Solid-State NMR and DFT Studies on the Formation of Well-Defined Silica-Supported Tantallaaziridines: From Synthesis to Catalytic Application

Article abstract of DOI:10.1002/chem.201504439

Single-site, well-defined, silica-supported tantallaaziridine intermediates [≡Si-O-Ta(η²-NRCH₂)(NMe₂)₂] [R=Me (2), Ph (3)] were prepared from silica-supported tetrakis(dimethylamido)tantalum [≡Si-O-Ta(NMe_2<

Full text of DOI:10.1002/chem.201504439

DOI: 10.1002/chem.201504439
Full Paper
&
Organometallic Chemistry
Solid-State NMR and DFT Studies on the Formation of Well-
Defined Silica-Supported Tantallaaziridines: From Synthesis to
Catalytic Application
Bilel Hamzaoui, JØrØmie D. A. Pelletier, Edy Abou-Hamad, Yin Chen, Mohamed El Eter,
Edrisse Chermak, Luigi Cavallo, and Jean-Marie Basset*[a]
Abstract: Single-site, well-defined, silica-supported tantallaa-
ziridine intermediates [ꢀSi-O-Ta(h2-NRCH2)(NMe2)2] [R=Me
(2), Ph (3)] were prepared from silica-supported tetrakis(di-
methylamido)tantalum [ꢀSi-O-Ta(NMe2)4] (1) and fully charac-
formation mechanism, by b-H abstraction, was investigated
by SS NMR spectroscopy and supported by DFT calculations.
The CÀH activation of the dimethylamide ligand is favored
for R=Ph. The results from catalytic testing in the hydroami-
noalkylation of alkenes were consistent with the N-alkyl aryl
amine substrates being more efficient than N-dialkyl amines.
1
terized by FTIR spectroscopy, elemental analysis, and H,13C
HETCOR and DQ TQ solid-state (SS) NMR spectroscopy. The
Introduction
early transition metal dialkylamido complexes [M(NR2)4] (M=Zr,
Nb, etc.; R=Me, Et).[2b] Deuterium incorporation at the a posi-
tion to the nitrogen atom was consistently observed and
proved the formation (by b-H elimination) of a transient metal-
laaziridine complex intermediate. Since then, metallaaziridine
complexes of various transition metals have been involved in
several organic transformations. Examples of such stoichiomet-
ric reactions have been reported by Buchwald et al.,[5] Norton
et al.,[6] Whitby et al.,[7] and Blagg et al.[7d,8] Typically, metallaa-
ziridines can generate, after workup, functionalized amines by
reaction with unsaturated CÀC bonds (olefins and acetylenes)
and CÀX bonds (aldehydes, carbonates, and imines).
Metallaaziridine complexes have attracted increasing attention
as models for key intermediates in reactions involving metal
amide species.[1] The formation mechanism of MÀC bonds ob-
served for complexes M(h2-NRCHR’) (R=alkyl, R’=H, alkyl)
from homoleptic M(NR2)n is a fascinating organometallic pro-
cess. It is particularly relevant to hydroaminoalkylation cataly-
sis, in which olefins are incorporated into the metal–carbon
bond of metallaaziridine complexes [Eq. (1)]. This reaction is re-
garded as a highly atom economical method for synthesizing
a broad range of substituted secondary amines under relatively
mild conditions.
These results fueled interest in a better understanding of
the reaction mechanism, with a view to improving the reactivi-
ty and selectivity of the catalysts. A series of mechanistic inves-
tigations have been reported including kinetic studies,[9] iso-
tope distribution studies,[2b] and solid-state (SS) NMR experi-
ments.[16] The formation of metallaaziridine intermediates was
postulated as the key step in the catalytic cycle. Such species
have been formally elucidated in only a few reports.[10] For in-
stance, [Cp*Ta(NMe2)Me3] undergoes intramolecular b-H migra-
tion to give [Cp*Ta(h2-CH2NMe)Me2] with release of methane.[11]
Similar azametallacyclopropane complexes have been isolated
and structurally elucidated by single-crystal X-ray crystallogra-
phy (Scheme 1).
Complexes of Groups 4,[2] 5,[2d,3] and 8[4] have been docu-
mented as efficient catalysts for hydroaminoalkylation. A gen-
eral mechanism was initially proposed by Nugent et al. on the
basis of deuterium exchange experiments with homoleptic
Herzon and Hartwig examined the reactivity of tantalum
amido derivatives as catalysts for the a-alkylation of alkyl aryl
amines and dialkyl amines with terminal olefin substrates.[3a]
They showed that the selectivity of hydroaminoalkylation reac-
tions with alkenes can be affected by the electronic properties
of the amine. N-Aryl substrates consistently led to higher yields
than the dialkyl analogues.[3a] Whitby and co-workers explained
this improved reactivity by the lone pair of nitrogen being less
available in N-aryl than in N-alkyl amines for donation to the
[a] B. Hamzaoui, Dr. J. D. A. Pelletier, Dr. E. Abou-Hamad, Dr. Y. Chen,
Dr. M. El Eter, Dr. E. Chermak, Prof. L. Cavallo, Prof. J.-M. Basset
King Abdullah University of Science and Technology (KAUST)
KAUST Catalysis Center (KCC), Thuwal, 23955-6900, (Saudi Arabia)
E-mail: jeanmarie.basset@kaust.edu.sa
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201504439.
Chem. Eur. J. 2016, 22, 3000 – 3008
3000
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Ta(h2ÀNRCH2)(NMe2)2] [R=Me (2), Ph (3)] (Scheme 2) starting
from [ꢀSi-O-Ta(NMe2)4] (1). All surface species were character-
ized by SS NMR spectroscopy, elemental analysis, and FTIR
spectroscopy.
Results and Discussion
Scheme 1. Previously reported molecular metallaaziridine complexes of
Group 4 and 5 transition metals.
Treating SiO2-(700) with pentakis(dimethylamido)tantalum in
pentane at ambient temperature for 1 h led to the formation
of 1 as a yellow powder. Comparison of FTIR spectra of SiO2-
metal center due to conjugation with the aromatic p system.[7a]
This observation is in agreement with the work by Herzon and
Hartwig in which [Ta(NMe2)5] and [{Cl3Ta(NMePh)2}2] were
tested as precatalysts for the reaction of N-(methyl-d3)-aniline
with 1-octene. Interestingly, the highest degree of incorpora-
tion of deuterium on the a-carbon atom was observed when
[{Cl3Ta(NMePh)2}2] was used as precatalyst.[12] Molecular metal-
laaziridine complexes have been documented for two decades,
but have only recently been explored within the framework of
surface organometallic chemistry. This field has a solid track
record in isolating and identifying transition metal catalysts
with unusual reactivities.[13] Recent advances include the isola-
tion of single-site, well-defined surface species such as [(ꢀSi-O-
Ta(=CH2)Me2] (an active catalyst for alkane metathesis)[14] and
[ꢀSi-O-Zr(HNMe2)(h2NMeCH2)(NMe2)].[15] The reactivity of the
latter species was harnessed, and the surface monohydride
[ꢀSi-O-Zr(HNMe2)(NMe2)2H] was selectively isolated and shown
to be an active catalyst for alkene hydrogenation.[15a] Addition-
al studies revealed that [ꢀSi-O-Zr(HNMe2)(h2-NMeCH2)(NMe2)]
was able to catalyze the hydroaminoalkylation of propylene
with dimethylamide, albeit in modest yields.[16]
and 1 indicated grafting of [Ta(NMe2)5] onto silica. The
(700)
signal corresponding to the isolated silanol groups in SiO2-(700)
(3745 cmÀ1 nOÀH) completely disappeared in 1. New bands ap-
peared in the regions corresponding to n(C-H) (3000–2800 cmÀ1
)
and dCÀH (1500–1300 cmÀ1). They can be assigned to dimethyl-
amine ligand vibrations (Figure 1).[15b] The combination and
overtone bands of the silica (1990, 1867, and 1639 cmÀ1) re-
mained unchanged (Figure 1). Elemental analysis of 1 gave
4.8% Ta, 2.7% C, and 1.6% N with a molar ratio of Ta/C/N=1/
8.4/4.2 (Æ0.3). Considering the amounts of grafted tantalum
and silanol groups available for grafting (0.27 and 0.30 mmol,
respectively) and the disappearance of vibrations due to silanol
groups, we can propose that for 1 a tantalum complex was
grafted to most of the isolated silanol groups. The Ta/N ratio
of 1/4.2 (Æ0.3) is consistent with a monopodal surface com-
1
plex with four NMe2 ligands. The H SS NMR spectrum of 1 ex-
hibits a single broad signal centered at 3.2 ppm (Supporting
Information, Figure S1); 13C SS NMR spectroscopy revealed one
major signal at 45 ppm (Supporting Information, Figure S2).
Both spectra are consistent with quasi-equivalent CH3 frag-
ments bound to nitrogen atoms.[11] A tetrakis-dimethylamido
monopodal structure [ꢀSi-O-Ta(NMe2)4] can be confirmed for
tantalum surface complex 1 (formally d0 10e). Interestingly,
a small peak at 35 ppm (13C SS NMR) and a shoulder at
2.5 ppm (1H SS NMR) can be interpreted as one NMe2 group in
axial position.[11] This result illustrates a slightly different reac-
tivity than that observed in our previous studies on the graft-
ing of [Zr(NMe2)4] on SBA15-(700) (formally d0 8e).[15b] In this case,
the azametallacyclopropane fragment is formed spontaneously
These results encouraged us to study surface metallaaziri-
dine complexes as other candidates for surface hydroaminoal-
kylation catalysts. Tantalum was considered first due to the
better performance of its molecular complexes over zirconium
counterparts.[2b,3a] The effect of the substrate employed (N-dia-
lkyl versus N-alkyl N-aryl) over the performance of active cata-
lytic species was also of interest (see above). Herein, we report
the preparation and the characterization of the first single-site,
well-defined,
silica-supported
tantallaaziridines
[ꢀSi-O-
at
room
temperature
to
afford
[ꢀSi-O-Zr(h2-
Scheme 2. Preparation pathways of 1–3. SiO2-(700): silica partially dehydroxylated at 7008C.
Chem. Eur. J. 2016, 22, 3000 – 3008 www.chemeurj.org
3001
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 1. FTIR spectra of SiO2-(700) and 1–3.
NMeCH2)(NHMe2)NMe2] (formally d0 10e, like 1). Both zirconium
and tantalum dimethylamido complexes exhibit similar ten-
dencies towards b-hydride elimination and metallacyclization.
The lower reactivity of zirconium complexes is explained by co-
ordination of one HNMe2 ligand released during grafting. It
was rationalized that the lower reactivity of [ꢀSi-O-Ta(NMe2)4]
regarding activation of the N-methyl ligand might be over-
come by controlled heating under vacuum. Thermal treatment
of 1 was conducted at 10À4 mbar (Scheme 2) to give 2. Initial
evacuation of 1 at room temperature left the IR spectrum un-
changed, although evacuation at a higher temperature (1508C)
resulted in diminishing intensity of signals in the alkyl vibration
region (Figure 1). The volatile substances identified by GC con-
sisted exclusively of HNMe2. Elemental analysis of 2 gave 4.8%
Ta, 2.0% C, and 1.2% N with a Ta/C/N ratio of 1/6.1/3.1 (Æ0.3).
This strongly suggests that one molecule of HNMe2 was elimi-
nated from 1.
group at 85 ppm. It is known that the signals of such methyl-
ene groups with highly restricted mobility are difficult to
detect by SS NMR.[17] The HETCOR spectrum revealed a clear
correlation between the proton signal at 2.3 ppm and the
carbon signals at 34 and 45 ppm (Figure 2E), which confirms
that all the signals observed in the 13C SS NMR spectrum can
1
be assigned to NMe carbon atoms. The H,1H double-quantum
(DQ) and triple-quantum (TQ) methods were applied to 2 to
confirm the presence of the TaCH2 fragment (Figure 2C and
2D). The peak at 3.2 ppm shows autocorrelation in the DQ
spectrum but not in the TQ spectrum, and it can be thus as-
signed to a TaCH2 fragment. Examples of NMR characterization
of related molecular complexes containing an MCH2 fragment
have been documented;[18] the corresponding 1H signal has
been observed in the range of 1.08–2.07 ppm, slightly lower
than that observed for 2.
The peak at 2.3 ppm displays a strong autocorrelation, vali-
dating the NMe2 assignment. Thus, we can propose with cer-
tainty the structure [ꢀSi-O-Ta(h2-NMeCH2)(NMe2)2] for 2. Its for-
mation can be explained by an intramolecular b-H abstraction
of one dimethylamido ligand (Scheme 2). This pathway has
Furthermore, a new band was observed at 1490 cmÀ1 and
the small band of 1 visible at 1422 cmÀ1 disappeared in 2.
These bands correspond to vibration of the CÀH bond of the
CH2 group formed in 2.[16] In comparison to [ꢀSi-O-
Zr(HNMe2)(h2-NMeCH2)(NMe2)], the wavenumber of this CH2
band is lower, as expected.
strong
similarities
to that
described
for
[ꢀSi-O-
Zr(HNMe2)(h2ÀNMeCH2)(NMe2)], but requires more intense con-
ditions (1508C versus RT). Taking in account the work of
Whitby et al.,[7a] we postulated that substituting an NMe(alkyl)
ligand by an NMePh ligand may improve the reactivity of the
corresponding grafted species toward the formation of the
tantallaaziridine complex. Hence,[ꢀSi-O-Ta(NMe2)4] (1) was
treated with one equivalent of methyl(phenyl)amine in pen-
tane at room temperature. After 4 h of reaction, the powder
was washed three times with pentane and the volatile sub-
stances were removed in vacuo. The resulting powder had
a brighter yellow color than 1. The exclusive gas-phase prod-
uct detected by GC was HNMe2. FTIR spectra of the solid prod-
1
Both H and 13C SS NMR spectra of 2 are notably more com-
1
plex than that of 1. The H NMR spectrum of 2 contains an in-
tense peak at 2.3 ppm and a broad peak at 3.2 ppm (Fig-
ure 2A). The 2.3 ppm signal is attributed to the NCH3 frag-
ments, as in 1. The resonance at 3.2 ppm can be assigned ten-
tatively to the TaCH2 moieties (the ZrCH2 analogue shows a sim-
ilar resonance at 3.2 ppm).[15b] The 13C SS NMR spectra of the
solid product shows two signals: a large one at 34 ppm and
a smaller one at 45 ppm (Figure 2B). These two peaks are rem-
iniscent of those previously seen for zirconium,[15b] except for
the absence of the broad signal corresponding to the CH2
Chem. Eur. J. 2016, 22, 3000 – 3008
www.chemeurj.org
3002
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
1
1
Figure 2. A) 1D H spin-echo MAS SS NMR spectrum of 2. B) 13C CP/MAS NMR spectrum of 2. C) 2D H,1H DQ/SQ, D) 1H,1H TQ/SQ, and E) 2D CP/MAS HETCOR
NMR spectra (see Supporting Information for details).
uct 3 showed that the bands characteristic of the aryl amide li-
gands are clearly visible (Figure 1). A very broad signal at
3061 cmÀ1 (aromatic CH bending) and another sharp one at
1594 cmÀ1 (nC ) could be detected. Elemental analysis of 3
chemical shift is smaller than that of 2 (3.2 ppm). This can be
explained by the presence of the aromatic ring in 3. Indeed,
the mesomeric effects of the aromatic ring (as an electron
donor) can induce a shift to lower frequency.[7a] Note that gen-
eration of an h2-imine complex by activation of the methyl
group of the amido ligand occurred at room temperature in
this case.
=
C
gave 5.1% Ta, 3.8% C, and 1.3% N with a Ta/C/N ratio of 1/
11.2/3.2 (Æ0.3).This indicates that methyl(phenyl)amine reacted
1
with 1. The H SS NMR spectrum of 3 (Figure 3A) exhibits pre-
dominantly one major signal at 3.4 ppm with two overlapping
minor ones at 2.0 and 6.9 ppm. The 13C SS NMR spectrum (Fig-
ure 3B) shows one very intense resonance at 45 ppm overlap-
ping with a weaker one at 35 ppm in addition to two small sig-
nals at 117 and 128 ppm. The two last-named resonances have
chemical shifts consistent with overlapping aromatic carbon
resonances, compatible with the resonance at 6.9 ppm in the
1H NMR (Figure 3A).
To shed light on the generation mechanism of the metalla-
aziridines [ꢀSi-O-Ta(h2-NMeCH2)(NMe2)2] (2) and [ꢀSi-O
Ta(h2NPhCH2)(NMe2)2] (3) from [ꢀSi-O-Ta(NMe2)4] (1), DFT calcu-
lations were conducted for their respective formation path-
ways including b-hydride elimination. Consistent with previous
work validating the cluster model,[20] the surface of the silica
support was modeled by a silsesquioxane Si8O12H8 cage struc-
ture with one [ꢀSi-O-Ta(NMe2)4] moiety grafted to one of the
SiOH groups (see Figure 4). The initial [ꢀSi-O-Ta(NMe2)4] species
This is confirmed by the correlation observed in the HETCOR
1
spectrum between these H and 13C signals (Figure 3E). More-
(1) was considered as the reference structure at 0 kcalmolÀ1
.
over, the DQ and TQ experiments showed no autocorrelation
for the 6.9 ppm resonance, as expected for a CH proton (Fig-
ure 3C and 3D). These observations are in line with the pres-
ence of an aromatic ligand in 3. The HETCOR spectrum indi-
cates a correlation between the signals at 35 and 45 ppm (13C
The formation of metallacyclic species 2 and 3 was found to
follow two different reaction pathways (Figure 5). For that
leading to methyl-substituted metallaaziridine
2 (blue in
Figure 5), a transition state (1–2 in Figure 5) was considered
corresponding to the transfer of a hydrogen atom from the
carbon atom of one of the NMe2 ligands to the nitrogen atom
of another NMe2 ligand.
1
SS NMR) and that at 3.4 ppm in the H SS NMR spectrum (Fig-
ure 3E). Furthermore, the DQ and TQ spectra both show auto-
correlation for the 3.4 ppm resonance, confirming the presence
of CH3 fragments (Figure 3C and D). Hence, the signals can be
assigned to NCH3 of the aminomethyl ligands. Finally, the
signal at 2.0 ppm (1H) autocorrelates in the DQ but not in the
TQ spectrum and can thus be attributed to a CH2 group.[19] Its
The overall geometry of the ligands around the tantalum
center in 1–2 is best described as octahedral. Transition state
1–2 is located 32.8 kcalmolÀ1 above the starting species 1. This
relatively high barrier is in agreement with the experimental
reaction requiring heating at 1508C. Converging the geometry
Chem. Eur. J. 2016, 22, 3000 – 3008
www.chemeurj.org
3003
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
1
1
Figure 3. A) 1D H spin-echo MAS SS NMR spectrum of 3. B) 13C CP/MAS NMR spectrum of 3. C) 2D H,1H DQ/SQ, D) 1H,1H TQ/SQ, and E) 2D CP/MAS HETCOR
NMR spectrum (see Supporting Information for details).
Ta(NMe2)3(NPhMe)] (1b), which has not been observed experi-
mentally. The second step consists of hydrogen transfer from
the NPhMe ligand to one of the NMe2 ligands, similar to the
conversion of 1 to 2. In the first step, transition state 1–1b has
an energy barrier of 21.2 kcalmolÀ1 relative to the initial struc-
ture 1. This relatively low barrier is consistent with the reaction
occurring at room temperature. Transition state 1–1b involves
hydrogen transfer from the nitrogen atom of the external
HN(Ph)(Me) ligand to the nitrogen atom of one of the NMe2 li-
gands. This hydrogen transfer leads to the dissociation of
HNMe2 and to the formation of intermediate 1b, which lies
above 1 by 2.9 kcalmolÀ1. The second transition state 1b–3
(Figure 5) is obtained by hydrogen transfer between the
carbon atom of the N(Ph)(Me) ligand to one of the NMe2 li-
gands, leading to metallacycle 3. Transition state 1b–3 is locat-
ed 30.1 and 27.2 kcalmolÀ1 above 1 and 1b, respectively. It is
noteworthy that the reaction was run under high vacuum, in
which the formed HNMe2 is removed. Hence, it can be as-
sumed that the most correct estimate for the overall barrier
corresponding to the formation of 3 is that of 27.2 kcalmolÀ1
corresponding to the energy difference between transition
state 1b–3 and intermediate 1b, rather than the largest
energy difference between the highest-energy transition state
1b–3 and the most stable intermediate 1 (30.1 kcalmolÀ1). In
this scenario, the barrier of 27.2 kcalmolÀ1 can nonetheless be
Figure 4. Structures of key transition states along the reaction pathway in-
volving HN(Ph)(Me) with some distances [Š].
of transition state 1–2 onto the products side leads then to
the dissociation of HNMe2 and the formation of metallacycle 2;
the total Gibbs free-energy cost is 18.0 kcalmolÀ1 relative to
the initial species 1.
For the pathway leading to phenyl-substituted metallaaziri-
dine (green in Figure 5), two steps were considered starting
with the ligand exchange between NMe2 and NPhMe followed
by the b-H elimination. HN(Ph)(Me) can replace a NMe2 ligand
on the tantalum center of 1 leading to intermediate [ꢀSi-O-
Chem. Eur. J. 2016, 22, 3000 – 3008
www.chemeurj.org
3004
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 5. DFT reaction profiles for the transformation of 1 into 2 or 3.
considered consistent with this reaction occurring at room
temperature. Finally, converging transition state 1b–3 on the
products side leads to the [ꢀSi-O-Ta(h2-NPhCH2)(NMe2)2] metal-
lacycle 3, 14.6 kcalmolÀ1 above 1 in terms of free energy. In
conclusion, our calculations show that the presence of an aryl
group on the NR2 ligand reduces the free-energy barrier for
metallacycle formation from 32.8 to 27.2 kcalmolÀ1, which fol-
lows the trend of the experimental observations.
Table 1. Hydroamination substrate testing with catalysts 1–3.
Entry
Aryl amine
Catalyst
Yield[a] [%]
24 h
3 d
38
6 d
41
Further, we also carried out DFT-based NMR calculations to
strengthen the assignments of the NMR peaks (Supporting In-
formation, Table S5). Corroborating the interpretation of the
experimental results, the calculated chemical shifts of the C
atom of the metallacycle TaCH2 moiety in 2 and 3 (respectively
61.0 and 58.2 ppm) are shifted by roughly 10 ppm relative to
the average chemical shift of the C atoms of the Me groups of
TaNMe2 in 2 and 3 (47.80 and 48.08 ppm, respectively).
Species 1–3 were considered as catalysts for hydroamination
reactions. The screening was conducted with selected sub-
strates such as dialkyl amines (HNMeR; R=Pr, iBu) and N-
methyl anilines (HNMe(4-R-Ar); R=H, OMe, OCF3, Cl). All runs
were carried out under similar conditions [1658C, 1-octene
(1.5 equiv), 4 mol% catalyst, toluene (0.4 mL)] and terminated
after 1, 3, or 6 d (Table 1). Since the branched alkylation prod-
ucts were the sole regioisomers detected, insertion occurring
in 1,2 fashion. Precatalysts 1–3 were screened with methyl-
(phenyl)amine and formed the branched hydroaminoalkylation
product (Table 1, entries 1–3). All precatalysts resulted in com-
parable yields (typically about 45% after 6 d) regardless of the
duration of the experiments. This indicates that the catalytical-
ly active species is likely to be the same in Table 1 (entries 1–3)
and assumed to be 3.
1
2
3
1
2
3
23
19
22
29
35
42
46
4
5
1
1
20
21
42
39
51
48
6
1
31
42
53
7
8
9
1
1
3
10
12
12
26
27
25
30
33
31
[a] Determined by GC-FID with amine as default product. The yields
based on the olefin substrate were similar.
Reactions with electron-rich N-methylanilines such as (4-
chlorophenyl)methylamine (Table 1, entry 4), (4-methoxyphe-
nyl)methylamine (Table 1, entry 5), and methyl[4-(trifluorome-
thoxy)phenyl]amine (Table 1, entry 6), occurred in slightly
higher yields (around 50% after 6 d). The introduction of an
electron-withdrawing substituent on the para position of the
N-aryl ligand seems to have a limited effect. Dialkyl amides
such as N,2-dimethyl-1-propanamine (Table 1, entry 7) and
methyl(propyl)amine (Table 1, entry 8) reacted in lower yields
(around 30% after 6 d). Thus, substituting the amine with elec-
tron-donating groups reduces the yield significantly. These re-
sults are in agreement with catalytic a-alkylation of amines
being more efficient with N-aryl alkyl amines than with dialkyl
amines (see above).
When 3 was employed as a catalyst for the reaction involv-
ing methyl(propyl)amine (Table 1, entry 9), we detected by GC/
Chem. Eur. J. 2016, 22, 3000 – 3008
www.chemeurj.org
3005
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
sponding to a line broadening of 80 Hz was applied prior to Fouri-
er transformation.
MS a small amount of hydroaminoalkylated product containing
an aromatic group (see Supporting Information, Figure S4). We
also noted the disappearance of the aromatic signals in the
FTIR spectrum of the catalyst after the reaction (see Support-
ing Information, Figure S6). Since all hydroaminoalkylated
products contain NÀH bonds, the possibility that the product
is adsorbed on the surface was checked by recording the FTIR
spectra of the catalysts after the reaction, and no new peak in
the region of the NH vibration was observed (see Supporting
Information, Figure S5).
1H,1H MQ NMR spectroscopy
Two-dimensional DQ experiments were recorded on Bruker
AVANCE III spectrometer with a conventional double-resonance
3.2 mm CP/MAS probe, according to the following general
scheme: excitation of DQ coherences, t1 evolution, Z filter, and de-
tection. The spectra were recorded in a rotor-synchronized fashion
in t1, that is, the t1 increment was set equal to one rotor period
(4.545 ms). One cycle of the standard back-to-back (BABA) recou-
pling sequence was used for the excitation and reconversion
period. Quadrature detection in w1 was achieved by using the
States-TPPI method. A spinning frequency of 22 KHz was used. The
908 proton pulse length was 2.5 ms, and a recycle delay of 5 s was
used. A total of 128 t1 increments with 32 scans each were record-
ed.
Conclusion
Single-site, well-defined, silica-supported tetrakis(dimethylami-
do)tantalum [ꢀSi-O-Ta(NMe2)4] (1) can undergo b-H elimination
of dimethylamine leading to the corresponding silica-support-
ed tantallaaziridine. By combining SS NMR spectroscopy of the
isolated intermediates and DFT studies on the reaction mecha-
nism, we can conclude that the electronic properties of the
amine direct the reactivity of the tantalum center. [ꢀSi-O-Ta(h2-
NMeCH2)(NMe2)2] (2) can be conveniently obtained by thermal
treatment of 1 (1508C, 24 h). In contrast, the tantallaaziridine
species [ꢀSi-O-Ta(h2-NPhCH2)(NMe2)2] (3) can be prepared at
room temperature by treatment of 1 with a methyl aryl amine.
Catalytic testing of a selection of amines with various electron-
ic properties was carried out. The a-alkylation of amines was
consistently more efficient with N-aryl N-alkyl amine substrates
than with their dialkyl amine counterparts.
Preparation of [ꢀSi-O-Ta(NMe2)4] (1)
In a double Schlenk vessel, 132 mg of Ta(NMe2)5 in slight excess
(1.1 equiv) with respect to the amount of surface-accessible silanol
groups (0.3 mmol per gram) was treated with 1 g of SiO2À700 at
room temperature in pentane for 1 h. After filtration and four
washing cycles, all volatile compounds were evaporated and the
white solid was dried for 1 h under dynamic vacuum
(<10À5 mbar).
Preparation of [ꢀSi-O-Ta(h2-NMeCH2)(NMe2)2] (2)
In a glass reactor (230 mL), 1 g of 1 was heated with a gradient
from room temperature to 1508C (18CminÀ1) and then at 1508C
for 24 h under dynamic vacuum (<10À5 mbar). The resulting
powder was then cooled to 228C and analyzed by FTIR and SS
NMR spectroscopy.
Experimental Section
General
Preparation of [ꢀSi-O-Ta(h2-NPhCH2)(NMe2)2] (3)
All experiments were carried out under an atmosphere of argon or
nitrogen. The surface species were handled by using high-vacuum
lines (<10À5 mbar) and glovebox techniques. Elemental analyses
were performed at the Microanalysis Labor Pascher and KAUST An-
alytical Corelab. All chemicals including [Ta(NMe2)5] were purchased
from Sigma-Aldrich. FTIR spectra were recorded with a Nicolet
6700 FTIR spectrometer equipped with a cell designed for in situ
experiments with CaF2 windows. Typically, sixteen scans were accu-
mulated for each spectrum (resolution of 4 cmÀ1).
In
a double Schlenk vessel, 0.14 mL of methyl(phenyl)amine
(1 equiv) was treated with 500 mg of 1 at room temperature in
pentane for 4 hr. After filtration and four washing cycles, all gas-
phase products were analyzed by GC, and the solid product was
dried for 1 h under dynamic vacuum (<10À5 mbar).
Alkylation of N-alkyl aryl amines and N-dialkyl amines with
1-octene
Three vacuum-sealed ampoules was charged sequentially with the
catalyst (0.04 equiv), toluene (400 mL), secondary amine (1 equiv),
a Teflon-coated stir bar, and 1-octene (1.50 equiv). The ampoules
were then removed from the glovebox and connected to a high-
vacuum line. The mixture was condensed by cooling with liquid ni-
trogen and the ampoules were evacuated (<10À5 mbar). After seal-
ing, the ampoules were placed in an oil bath (1658C). The reaction
mixture was heated for 1, 3, and 6 d, and each time after reaction
it was cooled to 228C. The product solution was filtered to remove
the catalyst and the remaining liquid product was analyzed by GC-
FID and GC-MS.
SS NMR spectroscopy
1
One dimensional H MAS and 13C CP-MAS solid state NMR spectra
were recorded with a Bruker AVANCE III spectrometer operating at
400 MHz for 1H with a conventional double-resonance 4 mm
CPMAS probe. The samples were introduced under argon into zir-
conia rotors, which were then tightly closed. The spinning frequen-
cy was set to 17 kHz for H and 10 kHz for 13C spectra. NMR chemi-
1
cal shifts are reported with respect to TMS as external reference for
1H and 13C. For CP/MAS 13C NMR, the following sequence was
used: 908 pulses on the proton (pulse length 2.4 s), a cross-polari-
zation step with a contact time of typically 2 ms, and finally acquis-
ition of the 13C signal under high-power proton decoupling. The
delay between the scans was set to 5 s to allow complete relaxa-
Computational details
All calculations were performed with Gaussian 09.[21] For geometry
optimizations, transition-state searches, and vibrational-frequencies
evaluations, we used the Perdew–Burke–Erznehof (PBE)[22] function-
1
tion of the H nuclei, and the number of scans was 3000–5000 for
13C and 32 for 1H. An apodization function (exponential) corre-
Chem. Eur. J. 2016, 22, 3000 – 3008
www.chemeurj.org
3006
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
260; e) J. A. Tunge, D. A. Gately, J. R. Norton, J. Am. Chem. Soc. 1999,
121, 4520–4521.
al along with the split valence polarization basis set of Ahlrichs for
main group atoms (basis set SVP in G09),[23] while for tantalum we
described the core electrons with the Stuttgart SDD[23] quasirelativ-
istic effective core potential with the associated triple-z basis set
for valence electrons. Minima and transition states were character-
ized by the appropriate number of negative eigenvalues in the
Hessian. We also checked that the imaginary frequency of all the
transition-state structure corresponds to a vibration along the reac-
tive path between reactants and products. We ran the final single-
point energy calculations over the PBE optimized geometries using
the hybrid PBE0 functional[24] with the empirical dispersion correc-
tion of Grimme (D3)[25] along with the triple-z valence plus one po-
larization function basis set of Ahlrichs (basis set TZVP in G09).[26]
We evaluated the final Gibbs free energies of all stationary point of
the potential-energy surface by adding the PBE0-D3/TZVP/SDD
single point energies to the thermal corrections at 298.15 K, includ-
ing zero-point vibrational energy at the PBE/SVP/SDD level and en-
tropy terms. Chemical shifts relative to TMS were evaluated on the
Gaussian 09 optimized geometries by using the ADF suite of pro-
grams.[27] The PBE0 functional in connection with the all-electron
triple-z plus one polarization function basis set TZVP was used.
Scalar relativistic effects were included through the ZORA Hamilto-
nian.[28]
[7] a) N. Coles, M. C. J. Harris, R. J. Whitby, J. Blagg, Organometallics 1994,
13, 190–199; b) J. M. Davis, R. J. Whitby, A. Jaxachamiec, Tetrahedron
Lett. 1992, 33, 5655–5658; c) J. M. Davis, R. J. Whitby, A. Jaxachamiec,
Tetrahedron Lett. 1994, 35, 1445–1448; d) M. C. J. Harris, R. J. Whitby, J.
Blagg, Tetrahedron Lett. 1995, 36, 4287–4290.
[8] a) N. Coles, R. J. Whitby, J. Blagg, Synlett 1990, 271–272; b) N. Coles,
R. J. Whitby, J. Blagg, Synlett 1992, 143–145.
[9] a) I. Prochnow, P. Zark, T. Muller, S. Doye, Angew. Chem. Int. Ed. 2011, 50,
6401–6405; Angew. Chem. 2011, 123, 6525–6529; b) A. L. Reznichenko,
K. C. Hultzsch, J. Am. Chem. Soc. 2012, 134, 3300–3311.
[10] a) L. W. Francisco, P. S. White, J. L. Templeton, Organometallics 1996, 15,
5127–5136; b) L. H. Polm, G. Vankoten, C. J. Elsevier, K. Vrieze, B. F. K.
Vansanten, C. H. Stam, J. Organomet. Chem. 1986, 304, 353–370; c) H.
Hoberg, V. Gotz, C. Kruger, Y. H. Tsay, J. Organomet. Chem. 1979, 169,
209–217; d) J. Browning, H. D. Empsall, M. Green, F. G. A. Stone, J.
Chem. Soc. Dalton Trans. 1973, 381–387.
[11] J. M. Mayer, C. J. Curtis, J. E. Bercaw, J Am Chem. Soc 1983, 105, 2651–
2660.
[12] S. B. Herzon, J. F. Hartwig, J. Am. Chem. Soc. 2008, 130, 14940.
[13] J. M. Basset, S. L. Scott, L. Choplin, A. Leconte, M. Quignard, F. Santini, C.
Theolier in Organometallic Chemistry as a Basis for Understanding Heter-
ogeneous Catalysis, Vol. 39: Elementary Reaction Steps in Heterogeneous
Catalysis (Eds.: R. W. Joyner and R. A. van Santen), Springer Science+
Business Media, Dordrecht, 1993, pp. 39–49.
[14] Y. Chen, E. Abou-hamad, A. Hamieh, B. Hamzaoui, L. Emsley, J. M.
Basset, J. Am. Chem. Soc. 2015, 137, 588–591.
Acknowledgements
[15] a) B. Hamzaoui, M. El Eter, E. Abou-Hamad, Y. Chen, J. D. A. Pelletier,
J. M. Basset, Chem. Eur. J. 2015, 21, 4294–4299; b) M. El Eter, B. Ham-
zaoui, E. Abou-Hamad, J. D. A. Pelletier, J. M. Basset, Chem. Commun.
2013, 49, 4616–4618.
[16] B. Hamzaoui, J. D. A. Pelletier, M. El Eter, Y. Chen, E. Abou-Hamad, J.-M.
Basset, Adv. Synth. Catal. 2015, 357, 3148–3154.
The authors wish to thank King Abdullah University of Science
and Technology for continuous support.
[17] M. Beaudoin, S. L. Scott, Organometallics 2001, 20, 237–239.
[18] S. L. Buchwald, B. T. Watson, M. W. Wannamaker, J. C. Dewan, J Am
Chem. Soc 1989, 111, 4486–4494.
Keywords: hydroamination · metallacycles · solid-state NMR
spectroscopy · supported catalysts · tantalum
[19] F. A. Pasha, L. Cavallo, J. M. Basset, ACS Catal. 2014, 4, 1868–1874.
[20] a) Y. Chen, R. Credendino, E. Callens, M. Atiqullah, M. A. Al-Harthi, L. Cav-
allo, J.-M. Basset, ACS Catal. 2013, 3, 1360–1364; b) B. Werghi, A. Bend-
jeriou-Sedjerari, J. Sofack-Kreutzer, A. Jedidi, E. Abou-Hamad, L. Cavallo,
J.-M. Basset, Chem. Sci. 2015, 6, 5456–5465; c) V. D’Elia, H. Dong, A. J.
Rossini, C. M. Widdifield, S. V. C. Vummaleti, Y. Minenkov, A. Poater, E.
Abou-Hamad, J. D. A. Pelletier, L. Cavallo, L. Emsley, J.-M. Basset, J. Am.
Chem. Soc. 2015, 137, 7728–7739; d) A. Hamieh, Y. Chen, S. Abdel-
Azeim, E. Abou-hamad, S. Goh, M. Samantaray, R. Dey, L. Cavallo, J. M.
Basset, ACS Catal. 2015, 5, 2164–2171; e) N. Riache, A. Dery, E. Callens,
A. Poater, M. Samantaray, R. Dey, J. Hong, K. Li, L. Cavallo, J.-M. Basset,
Organometallics 2015, 34, 690–695.
[21] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R.
Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H. Na-
katsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G.
Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Ha-
segawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven,
J. A. Montgomery Jr., J. E. Peralta, F. Ogliaro, M. J. Bearpark, J. Heyd,
E. N. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J. Normand, K.
Raghavachari, A. P. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi,
N. Rega, N. J. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C.
Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J.
Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma,
V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich,
A. D. Daniels, Ö. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski, D. J. Fox,
Gaussian, Inc., Wallingford, CT, USA, 2009.
[1] J. M. P. Lauzon, L. L. Schafer, Dalton Trans. 2012, 41, 11539–11550.
[2] a) M. G. Clerici, F. Maspero, Synthesis 1980, 305–306; b) W. A. Nugent,
D. W. Ovenall, S. J. Holmes, Organometallics 1983, 2, 161–162; c) J. Dçr-
fler, T. Preuss, A. Schischko, M. Schmidtmann, S. Doye, Angew. Chem. Int.
Ed. 2014, 53, 7918–7922; Angew. Chem. 2014, 126, 8052–8056; d) E.
Chong, L. L. Schafer, Org Lett. 2013, 15, 6002–6005; e) J. Dçrfler, S.
Doye, Angew. Chem. Int. Ed. 2013, 52, 1806–1809; Angew. Chem. 2013,
125, 1851–1854.
[3] a) S. B. Herzon, J. F. Hartwig, J Am Chem. Soc 2007, 129, 6690; b) P. W.
Roesky, Angew. Chem. Int. Ed. 2009, 48, 4892–4894; Angew. Chem.
2009, 121, 4988–4991; c) E. Chong, J. W. Brandt, L. L. Schafer, J Am
Chem. Soc 2014, 136, 10898–10901; d) P. Eisenberger, R. O. Ayinla,
J. M. P. Lauzon, L. L. Schafer, Angew. Chem. Int. Ed. 2009, 48, 8361–8365;
Angew. Chem. 2009, 121, 8511–8515; e) P. Garcia, Y. Y. Lau, M. R. Perry,
L. L. Schafer, Angew. Chem. Int. Ed. 2013, 52, 9144–9148; Angew. Chem.
2013, 125, 9314–9318; f) J. Dçrfler, S. Doye, Eur J Org Chem. 2014,
2014, 2790–2797; g) G. F. Zi, F. R. Zhang, H. B. Song, Chem. Commun.
2010, 46, 6296–6298; h) A. L. Reznichenko, T. J. Emge, S. Audorsch, E. G.
Klauber, K. C. Hultzsch, B. Schmidt, Organometallics 2011, 30, 921–924.
[4] a) D. C. Schmitt, J. Lee, A. M. R. Dechert-Schmitt, E. Yamaguchi, M. J. Kri-
sche, Chem. Commun. 2013, 49, 6096–6098; b) T. Y. Chen, R. Tsutsumi,
T. P. Montgomery, I. Volchkov, M. J. Krische, J. Am. Chem. Soc. 2015, 137,
1798–1801.
[5] a) R. B. Grossman, W. M. Davis, S. L. Buchwald, J. Am. Chem. Soc. 1991,
113, 2321–2322; b) S. L. Buchwald, R. B. Nielsen, J. Am. Chem. Soc. 1988,
110, 3171–3175; c) S. L. Buchwald, R. B. Nielsen, J. C. Dewan, Organome-
tallics 1988, 7, 2324–2328.
[22] J. P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett. 1996, 77, 3865–
3868.
[23] A. Schäfer, H. Horn, R. Ahlrichs, J. Chem. Phys. 1992, 97, 2571–2577.
[24] C. Adamo, V. Barone, J. Chem. Phys. 1999, 110, 6158–6170.
[25] S. Grimme, S. Ehrlich, L. Goerigk, J. Comput. Chem. 2011, 32, 1456–
1465.
[6] a) J. X. Chen, J. A. Tunge, J. R. Norton, J. Org. Chem. 2002, 67, 4366–
4369; b) D. A. Gately, J. R. Norton, P. A. Goodson, J. Am. Chem. Soc.
1995, 117, 986–996; c) T. V. Lubben, K. Plossl, J. R. Norton, M. M. Miller,
O. P. Anderson, Organometallics 1992, 11, 122–127; d) J. A. Tunge, C. J.
Czerwinski, D. A. Gately, J. R. Norton, Organometallics 2001, 20, 254–
[26] A. Schäfer, C. Huber, R. Ahlrichs, J. Chem. Phys. 1994, 100, 5829–5835.
Chem. Eur. J. 2016, 22, 3000 – 3008
www.chemeurj.org
3007
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[27] a) G. Te Velde, F. M. Bickelhaupt, E. J. Baerends, C. Fonseca Guerra, S. J.
van Gisbergen, J. G. Snijders, T. Ziegler, J. Comput. Chem. 2001, 22, 931–
967; b) G. Schreckenbach, T. Ziegler, Int. J. Quantum Chem. 1997, 61,
899–918; c) G. Schreckenbach, T. Ziegler, Int. J. Quantum Chem. 1996,
60, 753–766; d) G. Schreckenbach, T. Ziegler, J. Phys. Chem. 1995, 99,
606–611; e) J. Autschbach, S. Patchkovskii, B. Pritchard, J. Chem. Theory
Comput. 2011, 7, 2175–2188; f) J. Autschbach, E. Zurek, J. Phys. Chem. A
2003, 107, 4967–4972.
[28] a) E. van Lenthe, A. Ehlers, E.-J. Baerends, J. chem. phys. 1999, 110,
8943–8953; b) E. van Lenthe, E.-J. Baerends, J. G. Snijders, J. chem. phys.
1994, 101, 9783–9792; c) E. van Lenthe, E.-J. Baerends, J. G. Snijders, J.
chem. phys. 1993, 99, 4597–4610.
Received: November 4, 2015
Published online on January 27, 2016
Chem. Eur. J. 2016, 22, 3000 – 3008
www.chemeurj.org
3008
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim