1. Compound 1 has a C-C bond length
of 1.438 (5) Å for the ethylene ligand, which is ∼0.1 Å longer
than its value in free ethylene, indicating that complex 1 is best
described as intermediate between a titanium(II) olefin adduct
and a titanium(IV) metallacyclopropane.21
Thermal N2 Loss and Formation of Cp*FvTiCH2SiMe3
(Fv ) η5,η1-Me4C5CH2) (4). Complex 2 is thermally unstable
in both the solid and solution states. It decomposes at 25 °C in
the solid state with a half-life of ∼20 h to give a red powder.
This powder has not been identified, but on the basis of its low
solubility it is presumed to be polymeric in nature. In benzene
and toluene solution 2 decomposes with a half-life of 2 h at 32
°C to give fulvene complex 4 in 84% isolated yield (Scheme
2). Complex 4 was identified on the basis of its NMR spectral
data and elemental analysis. In addition, 4 was independently
synthesized as described previously.13,22 The inequivalent
methyl groups of the metalated Cp* ring are clearly visible in
of 2 d at room temperature to give the deep red metallacy-
clobutane complexes 6-9 (Scheme 3) in good yield. Loss of
dinitrogen was confirmed by elemental analysis. The 1H NMR
spectra of these materials show the expected features, including
inequivalent Cp* ligands. In the case of substituted olefins,
there are six distinct metallacycle isomers that may be formed
(Figure 2). We observe that alkene addition to 2 is regio- and
stereospecific, since we observe only one of these possible
isomers. This isomer was identified as R,â-disubstituted isomer
A (Figure 2) by a combination of one- and two-dimensional
NMR techniques as described below. 1-Butene was the most
sterically hindered alkene that still gave complete conversion
to the metallacycle product; reaction of 2 with the bulkier olefin
3-methylbutene gave a mixture of fulvene complex 4 and the
expected metallacyclobutane. Thermolysis of 2 in the presence
of even bulkier R-olefins such as 3,3-dimethylbutene gave
exclusively 4. Complex 2 does not react with internal olefins
or alkynes including the strained cyclic olefin norbornadiene.
NMR Determination of the Regio- and Stereochemistry
of Metallacyclobutane Complexes 6-9. The regiochemistry
of the metallacyclobutane complexes was determined as follows.
The 1H NMR spectra of 7-9 (Scheme 3) display several
distinguishing features, including diastereotopic Cp* resonances,
a trimethylsilyl group, and several strongly coupled resonances
between -0.9 and 2 ppm for the protons in the metallacycle
ring. In addition to peaks for the Cp* ligands and trimethylsilyl
groups, the 13C{1H} NMR spectra of 7-9 show two resonances
between 60 and 70 ppm, which is within the region associated
with R-carbons of titanacyclobutanes.25-30 By using standard
1
both the H and 13C{1H} NMR spectra of 4. The protons on
both methylene carbons are diastereotopic, and therefore the
1H NMR spectrum shows two sets of doublets for these protons.
The resonances due to the methylene protons of the trimethyl-
silylmethyl ligand show silicon satellites, and they can therefore
be distinguished from the two doublets of the fulvene methylene
1
group. A H-13C HMQC23 spectrum showed that the Me3-
SiCH2 methylene protons correlate with the methylene resonance
at 45.8 in the 13C{1H} NMR spectrum. The methylene carbon
of the fulvene ligand resonates at 76.1 ppm in the 13C{1H} NMR
spectrum. This is well within the typical range observed for
titanium fulvene complexes. For example, the CH2 group of
Cp*FvTiCH3 (5), prepared earlier from Cp*2TiMe2 (Scheme
2), resonates at 73.9 ppm in the 13C{1H} NMR spectrum.24
Reactivity of 2 toward Terminal Alkenes. Formation of
Metallacyclobutane Complexes. Compound 2 reacts with
terminal alkenes H2CdCHR (R ) H, Ph, Me, Et) over a period
(20) Allen, F. H.; Kennard, O.; Watson, D. G.; Brammer, L.; Orpen, A.
G.; Taylor, R. J. Chem. Soc., Perkin Trans. 1987, S1.
(21) Cohen, S. A.; Auburn, P. R.; Bercaw, J. E. J. Am. Chem. Soc. 1983,
105, 1136.
(22) Luinstra, G. A.; Brinkmann, P. H. P.; Teuben, J. H. J. Organomet.
Chem. 1997, 532, 125.
(23) Bax, A.; Subramanian, S. J. Magn. Reson. 1986, 67, 565.
(24) McDade, C.; Green, J. C.; Bercaw, J. E. Organometallics 1982, 1,
1629.
(25) Lee, J. B.; Gajda, G. J.; Schaefer, W. P.; Howard, T. R.; Ikariya,
T.; Straus, D. A.; Grubbs, R. H. J. Am. Chem. Soc. 1981, 103, 7358.
(26) Lee, J. B.; Ott, K. C.; Grubbs, R. H. J. Am. Chem. Soc. 1982, 104,
7491.
(27) Howard, T. R.; Lee, J. B.; Grubbs, R. H. J. Am. Chem. Soc. 1980,
102, 6876.
(28) Gilliom, L. R.; Grubbs, R. H. Organometallics 1986, 5, 721.
(29) Straus, D. A.; Grubbs, R. H. J. Mol. Catal. 1985, 28, 9.

η2-N2-Titanium Diazoalkane Complex
J. Am. Chem. Soc., Vol. 120, No. 25, 1998 6319
DEPT pulse sequences, these resonances were determined to
arise from one methine and one methylene carbon, indicating
that the metallacycles are R,â-disubstituted. (This restricts the
possible structures to isomers A-D in Figure 2.) The 1H NMR
spectra of 7-9 each display a doublet integrating to one proton
between 1.2 and 2 ppm. This proton signal correlates with the
1
downfield methine resonance in a one-bond H-13C HMQC
NMR experiment.23 In several cases, silicon satellites were
observed for this methine resonance in the 1H NMR spectrum,
indicating that the trimethylsilyl group is directly bonded to the
methine carbon. The metallacycle methines not bound to
titanium resonated between 15 and 30 ppm in the 13C{1H} NMR
spectrum. The 1H-1H COSY NMR spectra of 8 and 9 revealed
that the proton resonance that correlated with the central methine
carbon was strongly coupled to the protons of the methyl or
ethyl substituent. None of the other metallacyclobutane ring
protons were coupled to the methyl or ethyl substituents. Taken
together, these data indicate that the trimethylsilyl-substituted
carbon resonance is R to the metal center, limiting the choices
to isomers A and B in Figure 2.
Figure 3. ORTEP diagram of Cp*2Ti(CH(SiMe3)CH(Me)CH2) (8).
Table 3. Selected Bond Distances (Å) and Angles (deg) for 8
To determine the relative stereochemistry of the metallacycle
methine carbons in 7-9 (i.e., to distinguish between isomers
Ti-C21
2.155(4)
Ti-C23
2.192(4)
2.1600(6)
1.557(5)
1.533(5)
1
A and B), H-1H NOESY spectra were acquired as described
Ti-Cp*1
C21-C22
Si1-C23
2.1447(6) Ti-Cp*2
1.551(5)
1.873(4)
in the Experimental Section. The analysis was similar for each
metallacycle and will be described in detail for the 1-butene
adduct 9. The NOESY spectrum of 9, acquired using a mixing
time of 792 ms, is given as Supporting Information. The
substituent labeling used in the following discussion is shown
in Scheme 3. Cross-peaks were observed between Cp*a and
the Me3Si group, Hb and Hc. Cp*b showed NOE’s to Hd, the
ethyl group, and Ha. No NOE’s were observed between the
ring protons and the Cp* on the opposite side of the ring from
them. If the substituents were in a cis arrangement (isomer B,
Figure 2), the two ring methine resonances (Ha and Hb) would
be expected to show cross-peaks to the same Cp*. Instead, each
shows a cross-peak to a different Cp*. Only the trans isomer
is consistent with the observed pattern of cross-peaks. Together
with the HMQC and COSY NMR data, the NOESY experiments
indicate that the disubstituted metallacyclobutanes possess
structure A (Figure 2).
C22-C23
C22-C24
Cp*1-Ti-Cp*2 135.78(3)
C21-Ti-Cp*1
C23-Ti-Cp*1
C21-Ti-C23
Ti-C23-C22
Ti-C23-Si1
104.7(1)
C21-Ti-Cp*2
C23-Ti-Cp*2
Ti-C21-C22
C21-C22-C23
C23-C22-C24
106.4(1)
108.0(1)
87.6(2)
109.0(3)
113.5(3)
111.4(1)
71.2(1)
86.1(2)
145.9(2)
C21-C22-C24 111.5(3)
intramolecular bond lengths and angles for this compound. As
expected from the NMR experiments, the two substituents are
on adjacent carbons and are trans to one another. The structure
also shows that the trimethylsilyl-substituted carbon is R to the
metal center. The Ti-C21 and Ti-C23 distances are statisti-
cally different (2.155(4) and 2.192(4) Å, respectively) but
essentially equivalent and are within the typical range (2.14-
2.21 Å) for Ti-Csp bonds.31 The C-C distances in the
3
Only two isomers are possible for metallacycle 6, since in
this case isomers C-F are all equivalent, and A and B are
equivalent (Figure 2). The two possible isomers have the
trimethylsilyl group R or â to the metal center. In addition to
the expected trimethylsilyl and diastereotopic Cp* resonances,
the 13C{1H} NMR spectrum of 6 shows two resonances, at 71.1
and 73.2 ppm. Using standard DEPT pulse sequences, these
resonances were assigned to methylene and methine carbons,
respectively. As discussed above, these resonances are due to
the R carbons of the metallacyclobutane ring. This indicates
that, as in 7-9, the trimethylsilyl-substituted carbon is R to the
metal center in 6.
X-ray Structure of Cp*2Ti(CH(SiMe3)CH(Me)CH2) (8).
To confirm the NMR assignment of the structures of 6-9, a
single-crystal X-ray diffraction study of propene adduct 8 was
undertaken. The structure was solved by Patterson methods
and refined using standard least squares and Fourier techniques.
Crystal and data collection parameters are shown in Table 1,
and details of the structure determination are given in the
Experimental Section. Positional and thermal parameters are
given in Tables S7 and S8 of the Supporting Information. An
ORTEP diagram of 8 is shown in Figure 3. Table 3 provides
metallacycle ring are similar to those of previously characterized
titanacyclobutane complexes.25 The ring is puckered with a
dihedral angle of 28° between the C21TiC23 and C21C22C23
planes. Although the vast majority of structurally characterized
titanacyclobutanes are planar, puckering of the ring has been
observed previously in an R,â-disubstituted titanacyclobutane
10 (Scheme 3) derived from Tebbe’s reagent and a norbornene
diester.30 The distortion in the case of 8 is probably a result of
the extreme steric congestion between the metal center and the
two ring substituents.
To determine whether the observed isomer is the kinetic rather
than the thermodynamic product of the reaction of 2 with
alkenes a solution of styrene adduct 7 in C6D6 was heated to
45 °C for 2 d. No rearrangement of 7 was observed. However,
heating to 75 °C resulted in formation of fulvene complex 4.
This is likely the result of a retrocyclization of the metallacycle
to form a transient carbene complex which then reacts with a
C-H bond of the ligand (vide infra). Attempts to exchange
one alkene for another in the metallacyclobutanes were unsuc-
cessful. For instance, heating the 1-butene adduct 9 to 75 °C
in the presence of 8 equiv of styrene resulted only in formation
of 4 and release of butene. The released butene was a mixture
(30) Stille, J. R.; Santarsiero, B. D.; Grubbs, R. H. J. Org. Chem. 1990,
55, 843.
(31) Orpen, A. G.; Brammer, L.; Allen, F. H.; Kennard, O.; Watson, D.
G.; Taylor, R. J. Chem. Soc., Dalton Trans. 1989, S1.

6320 J. Am. Chem. Soc., Vol. 120, No. 25, 1998
Polse et al.
Figure 4. Plot of concentration vs time and exponential fit for the
formation of 7 from 2 and styrene at 32 °C.
Figure 6. Eyring plot (30-60 °C) for the reaction of 2 with styrene.
Scheme 4
detailed study of the electronic factors influencing dinitrogen
loss from phenyl-substituted diazoalkane complexes will be the
subject of a future publication.32
Reaction of 2 with Lewis Bases. Retention of N2. In an
attempt to intercept any intermediates in the thermolysis of 2
to 4, the reaction was run in the presence of 5 equiv of
trimethylphosphine. The added trimethylphosphine had no
effect on the reaction rate, and formation of 4 was observed
with no detectable intermediates. In contrast, when excess
pyridine or p-(dimethylamino)pyridine (DMAP) is added to a
benzene-d6 solution of 2, a color change from green to purple
is observed. The 1H NMR spectra of the purple solutions show
the presence of 2 and a new compound, assigned as the pyridine
(11) or DMAP (12) adduct of 2 (Scheme 4). The resonances
for free and bound pyridine are coalesced at room temperature,
indicating that exchange of free and bound pyridine is rapid on
the NMR time scale. Removal of the volatile materials under
dynamic vacuum leaves a green powder which, when redis-
Figure 5. Plot of kobs vs styrene concentration for the reaction of 2
with styrene.
of cis- and trans-2-butene, rather than the expected 1-butene;
the mechanism of isomerization of the butene is unknown at
this time.
Kinetic Study of Formation of 7 from 2 and Styrene. The
kinetics of the reaction of diazoalkane complex 2 with styrene
in toluene-d8 have been examined by 1H NMR spectroscopy at
a variety of temperatures and styrene concentrations. To ensure
that the reaction would be pseudo-first-order in 2, the lowest
styrene concentration examined was 10 times greater than the
concentration of 2. A typical plot of concentration vs time is
shown in Figure 4. The reaction follows first-order kinetics to
>5 half-lives. A plot of kobs vs [styrene] (Figure 5) demonstrates
that the reaction rate is independent of styrene concentration.
An Eyring plot (30-60 °C) (Figure 6) gives ∆Hq ) 22.5 ( 0.3
kcal/mol, ∆Sq ) -3.1 ( 1.0 eu, and ∆Gq ) 23.4 ( 0.4 kcal/
mol (calculated at 25 °C). To assess whether any charge buildup
occurs in the transition state, the rate of the reaction was also
measured at 45.7 °C in the more polar solvent, THF-d8. The
rate constant was 6.04 × 10-4 (( 0.01 × 10-4) s-1, which is
the same within error as that measured in toluene-d8 at the same
temperature. At 45 °C in toluene-d8, 2 rearranges to fulvene
1
solved in benzene-d6, gives a H NMR spectrum of pure 2,
confirming that the base binds reversibly. Solutions of 11 and
12 decompose slowly at 25 °C to give fulvene complex 4 and
free pyridine or DMAP. The rate of formation of 4 is markedly
slower in the presence of pyridine or DMAP; however, 11 and
12 could not be isolated due to loss of pyridine or DMAP. It
seems reasonable to assign these complexes as the pyridine or
DMAP adduct of an end-on diazoalkane complex (Scheme 4),
since a similar compound, Cp2Ti(N2CPh2)(PMe3) (13) (Scheme
5) is known. It is unlikely that 11 and 12 still possess side-
bound diazoalkane ligands with the pyridine or DMAP coor-
dinated to the metal, since the bulky decamethyltitanocene
complex 4 with a kobs of 5.17 × 10-4 (( 0.10 × 10-4) s-1
,
identical within experimental error to that for the reaction of 2
with styrene at 45 °C of 5.02 × 10-4 (( 0.04 × 10-4) s-1. The
formation of the fulvene complex 4 from 2 at 45.7 °C in THF-
d8 occurred with a kobs of 5.82 × 10-4 (( 0.04 × 10-4) s-1. A
(32) Kaplan, A. W.; Polse, J. L.; Ball, G. E.; Andersen, R. A.; Bergman,
R. G. Submitted for publication.

η2-N2-Titanium Diazoalkane Complex
J. Am. Chem. Soc., Vol. 120, No. 25, 1998 6321
Scheme 5
shifted to 6.33 ppm. The 13C NMR spectrum displays two
resonances at 180.2 and 133.3 ppm which were assigned as
methine carbons using standard DEPT pulse sequences. The
resonance at 180.2 is characteristic of an sp2 carbon bound to
Ti.35 If the diazoalkane complex had acted as a base to
deprotonate the acetylene and give a metal acetylide complex,
1
neither the strong coupling in the H NMR spectrum nor the
two methine resonances in the 13C NMR spectrum would be
expected.
Treatment of a benzene solution of 2 with 1 equiv of
phenylacetylene gives a red solution from which crystals of
azametallacyclobutene 17 (Scheme 6) can be isolated in 64%
yield. Retention of dinitrogen was confirmed by mass spec-
1
troscopy and elementary analysis. The H NMR spectrum of
Scheme 6
17 shows singlets for the metallacycle and diazoalkane methine
protons at 6.34 and 2.99 ppm, respectively. The 13C NMR
spectrum shows a methine resonance at 192.3 ppm, consistent
with an sp2 carbon bound to Ti. The chemical shifts for the
metallacyclobutene ring are comparable to those reported for
related titanium oxametallacyclobutene complexes.36 In addi-
tion, the shift is quite similar to that of the R-proton in 14 (180.2
ppm). As discussed above for 14, formulation of 17 as an
acetylide complex resulting from deprotonation of the alkynyl
proton by the acidic nitrogen of 2 is inconsistent with the
presence of a methine carbon resonance at >180 ppm. In
addition, the IR spectrum shows no N-H or acetylide CtC
stretches as would be expected from an amide acetylide
complex.
Treatment of 2 with allene at -195 °C results in an immediate
color change from green to red upon warming to room
temperature. Removal of the solvent gives a red powder which
may be crystallized from hexanes at -50 °C to give Cp*2Ti-
(NNdCHSiMe3)C(CH2)CH2 (18) in 78% yield (Scheme 6).
Retention of dinitrogen was confirmed by elemental analysis
and mass spectroscopy. In addition to two singlets for the
trimethylsilyl group and Cp* methyl groups, the 1H NMR
spectrum of 18 displays two doublets (3J ) 1.9 Hz) at 4.77 and
3.78 ppm corresponding to the exocyclic protons on the
metallacycle ring. The methylene protons R to the titanium
center resonate at 2.62 ppm, and the methine proton on the
diazoalkane fragment is shifted to 6.23 ppm. The equivalent
Cp* methyl groups and ring methylene protons are consistent
with either a planar azametallacyclobutane ring or a puckered
ring undergoing a rapid ring flip. Although the puckered
structure of the metallacycle ring was determined by X-ray
crystallography (vide infra), the resonances for the Cp* and
methylene protons remained singlets down to - 70 °C. To
determine the regiochemistry of azametallacyclobutane complex
Scheme 7
fragment strongly disfavors coordination of three ligands in the
metallocene wedge.
Reactivity toward Alkynes and Allene. Synthesis and
Structure of Azatitanacycle Complexes. Exposing a deep
green benzene solution of 2 to a stream of dry acetylene gas
results in an immediate color change to deep red. Removal of
the volatile materials followed by crystallization from hexanes
(-50 °C) yields red crystals of a new complex, identified as
azametallacyclobutene 14 (Scheme 6). The reaction time is
critical since longer reaction times led to precipitation of a black
solid which is presumed to be polyacetylene. Retention of
dinitrogen in 14 was confirmed by elemental analysis and mass
spectroscopy. In addition to two singlets for the trimethylsilyl
1
18, a H-1H NOESY spectrum was acquired as described in
the Experimental Section. The two possible isomers have the
exocyclic double bond R or â to the metal center. The proton
on the diazoalkane moiety and the Me3Si group both showed
cross-peaks to the exocyclic protons, but not to the ring protons.
This indicates that the exocyclic double bond is â to the metal
center.
To confirm the NMR assignment of 18, a single-crystal X-ray
diffraction study was undertaken. The structure was solved by
direct methods and refined using standard least-squares and
1
group and Cp* methyl groups, the H NMR spectrum of 14
displays two doublets (3J ) 9.4 Hz) at 7.29 and 6.52 ppm
attributable to the methine protons on the metallacycle ring. This
compares favorably with the methine resonances in related aza-
(15) and thiatitanacyclobutene (16) complexes (Scheme 7)
(15: δ 7.02, 7.08 ppm, 3J ) 9 Hz; 16: δ 6.52, 7.42 ppm, 3J )
9.9 Hz).33,34 The methine proton of the diazoalkane moiety is
(33) Sweeney, Z. K.; Bergman, R. G. Unpublished results.
(34) Polse, J. L.; Andersen, R. A.; Bergman, R. G. Manuscript in
preparation.
(35) Doxsee, K. M.; Juliette, J. J. J.; Weakley, T. J. R.; Zientara, K.
Inorg. Chim. Acta 1994, 222, 305 and references therein.
(36) Polse, J. L.; Andersen, R. A.; Bergman, R. G. J. Am. Chem. Soc.
1995, 117, 5393.

6322 J. Am. Chem. Soc., Vol. 120, No. 25, 1998
Polse et al.
1
at 8.09 possesses a 68.2 Hz J coupling to 15N. In addition to
resonances for the Me3Si and Cp* groups, the 13C NMR
spectrum shows only one other resonance at 177.1 ppm, assigned
as a quaternary carbon by standard DEPT pulse sequences. In
addition, the IR spectrum of 19 shows a stretch at 3290 cm-1
consistent with a NH stretch. On the basis of the above data,
complex 19 is assigned the structure shown in Scheme 8.
Interestingly, 19 also is obtained from oxotitanium complex
Cp*2Ti(O)py (20)37 and Me3SiC(H)N2.
Discussion
Synthesis and Structure of Cp*2Ti(N2CHSiMe3) (2). As
described in the Results section, treatment of 1 with (trimeth-
ylsilyl)diazomethane results in clean formation of diazoalkane
complex 2 and ethylene (eq 3). The retention of dinitrogen
was somewhat unexpected in light of similar reactions of 1
toward oxygen-atom and nitrene donors. As described previ-
ously, 1 reacts with N2O and organic azides to liberate ethylene
and dinitrogen, giving Cp*2Ti(O)py (20)38 and Cp*2TidNR
(21), respectively (Scheme 9).36,39 By analogy to the oxygen-
and nitrene-transfer reactions, carbene complex 22 (Scheme 9)
would be the expected product of the reaction of a diazoalkane
with 1. If N2O40-45 and azide complexes46,47 similar to 2 are
formed on the reaction path to 20 and 21, they lose nitrogen
much more rapidly than does 2. Complex 2 liberates dinitrogen
slowly to generate what is presumed to be a carbene complex
intermediate, which then reacts with a variety of substrates (vide
infra). It is possible that the instability of the carbene complex
leads to a higher barrier for N2 loss than was observed for the
reactions of 1 with N2O and organic azides.
Figure 7. ORTEP diagram of Cp*2Ti(N(NdC(H)SiMe3)C(CH2)CH2)
(16).
Table 4. Selected Bond Distances (Å) and Angles (deg) for 18
Ti-N1
2.065(2)
2.1308(4)
1.509(3)
1.352(4)
1.294(3)
Ti-C1
2.187(2)
2.1209(4)
1.388(3)
1.374(2)
Ti-C101
C1-C2
C2-C3
N2-C4
Ti-C102
N1-C2
N1-N2
N1-Ti-C1
N1-Ti-C101
C1-Ti-C101
Ti-N1-C2
N1-C2-C1
N1-C2-C3
N2-N1-C2
66.41(8)
108.55(5)
104.73(7)
95.8(1)
107.0(2)
127.3(3)
117.3(2)
C101-Ti-C102
N1-Ti-C102
C1-Ti-C102
Ti-C1-C2
138.61(2)
108.58(5)
106.37(7)
87.5(1)
143.5(1)
125.5(2)
Ti-N1-N2
C1-C2-C3
A possible mechanism for the formation of 2 is outlined in
Scheme 10. The mechanism postulates an initial insertion of
the diazoalkane into the Ti-C bond of 1 to give an azametal-
lacyclobutane complex (23). A similar insertion of diphenyl-
diazomethane into the Zr-C bond of Cp*2Zr(η2-PhCtCPh) to
give an azametallacyclobutene complex has been described
previously (eq 4).48 The intermediate azametallacyclobutane
Fourier techniques. Crystal and data collection parameters are
shown in Table 1 and details of the structure determination are
given in the Experimental Section. Positional and thermal
parameters are given as Supporting Information. An ORTEP
diagram of 18 is shown in Figure 7. Table 4 provides
intramolecular bond lengths and angles for this compound. As
expected, the exocyclic double bond is â to the metal center.
The structure also shows that the trimethylsilyl group is directed
out of the Cp* wedge. The Ti-C1 distance was found to be
2.187(2) Å, within the typical range (2.14-2.21 Å) for Ti-
Csp3 bonds.31 The ring deviates slightly from planarity; the
dihedral angle between the planes formed by Ti(1), C(1), and
N(1), and the plane formed by C(2), C(1), and N(1) is 161°.
Ti(1) lies 0.645 Å out of the least-squares plane formed by N(1),
C(1), C(2), and C(3).
23 can then undergo a cycloreversion reaction similar to the
one discussed for Cp*2Ti(N(Ph)CH2CH2) (eq 5)34 to liberate
To study the generality of its cycloaddition reactions, complex
2 was treated with N2O, which is isoelectronic with allene. When
2 was exposed to 1 equiv of N2O in benzene, the solution turned
from dark green to a deep red-black. Removal of the volatile
materials and crystallization from Et2O (-50 °C) gave crystals
of 19 (Scheme 8). Mass spectroscopy and elemental analysis
confirmed that 19 possessed an elemental composition corre-
sponding to retention of 1 equiv of N2, presumably from the
diazoalkane fragment. This rules out the possibility that 19 is
the product of a [2 + 2] cycloaddition of 2 with N2O as observed
for allene. Examination of the 1H and 13C NMR spectra of 19
showed that it did not contain a five-membered metallacycle
(Scheme 8), the result from oxygen atom transfer. The 1H NMR
spectrum shows a broad one-proton singlet at 8.09 ppm and a
singlet for equivalent Cp* methyl groups. Examination of the
15N-filtered 1H NMR spectrum of 19 showed that the resonance
(37) Smith, M. R., III; Matsunaga, P. T.; Andersen, R. A. J. Am. Chem.
Soc. 1993, 115, 7049.
(38) The reaction of 1 with N2O must be run in the presence of excess
pyridine, since 16 is only stable as a pyridine adduct.
(39) Smith, M. R., III; Ball, G. E.; Andersen, R. A. Manuscript in
preparation.
(40) Armor, J. N.; Taube, H. J. Am. Chem. Soc. 1969, 91, 6874.
(41) Bottomley, F.; Crawford, J. R. J. Chem. Soc., Chem. Commun. 1971,
200.
(42) Bottomley, F.; Crawford, J. R. J. Am. Chem. Soc. 1972, 94, 9092.
(43) Diamantis, A. A.; Sparrow, G. J. J. Chem. Soc., Chem. Commun.
1970, 819.
(44) Diamantis, A. A.; Sparrow, G. J.; Snow, M. R.; Norman, T. R. Aust.
J. Chem. 1975, 28, 1231.
(45) Tuan, D. F.; Hoffman, R. Inorg. Chem. 1985, 24, 871.
(46) Fickes, M. G.; Davis, W. M.; Cummins, C. C. J. Am. Chem. Soc.
1995, 117, 6384.
(47) Proulx, G.; Bergman, R. G. J. Am. Chem. Soc. 1995, 117, 6382.
(48) Vaughan, G. A.; Hillhouse, G. L.; Rheingold, A. L. J. Am. Chem.
Soc. 1990, 112, 7994.

η2-N2-Titanium Diazoalkane Complex
J. Am. Chem. Soc., Vol. 120, No. 25, 1998 6323
Scheme 8
Scheme 9
(Scheme 1) and Cp2Ti(N2CPh2)(PMe3) (13) (Scheme 5) do not
lose dinitrogen thermally or photochemically.15,49 It is possible
that an open coordination site at the metal center is required
for N2 loss to occur and that the tightly bound ancillary ligands
of 3 and 13 render them inert to this reaction.
There are several other examples of isolated diazoalkane
complexes that lose dinitrogen to give metal carbene
complexes.2,8-10,12,16,50 For example, the bridging molybdenum
diazoalkane complex 24 loses dinitrogen thermally to give
carbene complex 25 (eq 6).8 The authors postulate that carbene
Scheme 10
formation occurs via a cyclic transition state in which the
diazoalkane and the dimolybdenum center form a five membered
ring which then extrudes N2.8 Although the intimate mechanism
of dinitrogen loss from 2 is still largely unknown, we have been
able to obtain some preliminary information about it from a
kinetic study of the reaction of 2 with styrene (vide infra), and
an in-depth study of the electronic factors involved in this step
is underway.32
Reaction of 2 with r-Olefins. Mechanistic Discussion. We
have carried out a kinetic study of the reaction of 2 with styrene
in order to obtain further information on the mechanism of
nitrogen loss and metallacycle formation. The kinetic study
was carried out as described in the Results Section and
Experimental Section. The reaction was first order in [2], and
the rate showed no dependence on styrene concentration.
Monitoring the reaction at a variety of temperatures between
30 and 60 °C allowed measurement of the activation parameters
∆Hq ) 22.5 ( 0.3 kcal/mol, ∆Sq ) -3.1 ( 1.0 eu, and ∆Gq )
23.4 ( 0.4 kcal/mol (calculated at 25 °C). Because there is no
rate dependence on styrene concentration, any mechanism that
invokes an attack of the alkene prior to or during the rate
determining step of the reaction can be ruled out.
ethylene; coordination of the second nitrogen forms 2.
It is also possible to envision a mechanism in which ethylene
dissociation precedes diazoalkane coordination. Such a mech-
anism can be discounted, however, since ethylene does not
dissociate from 1 to give free Cp*2Ti.21
Thermal Loss of Dinitrogen from 2. Formation of
Cp*FvTiCH2SiMe3 (4). As was noted above, in the absence
of traps 2 loses dinitrogen thermally to give fulvene complex 4
(Scheme 2). The mild conditions under which 2 loses dinitrogen
are surprising, since (as pointed out earlier) few isolable
diazoalkane complexes release dinitrogen.2-4 For example, the
related titanium diazoalkane complexes Cp2Ti(DEDM) (3)
We favor a mechanism (Scheme 11) in which the rate-limiting
step involves rearrangement to lose dinitrogen and give carbene
(49) For examples of other structurally characterized Ti diazoalkane
complexes, see: Kool, L. B.; Rausch, M. D.; Alt, H. G.; Herberhold, M.;
Hill, A. F.; Thewalt, U.; Wolf, B. J. Chem. Soc., Chem. Commun. 1986,
408. See also ref 15.

6324 J. Am. Chem. Soc., Vol. 120, No. 25, 1998
Polse et al.
Scheme 11
free 2 and is consistent with our supposition that dinitrogen loss
requires a vacant coordination site at titanium (vide supra). We
have also attempted to induce dinitrogen loss photochemically
at low temperatures. Irradiation of 2 through Pyrex glass at
-10 °C alone or in the presence of pyridine gave no reaction.
Irradiation of 2 through quartz led to decomposition to an
intractable mixture of products.
Reactions of 2 that Retain Nitrogen. Azametallacycle
Formation. As mentioned in the Results section, when treated
with terminal alkynes, 2 undergoes cyclization reactions that
proceed with retention of N2 to give azametallacyclobutene
complexes. Hillhouse and co-workers have previously described
a zirconium azametallacyclobutene complex in which the
fragments are assembled in reverse order.52 The alkyne complex
Cp*2Zr(PhCtCPh) reacts with diphenyldiazomethane to form
the metallacycle shown in eq 4. The cyclization reactions with
alkynes are quite similar to those observed for group IV imido
and oxo complexes.53-63 In particular, the reaction of 2 with
acetylene to form 14 is similar to the reaction of the imido
complex 21 with acetylene to form metallacycle 15 (Scheme
7).34 In addition, the reactions of 2 with Lewis bases described
above are analogous to the reaction of transient group IV imido
complexes with Lewis bases to form stable adducts.55,56,65 This
suggests that at least in solution, there is some imide-like
character to the bonding in 2, which can be illustrated as shown:
complex intermediate 26. Complex 26 can then either react
with the C-H bond of the Cp* ligand to give 4, or it can
undergo a [2 + 2] cycloaddition with alkenes to give metalla-
cyclobutanes. The reaction of 26 with a Cp* methyl group to
give 4 is precisely analogous to the Cp* C-H activation in
transient Cp*2TidCH2 reported by Bercaw (Scheme 2).24 This
mechanism is consistent with the lack of rate dependence on
alkene concentration, and it also accounts for the observed
regiochemistry of the metallacyclobutanes. Approach of the
R-olefin to the carbene in a manner that would place the
substituent R to the metal center (transition state A, Scheme
11) is disfavored due to unfavorable steric interactions with the
Cp* rings. Similarly, the olefin approaches in such a way as
to minimize contact between the olefinic substituent and the
trimethylsilyl group of the carbene (transition state B, Scheme
11), and therefore the trans-R,â-disubstituted isomer is formed.
There are several possibilities for the structure of the transition
state leading to dinitrogen loss from 2. One possibility that is
often invoked in the reactivity of diazoalkanes to form metal
carbene complexes is rearrangement to an intermediate in which
the diazoalkane coordinates through the CdN rather than the
NdN bond. Then, in a reaction that is analogous to release of
CO from metal ketene complexes, the dinitrogen is extruded to
form carbene intermediate 26. Another possibility involves
coordination of carbon to titanium in a four-membered ring that
then loses dinitrogen. The latter pathway would be consistent
with the lack of a solvent effect on the reaction, since it
possesses no formal charge separation. These possible transition
states must be treated as conjecture, since we have little direct
information on the dinitrogen extrusion step. A detailed study
of substituent effects on dinitrogen extrusion from aryl-
substituted diazoalkane complexes is currently in progress.32
Attempts to trap carbene intermediate 26 with the TidC bond
intact have not yet been successful.51 Bases such as pyridine
and DMAP add to 2 to give adducts 11 and 12 which decompose
to 4 (Scheme 4). Interestingly, the rate of formation of 4 is
markedly slower in the presence of pyridine or DMAP,
undoubtedly because 2 is “tied up” by formation of substantial
amounts of 11 or 12 as a result of the rapid equilibrium between
11 or 12 and 2. This suggests that 11 and 12 decompose via
The reaction of 2 with phenylacetylene yields azametallacy-
clobutene complex 17 (Scheme 6). Despite its similarity to 14,
the formation of 17 was somewhat unexpected, since the
analogous reaction of imido complex 21 with phenylacetylene
yields the amide-acetylide rather than the metallacyclobutene
product (Scheme 12).34 Oxotitanium complex 20 reacts with a
variety of terminal alkynes to give the corresponding oxamet-
allacyclobutene complexes as shown in Scheme 12.36 The
oxametallacyclobutene complexes rearranged to the more
thermodynamically stable hydroxo acetylide complexes. To
determine whether it would undergo a similar rearrangement, a
benzene-d6 solution of 17 was heated to 45 °C for 7 days, during
(52) Vaughan, G. A.; Hillhouse, G. L.; Rheingold, A. L. J. Am. Chem.
Soc. 1990, 112, 7994-8001.
(53) Meyer, K. E.; Walsh, P. J.; Bergman, R. G. J. Am. Chem. Soc. 1995,
117, 3749.
(54) Walsh, P. J.; Baranger, A. M.; Bergman, R. G. J. Am. Chem. Soc.
1992, 114, 1708.
(55) Mountford, P. Chem. Commun. 1997, 2127.
(56) Walsh, P. J.; Hollander, F. J.; Bergman, R. G. J. Am. Chem. Soc.
1988, 110, 8729.
(57) Housmekerides, C. E.; Ramage, D. L.; Kretz, C. M.; Shontz, J. T.;
Pilato, R. S.; Geoffroy, G. L.; Rheingold, A. L.; Haggerty, B. S. Inorg.
Chem. 1992, 31, 4453.
(58) Carney, M. J.; Walsh, P. J.; Hollander, F. J.; Bergman, R. G. J.
Am. Chem. Soc. 1989, 111, 8751.
(59) Carney, M. J.; Walsh, P. J.; Hollander, F. J.; Bergman, R. G.
Organometallics 1992, 11, 761.
(60) Carney, M. J.; Walsh, P. J.; Bergman, R. G. J. Am. Chem. Soc.
1990, 112, 6426.
(61) Walsh, P. J.; Hollander, F. J.; Bergman, R. G. Organometallics 1993,
12, 3705.
(62) Meyer, K. E.; Walsh, P. J.; Bergman, R. G. J. Am. Chem. Soc. 1994,
116, 2669.
(63) Meyer, K. E.; Walsh, P. J.; Bergman, R. G. J. Am. Chem. Soc. 1995,
117, 974.
(50) Cowie, M.; McKeer, I. R.; Loeb, S. J.; Gauthier, M. D. Organo-
metallics 1986, 5, 860.
(51) Petasis, N. A.; Staszewski, J. P.; Fu, D. K. Tetrahedron Lett. 1995,
36, 3619.
(64) Bennett, J. L.; Wolczanski, P. T. J. Am. Chem. Soc. 1994, 116,
2179.
(65) Cummins, C. C.; Baxter, S. M.; Wolczanski, P. T. J. Am. Chem.
Soc. 1988, 110, 8731.

η2-N2-Titanium Diazoalkane Complex
J. Am. Chem. Soc., Vol. 120, No. 25, 1998 6325
Scheme 12
and trimethylsilyldiazomethane. Unlike most isolable diazoal-
kane complexes, 2 releases dinitrogen under mild conditions to
give Cp*FvTiCH2SiMe3 (4). Thermolysis of 2 in the presence
of R-olefins leads to the trans-R,â-disubstituted metallacyclobu-
tane complexes 6-9. The regiochemistry of the metallacy-
clobutane complexes was determined by a combination of one-
and two-dimensional NMR techniques and confirmed by X-ray
crystallography in the case of Cp*2Ti(CH(SiMe3)CH(Me)CH2)
(8). A kinetic study of the reaction of 2 with styrene supports
a mechanism involving a transient carbene complex which is
trapped by styrene to give the metallacyclobutane Cp*2Ti(CH-
(SiMe3)CH(Ph)CH2) (7). Addition of Lewis bases pyridine and
DMAP to 2 do not trap this transient carbene complex, but form
reversible adducts with the metal center which slows the
formation of 4. It seems likely that the open coordination site
in 2 is necessary for N2 loss to occur, and this may be a general
requirement for N2 loss from metal diazoalkane complexes.
We have also shown that 2 has a varied reactivity that
involves the retention of N2. The coordination mode of the
diazoalkane fragment may be changed by the addition of Lewis
bases. Terminal acetylenes and allene react with 2 to form
azametallacyclobutane complexes 14, 17, and 18. N2O formally
transfers an oxygen atom to 2 to form titanacycle 19. The
reactivity pattern of these reactions is similar to that exhibited
by complexes containing TidX multiple bonds, and suggests
that the bonding in 2 can be described as possessing both
imido-olefin-adduct components.
Scheme 13
which time it decomposed to an intractable mixture of products.
Apparently, small changes in the nature of the titanium-
heteroatom bond profoundly influence the relative stability of
the isomeric metallacyclobutene and acetylide complexes.
Diazoalkane complex 2 also forms an azametallacyclobutane
complex (18) when treated with allene; the product is stable up
to 45 °C. In contrast, when the imido complex 21 is treated
with allene, a metallacycle is observed spectroscopically,66 but
is not stable to isolation. Oxotitanium complex 20 reacts with
allene to give an analogous oxametallacyclobutane complex (27,
eq 7).67 Complex 27 shows regiochemistry identical to that of
Experimental Section
General Methods. Unless otherwise noted, all reactions and
manipulations were carried out in dry glassware under a nitrogen or
argon atmosphere at 20 °C in a Vacuum Atmospheres 553-2 drybox
equipped with a MO-40-2 inert gas purifier, or using standard Schlenk
techniques. The amount of O2 in the drybox atmosphere was monitored
with a Teledyne model no. 316 trace oxygen analyzer. Some reactions
were carried out in thick-walled glass vessels fused to vacuum
stopcocks; these are referred to in the procedures as glass bombs. The
other instrumentation and general procedures used have been described
previously.53
Unless otherwise specified, all reagents were purchased from
commercial suppliers and used without further purification. Dimethy-
oxybenzene and 4-(dimethylamino)pyridine (DMAP) were sublimed
prior to use. Styrene was stirred over activated 4 Å molecular sieves,
vacuum distilled, and then stored at -50 °C in contact with activated
4 Å molecular sieves. Pyridine was distilled from CaH2 or sodium
under nitrogen and was stored in contact with activated 4 Å molecular
sieves. Trimethylphosphine was distilled from sodium just prior to
use. Acetylene gas was purified by passage through two -78 °C traps
separated by a concentrated H2SO4 trap. Pentane and hexanes (UV
grade, alkene free) were distilled from sodium benzophenone ketyl/
tetraglyme under nitrogen. Cyclohexane was distilled from calcium
hydride under nitrogen. Deuterated solvents for NMR experiments were
dried in the same way as their protiated analogues but were vacuum
transferred from the drying agent. Cp*2Ti(C2H4) was prepared by the
literature method,21 except that Cp*2TiCl was used instead of Cp*2-
TiCl2.68
18 (the exocyclic double bond is â to the metal center), and in
the solid state 18 shows a slight pucker of the metallacycle ring.
However, flipping of the ring is fast in solution, as demonstrated
1
by a low-temperature H NMR study which showed singlets
for the Cp* and methylene resonances down to -70 °C as is
true for complex 27.67
Treatment of diazoalkane complex 2 with N2O yields
metallacycle 19, where N2O has formally transferred an oxygen
atom, and a hydrogen shifts from carbon to nitrogen (Scheme
8). Hillhouse and co-workers observed that the alkyne complex
Cp*2Zr(η2-PhCtPh) reacts with N2O to affect formal transfer
of the oxygen atom and the intermediate demonstrating insertion
of N2O into the Zr-C(alkyne) bond was isolated (Scheme 13).48
We observe no intermediate in the formation of 19, but it seems
reasonable that N2O inserts before the extrusion of N2 as in the
zirconium system. Also, as mentioned in the Results section,
complex 19 can also be formed by the reaction of oxotitanium
complex 20 with (trimethylsilyl)diazomethane. To obtain the
same complex from the reaction with 2 and N2O, the released
N2 almost certainly comes from the N2O fragment.
NMR Spectroscopy. NMR experiments were performed on a
Brucker AMX spectrometer resonating at 300.13 MHz for 1H and 75.42
MHz for 13C that was equipped with an inverse probe. One-dimensional
13C{1H} spectra and DEPT spectra were recorded on a Brucker AMX
spectrometer resonating at 400 MHz for 1H and 100 MHz for 13C that
Conclusions
1
was equipped with a QNP probe. 15N-filtered H NMR spectra were
We have synthesized a novel side-bound diazoalkane complex
of titanium, Cp*2Ti(N2CHSiMe3) (2), from Cp*2Ti(C2H4) (1)
obtained using a pulse sequence similar to that described by Griffey,69
(66) Polse, J. L.; Bergman, R. G. Polse, J. L.; Bergman, R. G.
Unpublished results.
(67) Schwartz, D. J.; Smith, M. R., III.; Andersen, R. A. Organometallics
1996, 15, 1446.
(68) Pattiasina, J. W.; Heeres, H. J.; Von Bolhuis, F.; Meetsma, A.;
Teuben, J. H. Organometallics 1987, 6, 1004.
(69) Bax, A.; Griffey, R. H.; Hawkins, B. L. J. Magn. Reson. 1988, 55,
301.

6326 J. Am. Chem. Soc., Vol. 120, No. 25, 1998
Polse et al.
1
modified to give only the one-dimensional H NMR spectrum. All
The ether solution was reduced to 2 mL under vacuum and cooled to
-50 °C to yield red crystals of 8 (54.5 mg, 78.8%). IR (cyclohex-
ane): 3001(m), 2715 (w), 1994 (m), 1378 (s), 1243 (s), 1099 (s), 1072
two-dimensional experiments were acquired between 295 and 300 K.
NOESY spectra were acquired in phase-sensitive mode using the
Brucker pulse program noesytp. A shifted sine-bell window function
was applied to the raw data set in both dimensions.
(m), 1025 (m), 950 (m), 894 (m), 827 (s), 711 (s), 588 cm-1 1H NMR
.
(C6D6): δ 0.19 (m, 1H), 0.28 (s, 9H), 1.18 (d, 3J ) 6.0 Hz, 3H), 1.31
(m, 2H), 1.52 (m, 1H), 1.71 (s, 15H), 1.77 (s, 15H) ppm. 13C{1H}
NMR (C6D6): δ 5.4 (CH3), 12.2 (CH3), 12.5 (CH3), 14.4 (CH), 22.7
(CH3), 75.0 (CH), 77.8 (CH2), 118.7 (C), 118.9 (C) ppm. Anal. Calcd
for C27H46SiTi: C, 72.59; H, 10.40. Found: C, 72.64; H, 10.64.
Cp*2Ti(CHSiMe3CHEtCH2) (9). A glass bomb was charged with
crystals of 2 (51.7 mg, 0.120 mmol) and 8 mL of C6H6. The solution
was frozen (-195 °C) and degassed under vacuum. 1-Butene (0.362
mmol) was condensed onto the frozen solution from a 29.0-mL bulb.
The bomb was immersed in a 45 °C bath for 3 h, during which time
the solution turned from green to red. The solution was stirred an
additional 12 h at 25 °C, and the volatile materials were removed under
vacuum. The resulting red powder was extracted into diethyl ether
and filtered. The volume of the filtrate was reduced to 1 mL under
vacuum, and the solution was cooled to -50 °C for 12 h to yield red
crystals of 9 (52.0 mg, 94.4%). IR (Nujol): 2721 (w), 1240 (s), 1099
(w), 1070 (w), 1018 (m), 929 (m), 898 (m), 848 (s), 825 (s), 744 (w),
Cp*2TiN2CHSiMe3 (2). (Trimethylsilyl)diazomethane (300 µL of
a 2 M solution in hexanes, 0.600 mmol) was added via syringe to a
stirred solution of 1 (209 mg, 0.602 mmol) in C6H6 (10 mL). Gas
evolution ensued, and the solution turned from lime to forest green
within 30 s. The solution was stirred for 10 min, and the volatile
materials were removed under vacuum. The dark green powder was
extracted into hexanes and filtered. The volume of the filtrate was
reduced to 3 mL, and cooled to -80 °C to yield 2 as dark green blocky
crystals (169 mg, 65%). IR (Nujol): 2723 (w), 1894 (br,w), 1324 (m),
1232 (m), 1091 (m), 1022 (m), 848 (s), 703 (m) cm-1
.
1H NMR
(C6D6): δ 0.44 (s, 9H), 1.71 (s, 30H), 4.11 (s, 1H) ppm. 13C{1H}
NMR (C6D6): δ -0.35 (CH3), 11.35 (CH3), 99.9 (CH), 122.2 (C) ppm.
MS-EI m/z ) 433 [M+]. Anal. Calcd for C24H40N2SiTi: C, 66.62; H,
9.34; N, 6.48. Found: C, 66.31; H, 9.30; N, 6.65.
Cp*FvTiCH2SiMe3 (4). A glass bomb was charged with crystals
of 2 (74.3 mg, 0.172 mmol) and 10 mL of C6H6. The solution was
stirred for 2 d at 25 °C, during which time it turned from forest green
to blue-green. The volatile materials were removed under vacuum and
the blue-green powder was extracted into hexanes and filtered. The
volume of the filtrate was reduced to 1 mL under vacuum and cooled
to -50 °C to yield aquamarine crystals of 4 (58.7 mg, 84.3%). IR
715 (w), 665 (m) cm-1
.
1H NMR (C6D6) δ -0.59 (m, 1H), 0.27 (s,
3
9H), 0.73 (m, 1H), 1.14 (m, 1H), 1.15 (t, J ) 7.26 Hz, 3H), 1.40 (d,
3J ) 13.1 Hz, 1H), 1.51 (m, 1H), 1.71 (s, 15H), 1.77 (s, 15H), 1.92
(m, 1H) ppm. 13C{1H} NMR (C6D6) δ 5.4 (CH3), 12.2 (CH3), 12.5
(CH3), 13.9 (CH3), 22.1 (CH), 30.5 (CH2), 74.4 (CH), 74.9 (CH2), 118.7
(C), 118.9 (C) ppm. Anal. Calcd for C28H48SiTi: C, 72.99; H, 10.52.
Found: C, 72.72; H, 10.53.
(Nujol): 2721 (w), 1238 (m), 1076 (w), 1022 (m), 881(m), 847 (s),
787 (s), 717 (m), 667 (m) cm-1
.
1H NMR (C6D6): δ -1.29 (d, J )
2
2
Kinetic Study of Formation of 7 from 2 and Styrene. Stock
solutions for runs performed at a constant styrene concentration were
prepared as follows. An oven-dried 2-mL volumetric flask was charged
with dimethoxybenzene (10.6 mg, 0.0771 mmol) and crystals of 2 (28.9
mg, 0.0668 mmol). A small amount of toluene-d8 was added, followed
by 250 µL (2.18 mmol) of styrene. The total volume was brought to
2 mL with toluene-d8. The stock solution was stored at -50 °C until
just prior to sample preparation. Samples were prepared by transferring
0.5 mL of the stock solution to an oven-dried J-Young NMR tube.
Stock solutions for runs performed with variable styrene concentra-
tions were prepared as follows. An oven-dried 2-mL volumetric flask
was charged with dimethoxybenzene (9.0 mg, 0.0651 mmol) and
crystals of 2 (28.9 mg, 0.0668 mmol). The total volume was brought
to 2 mL with toluene-d8. The stock solution was stored at -50 °C
until just prior to sample preparation. Samples were prepared by
transferring 0.5 mL of the stock solution to an oven-dried J-Young
NMR tube. The appropriate volume of styrene was added to each NMR
tube using a gastight syringe. The solution for the kinetic study of the
formation of 4 from 2 was prepared identically, except that no styrene
was added.
The samples were lowered into the preheated probe of a Brucker
AMX 300 spectrometer. The temperature in the probe was calibrated
by measuring the peak separation of a sample of neat ethylene glycol.70
Spectra were acquired periodically. To avoid any complications due
to different relaxation times for the various resonances, spectra were
recorded using one π/2 pulse. An automatic baseline correction was
applied to the data. Integrals were placed manually but not phased.
Absolute concentrations of each reagent were determined based on
integration relative to the dimethoxybenzene resonance.
Kinetic Study of Formation of 4 and 7 from 2 in THF-d8. The
stock solution for runs performed in THF-d8 was prepared as follows.
An oven-dried 1-mL volumetric flask was charged with dimethoxy-
benzene (5.2 mg, 0.0376 mmol) and crystals of 2 (14.6 mg, 0.0337
mmol). The total volume was brought to 1 mL with THF-d8. To an
oven-dried J-Young NMR tube was transferred 0.5 mL of the stock
solution, followed by 29 µL (0.253 mmol) of styrene. The remaining
0.5 mL of the stock solution was transferred to a second J-Young NMR
tube and frozen (-78 °C) until just prior to the kinetic study of the
formation of 4 from 2, and the spectra were obtained as described above.
Reaction of 2 with Pyridine. Generation of 11. An NMR tube
was charged with a solution of 2 (8.5 mg, 0.0197 mmol) in 0.5 mL of
10.8 Hz, 1H), 0.15 (s, 9H), 0.21 (d, J ) 10.8 Hz, 1H), 1.26 (s, 3H),
2
1.29 (d, J ) 4.0 Hz, 1H), 1.41 (s, 3H), 1.80 (s, 15H), 1.83 (s, 3H),
2.02 (d, 2J ) 4.0 Hz, 1H), 2.20 (s, 3H). 13C{1H} NMR (C6D6): δ 6.1
(CH3), 10.8 (CH3), 11.4 (CH3), 12.3 (CH3), 12.5 (CH3), 15.6 (CH3),
45.8 (CH2), 76.1 (CH2), 118.9 (C), 120.0 (C), 125.1 (C), 125.9 (C),
126.6 (C), 130.1 (C) ppm. MS-EI m/z ) 405 [M+]. Anal. Calcd for
C24H40SiTi: C, 71.24; H, 9.98. Found: C, 71.18; H, 10.24.
Cp*2Ti(CHSiMe3CH2CH2) (6). A glass bomb was charged with
1 (111 mg, 0.320 mmol) in 10 mL of C6H6. (Trimethylsilyl)-
diazomethane (160 µL of a 2 M solution in hexanes) was added using
a syringe. The bomb was quickly closed. The solution turned from
lime to dark green immediately. The bomb was immersed in a 45 °C
bath for 6 h, during which time the green solution gradually turned
bright red. The volatile materials were removed under vacuum, and
the red oily residue was extracted into diethyl ether and cooled to -50
°C for 12 h to yield dark red crystals of 6 (84.4 mg, 60.9%). IR
(Nujol): 2721 (m), 2049 (w), 1240 (s), 1020 (m), 887 (m), 829 (s),
665 (m) cm-1 1H NMR (C6D6): δ -0.53 (m, 1H), 0.28 (s, 9H), 1.06
.
(m, 1H), 1.54 (m, 1H), 1.70 (s, 15H), 1.73 (s, 15H), 1.77 (m, 2H) ppm.
13C{1H} (C6D6): δ 3.9 (CH3), 5.8 (CH2), 12.2 (CH3), 12.3 (CH3), 71.1
(CH2), 73.2 (CH), 116.9 (C), 117.3 (C) ppm. Anal. Calcd for C26H44-
SiTi: C, 72.17; H, 10.27. Found: C 71.78; H, 10.34.
Cp*2Ti(CHSiMe3CHPhCH2) (7). A solution of styrene (98.1 mg,
0.942 mmol) in C6H6 (2 mL) was added to a stirred solution of 2 (54.6
mg, 0.126 mmol) in C6H6 (10 mL). The solution was stirred for 48 h
at 25 °C, during which time it turned from green to red. The volatile
materials were removed under vacuum, and the resulting red powder
crystallized from diethyl ether at -50 °C to give red crystals of 7 (55.2
mg, 86.1%). 1H NMR (C6D6): δ 0.13 (s, 9H), 1.36 (m, 2H), 1.65 (m,
3
1H), 1.77 (s, 15H), 1.79 (s, 15H), 1.99 (d, J ) 14 Hz, 1H), 7.8 (m,
1H), 7.38 (m, 2H), 7.45 (m, 2H) ppm. 13C{1H} NMR (C6D6): δ 5.63
(CH3), 12.3 (CH3), 12.7 (CH3), 29.3 (CH), 62.0 (CH), 79.0 (CH2), 119.8
(C), 120.3 (C), 124.9 (CH), 128.3 (CH), 128.5 (CH), 149.9 (C) ppm.
Anal. Calcd for C32H48SiTi: C, 75.54; H, 9.53. Found: C,75.78; H,
9.51.
Cp*2Ti(CHSiMe3CHMeCH2) (8). A glass bomb was charged with
crystals of 2 (67.1 mg, 0.155 mmol) and 20 mL of C6H6. The solution
was frozen (-195 °C) and degassed under vacuum. Propene (0.479
mmol) was condensed onto the frozen solution from a 138-mL bulb.
The bomb was immersed in a 45 °C bath for 5 h, during which time
the solution turned from green to red. The solution was then stirred at
room temperature for 2 d. The volatile materials were removed under
vacuum, and the resulting red powder was extracted into diethyl ether.
(70) Amman, C.; Meier, P.; Merbach, A. E. J. Magn. Reson. 1982, 46,
319.

η2-N2-Titanium Diazoalkane Complex
J. Am. Chem. Soc., Vol. 120, No. 25, 1998 6327
1
toluene-d8. An initial H NMR spectrum was recorded, and pyridine
mL under vacuum and cooled to -50 °C to yield red microcrystals of
19 (48 mg, 54.0%). 1H NMR (C6D6): δ 8.09 (br, 1H), 1.74 (s, 30H),
0.46 (s, 9H) ppm. 13C{1H} (C6D6): δ 177.1 (C), 122.1 (C), 11.7 (CH3),
-0.4 (CH3) ppm. IR (Nujol): 3290 (w), 2214 (m), 2197 (m), 1959
(w), 1954 (w), 1948 (w), 1938 (w), 1917 (m), 1902 (m), 1514 (w),
1414 (w), 1402 (w), 1371 (m), 1321 (w), 1277 (m), 1244 (m), 1176
(w), 1163 (w), 1020 (m), 991 (w), 976 (w), 953 (w), 914 (m), 841
(m), 808 (w), 752 (m), 723 (w), 692 (w), 656 (w), 631 (m) cm-1; MS-
EI m/z ) 448 [M+]. Anal. Calcd for C24H40N2OSiTi: C, 64.26; H,
8.99; N, 6.24. Found: C 64.10; H, 9.20; N, 5.86.
X-ray Crystallographic Study of Cp*2Ti(N2CHSiMe3) (2). Crys-
tals of 2 suitable for X-ray diffraction were obtained by slow cooling
of a hexanes solution of 2 to -50 °C. A fragment of one of these
green blocky crystals having dimensions of 0.20 × 0.23 × 0.28 mm
was mounted on a glass fiber using Paratone N hydrocarbon oil. All
measurements were made on a Siemens SMART (Siemens Industrial
Automation, Inc.) diffractometer with a CCD area detector using
graphite monochromated Mo KR radiation. All calculations were
performed using the TEXSAN crystallographic software package of
Molecular Structure Corporation.
Cell constants and an orientation matrix obtained from a least-squares
refinement using the measured positions of 7153 reflections in the range
3.00 < 2θ < 45.00° corresponded to a primitive monoclinic cell with
dimensions given as Supporting Information. The final cell parameters
and specific data collection parameters for this data set are given in
Table 1 or as Supporting Information. The systematic absences of
h0l: h + l * 2n and 0k0: k * 2n uniquely determined the space group
to be P21/n (#14).
Data were integrated to a maximum 2θ value of 46.5 ° using the
program SAINT with box parameters of 1.6 × 1.6 × 0.6°. The data
were corrected for Lorentz and polarization effects. No decay or
absorption correction was applied. The 9838 integrated reflections were
(16 µL, 0.198 mmol) was added via a syringe. The solution turned
from green to reddish-orange upon addition of the pyridine. A 1H NMR
spectrum was recorded that showed a 55:45 mixture of 2 and 11. 1H
NMR spectral data for 11: δ 0.377 (s, 9H), 1.70 (s, 30H), 6.21 (s, 1H)
ppm. The resonances for free and bound pyridine were coalesced, and
gave broad peaks at δ 6.70, 6.99, and 8.47 ppm.
Reaction of 2 with DMAP. Generation of 12. An NMR tube
was charged with a solution of 2 (4.6 mg, 0.011 mmol) in 0.3 mL of
C6D6. DMAP was added to the tube as a solution in 0.2 mL of C6D6.
The solution turned from green to purple upon addition of the DMAP.
The 1H NMR spectrum of the purple solution shows that it is a mixture
of 2 and 12. 1H NMR spectral data for 12: δ 0.534 (s, 9H), 1.93 (s,
30 H), 6.40 (s, 1H) ppm. The resonances for free and bound DMAP
were coalesced, and gave broad peaks at δ 2.20, 6.10, and 8.46 ppm.
Cp*2Ti(N2C(H)(SiMe3)CHCH) (14). A Schlenk flask was charged
with 2 (200 mg, 0.462 mmol) in 10 mL of C6H6. Acetylene gas was
bubbled through the solution for 10 min. The solution turned from
dark green to red immediately, and the formation of polyacetylene was
observed. The volatile materials were removed under vacuum, and
the red oily residue was extracted into hexanes and cooled to -50 °C
for 12 h to yield red crystals of 14 (116 mg, 54%). IR (Nujol): 1618
(w), 1606 (w), 1491 (m), 1448 (m), 1441 (m), 1433 (m), 1421 (m),
1414 (m), 1402 (m), 1298 (m), 1242 (m), 1227 (m), 1176 (m), 1132
(w), 1020 (m), 858 (m), 833 (m), 804 (m), 735 (m), 663 (w), 613 (m)
cm-1 1H NMR (C6D6): δ 7.29 (d, J ) 9.4 Hz, 1H), 6.52 (d, J ) 9.4
.
Hz, 1H), 6.33 (s, 1H), 1.74 (s, 30H), 0.36 (s, 9H) ppm. 13C{1H}
(C6D6): δ 180.2 (CH), 133.3 (CH), 121.8 (C), 104.5 (CH), 12.4 (CH3),
-0.7 (CH3) ppm. MS-EI m/z ) 458 [M+]. HRMS (EI) m/z calcd for
C26H42N2SiTi: calcd 458.2597, obsd 458.2589.
Cp*2Ti(N(NdC(H)SiMe3)C(Ph)CH) (17). Phenylacetylene (98.6
mg, 0.965 mmol) was added to a stirred solution of 2 (417.6 mg, 0.965
mmol) in C6H6 (10 mL). The solution turned red upon addition and
was stirred for 15 min at 25 °C. The solution was concentrated to 5
mL and left at 25 °C for 5 d to yield red-black crystals of 17 (331 mg,
64%). IR (Nujol): 3062 (m), 3037 (m), 1591 (m), 1375 (s), 1323 (w),
1300 (w), 1265 (m), 1232 (s), 1171 (w), 1149 (w), 1097 (w) cm-1. 1H
NMR (C6D6): δ 8.05 (d, 2 H), 7.34 (t, 2 H), 7.13 (t, 1 H), 6.34 (s, 1
H), 2.99 (s, 1 H), 1.68 (s, 30 H), 0.51 (s, 9 H) ppm. 13C{1H} NMR
(C6D6): δ 192.3 (CH), 141.6 (C), 140.6 (C), 130.8 (CH), 128.1 (CH),
126.0 (CH), 125.5 (CH), 124.2 (C), 12.3 (CH3), -0.96 (CH3) ppm.
MS-EI m/z ) 534 [M+]. Anal. Calcd for C32H46N2SiTi: C, 71.88; H,
8.67; N, 5.24. Found: C, 71.73; H, 8.81; N, 5.06.
Cp*2Ti(NNdCHSiMe3)C(CH2)CH2 (18). A 50-mL glass bomb
was charged with crystals of 2 (119 mg, 0.275 mmol) and 10 mL of
toluene. The solution was frozen (-195 °C) and degassed under
vacuum. Allene (0.275 mmol) was condensed onto the frozen solution
from a 66-mL bulb. Upon warming to room temperature the solution
turned from dark green to bright red, and the solution was stirred for
30 min. The volatile materials were removed under vacuum and the
resulting red powder was extracted into hexanes. The volume of the
hexanes solution was reduced to 3 mL under vacuum and cooled to
-50 °C to yield red blocky crystals of 18 (98 mg, 77.8%). 1H NMR
(C6D6): δ 6.23 (s, 1H), 4.77 (d, J ) 1.9 Hz, 1H), 3.78 (d, J ) 1.9 Hz,
1H), 2.62 (s, 2H), 1.77 (s, 30H), 0.28 (s, 9H) ppm. 13C{1H} (C6D6):
δ 138.4 (C), 137.6 (C), 124.3 (C), 72.8 (CH2), 66.7 (CH2), 12.3 (CH3),
-1.2 (CH3) ppm. IR (Nujol): 1902 (w), 1576 (m), 1566 (w), 1512
(m), 1502 (m), 1493 (m), 1481 (m), 1466 (m), 1460 (m), 1456 (m),
1441 (m), 1425 (m), 1412 (m), 1402 (w), 1371 (m), 1348 (w), 1290
(m), 12464 (m), 1107 (m), 1022 (w), 1003 (w), 887 (m), 864 (m), 841
(m), 825 (m), 806 (m), 744 (m), 727 (m), 688 (w), 667 (w), 617 (m)
cm-1; MS-EI m/z ) 472 [M+]. Anal. Calcd for C27H44N2SiTi: C,
68.61; H, 9.38; N, 5.93. Found: C, 68.79; H, 9.54; N, 5.62.
averaged to 3629 unique data (including systematic absences) (Rint
)
0.068). The structure was solved by direct methods and expanded using
Fourier techniques. All non-hydrogen atoms were refined anisotropi-
cally, and hydrogen atoms were refined with isotropic thermal
parameters. A correction for secondary extinction was applied to the
data in the last cycles of least-squares (coefficient ) 1.27765 × 10
-6). One datum was removed from the least-squares refinement due
to evidence of multiple diffraction. The final cycle of full-matrix least-
squares refinement was based on 2904 observed reflections (I > 3.00σ-
(I)) and 414 variable parameters and converged (largest parameter shift
was 0.01 times its esd) with unweighted and weighted agreement factors
of R ) 0.045 and Rw ) 0.058. The standard deviation of an observation
of unit weight was 2.23. The weighting scheme was based on counting
statistics and included a factor (p ) 0.030) to downweight the intense
reflections. The maximum and minimum peaks on the final difference
Fourier map corresponded to 0.28 and -0.33 e-/Å3 respectively.
Selected bond lengths and angles are given in Table 2, and an ORTEP
is shown in Figure 1. Positional and anisotropic thermal parameters,
tables of bond lengths and angles, and complete crystal and data
collection parameters are provided as Supporting Information.
X-ray Crystal Structure of Cp*2Ti(CH(SiMe3)CH(Me)CH2) (8).
Red rodlike crystals of 8 suitable for X-ray diffraction were grown by
slow cooling of an ether solution of 8 to -50 °C. The crystal mounting
and data collection were performed as described for the X-ray
crystallographic analysis of 2. The specific data collection parameters
are given in Table 1 or as Supporting Information. Non-hydrogen atoms
were refined anisotropically. The hydrogen atom coordinates were
refined but their isotropic B’s were fixed. Selected bond lengths and
angles are given in Table 3, and an ORTEP is shown in Figure 3.
Positional and anisotropic thermal parameters, tables of bond lengths
and angles, and complete crystal and data collection parameters are
provided as Supporting Information.
Cp*2Ti(N(H)NdC(SiMe3)O) (19). A 60-mL glass bomb was
charged with crystals of 2 (116 mg, 0.267 mmol) and 10 mL of C6H6.
The solution was frozen (-195 °C) and degassed under vacuum. N2O
(0.267 mmol) was condensed onto the frozen solution from a 66-mL
bulb. Upon warming to room temperature the solution turned from
dark green to dark red, and the solution was stirred for 40 min. The
volatile materials were removed under vacuum and the resulting red
powder was extracted into ether. The ether solution was reduced to 8
X-ray Crystal Structure of Cp*2Ti(N(NdC(H)SiMe3)C(CH2)CH2)
(18). Crystals of 18 suitable for X-ray diffraction were obtained by
slow cooling of a hexanes solution of 18 to -50 °C. A red triangular-
shaped crystal having dimensions of 0.35 × 0.35 × 0.16 mm was
mounted on a glass fiber using Paratone N hydrocarbon oil. All
measurements were made on a Siemens SMART (Siemens Industrial
Automation, Inc.) diffractometer with a CCD area detector using

6328 J. Am. Chem. Soc., Vol. 120, No. 25, 1998
Polse et al.
graphite monochromated Mo KR radiation. All calculations were
performed using the TEXSAN crystallographic software package of
Molecular Structure Corporation.
weight was 2.15. The weighting scheme was based on counting
statistics and included a factor (p ) 0.030) to downweight the intense
reflections. The maximum and minimum peaks on the final difference
Fourier map corresponded to 0.26 and -0.32 e-/Å3 respectively.
Selected bond lengths and angles are given in Table 4, and an ORTEP
is shown in Figure 7. Positional and anisotropic thermal parameters,
tables of bond lengths and angles, and complete crystal and data
collection parameters are provided as Supporting Information.
Cell constants and an orientation matrix obtained from a least-squares
refinement using the measured positions of 5538 reflections in the range
3.00 < 2θ < 45.00° corresponded to a primitive triclinic cell with
dimensions given as Supporting Information. The final cell parameters
and specific data collection parameters for this data set are given in
Table 1 or as Supporting Information. On the basis of a statistical
analysis of intensity distribution, and the successful solution and
refinement of the structure, the space group was determined to be P1h
(#2).
Data were integrated to a maximum 2θ value of 51.9 ° using the
program SAINT with box parameters of 1.6 × 1.6 × 0.6°. The data
were corrected for Lorentz and polarization effects. No decay correction
was applied. An empirical absorption correction based on comparison
of redundant and equivalent data and an ellipsoidal model of the
absorption surface was applied using the program XPREP (Siemens
Industrial Automation, Inc.) where Tmax ) 0.91 and Tmin ) 0.88. The
7493 integrated and corrected reflections were averaged to 4601 unique
data (Rint ) 0.025). The structure was solved by direct methods and
expanded using Fourier techniques. All non-hydrogen atoms were
refined anisotropically, and hydrogen atoms were included but not
refined. The final cycle of full-matrix least-squares refinement was
based on 3941 observed reflections (I > 3.00σ(I)) and 280 variable
parameters and converged (largest parameter shift was 0.01 times its
esd) with unweighted and weighted agreement factors of R ) 0.039
and Rw ) 0.057. The standard deviation of an observation of unit
Acknowledgment. We thank the National Institutes of
Health (Grant No. GM-25459) and the Arthur C. Cope Fund
administered by the American Chemical Society for generous
financial support of this work. We thank Dr. F. J. Hollander,
director of the University of California Berkeley College of
Chemistry X-ray diffraction facility (CHEXRAY), and Dr. P.
Goodson for solving the crystal structures of 2, 8, and 18. This
paper is dedicated to Professor Warren T. Roper, for his many
ground-breaking contributions to organometallic chemistry, on
the occasion of his 60th birthday.
Supporting Information Available: Tables of positional
and thermal parameters and intramolecular bond lengths and
angles for the structures of 2, 8 and 18, NOESY spectrum of 9
(21 pages, print/PDF). See any current masthead page for
ordering information and Web access instructions.
JA974303D