Communications
doi.org/10.1002/ejic.202001152
The observed structure of (o-TMPÀ C H )AlH consists of a
pure HD. In an equilibrium H , HD and D exist in equimolar
2 2
6
4
2
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
hydride-bridged dimeric form. Therein, the AlÀ N distance of
.311(2) Å is significantly longer if compared to representative
amounts. These findings indicate that the À AlH -moiety adds
2
2
significant amounts of H to the given equilibrium. To validate
2
Al-N distances in similar compounds. Some selected representa-
that the hydrogen activation involves the AlH -moiety, a
2
tive compounds and their AlÀ N bond lengths are as follows,
complementary experiment using only D2 was performed,
which resulted in the formation of HD (Supporting Information).
Since a competitive reaction may involve the activation of
[
20]
[
Me NÀ C H À AlH ] 2.118(3) Å (hydride bridged dimer),
2
10
6
2 2
[21]
[22]
Me NÀ AlMe
2.045(2) Å,
(Me N) AlH
3
2.163(2) Å,
3
3
3
2
[15]
[
TEDAÀ AlH ] 2.161(2) Å (both AlÀ N distances are equal) and
deuterated solvent, a blind experiment only with H gas under
3
n
2
[23]
TEDA-(AlMe ) 2.065(8) Å. FLPs and classic Lewis pairs can be
the same conditions was performed. This did not result in the
formation of any HD, indicating that the reaction involves
gaseous hydrogen. Based on these observations, we assume a
hypervalent Al-species, like a six-coordinate [(o-TMPÀ C H )
3
2
discriminated by atomic distances. If the bond length is in the
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
[24]
order of the sum of atomic radii (d(AlÀ N)=1.97 Å) it is a Lewis
adduct. However, if the atomic distance is in the order of the
6
4
[25]
sum of crystallographic van der Waals radii (d(AlÀ N)=3.39 Å)
it indicates “Frustration”. Based on these geometrical approx-
imations, (o-TMPÀ C H )AlH should be described more convinc-
AlH D] or even a charged penta-coordinate D[(o-TMPÀ C H )
3
6
4
AlH ], to be involved in a plausible mechanism.
3
Unfortunately, due the complexity of the system (three-
phase system, diffusion, (gas) solubility and precise loadings)
we cannot provide quantitative information. In further studies,
we are planning to study the kinetics of the reaction.
6
4
2
ingly as an “active” Lewis pair than an “FLP” as reported by
[18]
Chen et al. Noteworthy, the AlÀ N distance of 2.311(2) Å in
o-TMPÀ C H )AlH is slightly longer than in BÀ P based “active”
(
6
4
2
Lewis pairs (d(BÀ P) ~2.18-2.20 Å), which exhibit classic FLP
Since the first hydrogen activation with the intramolecular
Al-N compound (o-TMPÀ C H )AlH was successful, we continued
[26]
reactivity, as described by Erker and coworkers. In contrast,
6
4
2
Repo and coworkers have reported (o-TMPÀ C H )B(C F ) as a
to use intermolecular pairs. These intermolecular pairs are
readily available by mixing alanes with amines. This offers
variable basicity of the amine and different sterically demand-
ing ligands on the aluminium. A selection of different amines
(including TEDA, lutidine, dimethylaniline) and readily available
alanes was tested without conclusive results. Similar non-
6
4
6 5 2
[19]
“
real FLP” with d(BÀ N)=3.031(4) Å.
Since (o-TMPÀ C H )AlH may describe the lower limit for
6
4
2
FLP-type reactivity for the Al-N system, it raises the questions
can it activate molecular hydrogen? Therefore, we chose the
common isotope exchange reaction of HD to H2 and D2 in
toluene-d to address this question.
conclusive results were obtained when only trisalkylalanes or
8
i
A slurry of (o-TMPÀ C H )AlH in toluene-d was charged with
HAlBu were used.
6
4
2
2
1
HD gas, and the isotope exchange reaction was followed via H
However, successful HD-exchange experiments with
i
NMR, Figure 1, Supporting Information. Initially, the HD triplet
N-methyldiphenylamine (Ph NMe) and AlR (R=Me, Et, Bu)
2
3
1
occurs at 4.46 ppm ( J=43 Hz), while the H singlet at 4.50 ppm
were conducted and monitored by NMR spectroscopy,
Scheme 2.
2
is increasing over time (~1 h) and gentle heating (additional
i
1
~
90 min, 60°C). This indicates the activation of HD, allowing
The reaction of AlBu À Ph NMe with HD was followed via H
3
2
2
the formation of H and D (D was not measured due to low
and H NMR spectroscopy. We observed the formation of H
2
2
2
2
sensitivity). In another, longer experiment, the HD-exchange
and D2 within an hour at ambient temperature, Figure 2,
1
reaction resulted in H amounts above the expected yields for
Supporting Information. Quantification via H NMR spectro-
2
[27]
scopy (corrected for the presence of para-hydrogen, but not
considering diffusion processes) resulted in a specific rate of
À 1 À 1
À 1 À 1
~
0.033 molmolcat
h
(~91 μmolgcat h ). This specific rate is
of the same order of magnitude as the hydrogenation reactions
[2b]
with common FLP catalysts.
Similar HD-exchange experi-
ments with AlR À Ph NMe (R=Me, Et) show that they convert
3
2
HD just at the detection level and therewith significantly slower
i
than AlBu À Ph NMe, Supporting Information. We could not
3
2
1
detect any change of the employed AlÀ N compounds using H
NMR spectroscopy, which points to a catalytic function.
Since the HD-exchange reactions proceed slower with less
bulky alkylalanes, we expect a correlation to steric reasons, such
1
Figure 1. H NMR spectra of the HD-exchange reaction with (o-TMPÀ C
6
H
4
)
Scheme 2. Hydrogen activation by HD isotope exchange using alane-amine
mixtures.
AlH
2
in Tol-d
8
, for the full-spectrum see Supporting Information.
Eur. J. Inorg. Chem. 2021, 1240–1243
1241
© 2021 The Authors. European Journal of Inorganic Chemistry published
by Wiley-VCH GmbH