Titanium-Catalyzed Hydroamination and Multicomponent Coupling with a Simple Silica-Supported Catalyst

Article abstract of DOI:10.1021/acs.organomet.8b00313

Hydroamination and multicomponent coupling reactions catalyzed by homogeneous Ti(IV) complexes can produce valuable imines, amines, and other nitrogen-containing organic building blocks. Typically catalysts for this transformation are very sensitive to an

Full text of DOI:10.1021/acs.organomet.8b00313

Article
pubs.acs.org/Organometallics
Cite This: Organometallics XXXX, XXX, XXX−XXX
Titanium-Catalyzed Hydroamination and Multicomponent Coupling
with a Simple Silica-Supported Catalyst
Kelly E. Aldrich and Aaron L. Odom*
Department of Chemistry, Michigan State University, 578 South Shaw Lane, East Lansing, Michigan 48823, United States
S
* Supporting Information
ABSTRACT: Hydroamination and multicomponent cou-
pling reactions catalyzed by homogeneous Ti(IV) complexes
can produce valuable imines, amines, and other nitrogen-
containing organic building blocks. Typically catalysts for this
transformation are very sensitive to ancillary ligand design and
often suffer from catalyst deactivation, necessitating use in a
single reaction. Here, we have attempted to address these
issues by moving toward a solid-supported Ti catalyst active
for these reaction types. We present initial results to
synthesize and probe the catalytic reactivity of silica-supported
titanium amide precatalysts. With minimal treatment of
200
commercially available fumed silica, Ti(NMe2)2/SiO2 can be isolated upon addition of Ti(NMe2)4. This species gives
high yields and high regioselectivity for hydroamination of a variety of alkynes with anilines. The solid-supported species is also
an active catalyst for the formation of tautomers of unsymmetrical 1,3-diimines via three-component coupling of bulky anilines,
alkynes, and isonitriles. Reusability and possible catalyst deactivation pathways are also discussed.
rates of many different titanium precatalysts were measured as
a function of donor ability and steric profile of the ancillary
ligands. The correlation demonstrated that electron-deficient
and small ancillary ligands enhance the rate of the catalyst,
quantitatively confirming previous qualitative observations.25,26
However, it was also observed in this study that, in making
sufficiently electron deficient or small ligands to enhance the
rate of these catalysts beyond the fastest known derivative,
pathways to side reactions or catalyst decomposition may be
introduced or amplified.25 Additionally, even with stable
precatalyst variants that perform HA and 3CC cleanly, the
catalyst would be difficult to recover from the crude reaction
mixture.
One possible method to circumvent some decomposition
pathways and allow catalyst recovery is to move towards a
solid-supported catalyst. Further, supported catalyst systems
would also be advantageous for incorporation of the chemistry
into flow systems.27,28 Solid-supported hydroamination and
related reactions have been demonstrated with other metals
capable of this transformation.25 For example, Basset et al. have
demonstrated that tantalum bound to silica, bearing
dimethylamide ligands, can perform hydroaminoalkylation of
alkenes through a proposed tantalaziridine species.29,30 Intra-
molecular hydroamination was also demonstrated by Le Roux
and co-workers using mesoporous silica-supported lanthanum
INTRODUCTION
■
Titanium- and zirconium-catalyzed hydroaminations (HA) are
well-established reactions that have been studied for several
decades.1−9 HA allows for the formation of new C−N bonds
across different types of C−C multiple bonds. This reaction
pathway also can be modified to generate multicomponent
coupling products by addition of a third reactant, which can
result in additional C−C or C−N bond formation.10−15
An example of a titanium-catalyzed multicomponent
coupling (Figure 1), iminoamination, involves the coupling
of an amine, alkyne, and isonitrile to form tautomers of 1,3-
diimines (3CC).10 These reactions are carried out with a
homogeneous catalyst, such as Ti(NMe2)2dpm, where dpm =
5,5-dimethyldipyrrolylmethane. A selection of titanium cata-
lysts active for hydroamination or related multicomponent
coupling reactions, such as Tonks and co-workers’ azo-based
reactions22 which use Mountford’s Ti(NPh)Cl2(py)3,23,24 are
also shown in Figure 1.
Some of the known homogeneous titanium HA catalyses are
fast; for example, Ti(NMe)2dpm catalyzes the HA of 1-hexyne
with aniline in ∼5 min, starting at room temperature, with 5
mol% loading.8 However, the catalysts can suffer from
dimerization processes and other deactivation pathways.25
These issues are frequently addressed using bulky ancillary
ligands and chelates to minimize loss of the active species
(Figure 1). Even with these common design principles,
prevention of these deactivation pathways and undesired
reactivity can still be challenging.
Special Issue: Organometallic Complexes of Electropositive Ele-
ments for Selective Synthesis
We recently reported a detailed analysis of ligand design
Received: May 14, 2018
issues with regard to titanium HA catalysis.25 In this study, the
© XXXX American Chemical Society
A
DOI: 10.1021/acs.organomet.8b00313
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
substrate scope accessible to the heterogeneous catalyst, and
catalyst reusability.
RESULTS AND DISCUSSION
■
Fumed silica was treated much as Beaudoin and Scott
described.38,40 The silica was partially dehydrogenated under
vacuum at 200 °C for 8 h to give SiO2200. They reported that
treatment with Ti(NMe2)4 gave a solid with 1.93 0.4 wt %
Ti with “0.49 Ti per hydroxyl (average of 12 experiments.) The
amount of HNMe2 liberated during grafting corresponds to
1.99 0.04 per Ti.”38 Similarly, we measured 2.33 0.12 wt %
Ti with our sample and 2.03
0.12 HNR2 per Ti(NR2)4
added (Scheme 1), suggesting a very similar material.
Scheme 1. Preparation and Some Properties of Ti(NR2)2/
200
SiO2 (R = Et, Me)
Figure 1. (top) Proposed catalytic cycle for homogeneous hydro-
amination of alkynes based on the Bergman mechanism and a three-
component coupling reaction, iminoamination. The reactions are
demonstrated with the known precatalyst Ti(NMe2)2dpm. (bottom)
Sampling of known Ti precatalysts for hydroamination and related
reactions.5,9,16−22
silylamides (Y, La, and Nd) under mild conditions.30,31
Additionally, hydroamination by heterogeneous catalysts has
been demonstrated by a few others, with a number of new
reports in the past few years, including late-metal nanoparticles
and gel-supported catalysts.32−35
There is precedence for the use of titanium bound to a solid
support material as an active catalytic species. For example,
solid-supported silica catalysts with surface-bound titanium
have been shown to be active catalysts for olefin epoxidation or
olefin polymerization (with cocatalysts) by Scott and co-
workers.36,37
200
The Ti(NEt2)4-based material Ti(NEt2)2/SiO2
was
prepared for ease of NMR titration of the HNEt2 byproduct
and determination of the number of hydroxyl sites added to
the metal center (Scheme 1). For the catalyses discussed
200
below, Ti(NMe2)2/SiO2 was used.
Hydroamination Activity of Ti(NMe2)2/SiO2200. Upon
addition of a primary amine to the Ti(NMe2)2/SiO2
200
material, a color change is observed in the material,
transitioning from pale yellow to deep reddish orange. This
is similar to the color change observed for homogeneous
precatalysts upon addition of a primary amine. Initially, 5 mol
% of the precatalyst at 110 °C in toluene was used; however,
hydroamination of 1-hexyne with aniline was incomplete after
48 h (Scheme 2, Conditions A), and the regioselectivity was
low at 4:1. At 140 °C (Conditions B), the same reaction was
complete in ∼16 h and, interestingly, gave a significantly higher
regioselectivity of 19:1. To investigate the effect of increasing
the temperature further, the solvent was changed to p-cymene,
and the reaction was run at 180 °C (Conditions C). Under
these conditions, the hydroamination of 1-octyne, which
replaced the lower-boiling 1-hexyne for this experiment, with
aniline is complete in <40 min and showed perfect
regioselectivity to the limit of our GC detector (>100:1).
In addition to HA reactions with terminal alkyl-substituted
alkynes, Conditions C (Scheme 2) provided the best outcomes
for HA of internal alkynes, as well. For example, HA of aniline
and diphenylacetylene takes more than 64 h to reach
In a separate report by Scott, they demonstrated that
Ti(NR2)4 (R = Me, Et) can bind to silica by protolytic
38
200
replacement of the amide ligands, yielding Ti(NR2)2/SiO2
.
This binding motif at the silica surface bears a strong electronic
resemblance to known homogeneous systems capable of HA
and 3CC chemistry. It was also observed that, under mild
conditions, titanium imide species can be formed from
titanium amides.38
Surface-bound titanium amides and imides on silica would
seem ideal species with which to translate titanium-catalyzed
HA and 3CC reactions to heterogeneous catalysis. This
prospect is especially attractive for the titanium systems
because of the thorough mechanistic work that has been done
to establish the interactions between substrate and metal in the
homogeneous systems. In addition to the mechanistic
foundation, there is also a precedence for recycling solid-
supported catalysts, which can easily be recovered from
catalytic reaction mixtures and reactivated with minimal
manipulation.39 Here, we present initial investigations into
solid-supported titanium species that can catalyze HA and
3CC related to our homogeneous systems, exploration of the
200
completion with Ti(NMe2)2/SiO2 at 140 °C (Conditions
B). At 180 °C, the reaction is complete in under 14 h. A similar
increase in both the regioselectivity and rate of HA was also
B
DOI: 10.1021/acs.organomet.8b00313
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
into the Ti−C bond of the azatitanacyclobutene intermediate
with these substrates and reaction conditions.
Scheme 2. Reaction Conditions Screened for Ti(NMe2)2/
SiO2200-Catalyzed Hydroamination and Resulting Product
Distributions
We attempted to manipulate these relative rates, to favor
3CC, by changing the electronics and sterics of the substrates.
Increasing the size of the isonitrile could potentially impede
the rate of formamidine formation. Unfortunately, many of the
large tertiary and aromatically substituted isonitriles examined
were unstable at these reaction temperatures, with masses for
isonitrile oligomers and HA product observed from these
reaction mixtures. We also examined the effects of using
smaller isonitriles; when n-octyl isonitrile was used, a result
similar to the that for the CyNC experiments was observed
with considerable amounts of both HA and FA products, but
essentially no 3CC product was observed.
Without evidence of improvement from manipulation of the
isonitrile, focus was shifted to the amine substrate. By
substantially increasing the size of the primary amine to 2,6-
dimethylaniline (1), the iminoamination reaction yielded 51%
of the 3CC product. HA and formamidine products (10−15%
each) were still observed by GC analysis of the reaction
mixture, but the distribution favored the desired product.
To further probe the steric effect of the primary amine, the
series of reactions in Table 2 was examined. The amine was
altered by two successive decreases in size near the nitrogen
with the amine substrates 2,5-dimethylaniline (2) and 3,5-
dimethylaniline (3), each used with 1-octyne and cyclohexyl
isonitrile in the iminoamination reaction. The results
demonstrated a marked difference in yield and product
distribution. Amine 2 gave the 3CC product as the most
abundant species generated in moderate overall yield (50%
yield); this is comparable to the 3CC yield obtained with 1. In
addition, similar to 1, iminoamination employing 2 shows
about 15% HA product and a similar amount of formamidine.
This accounts for roughly 85% of the starting primary amine in
the crude reaction mixture after 48 h for both 1 and 2.
Using 3,5-dimethylaniline (3), however, gave just 7% yield
for the 3CC product, close to stoichiometric. While HA and
formamidine products are observed by GC, the combination of
these products does not account for most of the initial mass of
the aniline added to the reaction, and many unknown masses
were observed by GC/MS, none in high abundance. In terms
of favoring the 3CC product, it appears that with Ti(NMe2)2/
a
b
Yields and isomer ratios determined by GC/FID. Due to high
reaction temperatures in conditions C, 1-hexyne was substituted with
1-octyne. Conditions C are the same as the General Hydroamination
Procedure in the Experimental Section.
c
noted with 1-phenylpropyne at these different reaction
temperatures. Conditions C generally provided the highest
regioselectivity, shortest reaction times, and comparable if not
improved yields relative to the milder temperatures examined.
Thus, Conditions C were applied to the screening of a variety
of substrates.
The substrate screening experiments indicate that the
catalyst is efficient for hydroamination of a variety of alkynes.
However, Ti(NMe2)2/SiO2 has demonstrated poor activity
with primary alkyl amines and electron-withdrawing anilines.
Additionally, with terminal alkynes, at these high reaction
temperatures, significant amounts of the alkyne trimerization
products can be observed by GC. Despite these inadequacies,
the catalyst seems to perform the hydroamination of most
anilines, including relatively bulky derivatives such as 2,6-
dimethylaniline. Again, this is accomplished with a variety of
alkynes in relatively short reaction times and high yields.
Overall, these reactions also demonstrate high regioselectivity,
comparable with that of many homogeneous catalysts (Table
1).8
Observing successful hydroamination with a surface-bound
titanium species was an important step toward the goal of
catalyzing the three-component coupling (3CC) iminoamina-
tion reaction with a surface-supported Ti catalyst. Namely, this
reactivity suggests that the surface-bound titanium can form an
imide and carry out [2 + 2]-cycloadditions with alkynes if the
same mechanism is operative.
200
200
SiO2 as the catalyst, substitution of the 2-position of the
aniline substrate, at a minimum, is required for a productive
iminoamination rate relative to side reactions.
Switching from a terminal to an internal alkyne, 1-
phenylpropyne, leads to results comparable with those for 1
and 2. The reaction shows a reduced propensity toward side
reactions (HA and FA) and a moderate yield of the 3CC
product (36%) in 48 h. The reaction proceeds quite slowly,
however, and there is a substantial amount of unreacted amine
observed after 48 h.
These results demonstrate that multicomponent coupling
reactions, such as iminoamination, are possible with a simple
silica-supported titanium catalyst. The easily prepared Ti-
200
Iminoamination Activity of Ti(NMe2)2/SiO2200. Despite
(NMe2)2/SiO2
catalyst required steric blocking on the
amine to inhibit side reactions observed in three-component
coupling but still provided the 3CC products in moderate
yield.
these promising HA results, the first attempts at 3CC with the
200
Ti(NMe2)2/SiO2
catalyst were largely unsuccessful. The
3CC of aniline, 1-octyne, and cyclohexylisonitrile resulted in
large amounts of HA product and formamidine (FA) (Figure
2), observed in roughly equivalent quantities by GC. Only a
trace amount of 3CC was observed. One can hypothesize that
the 1,1-insertion into the amide is much faster than insertion
Reusability Studies on Ti(NMe2)2/SiO2200. It was
discovered early in the catalytic activity studies that Ti-
200
(NMe2)2/SiO2 can be recovered and reused in subsequent
catalytic reactions by simply filtering the catalyst material from
C
DOI: 10.1021/acs.organomet.8b00313
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
a
200
Table 1. Hydroamination Reactions Examined with Ti(NMe2)2/SiO2
a
b
Reaction Conditions C (Scheme 1): 5 mol % Ti as Ti(NMe2)2/SiO2200, 180 °C, 40 min (entry 1) or 14 h (entries 2−14). Yields reported are
calibrated GC/FID yields relative to the isolated amine derivative or a close derivative; see the Supporting Information for more details. Reactions
went to full conversion, except entries 5, 6, 9, and 10. These products were not isolated to determine regioselectivity because of their poor yields;
c
the likely Markovnikov product is shown. See the Supporting Information for more details.
the reaction mixture, washing it with organic solvent, and
As is obvious from these results, the catalyst is in fact
undergoing deactivation in subsequent runs. In the first three
reactions, there is little deviation in yield; however, reactions 4
and 5 indicate a severe reduction in product formation under
identical conditions. While some of this may be attributed to
mass loss of the catalyst itself over the course of the five
reactions due to handling, it seemed likely that something else
removing residual volatiles under vacuum. Following this
200
simple processing step, the used Ti(NMe2)2/SiO2
was
directly employed in subsequent reactions. The activity in the
second and third catalytic runs was comparable to the initial
use in terms of product yield after the same reaction time. An
experiment to examine the limits of catalyst recycling was
200
conducted in which a sample of Ti(NMe2)2/SiO2 was used
is more detrimental to catalyst recycling.
Examination of the used Ti(NMe2)2/SiO2
200
five times to perform the reaction shown in entry 1 of Table 1.
catalyst by
The yields from each trial are shown in Table 3.
ICP-OES shows that titanium is lost from the surface of the
D
DOI: 10.1021/acs.organomet.8b00313
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
200
Table 3. Reusability of Ti(NMe2)2/SiO2 for HA of 1-
Octyne with Aniline
a
trial no.
yield (%)
regioselectivity ratio
1
2
3
4
5
98
96
90
44
23
71
>100:1
32:1
14:1
10:1
10:1
b
6
7.3:1
a
Yields are from GC-FID of the crude reaction mixture, using a
b
calibrated internal standard. Material was treated with Ti(NMe2)4 to
reactivate.
catalyst during hydroamination. A series of control reactions
were run to determine what reagent or other factor is primarily
responsible for leaching the titanium from the surface of the
silica support (Table 4). Simply heating Ti(NMe2)2/SiO2200 in
p-cymene at 180 °C for 16 h has no effect on the mass of
titanium in Ti(NMe2)2/SiO2200. Similarly, heating to 180 °C
with 20 equiv of NH2Ph has no effect on titanium
concentration on the silica surface.
Figure 2. Possible pathways for the three predominant reactions
during iminoamination: formation of formamidine (FA), hydro-
amination (HA), and iminoamination (3CC). The products from the
three reactions are labeled by their reaction of production.
200
Somewhat surprisingly, heating Ti(NMe2)2/SiO2 to 180
°C in the presence of 1-octyne (40 equiv) for 16 h resulted in a
loss of 11% of the titanium from Ti(NMe2)2/SiO2200. On the
a
200
Table 2. Three-Component Coupling Reactions Examined with Ti(NMe2)2/SiO2
a
b
Reaction conditions: 5 mol % Ti as Ti(NMe2)2/SiO2200, 1.5 equiv of cyclohexyl isonitrile, p-cymene, 180 °C, 48 h. Yields reported are calibrated
c
GC-FID yields relative to an isolated isomer of 3CC1. For more details, see the Supporting Information. Side products due to two-component
reactions: HA, hydroamination product; FA, formamidine product.
E
DOI: 10.1021/acs.organomet.8b00313
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
the solid catalyst was removed from the latter reaction by
filtration. The other reaction mixture was simply resealed after
sampling, with no further manipulation.
Table 4. Results of ICP-OES Analysis of
Ti(NMe2)2(SiO2)200 Samples after Different Treatments
treatment of
Ti found by ICP ( )
(wt %)
Ti retained after
treatment (%)
Ti(NMe2)2(SiO2)200
none
2.33 0.12
1.94 0.11
2.35 0.14
2.38 0.14
2.05 0.11
100
83
∼100
∼100
88
HA (entry 1, Table 1)
180 °C in p-cymene
180 °C, 20 equiv NH2Ph
180 °C, 40 equiv 1-
octyne
basis of these results, it seems that the loss of titanium from the
catalyst is related to reaction with the alkyne.
Interestingly, upon treatment of this deactivated Ti-
200
(NMe2)2/SiO2
with a slight excess of Ti(NMe2)4 and
thorough washing and drying the material under vacuum, a
substantial amount of the catalytic reactivity is regained. This
“regenerated” material gives a total HA yield of 71% under the
same conditions as the reactions given in Table 3.
Perhaps more interesting than the yields are the
regioselectivities in Table 3. The regioselectivity drops
consistently with use, which suggests that the catalytic species
in the system are changing with repeated use. Even with the
“regenerated” material, entry 6, the regioselectivity continues
to drop even as some of the yield was restored. Perhaps more
important than a practical application of this approach to
regenerate the catalyst material, this experiment suggests that,
in addition to some titanium leaching from the catalyst
material by substrates, the surface of the catalyst material may
be undergoing changes that lead to alterations in the titanium
environment. The ramifications of these changes can be
observed directly in the effects borne out in the catalysis.
Figure 3. Yield of hydroamination product vs time using the
conditions and substrates in entry 2 of Table 1. The “control”
experiment (solid circles) is as in entry 2 of Table 1. The “filtration”
experiment (solid squares) had the solid catalyst removed by filtration
at 20 min. The reactions indicated by diamonds and triangles have 0.8
and 1.2 equiv of pyrrole added as a catalyst poison, respectively. See
the text for more details.
Both reaction mixtures were then heated for an additional 20
min. After heating was carried out a second time, another GC-
FID analysis was performed. The reaction from which the
catalyst was filtered showed no change in product concen-
tration within experimental error, 2%. The control experi-
ment, which still contained the solid catalyst, went to 100%
conversion after the second round of heating (Figure 3). These
observations are consistent with a heterogeneous catalyst being
responsible for the reactivity observed.
Several control experiments were carried out to assess which
200
component of the Ti(NMe2)2/SiO2
catalyst system is
responsible for the catalytic activity observed: the bound
titanium, the free titanium leached into solution, or the solid
support itself. If the leached titanium is responsible for the
catalysis, the regioselectivity and activity should be similar to
those of Ti(NMe2)4, where all of the ligands are protolytically
labile.25 Thus, HA and 3CC reactions were screened with 5
mol% of Ti(NMe2)4 and with SiO2200 without titanium ligated
to the surface. While Ti(NMe2)4 is a catalyst for HA under
To further explore the nature of the catalytically active
species relative to known homogeneous systems, we explored
catalyst poisoning experiments. One specific aspect of
200
these reaction conditions, the regioselectivities for its products
Ti(NMe2)2/SiO2 that we wanted to scrutinize further was
200
200
versus those of the products when Ti(NMe2)2/SiO2
is
the local Ti environment on the SiO2 surface (Scheme 1)
employed are very different. For example, hydroamination of
and its mechanistic similarity to homogeneous species.
Characterization of the material suggests that two protolytically
cleavable sites remain on the titanium in the form of two
dimethylamide ligands. Two protolytically cleavable sites are
required for imide formation and catalytic activity through the
pathways in Scheme 1. Two experiments were conducted, in
which two different concentrations of pyrrole were added to
the Ti(NMe2)2/SiO2200. The first experiment was conducted
with a substoichiometric amount of pyrrole (0.8 equiv) relative
to titanium, and the second experiment used a slight excess
(1.2 equiv). The results of these studies are shown in Figure 3.
In both catalyst poisoning experiments, a substantial
decrease in catalytic activity was noted, relative to the control
reaction. The pyrrole-containing reactions appear <30% as
1-phenylpropyne with aniline using Ti(NMe2)4 and Ti-
200
(NMe2)2/SiO2
gave products with regioselectivities of
>100:1 and 8.7:1, respectively. Additionally, Ti(NMe2)4
demonstrates only a small amount of 3CC product formation
(Table 2, entry 1) under the same 3CC conditions (17%) as
200
for Ti(NMe2)2/SiO2 (52%), again with different regiose-
lectivity (see the Supporting Information for additional
details).
Another set of experiments was conducted to verify that the
catalytic activity is in fact from Ti(NMe2)2/SiO2200. Two
catalytic reactions were run side by side following entry 2,
Table 1. These reactions were prepared by following the
general procedure but were then heated for only 20 min. After
20 min of heating, the reaction mixtures were cooled to room
temperature, centrifuged, and transferred back into a glovebox.
GC-FID analysis of the crude reaction solutions showed 74%
and 67% conversion to HA2, with the same regioisomer
distributions of 8.6:1 and 8.8:1, respectively (Figure 3). Next,
200
active as Ti(NMe2)2/SiO2 itself. In both pyrrole-containing
reactions, it was noted that similar regioisomer ratios were
observed versus the reaction without pyrrole (∼9:1),
suggesting the same active species is facilitating the catalytic
transformation.
F
DOI: 10.1021/acs.organomet.8b00313
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Secondary amines appear to be inactive for hydroamination,
also consistent with imido formation being necessary.
species may provide the substrate scope, tolerance, and surface
stability needed to achieve a reusable and selective titanium
catalyst for multicomponent couplings.
Attempted hydroamination of 1-octyne and 1-phenylpropyne
200
with N-methylaniline using Ti(NMe2)2/SiO2
as catalyst
resulted in no product formation that was detectible by GC-
MS.
EXPERIMENTAL SECTION
■
General Procedures. All manipulations were carried out using
standard Schlenk techniques or in a glovebox with a purified nitrogen
atmosphere, unless otherwise specified. Anilines, amines, and alkynes
were purchased from commercial sources and distilled under N2 prior
to use. Cyclohexyl isonitrile was prepared according to literature
methods.47 Ti(NMe2)4 and Ti(NEt2)4 were purchased from Gelest
and used as received. Toluene and pentane(s) were sparged with N2
and dried by passing over a column of activated alumina prior to use.
p-Cymene was dried over CaH2 and distilled under reduced pressure
prior to use. Tetrahydrofuran was dried over sodium and distilled
under dinitrogen prior to use. NMR solvents were purchased from
Cambridge Isotope Laboratories and distilled before use. Sodium
cyanoborohydride and p-toluenesulfonic acid were purchased
commercially and used as received.
Overall, the control experiments strongly suggest that the
200
reactivity observed with the Ti(NMe2)2/SiO2 is due to the
bound titanium species, not the titanium that is leached from
200
the surface during the catalysis. The SiO2 demonstrates no
catalytic activity for either HA or 3CC products, which also
supports the assignment of catalytic activity to bound titanium
species. Additionally, the severe reduction in catalytic activity
noted when pyrrole is included as a poison in reaction
mixtures, as well as the lack of reactivity with secondary amine,
is consistent with a hydroamination mechanism involving a
titanium imido species, although it is certainly not conclusive
proof of such a pathway.
200
Preparation of SiO2200. Fumed silica (Sigma-Aldrich, 200 25
m2/g, lot no. SLBT0198) was treated similarly to the published
preparation by Scott and co-workers.38 Silica (20 g) was placed in a 1
L beaker and mixed with deionized water to form a slurry. This slurry
was air-dried at room temperature for 48 h and dried further in a 100
°C glassware oven for 48 h. The dried silica was finely divided using a
mortar and pestle until the material was free-flowing and had an even
appearance. The silica was transferred to a 200 mL beaker and dried
for 8 h in a vacuum oven at 200 °C. This silica was transferred to an
inert-atmosphere glovebox, where it was stored in a sealed container
until use. This resulted in silica (SiO2200) with a surface hydroxyl
On the basis of these results, Ti(NMe2)2/SiO2 is readily
recoverable and reusable two to three times with yields
comparable to that of freshly prepared catalyst. Additional
reuse results in a precipitous decline in performance.
CONCLUSIONS
■
In this study, we have demonstrated that titanium bound to
silica can perform some of the same hydroamination and
multicomponent coupling reactions catalyzed by homogeneous
titanium complexes. The catalyst itself is reusable several times
for hydroamination but eventually succumbs to surface
leaching of titanium that appears to be due to side reactions
with the alkyne.
group density of 0.52 0.03 mmol/g (1.58 0.2 sites/Å3) by H
1
NMR titration of the silica with Ti(NEt2)4 as judged by released
HNEt2.
Preparation of Ti(NMe2)2/SiO2200. In an inert-atmosphere
Aside from actual loss of the catalyst from the surface,
additional interactions with substrate and the silica surface
could be deactivating the catalyst. For example, in a manner
similar to that for homogeneous analogues, it could be possible
that some stable intermediate is achieved through side
reactions, involving irreversible binding of substrates to the
titanium itself. An example of this was presented by Mountford
and co-workers, whereby their titanium catalyst forms a stable
six-membered ring upon a second insertion of alkyne into the
azatitanacyclobutene.41 While the solid support inherently aids
in the prevention of pathways such as dimerization,
deactivation via substrate side reactions is obviously still
possible. Deactivation pathways such as these bear further
investigation as improvements to catalyst design are pursued.
200
glovebox, a 125 mL Erlenmeyer flask was charged with SiO2 (4
g, 2.08 mmol OH, 1 equiv) and 30 mL of pentanes. The slurry was
stirred, and Ti(NMe2)4 (800 mg, 1.2 equiv, 2.50 mmol) was added
dropwise. The mixture was stirred for 4 h at room temperature, during
which time the silica turned pale yellow. The solids were collected by
filtration, rinsed with 20 mL of benzene, and briefly dried in vacuo.
The material was stored in the glovebox. A separate experiment with
Ti(NEt2)4, which was used because of easier quantitation of HNEt2
relative to gaseous HNMe2, liberated 2.03 ( 0.12) molecules of
HNEt2 per molecule of Ti(NEt2)4 consumed, which correlates to 2.32
0.14 wt % Ti (see the Supporting Information for additional details
on this Ti(NEt2)4 experiment). These results agree with a doubly
bound titanium motif at the SiO2 surface as reported by Scott.38
200
200
ICP analysis for titanium in the Ti(NMe2)2/SiO2 material agrees
with the NMR titrations, giving a matching 2.33
0.12 wt % for
200
titanium. Obviously, the above gives an average of two attachment
sites for the silica to titanium, but there could be a distribution of
singly, doubly, and triply attached species along with other minor
variations, such as metal centers with HNR2 as a dative ligand.
However, the data are consistent with the doubly bound titanium on
silica variant as the major species on the surface. Additionally, in light
of the reactivity studies (vide infra), the presence of, on average, two
protolytically cleavable sites is a predominant feature of the catalyst
material. Again, this does not preclude the presence of a mixture of
bound products.
General Hydroamination Procedure. A pressure tube (15 mL)
was charged with 5 mol % of Ti(NMe2)2/SiO2200 (100 mg, 0.05 mmol
of Ti) and a Teflon stir bar. A solution containing NH2R (1 mmol)
and alkyne (1−2 mmol) in p-cymene was prepared in a 1 mL
volumetric flask. This solution was then placed in the pressure tube,
which was sealed and transferred from the glovebox to a preheated
aluminum well plate (180 °C). For terminal alkynes, an excess of the
alkyne was needed for the reactions to go to completion, as judged by
amine conversion in GC-FID; thus, 2 mmol was used in these cases.
For terminal alkynes, a larger amount of alkyne trimerization product
The heterogeneous catalyst Ti(NMe2)2/SiO2 was found
to be readily synthesized, allowed very fast to moderate
catalysis at 5 mol % loading with good stability at the
temperatures investigated (up to 180 °C), and was very
inexpensive to prepare. Thus, the catalytic activity of
200
Ti(NMe2)2/SiO2 shows promise for silica-based titanium
HA and multicomponent coupling reactions. As-prepared
Ti(NMe2)2/SiO2200 catalyzes HA, in some cases, with excellent
regioselectivity and high yield. The biggest drawback of
200
Ti(NMe2)2/SiO2 as an HA catalyst is the lack of tolerance
to alkyl amines and the undesired formation of alkyne trimers
noted with terminal alkynes.
The catalyst’s ability to successfully produce modest yields
of the 3CC products is significant. While there does appear to
be a steric demand that limits the choice of amine to ortho-
substituted anilines, this catalyst still provides a route to a
versatile organic building block.10−13,42−46 These results
suggest that further examination of surface-bound titanium
G
DOI: 10.1021/acs.organomet.8b00313
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
was observed. The pressure tube was heated for 14 h with magnetic
stirring. The pressure tube was ambiently cooled and then
centrifuged, compacting the Ti(NMe2)2/SiO2 into an intensely
colored orange to brown pellet at the bottom of the tube. This left a
transparent yellow to orange solution. See the Supporting Information
for characterization details of the organic products.
a bis(phosphinoselenoic amide) supported titanium(iv) complex.
Dalton Transactions 2016, 45, 17824−17832.
200
(8) Odom, A. L. New C-N and C-C bond forming reactions
catalyzed by titanium complexes. Dalton Transactions 2005, 225−233.
(9) Severin, R.; Doye, S. The catalytic hydroamination of alkynes.
Chem. Soc. Rev. 2007, 36, 1407−1420.
General Iminoamination Procedure. A pressure tube (15 mL)
was charged with 5 mol % of Ti(NMe2)2/SiO2200 (100 mg, 0.05 mmol
of Ti) and a Teflon stir bar. A solution of NH2R (1 mmol), alkyne
(1−2 mmol), and CyNC (1.5 mmol) was prepared in a 1.5 mL
volumetric flask, diluted to this volume with p-cymene. This solution
was then placed in the pressure tube, which was sealed and transferred
from the glovebox to a preheated (180 °C) aluminum well plate. The
pressure tube was heated for 48 h with magnetic stirring. The pressure
tube was cooled to ambient temperature and was centrifuged,
(10) Cao, C.; Shi, Y.; Odom, A. L. A Titanium-Catalyzed Three-
Component Coupling To Generate α,β-Unsaturated β-Iminoamines.
J. Am. Chem. Soc. 2003, 125, 2880−2881.
(11) Majumder, S.; Odom, A. L. Titanium catalyzed one-pot
multicomponent coupling reactions for direct access to substituted
pyrimidines. Tetrahedron 2010, 66, 3152−3158.
(12) Odom, A. L.; McDaniel, T. J. Titanium-Catalyzed Multi-
component Couplings: Efficient One-Pot Syntheses of Nitrogen
Heterocycles. Acc. Chem. Res. 2015, 48, 2822−2833.
200
compacting the Ti(NMe2)2/SiO2 into a dark brown pellet at the
(13) Majumder, S.; Gipson, K. R.; Staples, R. J.; Odom, A. L.
Pyrazole Synthesis Using a Titanium-Catalyzed Multicomponent
Coupling Reaction and Synthesis of Withasomnine. Adv. Synth. Catal.
2009, 351, 2013−2023.
bottom of the tube. This left a dark orange to opaque brown solution.
See the Supporting Information for characterization details of the
organic products.
(14) Vujkovic, N.; Fillol, J. L.; Ward, B. D.; Wadepohl, H.;
Mountford, P.; Gade, L. H. Insertions into Azatitanacyclobutenes:
New Insights into Three-Component Coupling Reactions Involving
Imidotitanium Intermediates. Organometallics 2008, 27, 2518−2528.
(15) Straub, B. F.; Bergman, R. G. The Mechanism of Hydro-
amination of Allenes, Alkynes, and Alkenes Catalyzed by Cyclo-
pentadienyltitanium−Imido Complexes: A Density Functional Study.
Angew. Chem., Int. Ed. 2001, 40, 4632−4635.
(16) Yim, J. C. H.; Bexrud, J. A.; Ayinla, R. O.; Leitch, D. C.;
Schafer, L. L. Bis(amidate)bis(amido) Titanium Complex: A
Regioselective Intermolecular Alkyne Hydroamination Catalyst. J.
Org. Chem. 2014, 79, 2015−2028.
(17) Shi, Y.; Hall, C.; Ciszewski, J. T.; Cao, C.; Odom, A. L.
Titanium dipyrrolylmethane derivatives: rapid intermolecular alkyne
hydroamination. Chem. Commun. 2003, 586−587.
(18) Ackermann, L. TiCl4-Catalyzed Intermolecular Hydroamina-
tion Reactions. Organometallics 2003, 22, 4367−4368.
(19) Ackermann, L.; Born, R. TiCl4/t-BuNH2 as the sole catalyst
for a hydroamination-based Fischer indole synthesis. Tetrahedron Lett.
2004, 45, 9541−9544.
(20) Bexrud, J. A.; Beard, J. D.; Leitch, D. C.; Schafer, L. L.
Intramolecular Hydroamination of Unactived Olefins with Ti-
(NMe2)4 as a Precatalyst. Org. Lett. 2005, 7, 1959−1962.
(21) Li, C.; Thomson, R. K.; Gillon, B.; Patrick, B. O.; Schafer, L. L.
Amidate complexes of titanium and zirconium: a new class of tunable
precatalysts for the hydroamination of alkynes. Chem. Commun. 2003,
2462−2463.
(22) Gilbert, Z. W.; Hue, R. J.; Tonks, I. A. Catalytic formal [2 +
2+1] synthesis of pyrroles from alkynes and diazenes via TiII/TiIV
redox catalysis. Nat. Chem. 2016, 8, 63.
(23) Blake, A. J.; Collier, P. E.; Dunn, S. C.; Li, W. S.; Mountford, P.;
Shishkin, O. V. Synthesis and imido-group exchange reactions of tert-
butylimidotitanium complexes. J. Chem. Soc., Dalton Trans. 1997,
1549−1558.
(24) Mountford, P. New titanium imido chemistry. Chem. Commun.
1997, 2127−2134.
(25) Billow, B. S.; McDaniel, T. J.; Odom, A. L. Quantifying ligand
effects in high-oxidation-state metal catalysis. Nat. Chem. 2017, 9, 837.
(26) Swartz, D. L.; Odom, A. L. Synthesis, Structure, and
Hydroamination Kinetics of (2,2′-Diaryldipyrrolylmethane)- and
Bis(2-arylpyrrolyl)titanium Complexes. Organometallics 2006, 25,
6125−6133.
(27) Britton, J.; Jamison, T. F. The Assembly and Use of Continuous
Flow Systems for Chemical Synthesis. Nat. Protoc. 2017, 12, 2423−
2446.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.organo-
met.8b00313.
■
S
Additional experimental and synthetic details and
spectral data for isolated products (PDF)
AUTHOR INFORMATION
Corresponding Author
*E-mail for A.L.O.: odom@chemistry.msu.edu.
ORCID
Aaron L. Odom: 0000-0001-8530-4561
Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS
■
The authors thank Aaron Sadow for helpful discussions and
the National Science Foundation for generous support (CHE-
1562140).
REFERENCES
■
(1) Walsh, P. J.; Hollander, F. J.; Bergman, R. G. Generation, alkyne
cycloaddition, arene carbon-hydrogen activation, nitrogen-hydrogen
activation and dative ligand trapping reactions of the first monomeric
imidozirconocene (Cp2Zr:NR) complexes. J. Am. Chem. Soc. 1988,
110, 8729−8731.
(2) Walsh, P. J.; Baranger, A. M.; Bergman, R. G. Stoichiometric and
catalytic hydroamination of alkynes and allene by zirconium bisamides
Cp2Zr(NHR)2. J. Am. Chem. Soc. 1992, 114, 1708−1719.
(3) Johnson, J. S.; Bergman, R. G. Imidotitanium Complexes as
Hydroamination Catalysts: Substantially Enhanced Reactivity from an
Unexpected Cyclopentadienide/Amide Ligand Exchange. J. Am.
Chem. Soc. 2001, 123, 2923−2924.
(4) Pohlki, F.; Doye, S. The Mechanism of the [Cp2TiMe2]-
Catalyzed Intermolecular Hydroamination of Alkynes. Angew. Chem.,
Int. Ed. 2001, 40, 2305−2308.
(5) Khedkar, V.; Tillack, A.; Beller, M. A Dramatic Effect of Aryloxo
Ligands on the Titanium-Catalyzed Hydroamination of Alkynes. Org.
Lett. 2003, 5, 4767−4770.
(6) Thomson, R. K.; Bexrud, J. A.; Schafer, L. L. A Pentagonal
Pyramidal Zirconium Imido Complex for Catalytic Hydroamination
of Unactivated Alkenes. Organometallics 2006, 25, 4069−4071.
(7) Bhattacharjee, J.; Das, S.; Kottalanka, R. K.; Panda, T. K.
Hydroamination of carbodiimides, isocyanates, and isothiocyanates by
(28) Movsisyan, M.; Delbeke, E. I. P.; Berton, J. K. E. T.;
Battilocchio, C.; Ley, S. V.; Stevens, C. V. Taming Hazardous
Chemistry by Continuous Flow Technology. Chem. Soc. Rev. 2016,
45, 4892−4928.
H
DOI: 10.1021/acs.organomet.8b00313
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
(29) Hamzaoui, B.; Pelletier, J. D. A.; Abou-Hamad, E.; Chen, Y.;
Eter, M. E.; Chermak, E.; Cavallo, L.; Basset, J. M. Solid-State NMR
and DFT Studies on the Formation of Well-Defined Silica-Supported
Tantallaaziridines: From Synthesis to Catalytic Application. Chem. -
Eur. J. 2016, 22, 3000−3008.
́
(30) Coperet, C.; Comas-Vives, A.; Conley, M. P.; Estes, D. P.;
́
̃
Fedorov, A.; Mougel, V.; Nagae, H.; Nunez-Zarur, F.; Zhizhko, P. A.
Surface Organometallic and Coordination Chemistry toward Single-
Site Heterogeneous Catalysts: Strategies, Methods, Structures, and
Activities. Chem. Rev. 2016, 116, 323−421.
(31) Le Roux, E.; Liang, Y.; Storz, M. P.; Anwander, R.
Intramolecular Hydroamination/Cyclization of Aminoalkenes Cata-
lyzed by Ln[N(SiMe3)2]3 Grafted onto Periodic Mesoporous Silicas.
J. Am. Chem. Soc. 2010, 132, 16368−16371.
(32) Ng, E.-P.; Law, S.-P.; Mukti, R. R.; Juan, J.-C.; Adam, F.
Hydroamination of cyclohexene enhanced by ZnCl2 nanoparticles
supported on chiral mesoporous silica. Chem. Eng. J. 2014, 243, 99−
107.
(33) Davaasuren, B.; Emwas, A.-H.; Rothenberger, A. MAu2GeS4-
Chalcogel (M = Co, Ni): Heterogeneous Intra- and Intermolecular
Hydroamination Catalysts. Inorg. Chem. 2017, 56, 9609−9616.
(34) Sengupta, M.; Bag, A.; Das, S.; Shukla, A.; Konathala, L. N. S.;
Naidu, C. A.; Bordoloi, A. Reaction and Mechanistic Studies of
Heterogeneous Hydroamination over Support-Stabilized Gold Nano-
particles. ChemCatChem 2016, 8, 3121−3130.
́
(35) Cook, A. K.; Coperet, C. Alkyne Hydroamination Catalyzed by
Silica-Supported Isolated Zn(II) Sites. Organometallics 2018, 37,
1342.
(36) Sensarma, S.; Bouh, A. O.; Scott, S. L.; Alper, H. Olefin
epoxidation catalyzed by an air stable-supported titanium catalyst. J.
Mol. Catal. A: Chem. 2003, 203, 145−152.
(37) Scott, S. L.; Church, T. L.; Nguyen, D. H.; Mader, E. A.;
Moran, J. An investigation of catalyst/cocatalyst/support interactions
in silica-supported olefin polymerization catalysts based on
Cp*TiMe3*. Top. Catal. 2005, 34, 109−120.
(38) Beaudoin, M.; Scott, S. L. Spontaneous Evolution of Silica-
Supported Ti Amide Fragments to Imine and Imido Complexes.
Organometallics 2001, 20, 237−239.
(39) Eedugurala, N.; Wang, Z.; Chaudhary, U.; Nelson, N.; Kandel,
K.; Kobayashi, T.; Slowing, I. I.; Pruski, M.; Sadow, A. D. Mesoporous
Silica-Supported Amidozirconium-Catalyzed Carbonyl Hydrobora-
tion. ACS Catal. 2015, 5, 7399−7414.
(40) Bouh, A. O.; Rice, G. L.; Scott, S. L. Mono- and Dinuclear
Silica-Supported Titanium(IV) Complexes and the Effect of TiOTi
Connectivity on Reactivity. J. Am. Chem. Soc. 1999, 121, 7201−7210.
(41) Vujkovic, N.; Ward, B. D.; Maisse-Franco̧ is, A.; Wadepohl, H.;
Mountford, P.; Gade, L. H. Imido-Alkyne Coupling in Titanium
Complexes: New Insights into the Alkyne Hydroamination Reaction.
Organometallics 2007, 26, 5522−5534.
(42) Pasko, C. M.; Dissanayake, A. A.; Billow, B. S.; Odom, A. L.
One-pot synthesis of pyrroles using a titanium-catalyzed multi-
component coupling procedure. Tetrahedron 2016, 72, 1168−1176.
(43) Dissanayake, A. A.; Staples, R. J.; Odom, A. L. Titanium-
Catalyzed, One-Pot Synthesis of 2-Amino-3-cyano- pyridines. Adv.
Synth. Catal. 2014, 356, 1811−1822.
(44) Dissanayake, A. A.; Odom, A. L. Regioselective conversion of
alkynes to 4-substituted and 3,4-disubstituted isoxazoles using
titanium-catalyzed multicomponent coupling reactions. Tetrahedron
2012, 68, 807−812.
(45) Barnea, E.; Majumder, S.; Staples, R. J.; Odom, A. L. One-Step
Route to 2,3-Diaminopyrroles Using a Titanium-Catalyzed Four-
Component Coupling. Organometallics 2009, 28, 3876−3881.
(46) Majumder, S.; Gipson, K. R.; Odom, A. L. A Multicomponent
Coupling Sequence for Direct Access to Substituted Quinolines. Org.
Lett. 2009, 11, 4720−4723.
(47) Ugi, I.; Meyer, R.; Lipinski, M.; Bodesheim, F.; Rosendahl, F.
Cyclohexyl Isocyanide. Org. Synth. 1961, 41, 13.
I
DOI: 10.1021/acs.organomet.8b00313
Organometallics XXXX, XXX, XXX−XXX