Role of trace amine in the metathesis of imines by CpTa(=NR)Cl2

Article abstract of DOI:10.1021/om011085v

Tantalum imido compound CpTa(=NBu^t)Cl₂ (1) reacts with imines to give NR exchange (metathesis) products. Addition of <5 mol % of compound 1 to a 1:1 mixture of two different imines resulted in catalytic cross metathesis of the two imines to give an equilibrium mixture of imines. Kinetic studies of the reaction of 1 with excess imine (Bu^t)HC=N(p-tolyl) (2a) gave an average k_obs of 1.29 × 10^-4 min^-1 (σ 0.28). Pseudo-first-order behavior was observed for 1. The dependence of k_obs on the concentration of 2a was found to be nonlinear and dependent on catalytic amounts of contaminant amine p-tolylamine (9). Reaction of 1 with tert-butylamine (6) results in the formation of the imide/amide CpTa(=NBu^t)(NHBu^t)Cl (7) and HCl. Compound 7 did not react with imines to give exchange products, so it is proposed that the reaction of 1 with imines is significantly accelerated by the presence of HCl.

Full text of DOI:10.1021/om011085v

Organometallics 2002, 21, 1933-1941
1933
Role of Tr a ce Am in e in th e Meta th esis of Im in es by
Cp Ta (dNR)Cl2
Matthew C. Burland, Timothy W. Pontz, and Tara Y. Meyer*
Department of Chemistry, University of Pittsburgh, Pittsburgh, Pennsylvania 15260
Received December 21, 2001
Tantalum imido compound CpTa(dNBut)Cl2 (1) reacts with imines to give NR exchange
(metathesis) products. Addition of <5 mol % of compound 1 to a 1:1 mixture of two different
imines resulted in catalytic cross metathesis of the two imines to give an equilibrium mixture
of imines. Kinetic studies of the reaction of 1 with excess imine (But)HCdN(p-tolyl) (2a )
gave an average kobs of 1.29 × 10-4 min-1 (σ 0.28). Pseudo-first-order behavior was observed
for 1. The dependence of kobs on the concentration of 2a was found to be nonlinear and
dependent on catalytic amounts of contaminant amine p-tolylamine (9). Reaction of 1 with
tert-butylamine (6) results in the formation of the imide/amide CpTa(dNBut)(NHBut)Cl (7)
and HCl. Compound 7 did not react with imines to give exchange products, so it is proposed
that the reaction of 1 with imines is significantly accelerated by the presence of HCl.
In tr od u ction
We have discovered that compounds of the general
formula CpTa(dNR)Cl2 represent a new class of imine
F igu r e 1. Chauvin-type mechanism for imine metathesis
metathesis catalysts. The impetus for these studies
originated in our previous investigation of the imine
reactivity of the group VI bis(imide) complexes (DME)Cl2-
Mo(dNR)2.1-3 Although we had gained a significant
understanding of the overall catalytic cycle, the com-
plexity inherent in the dual-active site catalysts pre-
cluded the detailed mechanistic studies that are neces-
sary to understand the system and enable the eventual
application of the imine metathesis strategy to the
catalytic synthesis of small molecules and polymers.
Interestingly, the two other examples of imine/imide
exchange for which mechanistic data were reported
apparently proceed by different pathways. The viability
of a Chauvin-type4 mechanism (Figure 1) was estab-
lished by Bergman and co-workers in their observation
of isolable diazametallacyclic intermediates in the reac-
tion of CpCp′Zr(dNR)(THF) (Cp′ ) Cp, Cp*) with
imines.5-7 In contrast, Mountford and co-workers ob-
served neither diazametallacycles nor kinetics consis-
tent with a [2+2] mechanism and proposed an amine-
catalyzed process to explain the reaction of (py)3Cl2Ti-
(dNR) with imines.8,9 Other examples of imine/imide
and imine/carbodiimide metathesis are known, but little
mechanistic data has been reported for these sys-
at an electrophilic metal center.
tems.10-13 Given the variance in precedents, the deter-
mination of mechanism in our system is of particular
importance to us as well as others.
Since the dual-site molybdenum catalysts proved
unsuitable for kinetic studies, we sought to discover a
new single-site catalyst. Metal imides of the type CpTa-
(dNR)Cl2 were selected because they possess a single
imide functional group while being approximately iso-
lobal with the molybdenum bis(imide) complex sans
DME. Herein, we describe both the stoichiometric and
catalytic metathetic reactions of these tantalum mono-
(imide) complexes with imines and, on the basis of
kinetic analyses and 15N-labeling studies, propose an
acid-catalyzed mechanism.
Resu lts
Im id e/Im in e Meta th esis. CpTa(dNBut)Cl2, 1, re-
acted with 15 equiv of (But)HCdN(p-tolyl), 2a , in C6D6
at 70 °C to yield imine (But)HCdN(But), 4a , as shown
1
by comparison of the H NMR spectroscopic resonances
with those of an authentic sample (Scheme 1). A new
Cp-containing compound, assigned as p-tolyl imide 3a ,
was also produced, as evidenced by distinctive 1H NMR
resonances at δ 5.72 (Cp) and 2.19 (tolyl CH3). No other
species were observed. Within 3-4 days nearly all of
the starting imide 1 was converted to the new product
3a . No precipitates formed and integration versus an
(1) Cantrell, G. K.; Meyer, T. Y. J . Chem. Soc., Chem. Commun.
1997, 1551-1552.
(2) Cantrell, G. K.; Meyer, T. Y. Organometallics 1997, 16, 5381-
5383.
(3) Cantrell, G. K.; Meyer, T. Y. J . Am. Chem. Soc. 1998, 120, 8035-
8042.
(4) He´risson, J .-L.; Chauvin, Y. Macromol. Chem. 1970, 141, 161.
(5) Krska, S. W.; Zuckerman, R. L.; Bergman, R. G. J . Am. Chem.
Soc. 1998, 120, 11828-11829.
(6) Meyer, K. E.; Walsh, P. J .; Bergman, R. G. J . Am. Chem. Soc.
1994, 116, 2669-2670.
(10) Bruno, J . W.; Liu, X. J . Organometallics 2000, 19, 4672-4674.
(11) Royo, P.; Sa´nchez-Nieves, J . J . Organomet. Chem. 2000, 597,
(7) Meyer, K. E.; Walsh, P. J .; Bergman, R. G. J . Am. Chem. Soc.
1995, 117, 974-985.
61-68.
(12) Birdwhistell, K. R.; Lanza, J .; Pasos, J . J . Organomet. Chem.
(8) McInnes, J . M.; Mountford, P. J . Chem. Soc., Chem. Commun.
1998, 1669-1670.
1999, 584, 200-205.
(13) Weiss, K.; Kindl, P. Angew. Chem., Int. Ed. Engl. 1984, 23, 629-
(9) Mountford, P. J . Chem. Soc., Chem. Commun. 1997, 2127-2134.
631.
10.1021/om011085v CCC: $22.00 © 2002 American Chemical Society
Publication on Web 04/05/2002

1934 Organometallics, Vol. 21, No. 9, 2002
Burland et al.
Sch em e 1
internal standard confirmed that all species remained
soluble.
Tantalum imide 1 also reacted metathetically in C6D6
with excess (Ph)HCdN(Prn), 2b, to give imine (Ph)HCd
N(But), 4b, and a product assigned as CpTa(dNPrn)-
Cl2, 3b. Imine 4b was identified in the reaction mixture
by comparison with independently prepared compound.
The assignment of the exchange product was made on
the basis of the Cp resonance at δ 5.74 and a multiplet
at δ 4.23 that is characteristic of the R-methylene of an
N-alkyl group. The resonances for the hydrogens on the
â- and γ-carbons of this new species appear to overlap
with those of the starting material imide. After 1 day
at 55 °C a third species, 5, appeared. Compound 5 was
identifiable only from a new Cp resonance at δ 6.10.
Either this compound has no N-alkyl group or its
resonances overlap with those of 1 and/or 3b. We
speculate that 5 may be the product of â-hydride
elimination, which is unable to occur in the case of the
tert-butyl imide, 1, and the p-tolyl imide 3a . After 10
days of heating there were no new species present by
1H NMR spectroscopy, but precipitate was evident and
the integration of all species against an internal stan-
dard showed a confirming decrease in the concentration
of Cp-containing complexes.
F igu r e 2. Pseudo-first-order disappearance of CpTa-
(dNBut)Cl2, 1, in the presence of 15 equiv of (But)HCd
NTol, 2a (kobs ) 2.6 × 10-4 min-1). Corresponds to experi-
ment 5 in Table 1.
Ca ta lytic Im in e Meta th esis. CpTa(dNBut)Cl2, 1,
acts as an imine metathesis catalyst. Addition of <5 mol
% of 1 to a 1:1 mixture of imines 2a and 2b produces
the metathesis products (But)HCdN(Prn) and (Ph)HCd
N(p-tolyl) within hours at 70 °C (Scheme 2). The
reaction is slow, reaching equilibrium after >1 week.
The imines produced by imide/imine exchange and the
imide products, 3a and 3b, were also observed in the
reaction mixture. At longer reaction times (>2 weeks),
precipitate was observed.
Kin etic Stu d ies. The reaction of 1 with N-p-tolyl
1
imine 2a was monitored kinetically by H NMR spec-
troscopy (Scheme 3). In the presence of excess imine,
pseudo-first-order behavior was observed for 1. The
validity of a first-order analysis was confirmed by
following several reactions for more than 5 half-lives and
by the independence of the rate constant on the con-
centration of 1. The observed pseudo-first-order rate
constants, kobs, were obtained by fitting the [Imide 1]t
- [Imide 1]eq to a single-exponential function (Figure
2). An average Keq, defined as {[Imide 3a ][Imine 4a ]}/
{[Imide 1][Imine 2a ]}, of 1.29 (σ ) 0.28) was calculated
from the data for experiments 2-9 (experiment 1
omitted because only a 6.3:1 excess was used).
To determine the order of the imine in the reaction,
the dependence of kobs on imine was examined. The plot
of kobs versus [Imine] for experiments 1-9 did not show
a simple linear dependence (Figure 3). Moreover, two
control experiments established that the reaction de-
pended, at least partially, on a catalytic contaminant
F igu r e 3. Dependence of the pseudo-first-order rate
constant (kobs) for the disappearance of 1 on imine concen-
tration: Experiments 1-9 were performed with dried and
distilled imine (Imine1×); experiment 10 used doubly
purified imine (Imine2×). The imine1× data have been
fitted with a polynomial to highlight the nonlinear trend.
The curve does not represent a specific rate law.
present in the imine. In the first control experiment,
the original kinetic experiment (Scheme 3) was repeated
with doubly purified imine; the kobs obtained was
significantly smaller than the rate constant that would
have been predicted on the basis of data obtained from
the singly purified imine (Table 1, experiment 10, Figure
2). In the second control experiment, the relative ratios
of imine to imide were reversed such that there was an
excess of imide 1. The resulting reaction was extremely
slow, with only trace amounts of product visible after 1
week at 70 °C (not included in table).

Metathesis of Imines by CpTa(dNR)Cl2
Organometallics, Vol. 21, No. 9, 2002 1935
Sch em e 2
Sch em e 3
Sch em e 4
Ta ble 1. Kin etic Da ta for Im id e/Im in e Exch a n ge
Rea ction s
Since the most likely catalytically active contaminant
of the imine is amine, the reaction of 1 with excess
p-tolylimine 2a was repeated in the presence of added
p-tolylamine 9 (0.1-0.6 equiv with respect to 1, Scheme
4). A plot of the kobs obtained from these experiments
showed a direct dependence on amine concentration
(Table 1, experiments 11-13, Figure 4). Although the
trend appears linear, implying a first-order dependence
on amine, the experiments covered a relatively small
range of concentrations and the relationship is likely
not one of simple proportionality. Experimental difficul-
ties precluded studies of significantly lower or higher
concentrations.
Several attempts to remove the amine from the
reaction mixture failed. Any reagent sufficiently acidic
or basic enough to completely sequester the amine also
reacted in some fashion with the imide complex or with
the imine. The amine must be accepted as a necessary
component of any reaction mixture containing excess
imine.
Am in e/Im id e Rea ction s. To probe the role of amine
in the imide/imine metatheses, the direct reaction of
tert-butyl imide 1 with tert-butylamine, 6, was examined
(Scheme 5). By 1H NMR spectroscopy of the reaction
mixtures, the instantaneous formation of a new imide/
amide complex, CpTa(dNBut)(NHBut)Cl (7), was ob-
served. Formation of the ammonium salt NH3ButCl (8)
was also noted during the course of the reaction. The
imide/amide was the only soluble product observed,
independent of the ratio of amine to imide; addition of
a single equivalent of amine gave 0.5 equiv of compound
7. Although we could not isolate imide/amide 7 in
analytically pure form nor could we obtain X-ray quality
crystals, its formulation was consistent with all spec-
troscopic data and an analogous Cp* derivative is
known.11 A 2D heteronuclear multiple-bond correlation
(HMBC) NMR spectrum showed the expected long-
range coupling between the amide proton and the ipso
carbon of the attached tert-butyl group. Also, the 15N
kobs (10-4
/
[Imide 1]/M [Imine 2a ]init/M [Amine 9]init/M
Imide + Excess Imined
min-1
)
1
2
3
4
5
6
7
8
9
0.027
0.022
0.026
0.022
0.022
0.022
0.028
0.039
0.022
0.17
0.22
0.25
0.32
0.32
0.36
0.42
0.58
0.43
<0.002b
<0.002b
<0.002b
<0.002b
<0.002b
<0.002b
<0.004b
<0.004b
<0.004b
1.2a
1.2
1.7
3.3
2.6
3.7
4.5
5.7
4.7
Imide + Excess Doubly Purified Imined
10
0.028
0.42c <0.002b
1.2
Imide + Excess Doubly Purified Imine + Aminee
11
12
13
14
0.018
0.018
0.019
0.018
0.27c
0.27c
0.28c
0.27c
0.0018
0.0037
0.0077
0.011
9.6
10.5
12.6
14.5
a
kobs for this experiment is less certain due to an insufficient
b
excess of imine to imide (6.3:1). Upper limit based on NMR
sensitivity and/or integration of observed baseline signals that are
associated with trace amine. c Doubly purified imine. Estimated
d
uncertainty in kobs ) (0.6 × 10-4 based on uncertainties in
concentration, integration, and [Imine 2a ]eq.
e Estimated uncer-
tainty in kobs ) (2.5 × 10-4 based on uncertainties in concentra-
tion, integration, and [Imine 2a ]eq. Amine 9 ) p-tolylamine.
NMR spectra of the partially labeled and fully labeled
derivatives CpTa(dNBut)(15NHBut)Cl (7N1a) and CpTa-
(15NBut)(15NHBut)Cl (7N2) were consistent (vide supra).
The reaction of tert-butyl imide 1 with p-tolylamine
9 was also examined by 1H NMR (Scheme 6). The
reaction mixture was significantly more complex than
the reaction of 1 with tert-butylamine 6. At equilibrium,
at least five species were present including starting
imide (20%) and the amine/imide exchange product,
CpTa(dNTol)Cl2 3a (15%). Significant precipitate, along
with a broad resonance at δ 8.0, was consistent with
the formation of ammonium salts. Although we were

1936 Organometallics, Vol. 21, No. 9, 2002
Burland et al.
equivalent amine 6N to the imide in C6D6, a 1H NMR
spectrum showed that half of the starting material had
been converted to the partially labeled imide/amide
7N1a . The amine was completely consumed, and pre-
cipitate, presumably ammonium salt, was visible. Long-
3
range J NH coupling of 2.2 Hz was observed from the
amide nitrogen to the methyl peaks of the attached tert-
butyl group. No such coupling was observed to the imide
tert-butyl of 7N1a or in the remaining starting material
1. After heating for 2 h at 80 °C, the label began to wash
into the imide positions, as evidenced by the broadening
and splitting of the tert-butyl peaks. In the 15N NMR
spectra, parallel behavior was observed. Immediately
after addition, only the amide resonance at δ 116.7 was
easily detectable. No resonance for the labeled free amine
6N was observed, indicating that it had been completely
consumed. After heating, the resonances for the imide
groups for both 7N2 and imide complex 1N were easily
visible. Presumably the monolabeled compound CpTa-
(15NBut)(NHBut)Cl (7N1b) is also present, but it cannot
be spectroscopically differentiated in mixtures contain-
ing 7N2.
F igu r e 4. Dependence of the pseudo-first-order rate
constant (kobs) for the disappearance of 1 vs amine concen-
tration for a ternary reaction of 1, 15 equiv of imine 2a ,
and p-tolylamine 9.
Sch em e 5
The ternary experiment involving stoichiometric im-
ide 1, (Ph)HCdN(But) (4b), and labeled amine 6N was
also followed by NMR. Initial 15N NMR spectra showed
only the partially labeled imide/amide 7N1a . With
heating, resonances for the imide functionalities of 1N
and 7N2 became visible, as did the imine nitrogen for
4bN. Again, no free amine was observed. The 15N
spectra were collected with an NOE-suppressing pulse
sequence in order to make it possible to integrate the
peaks. Although the differences were not dramatic, it
appeared as if the label washed into the imide positions
faster than into the imine (Scheme 10). A control
experiment with the imine 4b and the labeled amine
6N showed no scrambling of the isotopic label, even after
heating at 80 °C for 1 h.
unable to identify all species in solution, by analogy to
the tert-butylamine reactions, some or all of the imide/
amide complexes are likely to be present: CpTa-
(dNBut)(NHTol)Cl (10a), CpTa(dNTol)(NHBut)Cl (10b),
CpTa(dNBut)(NHBut)Cl (7), and CpTa(dNTol)(NHTol)-
Cl (10c). The ratios of all species were found to vary
reversibly as a function of temperature.
The ternary reaction of tantalum imide 1, p-tolyl-
amine 9 (0.5 equiv), and N-tolylimine 2a (15 equiv) was
also studied by 1H NMR and, initially, was found to
contain all of the species seen in the binary imide/amine
reaction (Scheme 7). After 6 days at 70 °C, the reaction
had reached equilibrium and the spectrum was much
simplified. Only three Cp-containing species remained:
the product 3a (87%), starting tert-butylimide 1 (4%),
and a compound whose Cp resonance corresponded
exactly with the one observed in the binary imide/
tolyamine reaction (Scheme 6) for the most abundant
of the putative imide/amide compounds 10a -c. Imide/
amide 7 can be ruled out on the basis of the known shift.
Rea ction of Im id e/Am id e (7) w ith Im in e. To
determine if imide/amide 7 generated in situ might be
responsible for the amine-accelerated rates of imine/
imide exchange, we examined the reaction of isolated
imide/amide 7 with 15 equiv of imine (But)CHdN(P-
tolyl) (2a ) (Scheme 8). Interestingly, no new imine
product was detected after 2 days at 70 °C. The control,
consisting of imide 1 and 15 equiv of the same imine,
had proceeded to completion in the same period.
Reaction of the labeled imine 4bN with imide 1 and
unlabeled amine 6 was also followed by 15N and 1H
NMR spectroscopy (Scheme 11). After initial formation
of imide/amide 7, the resonances associated with the
imide 1N and the imide/amide 7N1a and 7N2 were seen
to grow in, while the imine resonance for 4bN was
reduced. It appeared that the 15N label was incorporated
into the imide 1N at a faster rate than into the imide/
amide 7N2. No free amine was seen at any point during
the reaction.
To determine if the precipitated ammonium salt
played a role in our reaction, or whether the nonpolar
nature of the solvent excluded its participation, we
conducted experiments to determine if NR exchange
between solution phase species and the precipitated
ammonium salt was possible (Scheme 12). We heated a
sample of imide 1 in C6D6 with 2 equiv of solid, 15N-
t
labeled ammonium salt BuNH3Cl, 8N. Although the
solid did not appear to dissolve after several days of
heating at 70 °C, a 15N NMR spectrum of the sample
clearly showed the presence of labeled imide 1N. A
similar experiment with the tert-butylamine 6 and the
labeled ammonium salt 8N also showed incorporation
of the label despite the nonpolar nature of the solvent.
As a comparison, a sample of unlabeled amine showed
no detectable signal in the 15N NMR spectrum after a
similar period of data collection.
15N-La belin g Stu d ies. Studies involving 15N-lableled
substrates provided further insight into these complex
reaction systems. We prepared 15NH2But (6N) and (Ph)-
HCd15N(But) (4bN) using standard methods. The reac-
tion of the labeled amine 6N with imide 1 was studied
over time by 1H NMR and 15N NMR spectroscopies
(Scheme 9). Within a minute of the addition of a single

Metathesis of Imines by CpTa(dNR)Cl2
Organometallics, Vol. 21, No. 9, 2002 1937
Sch em e 6
Sch em e 7
Sch em e 8
It is, at this point, relevant to discuss in more detail
the mechanism proposed by Philip Mountford to explain
the zero-order dependence on (py)3Cl2Ti(dNBut) of the
dNR exchange reaction with imine.8 In the first step,
Mountford postulates that a small amount of tert-
butylamine reacts with the tolylimine to produce free
p-tolylamine. Mountford and co-workers had previously
established that p-tolylamine reacts with the butylimide
titanium complex to regenerate the tert-butylamine and
the product p-tolylimide complex (Scheme 13). If these
two steps are combined, the overall reaction is catalytic
in amine, and if the first step is rate determining, the
zero-order dependence on imide complex is explained.
In addition, the titanium complexes are available to act
as Lewis acid catalysts for the amine/imine reaction.
Sch em e 9
An analogous mechanism can be proposed for the
reaction of 1 with imine (Scheme 14). In this case,
however, the relative rates of the two steps must be
reversed in order to explain the first-order dependence
of the reaction rate on 1. It is also necessary to postulate
that, in contrast with the Mountford system, step 2, the
amine/imide exchange, is reversible.
In addition to the amine-catalyzed pathway, two
alternative mechanisms can be postulated if, instead of
assuming that the amine acts as catalyst as depicted
in Scheme 14, it is hypothesized that one of the products
of the amine reaction with imide 1 is the mediator
(Scheme 15). The two products from the reaction of tert-
butylamine 6 with imide 1 are the imide/amide 7 and
the HCl that is sequestered as the ammonium salt, NH3-
ButCl 8. The amide functional group of 7 could conceiv-
ably act as the mediator for the NR exchange reaction
as shown in Scheme 16. The 1,3-proton transfer re-
quired could be intermolecular or intramolecular.
Discu ssion
CpTa(dNBut)Cl2, 1, undergoes imide/imine exchange
and catalyzes the metathesis of imines, but the primary
mechanism appears to involve trace amine. The fact
that the reaction is first order in 1 and that the rate
depends on amine suggests that the reaction of these
two components may be the rate-determining step of the
reaction. These data rule out a Chauvin-type mecha-
nism, which would be expected to be zero order in
amine.
HCl as catalyst is also intriguing. Although it is
mostly consumed by reaction with the basic tert-butyl-
amine 6, it is possible that some remains present in the
broad reaction mixture. The observed exchange of 15N

1938 Organometallics, Vol. 21, No. 9, 2002
Burland et al.
Sch em e 10
Sch em e 11
Sch em e 12
imide/amine, amine/imine, imide/imine, amide/amine,
and imine/imine (Scheme 17).
In considering these three mechanisms, we currently
favor the third, HCl catalysis, as the most significant.
The first mechanism, the amine-mediated pathway
postulated by Mountford, is likely to occur at some rate
in the reaction mixture. Both of the constituent reaction
steps have been observed. However, this mechanism
alone cannot explain some of our observations. First,
the reactions of imide 1 with amines are not straight-
forward imide/amine exchanges. In contrast with the
titanium complex studied by Mountford and co-workers,
which gives simple amine/imide exchange, the reaction
of amine with CpTa(dNBut)Cl2, 1, gives multiple prod-
ucts, including at least one stable imide/amide. Second,
the strongly Lewis acidic 1 appears to nearly consume
added amine. The concentration, therefore, of amine in
solution when it originates as a trace contaminant must
be very low. It is worth noting that the pyridine ligands
coordinated to the titanium would be expected to
diminish the Lewis acidity of the Mountford complex
label from ammonium salt into soluble imide and imine
suggests that the amine-bound HCl can interact with
the other species in solution. Moreover, HCl could
theoretically act as a catalyst for virtually any reaction
combination in the mixture: imide/imide, imide/amide,

Metathesis of Imines by CpTa(dNR)Cl2
Organometallics, Vol. 21, No. 9, 2002 1939
Sch em e 13
Sch em e 16
Sch em e 14
Sch em e 15
Sch em e 17
and may explain the divergence in behavior for the two
systems. Finally, we observed that the 15N label from
added labeled amine moved more quickly to the imide
positions than to the imine. This final observation is
inconsistent with the fact that imide 1 is involved in
the slowest rate-determining step. If we assumed the
amine pathway dominates, then we would expect the
label to move into the imine position faster. The second
mechanism we postulated, mediation by imide/amide 7,
can be discounted. The imide/amide, when isolated, does
not react with imines. Therefore, the compound cannot
be the catalyst.
The third mechanism, mediation by HCl, is consistent
with our data. The reaction of amine with imide 1 to
produce HCl is rapid and quantitative in the case of tert-
butylamine 6. Although we are working in nonpolar
solvents, the imide and imine functional groups are
quite basic and could be protonated to a small extent.
After protonation, one could postulate a catalyzed
addition/elimination sequence that depends on original
imide and also on imine. Concurrently, it must be
assumed that some of the amine-catalyzed processes
proposed in the first mechanism are occurring. As we
stated earlier, the amine-mediated pathway does not
explain all of our data, but it probably is active.
Although the idea of an acid-catalyzed stoichiometric
addition/elimination reaction between an imide and an
imine is intriguing and could be exploited, the presence
of trace acid in any catalytic mixture of two imines and
a transition metal catalyst is not tolerable. The acid-
catalyzed imine exchange is quite facile and would
compete with the metal-centered reaction. The benefits
of metal catalysis, selectivity, and activation and the
potential for metal-centered chain polymerization would
be lost.
were observed, and Royo reported that switching solvent
from C6D6 to CDCl3 increased the reaction rate. All of
these facts are suggestive of an HCl-catalyzed reaction
similar to that observed in our system. Also analogous
in composition and behavior are certain catalysts for NR
exchange such as Cl3V(dNTol) and (DME)Cl3Nb(dNPh)
reported by Birdwhistell12 and Bruno,10 respectively. In
particular, these group V compounds are extremely
electrophilic with potentially labile chloride supporting
ligands. These similarities suggest that HCl catalysis
could also be an important pathway for these systems.
It is also possible that the (DME)Cl2Mo(dNR)2 system
reported earlier by our group1-3 is subject to this type
of reactivity. Unfortunately, in this specific case the
complexity of the dual active site catalyst precludes the
detailed mechanistic studies necessary to confirm or
deny this conjecture.
Con clu sion s. We have discovered that CpTa(dNR)-
Cl3 reacts metathetically with amines to give the dNR
exchanged products and that the mono(imide) can act
as an imine metathesis catalyst. Kinetic analysis of the
reaction suggests that amine plays a primary role in
the reaction mechanism. 15N-labeling studies are most
consistent with a mechanism involving initial attack on
the Lewis acidic metal center, giving an inactive metal
complex and an equivalent of HCl. It is postulated that
the HCl catalyzes a variety of processes in the reaction
including imide/imine metathesis.
Exp er im en ta l Section
Gen er a l P r oced u r es. All manipulations were performed
under inert atmosphere using standard glovebox and Schlenk
techniques. Solvents were distilled from appropriate purifying
agents (noted in parentheses) prior to use: tetrahydrofuran
(sodium/benzophenone ketyl), toluene (sodium/benzophenone),
hexane (sodium/benzophenone), dichloromethane (P2O5), and
1,2-dichloroethane (P2O5). tert-Butylamine and p-tolylamine
were distilled from CaH2. The compounds CpTaCl4,14 trimeth-
ylsilylamines,15 15N-labeled ammonium salt,16 and imines17
were prepared according to established methods. All other
Our observations may have implications for other
systems that have exhibited imine or carbodiimide
metathesis activity. Royo, for example, reported imine
and carbodiimide metathesis in the closely related
pentamethylcyclopentadienyl system Cp*Ta(dNBut)-
Cl2.11 The reaction of 1 equiv of imine, (Ph)HCdN(Ph),
with the Cp* imide was found to be slow at 165 °C but
proceeded at lower temperatures when an excess of
imine was used. No diazametallacyclic intermediates
1
chemicals were used as received unless otherwise noted. H,
(14) Cardosa, A. M.; Clark, R. J . H.; Moorehouse, S. J . Chem. Soc.,
Dalton Trans. 1980, 1156-1160.

1940 Organometallics, Vol. 21, No. 9, 2002
Burland et al.
13C, and 15N NMR spectra were recorded with a Bruker AF500
or AF300 NMR spectrometer. Chemical shifts for 1H and 13C
are referenced to the residual protio impurity in the deuterated
solvent. 15N chemical shifts are referenced to formamide used
as an external standard. 14N NMR spectra were recorded by
Dr. M. Minelli, Grinnell College, Iowa. They were measured
on a Bruker AC 300 MHz NMR spectrometer with a 10 mm
broadband probehead (109Ag-31P) with digital tuning. Ni-
tromethane neat was used as external reference (0 ppm).
Cp Ta (dNBu t)Cl2 (1) was prepared according to the method
of Gibson for CpCl2Nb(dNBut) with slight modification.18 To
a stirred, room-temperature suspension of CpTaCl4 (1.38 g,
3.57 mmol) in 1,2-dichloroethane (50 mL) in a 100 mL Schlenk
flask was slowly added (TMS)NHBut (1.04 g, 7.17 mmol). The
flask was fitted with a reflux condenser, and the suspension
was refluxed under nitrogen overnight. The solvent was
removed under vacuum, and the resulting yellow-brown solid
was redissolved in hexane. The solution was filtered to remove
an insoluble tan powder, and the filtrate was stored at -35
°C overnight. The product crystals were isolated as golden
yellow needles by filtration in 72% yield (0.99 g, 2.55 mmol).
1H NMR (300.13 MHz, C6D6, 25 °C): δ 1.10 (s, 9H, C(CH3)3),
5.93 (s, 5H, C5H5). 13C{1H} NMR (75.4 MHz, CDCl3, 25 °C): δ
32.6 (C(CH3)3), 66.7 (C(CH3)3), 111.6 (C5H5). 14N{1H} NMR
(21.69 MHz, CH2Cl2, 25 °C): δ 30 (NC(CH3)3). 15N NMR of
CpTa(d15NBut)Cl2 (1N) (50.68 MHz, C6D6, 25 °C): δ 292.4 (NC-
(CH3)3).
Cp Ta (dNBu t)(NH Bu t)Cl (7). To a solution of CpTa-
(dNBut)Cl2 (0.49 g, 1.27 mmol) in toluene (5 mL) was added
H2NBut (300 µL, 2.85 mmol). Precipitate formed immediately,
and the solution was filtered through Celite to remove the
ammonium salt. The solvent was removed from the filtrate
under vacuum to give the product as an off-white powder in
86% yield (0.46 g, 1.09 mmol). Repeated recrystallizations did
not give material suitable for chemical analysis nor for X-ray
diffraction. 1H NMR (300.13 MHz, C6D6, 25 °C): δ 1.05 (s, 9H,
NC(CH3)3), 1.13 (s, 9H, NHC(CH3)3), 5.88 (s, 5H, C5H5), 6.40
(s, 1H, NHC(CH3)3). 13C{1H} NMR (75.4 MHz, CDCl3, 25 °C):
δ 33.6 (NC(CH3)3), 34.5 (NHC(CH3)3), 56.5 (NHC(CH3)3), 65.5
(NC(CH3)3), 108.5 (C5H5). 15N NMR of CpTa(d15NBut)(15NHBut)-
Cl (7N2) (50.68 MHz, C6D6, 25 °C): δ 116.7 (NHC(CH3)3), 267.0
(NC(CH3)3).
(P h )HCd15N(Bu t) (4bN) was prepared by standard meth-
ods.17 1H NMR (300.13 MHz, C6D6, 25 °C): δ 8.12 (d, 1 H, 15Nd
CHPh, J NH ) 3.8 Hz), 7.80 (d, 2 H, CHPh), 7.06-7.26 (m, 3 H,
CHPh), 1.23 (d, 9 H, 15NC(CH3)3, J NH ) 1.3 Hz). 15N NMR
(50.68 MHz, C6D6, 25 °C): δ 244.1.
Gen er a l P r oced u r e for NMR Exp er im en ts. Imine and
amine standard solutions were prepared by weighing the imine
into a volumetric flask and diluting with C6D6. Likewise,
standard solutions of the internal standards were prepared
using volumetric flasks. The necessary amounts of these
solutions were then transferred into an NMR tube equipped
with a Teflon stopcock. The catalyst was either weighed into
a vial, dissolved in 100 µL of C6D6, and transferred into the
NMR tube or taken from a standard solution prepared in a
volumetric flask. If needed, additional C6D6 was added to the
NMR tube so that concentration remained constant between
analogous experiments. When two imines were used, relative
stoichiometry was checked by integration prior to heating.
Rea ction of 1 a n d (Bu t)HCdN(p-tolyl), 2a . Compound
1 (0.0066 g, 0.017 mmol, 1 equiv), (But)HCdN(p-tolyl) (0.171
mol, 10 equiv), and anisole internal standard were added to
an NMR tube. The reaction was monitored at 55 °C for 45 days.
The progress of the reaction was monitored by 1H NMR
spectroscopy. Resonances for 3a and 4a were observed.
Rea ction of 1 a n d (P h )HCdN(P r n ), 2b. Compound 1
(0.0066 g, 0.017 mmol, 1 equiv), (Ph)HCdN(Prn) (0.171 mol,
10 equiv), and hexamethylbenzene internal standard were
added to an NMR tube. The reaction was monitored at 55 °C
for 45 days. The progress of the reaction was monitored by 1H
NMR spectroscopy. Resonances for 3b and 4b were observed
initially. After 1 day resonances for compound 5 were observed.
Rea ction of 1, (Bu t)HCdN(p-tolyl), 2a , a n d (P h )HCd
N(P r n ), 2b. Compound 1 (0.0066 g, 0.017 mmol, 1 equiv), (Ph)-
HCdN(Prn) (0.171 mmol, 10 equiv), (But)HCdN(p-tolyl) (0.171
mmol, 10 equiv), and hexamethylbenzene internal standard
were added to an NMR tube. The reaction was monitored at
25 °C for 25 days, then 55 °C for 35 days. The progress of the
reaction was monitored by 1H NMR spectroscopy. Resonances
for the metathesis products (Ph)HCdN(p-tolyl) and (But)CHd
N(But) were observed initially. After >2 weeks, precipitate was
observed.
Rea ction of Im id e 1 a n d La beled Am in e 6N. A 20 µL
portion of a 0.626 M (12.5 µmol) standard solution of labeled
amine 6N, in C6D6, was added to 100 µL of a 0.0912 M (9.12
µmol) standard solution of imide 1 in C6D6, in a screw top NMR
tube. A fine precipitate of ammonium salt formed immediately.
The progress of the reaction was monitored by 1H and 15N
NMR spectroscopy at 80 °C for 3 days. Resonances for the
labeled imide/amide 7N1a were observed initially. Over the
course of the reaction, resonances for 1N and 7N2 were
observed.
15NH2Bu t (6N). To a slurry 15NH3ButCl (0.30 g, 2.78 mmol)
in 1.75 mL of C6D6 in a three-necked round-bottom flask under
N2 was added solid NaOH (0.2 g, 5 mmol, 1.8 equiv). Deionizied
water (250 µL) was added via syringe to help dissolve the
solids. Some solid remained undissolved. The contents of the
flask were stirred for 2 h and then vacuum transferred into a
round-bottom flask containing CaH2. After stirring for 24 h,
the solution was vacuum transferred again into a clean flask
containing more CaH2 and stirred overnight before being
vacuum transferred into a clean, dry flask. The dry C6D6
solution of the labeled amine was diluted to a total volume of
3 mL by addition of C6D6. An 80 µL aliquot was added to a
solution of 50 µL of anisole (0.73 mmol) in C6D6 in an NMR
tube. Integration of the 1H NMR spectrum established that
the concentration of the bulk solution was 0.626 M (68% yield).
15N NMR spectroscopy determined that the label had been
incorporated into the product. 1H NMR (300.13 MHz, C6D6,
Rea ction of Im id e 1, Im in e 4b, a n d La beled Am in e 6N.
A 100 µL portion of a 0.626 M (62.5 µmol) standard solution
of labeled amine 6N, in C6D6, was added to a screw top NMR
tube containing 400 µL of a 0.0912 M (36.5 µmol) standard
solution of imide 1 and 100 µL of a 0.33 M (33.0 µmol) standard
solution of imine 4b. The progress of the reaction was
monitored by 1H and 15N NMR spectroscopy at 80 °C for 3 days.
Resonances for the labeled imide/amide 7N1a were observed
initially. Over the course of the reaction, resonance for 1N,
4bN, and 7N2 were observed.
3
25 °C): δ 0.98 (br, 2H, 15NH2), 3.29 (d, 9H, CH3, J NH ) 2.3
Hz). 15N NMR (50.68 MHz, C6D6, 25 °C): δ -53.6 (NH2).
(15) TMS amines were synthesized by either (a) the addition of
TMSCl to the parent lithium amine or (b) the reaction of TMSCl with
the parent amine in basic solution.
(16) Glueck, D. S.; Wu, J .; Hollander, F. J .; Bergman, R. G. J . Am.
Chem. Soc. 1991, 113, 2041-2054.
(17) Imines were prepared by stirring a benzene solution of the
corresponding amine and aldehyde over molecular sieves, followed by
filtration, removal of solvent in vacuo, and vacuum distillation.
(18) Williams, D. N.; Mitchell, J . P.; Poole, A. D.; Siemeling, U.;
Clegg, W.; Hockless, D. C. R.; O’Neil, P. A.; Gibson, V. C. J . Chem.
Soc., Dalton Trans. 1992, 739-751.
Rea ction of Im id e 1, La beled Im in e 4bN, a n d Am in e
6. A 200 µL portion of a 0.472 M (94.4 µmol) standard solution
of amine in C6D6 was added to a screw top NMR tube
containing 97.4 mg (251 µmol) of imide 1 and 200 µL of a 1.24
M (248 µmol) standard solution of imine 4bN. The progress of
1
the reaction was monitored by H and 15N NMR spectroscopy
at 68 °C for 2 days. Resonances for the labeled imide 1N were
observed initially. Over the course of the reaction, resonances
for 7N1a and 7N2 were observed.

Metathesis of Imines by CpTa(dNR)Cl2
Organometallics, Vol. 21, No. 9, 2002 1941
Rea ction of Im id e 1 w ith La beled Am m on iu m Sa lt 8N.
A 74.4 mg (0.19 mmol) portion of imide 1 and 42.9 mg (0.39
mmol) of labeled ammonium salt 8N were added to a screw
top NMR tube. Approximately 0.5 mL of C6D6 was added to
give a yellow solution with a white precipitate. The sample
was heated at 70 °C for 2 weeks. 15N NMR spectroscopy
revealed a resonance for labeled imide 1N.
Rea ction of Am in e 6 w ith La beled Am m on iu m Sa lt
8N. A 55.8 mg (0.50 mmol) portion of labeled ammonium salt
was added to 0.5 mL of a 0.5 M (0.25 mmol) solution of amine
6. The sample was heated at 70 °C for 2 weeks. 15N NMR
spectroscopy revealed a resonance for labeled amine 6N.
Ack n ow led gm en t. This research was supported by
NSF award CHE-0091400. We thank Dr. Martin Minelli
and Dr. Fu-Tyan Lin for help in acquiring 14N and 15N
NMR spectral data.
OM011085V