Mechanism of insertion of carbodiimides into the Zr-C bonds of zirconaaziridines. Formation of α-amino amidines

Article abstract of DOI:10.1021/om000768s

Treatment of zirconaaziridines Cp₂Zr-η²-[N(R¹)CH(R²)](THF) with carbodiimides results in the insertion of the carbodiimide into the Zr-C bonds. The insertion of bis(trimethylsilyl)-carbodiimide is reversible, wh

Full text of DOI:10.1021/om000768s

254
Organometallics 2001, 20, 254-260
Mech a n ism of In ser tion of Ca r bod iim id es in to th e Zr -C
Bon d s of Zir con a a zir id in es. F or m a tion of r-Am in o
Am id in es
J on A. Tunge,† Curtis J . Czerwinski,‡ Daniel A. Gately,‡ and J ack R. Norton*,†
Department of Chemistry, Columbia University, New York, New York 10027, and
Department of Chemistry, Colorado State University, Fort Collins, Colorado 80523
Received September 6, 2000
Treatment of zirconaaziridines Cp2Zr-η2-[N(R1)CH(R2)](THF) with carbodiimides results
in the insertion of the carbodiimide into the Zr-C bonds. The insertion of bis(trimethylsilyl)-
carbodiimide is reversible, which becomes significant at high THF concentrations. Kinetic
data indicate that the THF ligand must dissociate prior to carbodiimide insertion. Protic
cleavage of the organic fragment from zirconium results in formation of R-amino amidines.
In tr od u ction
the potential pharmaceutical applications and utility in
peptide modification,10 few syntheses of R-substituted
R-amino amidines or iminopeptides have been devel-
oped.11,12 Most use the method of Mengelberg, which
utilizes N′-protected R-amino iminoethers as intermedi-
ates (eq 1).
Zirconocene η2-imine complexes, also known as zir-
conaaziridines, are the synthetic equivalents of R-amino
carbanions (Scheme 1).1-3 Previous investigations have
shown that heterocumulenes such as isocyanates, and
to some extent CO2, insert into the Zr-C bond of zir-
conaaziridines;2 likewise, cyclic carbonates insert into
zirconaaziridines, yielding R-amino acid esters after
treatment with methanol.2,3 These studies have led to
the current investigation of carbodiimide4 insertion,
which should yield the analogous R-amino amidines
(Scheme 1).5
R-Amino amidines can be found in the antitumor
antibiotic bleomycin6 and in the iminopeptide antibiotic
bottromycin7 and are intermediates in the biosynthesis
of purines.8 Furthermore, amino amidines have recently
attracted interest as NO synthase inhibitors.9 Despite
Herein we report that carbodiimides insert into the
Zr-C bond of zirconaaziridines, leading to R-amino
amidines 5 after protic cleavage (Scheme 2).
* Corresponding author. E-mail: jnorton@chem.columbia.edu.
Fax: 212-854-7660.
Resu lts
† Columbia University.
Syn th esis of r-Am in o Am id in es. Bis(trimethylsi-
lyl)carbodiimide inserts rapidly and quantitatively into
the Zr-C bonds of zirconaaziridines in aromatic sol-
vents. Addition of HCl/Et2O results in precipitation of
the hydrochlorides of desilylated R-aminoamidines, 5 (R1
) Ph, H; R2 ) Ar, Bn, Pr), in good yield (Table 1).
Similarly, 1,3-di-p-tolyl carbodiimide inserts into zir-
conaaziridine 1c to give 4c in ca. 80% yield.13 Although
no products derived from Zr-N insertion have been
‡ Colorado State University.
(1) Buchwald, S. L.; Watson, B. T.; Wannamaker, M. W.; Dewan, J .
C. J . Am. Chem. Soc. 1989, 111, 4486-4494. (b) Grossman, R. B.;
Davis, W. M.; Buchwald, S. L. J . Am. Chem. Soc. 1991, 113, 2321-
2322. (c) Coles, N.; Harris, M. C. J .; Whitby, R. J .; Blagg, J . Organo-
metallics 1994, 13, 190-199.
(2) Gately, D. A.; Norton, J . R.; Goodson, P. A. J . Am. Chem. Soc.
1995, 117, 986. (b) Gately, D. A. Norton, J . R. J . Am. Chem. Soc. 1996,
118, 3479.
(3) Tunge, J . A.; Gately, D. A.; Norton, J . R. J . Am. Chem. Soc. 1999,
121, 4520-4521.
(4) Williams, A.; Ibrahim, I. T. Chem. Rev. 1981, 81, 589. (b)
Mikolajczyk, M.; Kielbasinski, P. Tetrahedron 1981, 37, 233.
(5) For syntheses and pharmaceutical uses of amidines see: (a)
Patai, S. The Chemistry of Amidines and Imidates; J ohn Wiley and
Sons: New York, 1975. (b) Patai, S. The Chemistry of Amidines and
Imidates, Vol 2.; J ohn Wiley and Sons: New York, 1991.
(6) Umezawa, H.; Takita, T.; Sugiura, Y.; Otsuka, M.; Kobayashi,
S.; Ohno, M. Tetrahedron 1984, 40, 501-509.
(9) Patents include: (a) G.D. Searle & Co. PCT Int. Appl. WO99/
46240, 1999. (b) Merck & Co., Inc. Patent 5,629,322, 1997. (c) G.D.
Searle & Co. 9635677, 1996. WO95/11232; Fisons Co., Patent WO95/
05363, 1995. (d) G.D. Searle & Co. Patent WO95/11014, 1995.
(10) Sauve, G.; Rao, V. S.; Lajoie, G.; Belleau, B. Can. J . Chem. 1985,
63, 3089.
(7) Schipper, D.; Gist-Brocades, V.; Delft, N. J . Antibiot. 1983, 36,
1076-1077. (b) Yamada, T.; Takashima, K.; Miyazawa, T.; Kuwata,
S.; Watanabe, H. Bull. Chem. Soc. J pn. 1978, 51, 878-883. (c) Yamada,
T.; Miyazawa, T.; Kuwata, S.; Watanabe, H. Bull. Chem. Soc. J pn.
1977, 50, 1827-1830. (d) Yamada, T.; Suegane, K.; Kuwata, S.;
Watanabe, H. Bull. Chem. Soc. J pn. 1977, 50, 1088-1093.
(8) Korshin, E. E.; Soboleva, G. I.; Levin, Y. A.; Zakharova, L. G.;
Litvinov, I. A.; (b) Naumov, V. A.; Podval’nyi, E. A.; Efremov, Y. Y.
Zh. Org. Khim. 1992, 28, 1242-1256. (c) Goto, T.; Isobe, M.; Covielle,
D. A.; Kishi, Y.; Inoue, S. Tetrahedron 1973, 29, 2035-2039.
(11) Barluenga, J .; Aznar, F.; Liz, R. J . Chem. Soc., Chem. Comm.
1981, 1181. (b) Barluenga, J .; Aznar, F.; Liz, R. J . Chem. Soc., Perkin
Trans. 1 1983, 1093. (c) Mengelberg, M. Chem. Ber. 1960, 2230.
(12) For references related to glycine amidines see: (a) Korshin, E.
E.; Soboleva, G. I.; Levin, Y. A.; Podval’nyi, E. A.; Efremov, Y. Y. Zh.
Org. Khim. 1993, 29, 577-587. (b) Korshin, E. E.; Eliseenkova, R. M.;
Zyablikova, L. G.; Zakharov, L. G.; Akhmadullin, A. G.; Nusinovich,
V. D.; Levin, Y. A. Russ. Chem. Bull. 1993, 42, 1071-1077. (c) Ueda,
T.; Okamoto, Y.; Tsuji, T.; Muraoka, M. Chem. Pharm. Bull. 1968, 26,
2355-2361.
10.1021/om000768s CCC: $20.00 © 2001 American Chemical Society
Publication on Web 12/07/2000

Formation of R-Amino Amidines
Organometallics, Vol. 20, No. 2, 2001 255
Sch em e 1
Sch em e 2
Ta ble 1. Isola ted Yield s of r-Am in o Am id in esa
Mech an ism of Car bodiim ide In ser tion in to Zr -C
Bon d s. Insertion of carbodiimides into main group14
and early M-C15 bonds is an important method for the
preparation of metal amidinate olefin polymerization
catalysts.14-17 While much is known about the mecha-
nisms of insertion of olefins, acetylenes, and CO into
early transition metal-carbon bonds,18 far less is known
about the mechanism of insertion of heterocumulenes.19
Braunstein has suggested that “precoordination of het-
erocumulenes [is] necessary” prior to insertion into
Zr-C bonds.20 The proposal of an intermediate complex
R-amino amidine
R1
R2
R3
yield (%)
5a b
5b
5c
Ph
Ph
Ph
o-anisyl
Ph
H
p-tBuC5H4
p-tBuC5H4
Ph
Ph
n-Pr
H
91
53
35
59
74
85
p-tolyl
p-tolyl
p-tolyl
H
5d
5eb,c
5fb,c
CH2Ph
H
a
Prepared from a 1:1 mix of carbodiimide and zirconaaziridine
b
in C6H6 and cleaved with H2O (5b-d ) or HCl. Isolated as the
HCl salts. c Zirconaaziridines were generated in situ by the
thermolysis of Cp2Zr[N(R1)CH2(R2)]Me.
(14) Coles, M. P.; Swenson, D. C.; J ordan, R. F.; Young J r.; V. G.
Organometallics 1998, 17, 4042-4048. (b) Coles, M. P.; Swenson, D.
C.; J ordan, R. F. Organometallics 1997, 16, 5183-5194. (c) Chang, C.;
Hsiung, C.; Su, H,; Srinivas, B.; Chiang, M. Y.; Lee, G.; Wang, Y.
Organometallics 1998, 17, 1595-1601.
(15) J ayaratne, K. C.; Sita, L. R. J . Am. Chem. Soc. 2000, 122, 958-
959. (b) Koterwas, L. A.; Fetinger, J . C.; Sita, L. R. Organometallics
1999, 18, 4183. (c) Sita, L. R.; Babcock, J . R. Organometallics 1998,
17, 5228-5230.
(16) Volkis, V.; Shmulinson, M.; Averbuj, C.; Lisovskii, A,; Edel-
mann, F. T.; Eisen, M. S. Organometallics 1998, 17, 3155-3157. (b)
Averbuj, C.; Tish, E.; Eisen, M. S. J . Am. Chem. Soc. 1998, 120, 8640-
8646.
(17) Patents include: (a) J ordan, R. F.; Coles, M. P. Patent 5,973,-
088, 1999. (b) J ordan, R. F.; Coles, M. P. Patent 5,777,120, 1998. (c)
Schlund, R.; Lux, M.; Edelmann, F.; Reissmann, U.; Rhode, W. BASF
Patent 5,707,913, 1998. (d) Flores, J . C.; Rausch, M. D. Patent 5,502,-
128, 1996.
isolated, competitive Zr-N insertion is the likely reason
for lowered yield of 4c. Such a side reaction was
observed with isocyanate insertion and suppressed by
the use of o-anisyl substituents on nitrogen (e.g., 1d ).2
Indeed, treatment of 1d with di-p-tolyl carbodiimide
gives metallacycle 4d in >95% yield by 1H NMR.
Treatment of the p-tolyl-substituted metallacycles 4
with H2O liberates the amino amidines 5, which can be
purified by chromatography (Table 1). In addition to
isolable zirconaaziridines 1a -d , zirconaaziridines gen-
erated in situ from Cp2ZrMe(OTf) and lithium amides
(1e,f) are efficiently trapped by carbodiimides. These
results suggest that a wide variety of R-amino amidines
should be accessible by our method.
(18) Collman, J . P.; Hegedus, L. S.; Norton, J . R.; Finke, R. G.
Principles and Applications of Organotransition Metal Chemistry, 1st
ed.; University Science Books: Mill Valley, CA, 1987.
(19) Gambarotta, S.; Strologo, S.; Floriani, C.; Chiesi-Villa, A.;
Guastini, C. Inorg. Chem. 1985, 24, 654.
(13) NMR yields were determined by integration vs hexamethylcy-
clotrisiloxane internal standard.

256 Organometallics, Vol. 20, No. 2, 2001
Tunge et al.
Sch em e 3
F igu r e 1. Plot of 1/kapp vs [THF] for the reaction of 0.035
M 1a with 0.41 M NCN in solutions containing 0.54-1.34
M [THF] at 265 K.
3 raises the issue of whether ligand dissociation from 1
is required before such an electrophile can insert into
the Zr-C bond of zirconaaziridines (eq 2). To this end
we have investigated the mechanism of such insertions
by carbodiimides.
centration range plots of 1/kapp vs [THF] were linear
(Figure 1), suggesting that dissociation of THF is
required prior to insertion.
A dissociative mechanism such as path B in Scheme
3 can be described by two sets of simplifying kinetic
assumptions.21 The rate law in eq 3 is obtained if we
assume rapid equilibration of 1a and 2a followed by
rate-determining insertion of carbodiimide. If we as-
sume instead that 2a is highly reactive and thus does
not accumulate as the reaction progresses, the rate law
given is the steady-state one in eq 4. If k1,k-1[THF] .
k2
K1k2[NCN]([1] + [2])
[THF] + K1
d[1]
dt
-
)
(3)
where K1 ) k1/k-1. If (k-1[THF] + k2[NCN]) . k1
Treatment of an orange toluene solution of 1a with 1
equiv of bis(trimethylsilyl)carbodiimide (NCN) results
in quantitative formation of the dark purple complex
4a . The obvious mechanisms are A and B. (A) The five-
membered metallacycle 4a may be formed through an
associative mechanism in which 1a reacts directly with
carbodiimide (Scheme 3). (B) If heterocumulene coor-
dination is required before insertion, a coordinatively
unsaturated intermediate such as 2a may be necessary
(Scheme 3).
In principle, associative (path A) and dissociative
(path B) mechanisms can be distinguished by the
dependence of the reaction rate on [THF]. The rate of
associative insertion should be insensitive to [THF],
whereas the kinetics of a dissociative mechanism should
exhibit an inverse relation between the rate and [THF].
To determine whether carbodiimide insertion occurs
through an associative or dissociative mechanism, the
rates of insertion were measured with various THF
concentrations at 265 K. The disappearance of 1a ,
followed in the presence of >10-fold excess of NCN to
ensure pseudo-first-order conditions, gave the apparent
rate constants, kapp, at individual THF concentrations.
Varying [THF] from 0.54 to 1.34 M showed an inverse
relationship between kapp and [THF]. Within this con-
k1k2[NCN][1]
d[1]
-
)
(4)
dt
k-1[THF] + k2[NCN]
The two mechanisms can be distinguished by the
dependence of the reaction rates on carbodiimide con-
centration. Equation 4 predicts that at high [NCN] (k2-
[NCN] . k-1[THF]) the rate of insertion will approach
saturation (rate constant k1); however, preequilibrium
kinetics (eq 3) should not exhibit this behavior. At the
carbodiimide concentrations investigated (0.41-1.4 M)
the reaction is first order in [NCN], consistent with
either rate law and set of assumptions.
Kin etics a n d Equ ilibr iu m of THF Dissocia tion
fr om Zir con a a zir id in e 1a . To determine whether eq
3 or eq 4 is correct, we studied the kinetics and
equilibrium of THF dissociation from 1a . Previously we
had studied the kinetics of THF dissociation from 1c.2a
1
Using a similar procedure, H NMR spectra of 1a were
collected in a 1.0 M THF solution. These spectra show
separate resonances for free and coordinated THF below
273 K; at 230 K the THF resonances are no longer
broadened by exchange. No evidence for 2a is observed
(21) Espenson, J . H Chemical Kinetics and Reaction Mechanisms,
2nd ed.; McGraw-Hill: New York, 1995; pp 77-90.
(20) Braunstein, P.; Nobel, D. Chem. Rev. 1989, 89, 1927.

Formation of R-Amino Amidines
Organometallics, Vol. 20, No. 2, 2001 257
Sch em e 4
lower temperatures favor insertion of carbodiimide to
give 4a .
F igu r e 2. Plot of 1/kapp vs [THF] for the reaction of 0.037
M 1a with 0.41 M NCN in solutions containing 0.79-3.08
M [THF] at 265 K.
in these spectra or in others collected at low tempera-
ture in the absence of added THF. Thus the equilibrium
between THF-coordinated zirconaaziridine 1a and ligand-
free zirconaaziridine 2a favors the former. Line-
broadening data for the two separate R-proton reso-
nances of the coordinated THF ligand were collected
over a 250-270 K temperature range.22 Eyring analysis
gives ∆Hq ) 16.1(6) kcal/mol and ∆Sq ) 11(2) eu,
suggesting that THF exchange occurs through a dis-
sociative mechanism. Confirming this hypothesis, the
line broadening was found to be independent of [THF-
free]. Therefore, the line broadening is a direct measure
of the rate of THF dissociation, k1; for 1a , k1 at 265 K is
79 s-1. Since the equilibrium between 1a and 2a lies
far to the left (k-1[THF] . k1), and k1 is >4 orders of
magnitude larger than the apparent rates of carbodi-
imide insertion, 1a must be rapidly equilibrating with
small concentrations of 2a on the time scale of insertion,
so k1,k-1[THF] . k2[NCN] and eqs 3 and 4 are indis-
tinguishable.
Rever sibility of Ca r bod iim id e In ser tion . At THF
concentrations over 2 M significant, reproducible, cur-
vature is observed in the 1/kapp vs [THF] plots (Figure
2). This prompted investigation of the effect of high
[THF] on the reaction kinetics. When complex 1a was
treated with bis(trimethylsilyl)carbodiimide in 12 M
THF-d8, the reaction went to only 15% of completion,
suggesting that carbodiimide insertion is reversible.
Confirming this hypothesis, treatment of toluene solu-
tions of pure 4a with PMe3 or tert-BuNCO resulted in
carbodiimide extrusion and formation of new zirconium
products expected from 2a (Scheme 4).2
The thermodynamics of carbodiimide insertion reveal
that while the reaction goes to >98% completion during
the well-behaved kinetics obtained at low [THF] (Figure
1), reversibility of carbodiimide insertion becomes sig-
nificant at higher THF concentrations. For example, at
3.0 M THF, under the conditions used to generate the
curve in Figure 2 (Keq ) 150), the insertion proceeds to
only 95% completion. The curvature originally observed
at high [THF] resulted from treating the high [THF]
kinetics as irreversible rather than as an approach to
equilibrium. The equilibrium kinetic situation is pic-
tured in eq 5, and the rate law for approach to equilib-
rium is given in eq 6. Fitting the data in Figure 2 as an
approach to equilibrium using a floating endpoint
exponential21 results in the corrected curve shown in
Figure 3. The improved linear fit of the corrected data
indicates that carbodiimide extrusion, k-2, is significant
at high [THF].
k1k2[NCN] + k-1k-2[THF]
k-1[THF] + k2[NCN]
d[∆1]
dt
-
)
[∆1] (6)
To investigate the thermodynamics of carbodiimide
insertion, a solution prepared from 0.04 M 1a and 0.06
M NCN in 11 M THF was allowed to equilibrate at
An a lysis of Ca r bod iim id e In ser tion Da ta . The
investigation of THF dissociation has already shown
that k-1 . k2, so eq 6 can be simplified to give the kobs
in eq 7. From eq 7 it is apparent that plots of kobs vs
1/[THF] should be linear with a slope of K1k2[NCN] (K1
) k1/k-1) and an intercept of k-2. These plots are indeed
linear and give K1k2 ) 1.00(2) × 10-3 s-1 at 265 K and
K1k2 ) 2.03(5) × 10-3 s-1 at 273 K. Using the observed
K1k2 and the value of k1, we can determine that k-1/k2
) 7.9 × 104 at 265 K and 7.6 × 104 at 273 K, so
intermediate 2a is trapped much more rapidly by THF
1
273 K overnight and the H NMR spectrum recorded.
Integration of the solution species relative to an in-
ternal standard gave Keq ) 150 at 273 K. Monitoring
the equilibrium between 1a + NCN and 4a + THF (eq
5) as a function of temperature resulted in a linear
van’t Hoff plot, giving ∆H° ) -4.3(2) kcal/mol and
∆So ) -5.8(5) eu. Importantly, these data show that
(22) Sandstrom, J . Dynamic NMR Spectroscopy; Academic Press:
New York, 1982.

258 Organometallics, Vol. 20, No. 2, 2001
Tunge et al.
well known24 and has been exploited synthetically by
Takahashi.25 The activation energies of alkyne extrusion
are highly dependent on the size of the alkyne substit-
uents; trimethylsilylacetylenes are easily substituted by
various unsaturated compounds (eq 9).
F igu r e 3. (b) Data in Figure 2 (r2 ) 0.86). ([) Corrected
data obtained by exponential fits of kinetic traces using
floating infinity point (r2 ) 0.98).21
To test whether TMS substituents promote carbodi-
imide loss from 4, we have compared the equilibria for
bis(trimethylsilyl)carbodiimide and di-p-tolyl carbodi-
imide insertions into zirconaaziridine 1d . While di-p-
tolyl carbodiimide inserts quantitatively, the insertion
of bis(trimethylsilyl)carbodiimide is less than 5% at
equilibrium (eq 10), indicating the steric effect noted by
Takahashi may apply to carbodiimide insertion also.
than by carbodiimide. The values of k-2 cannot be
determined directly from the intercepts (which are too
close to zero to be determined with precision), but
dividing K1k2 by Keq gives k-2 as 5.3 × 10-6 s-1 at 265
K and 1.4 × 10-5 s-1 at 273 K.
k1k2[NCN] + k-1k-2[THF]
k-1[THF]
d[∆1]
dt
-
)
[∆1]
k1k2[NCN]
k-1[THF]
kobs
)
+ k-2 (7)
Kin etics of Ca r bod iim id e Extr u sion fr om Com -
p lex 4a . Treatment of 4a with 11 M THF in the absence
of free NCN (k2[NCN] , k-2) results in complete
conversion to 1a (eq 8). The disappearance of the 509
nm visible absorbance of complex 4a shows good first-
order behavior. At 298 K the rate of carbodiimide
extrusion to form 1a is 3.1 × 10-4 s-1 at 11 M THF and
3.3 × 10-4 s-1 at 6.7 M THF. This result suggests that
the reaction is zero order in [THF], so k-2 is rate limiting
and the observed rate of disappearance of 4a is equal
to k-2. The temperature dependence of k-2 shows good
Eyring behavior, giving ∆Hq ) 20.0(3) kcal/mol and ∆Sq
) -7(1) eu.23 Extrapolation of these data to 265 K gives
k-2 ) 5.6(4) × 10-6 s-1, in good agreement with that
calculated above from the kinetics of carbodiimide
insertion.
Con clu sion s. We have investigated the formation of
amino amidines from zirconaaziridines and carbodiim-
ides. A variety of R-amino amidines can be produced by
simply varying the reactant amine and the carbodiim-
ide. The rate of insertion of carbodiimides into the Zr-C
bonds of zirconaaziridines is inversely proportional to
[THF], so the insertion of carbodiimides occurs after
dissociation of THF from 1. At high [THF] the revers-
ibility of bis(trimethylsilyl))carbodiimide insertion be-
comes apparent, but the insertion of di-p-tolyl carbodi-
imide appears unaffected. The thermolysis of the amino
amidine metallacyle 4a can be used to generate the
ligand-free zirconaaziridine 2a , which can react with
other electrophiles. Braunstein’s suggestion20 that co-
(23) For an elementary step in which 4a forms 2a and free
carbodiimide, the entropy of activation would be expected to be positive.
The observed negative entropy of activation suggests that a coordina-
tion complex such as 3 (eq 2) may be the direct product of k-2, although
dissociative reactions can have negative entropies of activation.
(24) McDade, C.; Bercaw, J . E. J . Organomet. Chem. 1985, 279,
281-315. (b) Erker, G.; Dorf, U.; Rheingold, A. L. Organometallics
1988, 7, 138-143. (c) Grubbs, R. H.; Miyashita, A. J . Am. Chem. Soc.
1978, 100, 1300-1303.
(25) Takahashi, T.; Xi, C.; Xi, Z.; Kageyama, M.; Fischer, R.;
Nakajima, K.; Negishi, E. J . Org. Chem. 1998, 63, 6802-6806. (b)
Takahashi, T.; Kotora, M.; Hara, R.; Xi, Z. Bull. Chem. Soc. J pn. 1999,
72, 2591-2603, and references therein.
The process of carbodiimide extrusion is formally the
activation of the â-C-C bond of 4a . Cleavage of the
equivalent bond in five-membered zirconacarbocycles is

Formation of R-Amino Amidines
Organometallics, Vol. 20, No. 2, 2001 259
filtered. The lithium triflate was further washed with 2 × 5
mL of hexanes. Removal of the hexanes produced an orange
solid, which was dried under vacuum. Yield: 910 mg (91%).
1H NMR (C6D6): δ 7.21 (t, 1H), 6.92 (t, 2H), 6.67 (d, 2H), 5.67
(s, 10H), 3.10 (m, 2H), 1.32 (m, 2H) 1.16 (s, 2H), 0.81 (t, J )
7.2 Hz, 3H) 0.29 (s, 3H). 13C NMR (C6D6): δ 158.3, 128.6, 123.4,
ordination of heterocumulenes is a prerequisite for
insertion is supported by the fact that THF must
dissociate before carbodiimide insertion.
Exp er im en ta l Section
120.6, 110.3, 51.8, 32.2, 23.5, 20.9, 14.3. Anal. Calcd (C21H27
NZr): C, 65.57; H, 7.07; N, 3.64. Found: C, 65.23; H, 6.82; N,
3.13.
-
Ma ter ia ls. All air-sensitive compounds were prepared and
handled under a N2 atmosphere using standard Schlenk and
inert-atmosphere box techniques. Hexanes, benzene, THF, and
toluene were distilled under nitrogen from sodium benzophe-
none ketyl. Cp2ZrCl2, generously donated by Boulder Scientific,
was converted to Cp2ZrMe(OTf) as previously described.1,2
Cp2Zr[η2-N(Ph)CHPh]‚THF,2 Cp2Zr[η2-N(o-anisyl)CHPh],2 and
Cp 2Zr (TMSNCH2CH2P h )CH3 (1f). To a solution of 0.33
mL (2.6 mmol) of phenethylamine in 10 mL of Et2O at 0 °C
was added 1.6 mL of 1.6 M BuLi/hexane solution. After 10 min
this was followed by 0.33 mL of TMSCl. This mixture was
stirred for 10 min before addition of 1.6 mL of 1.6 M BuLi.
The resulting mixture was stirred 10 min, then added to a
solution of 1.0 g of Cp2ZrMeOTf (2.6 mmol) in 25 mL of THF
at -50 °C. After stirring for 1 h, the solution was warmed to
room temperature, and the solvent removed. A 15 mL portion
of hexanes was added and the solid filtered. The solid was
further washed with 2 × 5 mL of hexanes. Removal of the
hexanes produced a white solid, which was dried under
vacuum. Yield: 860 mg (78%). 1H NMR (C6D6): δ 7.21 (m, 5H),
5.75 (s, 10H), 2.75 (m, 2H), 2.41 (m, 2H), 0.35 (s, 3H), 0.19 (s,
9H). 13C NMR (C6D6): δ 139.8, 129.1, 128.3, 126.3, 110.9, 46.1,
42.7, 24.9, 3.5.
3
Cp2Zr(TMSNCH2CH2Ph)CH3 were prepared as previously
1
described. H NMR were recorded at 300 MHz, and 13C NMR
were recorded at 75 MHz unless otherwise noted. NMR
solution concentrations were determined from integration vs
a
hexamethylcyclotrisiloxane standard. All kinetics were
monitored for >3t1/2, and temperatures were measured using
a MeOH standard.26 Chromatography was done on a Chro-
matotron (Harrison Research Inc.) with silica gel (Merck, TLC
grade 7749) as the adsorbent.
N-P h en yl-4-ter t-bu tylben zyla m in e. Aniline (1.1 mL, 12
mmol) and ca. 1 g MgSO4 were added to a C6H6 (30 mL)
solution of 4-tert-butylbenzaldehyde (2 mL, 12 mmol). After
stirring for 3 h the mixture was filtered. EtOH (5 mL) was
added to the filtrate, which was then treated with NaBH4 (1
g, 26 mmol). The mixture was stirred for 30 min, at which
point 1 M HCl was added slowly until there was no visible
reaction. The mixture was made basic with NaHCO3, and the
aqueous layer extracted 2 × 5 mL C6H6. The combined organic
layers were dried (MgSO4) and removed, leaving a white solid.
P r epar ation ofCp2Zr [N(TMS)C(N-TMS)CH(4-tBu C6H4)N-
(C6H5)] (4a ) in Situ . A C6D6 (0.7 mL) solution of zirconaaziri-
dine 1a (30 mg, 56 µmol) was treated with bis(trimethylsilyl)-
carbodiimide (15 µL, 67 µmol), and the solution changed from
orange to dark purple. Yield: >95% yield by 1H NMR. 1H NMR
(C6D6): δ 7.45 (d, 2H), 7.21 (m, 4H), 6.90 (t, 1H), 6.59 (d, 2H),
6.00 (s, 1H), 5.89 (s, 5H), 5.84 (s, 5H), 1.16 (s, 9H), 0.45 (s,
9H), 0.28 (s, 9H). 13C (C6D6): δ 171.1, 157.4, 150.1, 140.2, 129.4,
125.0, 120.9, 120.5, 114.3, 80.2, 34.4, 31.4, 3.2, 2.3. HRMS for
(C34H48N3Si2Zr): [M + H] calcd 644.2434, found 644.2451.
P r ep a r a tion of Cp 2Zr [N(TMS)C(N-TMS)CH(CH2P h )N-
(TMS)] (4f) in Situ . A C7D8 (0.7 mL) solution of Cp2Zr(Me)N-
(TMS)CH2CH2Ph3 (30 mg, 70 µmol) was treated with bis-
(trimethylsilyl)carbodiimide (18 µL, 86 µmol) and heated at
60 °C for 75 min, resulting in a bright orange solution. Yield:
1
Yield: 2.44 g, 85%. H NMR (CDCl3): δ 7.26 (d, 2H), 7.21 (d,
2H), 7.08 (t, 2H), 6.62 (t, 1H), 6.54 (d, 1H), 4.18 (s, 2H), 3.95
(br s, 1H), 1.22 (s, 9H). 13C NMR (CDCl3): δ 150.2, 148.2, 136.3,
129.2, 127.4, 125.5, 117.5, 112.8, 48.0, 34.5, 31.4. HRMS for
(C17H22N) [M + H] calcd 240.1752, found 240.1740.
Cp 2Zr (η2-P h NC(H)C6H4C(CH3)3)THF (1a ). An Et2O (10
mL) solution of N-phenyl-4-tert-butylbenzylamine (940 mg, 3.9
mmol) was cooled to -10 °C and treated with BuLi (2.45 mL,
3.9 mmol). After this solution was stirred for 10 min, it was
added to a THF (20 mL) solution of Cp2ZrMe(OTf) (1.50 g, 3.9
mmol) at -60 °C. The solution, which immediately turned
bright yellow, was stirred at this temperature for 30 min,
warmed slowly to room temperature, and stirred for 4 h to
produce an orange solution. The solvent was removed in vacuo,
leaving an orange solid, which was treated with C6H6 (25 mL).
The solution was filtered to remove the LiOTf, which was
washed with an additional 10 mL of C6H6. The solvent was
removed to leave an orange solid, which was dried under
1
1
>95% yield by H NMR. H NMR (C6D6): δ 7.27 (d, 2H), 7.19
(t, 2H), 7.07 (m, 1H), 6.14 (s, 5H), 6.08 (s, 5H), 3.97 (dd, J )
10.1, 2.1 Hz, 1H), 3.37 (dd, J ) 12.8, 10.3 Hz, 1H), 2.94 (dd, J
) 12.8, 2.1 Hz, 1H), 0.34 (s, 9H), 0.11 (s, 9H), -0.08 (s, 9H).
13C (C7D8): δ 168.9, 140.0, 130.6, 128.4, 126.5, 115.0, 114.7,
79.5, 47.3, 3.5, 2.9, 2.1.
P r ep a r a tion of Cp 2Zr [N(TMS)C(N-TMS)CH(n -p r op yl-
)N(P h )] (4e) in Situ . A C7D8 (0.7 mL) solution of Cp2Zr(Me)N-
3
(Ph)CH2CH2CH2CH3 (30 mg, 78 µmol) was treated with
bis(trimethylsilyl)carbodiimide (18 µL, 86 µmmol) and heated
1
at 60 °C for 75 min, resulting in a dark purple solution. Yield:
vacuum. Yield: 1.71 g (82%), 95% pure by H NMR. The solid
1
>95% yield by H NMR. 1H NMR (C7D8): δ 7.11 (m, 2H), 6.77
was recrystallized from a solution of 6 mL of THF and 35 mL
of hexanes, which gave 1.06 g (62% recovery). 1H NMR
(C6D6): δ 7.53 (d, 2H), 7.26 (t, 3H), 6.83 (m, 3H), 5.60 (s, 5H),
5.47 (s, 5H), 3.77 (br s, 1H), 3.35 (br s, 4H), 1.41 (s, 9H), 1.22
(br s, 4H). 13C NMR (C6D6): δ 129.3, 125.4, 118.9, 116.6, 109.5,
107.8, br s 73, 34.5, 32.1, 25.7. The R-THF protons and the
zirconaaziridine methine are broadened considerably by ex-
change. Anal. Calcd for (C31H37NOZr): C, 70.14; H, 7.03; N,
2.64. Found: C, 67.36; H, 6.79; N, 2.54 (not reanalyzed in
view of the straightforward relationship to known zircona-
aziridines).1-3
(m, 1H), 6.33 (m, 2H), 6.13 (s, 5H), 5.86 (s, 5H), 4.61 (dd, J )
9.5, 4.0 Hz, 1H), 2.10 (m, 1H), 1.87 (m, 1H), 1.39 (m, 1H), 1.22
(m, 1H), 0.81 (dd, J ) 9.7, 7.3 Hz, 3H), 0.37 (s, 9H), 0.30 (s,
9H). 13C (C7D8): δ 173.6, 156.6, 129.04, 119.6, 119.2, 114.6,
114.4, 78.0, 39.3, 20.8, 14.7, 3.3, 2.7.
r-Am in o Am id in es 5. Amino amidines 5a -c were prone
to decomposition; over the period of months gummy yellow
solids were formed. The product of decomposition is not yet
known. Remaining amino amidines in these mixtures could
be recovered by titurating in hexane.
Cp 2Zr (P h NCH2CH2CH2CH3)CH3 (1e). To a solution of
N-phenyl butylamine (2.6 mmol) in 7 mL of Et2O at 0 °C was
added 1.6 mL of 1.6 M BuLi/hexane solution. This was stirred
for 10 min, then added to a solution of 1.03 g of Cp2ZrMeOTf
(2.6 mmol) in 25 mL of THF at -50 °C. After stirring for 1 h,
the solution was warmed to room temperature and the solvent
removed. A 15 mL sample of hexanes was added and the solid
H2NC(NH)CH(4-tBu C6H4)NH(C6H5)‚HCl (5a ). A toluene
(5 mL) solution of 1a (250 mg, 0.47 mmol) was treated with
1,3-bis(trimethylsilyl)carbodiimide (95 µL, 0.52 mmol). After
15 min, 1 M HCl/Et2O (4 mL) was added and immediately
filtered. The precipitate was washed with Et2O (4 mL). Yield:
135 mg (91%). IR (Nujol): ν(NH) 3500-2500, ν(CdN) 1684,
ν(CdC) 1601 cm-1 1H NMR (D2O): δ 7.45 (d, 2H), 7.32 (d,
.
2H), 7.14 (t, 2H), 5.11 (s, 1H), 1.17 (s, 9H) 13C NMR (D2O): δ
(26) Van Geet, A. L. Anal. Chem. 1970, 42, 679-680.
153.8, 146.2, 132.8, 130.0, 127.7, 126.9, 119.9, 118.8, 113.8,

260 Organometallics, Vol. 20, No. 2, 2001
Tunge et al.
59.9, 34.4, 30.7. HRMS for (C18H24N3): calcd 282.1970, found
282.1975.
in CH2Cl2 and spotted on chromatotron. Eluting with 30:1
hexane/EtOAc gave a white solid. Yield: 192 mg (59%). 1H
NMR (CDCl3): δ 8.29 (s, 1H), 7.75 (d, 2H), 7.15 (m, 3H), 6.95
(m, 5H), 6.87 (m, 3H), 6.73 (m, 3H), 6.48 (d, 1H), 5.33 (s, 1H),
4.68 (s, 1H), 3.73 (s, 3H), 2.11 (s, 3H), 2.05 (s, 3H). 13C
(CDCl3): δ 154.15, 147.42, 147.19, 140.03, 137.72, 137.20,
132.35, 131.32, 129.60, 129.45, 129.21, 128.70, 128.45, 121.94,
121.80, 119.96, 119.55, 112.31, 109.81, 60.32, 56.67, 21.12.
Anal. Calcd for (C29H29N3O): C, 79.97; H, 6.71; N, 9.65.
Found: C, 79.83; H, 6.67; N, 10.00.
H2NC(NH)CH(n -p r op yl)NH(C6H5)‚HCl (5e). A toluene (5
mL) solution of N-butyl aniline (320 µL, 2.0 mmol) at -10 °C
was treated with BuLi (2 mL, 2.0 mmol). This was stirred 5
min, then added to a toluene (5 mL) solution of Cp2ZrMeOTf
(770 mg, 2.0 mmol) at -60 °C. The resulting orange mixture
was stirred at -10 °C for 30 min. Bis(trimethylsilyl)carbodi-
imide (420µL, 2.0 mmol) was then added and the solution
heated to 60 °C for 75 min, turning dark purple. This mixture
was filtered, then added to 10 mL of 1 M HCl/Et2O. The
resulting precipitate was immediately collected and washed
with 5 × 1 mL of CH3CN. The remaining white solid was dried
under vacuum. Yield: 337 mg, 74%. IR (Nujol): ν(NH) 3800-
2200, ν(CdN) 1691. 1H NMR (D2O): δ 7.17 (t, 2H), 6.75 (t,
1H), 6.57 (d, 2H) 4.04 (t, 1H), 1.78 (q, 2H), 1.39 (sp, 2H) 0.82
(t, 3H). 13C NMR (D2O): δ 173.07, 146.38, 130.05, 119.46,
113.42, 55.60, 35.72, 18.85, 13.07. HRMS for (C11H18N3): calcd
192.1501, found 192.1494.
4-Me -C6H 4(H )N C (N -4-Me -C6H 4)C H (4-t B u -C 6H 4)N H -
(C6H5) (5b). A toluene (5 mL) solution of 1a (150 mg, 0.28
mmol) was treated with 1,3-di-p-tolylcarbodiimide (63 mg, 0.28
mmol). After 30 min, 0.5 mL of H2O was added, stirred 15 min,
and dried over MgSO4. The mixture was filtered and the
solvent removed. The residue was taken up in CH2Cl2 and
spotted on a chromatotron plate. Eluting with 5:1 pentane/
EtOAc gave a white solid. Yield: 68.3 mg (53%). IR (Nujol):
1
ν(NH) 3328, 3241, ν(CdN) 1633, ν(CdC) 1605 cm-1. H NMR
((CD3)2SO): δ. 8.40 (s, 1H), 7.62 (br d, 2H), 7.42 (d, 2H), 7.32
(d, 2H), 7.12 (t, 2H), 7.03 (t, 4H), 6.67-6.54 (m, 5H), 6.13
(d, J ) 8.0 Hz, 1H), 5.20 (d, J ) 8.0 Hz), 2.23 (s, 3H), 2.22
(s, 3H). 13C NMR (CD3)2SO): δ 154.9, 151.2, 147.8, 147.7,
139.0, 137.3, 131.5, 131.4, 130.1, 129.8, 129.6, 128.2, 126.2,
121.5, 120.0, 118.5, 114.0, 57.1, 35.1, 32.0, 21.2. HRMS for
(C32H36N3): [M + H] calcd 462.2909, found 462.2890.
Alter n a tive P r ep a r a tion of 5b Dir ectly fr om Am in e.
A toluene (5 mL) solution of N-phenyl-4-tert-butylbenzylamine
(310 mg, 1.3 mmol) was cooled to -10 °C and treated with
BuLi (0.82 mL, 1.3 mmol). After this solution was stirred for
10 min, it was added to a toluene (10 mL) solution of Cp2ZrMe-
(OTf) (500 mg, 1.3 mmol) at -60 °C and stirred 20 min; di-p-
tolylcarbodiimide (300 mg, 1.3 mmol) was added and the
resulting solution warmed to room temperature. After 2 h, 1
mL of H2O was added and the solution stirred 1/2 h and dried
over MgSO4. The mixture was filtered, and the solvent
removed under vacuum, leaving a white solid, which was
washed with 3 mL of Et2O. Yield: 216 mg (36%).
H2NC(NH)CH(CH2P h )NH2‚HCl (5f). A toluene (4 mL)
solution of Cp2Zr(Me)[N(TMS)CH2CH2Ph]3 (325 mg, 0.76
mmol) at -10 °C was treated with bis(trimethylsilyl)carbodi-
imide (162 mg, 0.87 mmol) and heated to 60 °C for 100 min,
turning orange. It was filtered, then added to 10 mL of 1 M
HCl/Et2O. The resulting precipitate was immediately collected
and washed with 5 × 1 mL of CH3CN. The remaining solid
was recrystallized from ethanol, giving white needles. Yield:
129 mg (85%). IR (Nujol): ν(NH) 3534, 3500-2443, ν(CdN)
4-Me-C6H4(H)NC(N-4-Me-C6H4)CH(C6H5)NH(C6H5) (5c).
A C6H6 (5 mL) solution of Cp2Zr[η2-N(Ph)CHPh]‚THF (250 mg,
0.53 mmol) was treated with 1,3-di-p-tolylcarbodiimide (118
mg, 0.53 mmol). After 30 min, 1 mL of H2O was added and
the solution stirred 1/2 h and dried over MgSO4. The mixture
was filtered and the solvent removed. The residue was taken
up in CH2Cl2 and placed on a chromatotron plate. Eluting with
20:1 hexane/EtOAc gave a white solid. Yield: 74 mg (35%).
1
1697 cm-1. H NMR (D2O): δ 7.39-7.18 (m, 5H), 4.35 (dd, J
) 8.8, 6.9 Hz, 1H), 3.28 (dd, J ) 14.0, 6.9, 1H), 3.15 (dd, J )
13.9, 8.9, 1H). 13C NMR (D2O): δ 165.5, 132.7, 129.7, 128.8,
53.5, 37.0. HRMS for (C9H14N3): calcd 164.1188, found 164.1183.
1
IR (Nujol): ν(NH) 3347, ν(CdN) 1638, ν(CdC) 1601 cm-1. H
NMR (CD3)2CO): δ 8.31 (br s, 1H), 7.70 (br s, 2H), 7.32 (s,
5H), 7.17 (t, 2H), 7.05-6.95 (br m, 4H), 6.79 (d, 2H), 6.71 (t,
1H) 6.52 (br d, 2H) 5.36 (br s, 1H), 5.22 (d, 1H, J ) 4.02 Hz),
2.23 (s, 6H). 13C NMR (CD3)2CO): δ 154.4, 147.8, 140.0, 131.6,
131.5, 129.5, 129.2, 129.0, 128.5, 121.6, 119.8, 119.0, 118.8,
113.9, 59.2, 20.3. HRMS for (C28H28N3): [M + H] calcd
406.2283, found 406.2281.
Ack n ow led gm en t. Financial support for this work
was provided by NSF Grant CHE-98-96151. The au-
thors thank Boulder Scientific for generously supplying
Cp2ZrCl2.
4-Me-C6H 4(H )NC(N-4-Me-C6H 4)CH (C6H 5)NH (2-MeO-
C6H4) (5d ). A C6H6 (5 mL) solution of Cp2Zr[η2-N(o-anisyl)-
CHPh] (324 mg, 0.75 mmol) was treated with di-p-tolylcarbo-
diimide (167 mg, 0.75 mmol). After 30 min, 1 mL of H2O was
added, stirred 1/2 h, and dried over MgSO4. The mixture was
filtered and the solvent removed. The residue was taken up
Su p p or tin g In for m a tion Ava ila ble: Kinetic data, plots
of these data, kinetic derivations, and selected NMR spectra
are available. This material is available free of charge via the
Internet at http://pubs.acs.org.
OM000768S