Isolation and structural elucidation of a key aluminoaromatic intermediate and evidence for dismutation phenomena in TMP-alumination chemistry

Article abstract of DOI:10.1039/b713913f

Lithium TMP-aluminate ⁱBu₃Al(TMP)Li undergoes dismutation in THF solution to precipitate the tetraalkylaluminate [{Li·(THF)₄}⁺{Al(ⁱBu)₄} ^-], but reacts kinetically as a TMP base towards N,N- diisopropylbenzamide to afford the crystalline ortho-aluminated species [(THF)₃·Li{O(=C)N(ⁱPr)₂(C ₆H₄)}Al(ⁱBu)₃] and TMPH. The Royal Society of Chemistry.

Full text of DOI:10.1039/b713913f

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Isolation and structural elucidation of a key aluminoaromatic
intermediate and evidence for dismutation phenomena in
TMP-alumination chemistry{
´
Ben Conway, Eva Hevia, Joaqu´ın Garc´ıa-Alvarez,* David V. Graham, Alan R. Kennedy and
Robert E. Mulvey*
Received (in Cambridge, UK) 11th September 2007, Accepted 5th October 2007
First published as an Advance Article on the web 12th October 2007
DOI: 10.1039/b713913f
Lithium TMP-aluminate ‘‘iBu3Al(TMP)Li’’ undergoes dismu-
tation in THF solution to precipitate the tetraalkylaluminate
[{Li?(THF)4}+{Al(iBu)4}2], but reacts kinetically as a TMP base
towards N,N-diisopropylbenzamide to afford the crystalline
ortho-aluminated species [(THF)3?Li{O(LC)N(iPr)2(C6H4)}Al-
(iBu)3] and TMPH.
that crystalline 1 was obtained from bulk THF solution, whereas
the supporting information indicated it was prepared in hexane
solution with a stoichiometric quantity of THF. This report
i
stimulated us to revisit the reaction between Bu3Al(TMP)Li and
[PhC(LO)NiPr2] to attempt to extract structural information from
bulk THF solutions, to add to the knowledge gleamed previously
from hexane solutions. As revealed herein, we can now report
the first successful isolation and spectroscopic/crystallographic
characterization of an aluminoaromatic intermediate5 obtained via
the actual experimental conditions used to effect direct alumination
of the benzamide. Our study also sheds new and surprising light
on the constitution of the base in solution, which points to an
inherently more complicated chemistry than that previously
described.
First reported in 2004 by Uchiyama et al., lithium TMP-aluminate
‘‘iBu3Al(TMP)Li’’ (where TMP is the amide 2,2,6,6-tetramethyl-
piperidide) is an excellent reagent in THF solution for selectively
deprotonating and concomitantly aluminating a wide range of
functionalized aromatics.1 This new direct alumination method
possesses many advantages over the indirect two-step metathesis
approach (for example, synthesis of an aromatic lithium or
Grignard species, followed by reaction with an aluminium salt)
usually employed for preparing aromatic aluminium compounds.2
We began the study by re-preparing the ‘‘base’’ 1 following the
literature procedures. In our hands, 1 failed to crystallise (or even
deposit as a solid) from neat THF solution despite several
attempts; but was readily produced via a hexane solution
containing a stoichiometric quantity of THF. To our surprise, 1
did not metallate the benzamide to any appreciable extent in
hexane or benzene solution. Taken together with previous findings,
the implication of this unexpected failure is that excess THF is
necessary for metallation efficacy, thus casting doubt on the case
for 1 being the true active base. We therefore switched our focus to
bulk THF solutions. Following exactly the original literature pro-
cedure,1 we treated the benzamide with 2.2 molar equivalents of
iBu3Al(TMP)Li in neat THF solution but omitted any subsequent
electrophilic trapping step. Initially this gave a homogeneous
yellow solution, but after 1 h deposited a colourless crystalline
i
In formal equation terms, these reactions of Bu3Al(TMP)Li
represent a simple exchange of an aromatic hydrogen for a
triorganoaluminium–lithium (R3Al, Li+) fragment, but in reality
they can be extraordinarily complex. This is because the base itself
is bimetallic, multi-Lewis acidic, heteroleptic and highly coordi-
nated about the (ultimately) active anionic Al centre, combined
with the fact that organoaluminium compounds have a strong
propensity for undergoing ligand exchange, aggregation, and
solvation phenomena.2 Structural information on the nature of the
base in solution and in the solid-state is therefore vital to help
unravel this inherent complexity, which is exacerbated on addition
of the aromatic molecule to be aluminated. We started along this
path by reporting3 the crystal structure of the Lewis base-stabilized
derivative [L?Li(m-TMP)(m-iBu)Al(iBu)2] (L = N,N-diisopropyl-
benzamide = [PhC(LO)NiPr2]), the synthesis of which revealed
the marked solvent dependency of the basic performance of
iBu3Al(TMP)Li: it metallates the benzamide essentially quantita-
tively (starting with 2.2 molar equivalents of base) in bulk THF,
negligibly in a stoichiometric quantity of THF, and essentially not
at all in bulk hexane. Most recently, Uchiyama and co-workers
carried out a combined experimental and theoretical study4 which
concluded that the active base of lithium TMP-aluminate is the
mono-THF solvate [THF?Li(m-TMP)(m-iBu)Al(iBu)2], 1, though
there was some inconsistency as to its origin with the paper stating
1
7
solid identified directly by H, Li and 13C NMR spectroscopy
(and indirectly by comparison with its known crystallographically
characterised dioxane analogue [{Li?(dioxane)4}+{Al(iBu)4}2] I,6
Fig. 1) as the solvent-separated ionic aluminate [{Li?(THF)4}+-
{Al(iBu)4}2] 2.{ Precipitation of 2 was also observed in THF
i
solutions of Bu3Al(TMP)Li in the absence of the benzamide. On
WestCHEM, Department of Pure and Applied Chemistry, University of
Strathclyde, Glasgow, UK G1 1XL. E-mail: r.e.mulvey@strath.ac.uk
{ The HTML version of this article has been enhanced with colour
images.
Fig. 1 Chemdraw representation of [{Li?(dioxane)4}+{Al(iBu)4}2] I.
Chem. Commun., 2007, 5241–5243 | 5241
This journal is ß The Royal Society of Chemistry 2007

View Article Online
Scheme 1 Dismutation of putative iBu3Al(TMP)Li in neat THF
solution.
Fig. 2 Molecular structure of
3
with thermal ellipsoids at 50%
probability level. The hydrocarbon backbones of the three THF molecules
˚
and all hydrogen atoms are omitted for clarity: Selected bond lengths (A)
and bond angles (u): Li1–O1 1.860(7), Li1–O2 1.954(7), Li1–O3 1.956(7),
Li1–O4 1.945(7), Al1–C7 2.050(4), Al1–C8 2.013(4), Al1–C12 2.017(4),
Al1–C16 2.023(4); O1–Li1–O2 113.4(4), O1–Li1–O3 110.4(3), O1–Li1–O4
108.1(3), O2–Li1–O3 107.2(3) O2–Li1–O4 110.2(3), O3–Li1–O4
107.4(3), C7–Al1–C8 109.84(16), C7–Al1–C12 105.44(16), C7–Al1–C16
110.16(16), C8–Al1–C12 103.26(17), C8–Al1–C16 113.73(17), C12–Al1–
C16 113.90(17).
Scheme 2 How the identity of the aluminate reagent greatly influences
the course of the direct alumination reaction.
this evidence, dismutation, previously unconsidered,4 depicted in
its simplest possible form in Scheme 1, must clearly contribute to
the solution chemistry of the putative iBu3Al(TMP)Li formulated
reagent.
(Al–H exchange) nature of the reaction,8 the Al resides almost
Interestingly, isolated 2 is inert towards the benzamide in neat
THF solution (Scheme 2). This inactive component of the
iBu3Al(TMP)Li THF-solution could explain why 2.2 molar
equivalents of the base are required to effect 94% metallation
of the benzamide as determined by electrophilic quenching with
coplanar with the aromatic plane [175.9(3)u] with the newly
˚
developed Al–arylC7 bond [2.050(4) A] modestly longer than the
˚
Al–alkyl bonds (mean length, 2.018 A).
On adding the new substantial information gleaned from
unearthing 2 and 3 to the existing body of evidence, a more
realistic picture of the constitution of the base and its reaction with
the benzamide begins to emerge. With 1 ruled out as a candidate
for the active base, Scheme 3 shows the likely other species
involved with 2 and 3, the participation of which has been
established. The existence of molecular ‘‘iBu3Al(TMP)Li’’ can be
inferred from its trapping by benzamide L in the Lewis-base
i
iodine.1 Returning to our investigation of the Bu3Al(TMP)Li,
benzamide, THF mixture, after isolating 2 from the solution, the
filtrate deposited a second crystalline product, characterised by 1H,
7Li and 13C NMR spectroscopy, and X-ray crystallography, as
the tris(THF)-solvated lithium trialkyl-monoorthoarylaluminate
[(THF)3?Li{O(LC)N(iPr)2(C6H4)}Al(iBu)3], 3.{ This complex
represents the first tangible metallo intermediate of a direct
alumination reaction (performed in THF solution) of a benzamide,
or indeed of any functionalised aromatic compound, to be struc-
turally defined and well characterised.7 Its molecular structure§
(Fig. 2) can be considered a contacted ion pair. The anionic moiety
comprises a distorted tetrahedral Al centre made up of four C
atoms, three from terminal iBu ligands and one from the
deprotonated ortho position of the benzamide fragment. Contact
to the cationic moiety is through the carbonyl O atom which binds
terminally to Li to complete a distorted tetrahedral coordination
sphere of O atoms, the remaining three of which belong to THF
ligands. Key bond lengths and bond angles are listed in the figure
legend. There is no contact between the Li and the deprotonated
ortho C atom of the benzamide fragment (separation distance,
adduct [L?Li(m-TMP)(m-iBu)Al(iBu)2].3 Within
a bulk THF
medium, this heteroleptic trisalkyl-monoamido coordination of
Al is likely to be maintained (at least transiently) and probably
to be accompanied by a tetra (THF)-solvated Li+, a commonly
found cation.9 The detection of 2 proves unequivocally that
[{Li?(THF)4}+{Al(TMP)(iBu)3}2] must undergo a dismutation
process, the third component of which can be formulated formally
as [{Li?(THF)n}+{Al(TMP)2(iBu)2}2] from stoichiometric balance.
…
Li C7 4.759(8) A). In contrast, the Li–O1 (carbonyl) bond length
˚
˚
is short at 1.860(7) A, though essentially equidistant to that in
i
i
3
˚
aforementioned [L?Li(m-TMP)(m-Bu)Al( Bu)2] [1.872(4) A], where
the benzamide is neutral, not anionic. To make this short Li–O1
bond the carbonyl function lies almost perpendicular [C7C2C1O1
99.7(4)u] to the aromatic plane. Consistent with the alumination
Scheme 3 Postulated pathway for the formation of 3.
5242 | Chem. Commun., 2007, 5241–5243
This journal is ß The Royal Society of Chemistry 2007

View Article Online
278 uC and the mixture was stirred for 30 min at 0 uC. N,N-
Diisopropylbenzamide (2.5 mmol, 0.53 g) was added at 278 uC and the
mixture was stirred during 3 h at r.t. to give a yellow solution and a white
solid. Heating the solution to refluxing temperature was needed to form a
clear solution. Bench cooling of this solution afforded again colourless
crystals of 2 (0.71 g, 23%). All the solvent of the filtrate was removed
in vacuo, followed by the addition of 10 mL of hexane to form a yellow
solution. Freezer cooling of this solution at 227 uC afforded colourless
As it has been demonstrated that 2 is inert towards the benzamide,
alumination of the benzamide must be effected by one (or
potentially both) of the TMP-aluminates. Proof of the TMP
ligand transfer selectivity occurring within the reaction comes
from the detection of TMPH in filtrates following the isolation
of crystalline product 3 from THF solutions. The TMP-free,
tris(alkyl) composition of 3 combined with its high isolated yield
(54% with respect to the benzamide) would appear to provide
clinching evidence that the dominant active base within the
mixture is a tris(alkyl) composed, mono-TMP aluminate with a
stoichiometric excess (with respect to lithium) of THF, namely
[Li?(THF)xAl(TMP)(iBu)3]. Tetra-solvated, solvent separated
[{Li?(THF)4}+{Al(TMP)(iBu)3}2] is the most likely candidate,
but a kinetically labile, lower-solvated (x = 2 or 3), contacted ion-
pair variant cannot be unequivocally ruled out.
1
crystals of 3 (0.84 g, 54% based on benzamide). H NMR (400.13 MHz,
3
d6-benzene, 293 K): d 8.45 (d, 1H, JHH = 7.5 Hz, 1H m-C6H4), 7.25
3
(m, 1H, p-C6H4), 7.08 (m, 1H, m*-C6H4), 6.88 (d, 1H, JHH = 7.5 Hz,
o*-C6H4), 4.07 and 3.19 (sept, 1H each, 3JHH = 6.7 Hz, NCH(CH3)2), 3.23
(m, 12H, OCH2 THF), 2.45 (sept, 3H, JHH = 6.6 Hz, CH2CH(CH3)2),
3
1.43 (m, 24H, 18H CH2CH(CH3)2 and 6H NCH(CH3)2), 1.32 (m, 12H,
CH2 THF), 1.71, 1.09 and 0.85 (d, 6H each, 3JHH = 6.7 Hz, NCH(CH3)2),
0.28 (m, 6H, CH2CH(CH3)2). 13C{H} NMR (100.63 MHz, d6-benzene,
293 K): d 177.98 (CLO), 160.57 (o-C6H4), 144.43 (i-C6H4), 141.18
(m-C6H4), 125.59 (p-C6H4), 124.05 (m*-C6H4), 122.67 (o*-C6H4), 67.69
i
i
(OCH2 THF), 51.36 and 45.42 (CH Pr), 29.13 and 28.88 (CH3 of Bu),
27.67 (CH of iBu), 24.62 (CH2 THF), 20.64 and 19.57 (CH3 iPr), 19.45 (2C,
CH3 iPr). Signal for Al–CH2 of iBu was not observed. 7Li NMR
(155.50 MHz, d6-benzene, 293 K, reference LiCl in D2O at 0.00 ppm): d
In conclusion, the solution and structural chemistry of
‘‘iBu3Al(TMP)Li’’ has now been shown to be decidedly more
intricate than previously thought.4 Based on our synthetic,
reactivity, NMR spectroscopic and X-ray crystallographic studies,
a reaction scheme involving previously unconsidered solvent-
separated ion-pair species and a dismutation process can be
postulated. In the presence of the benzamide, or another strongly
Lewis basic aromatic substrate susceptible to metallation, pathway
A will be predominant, whereas if the base is left alone dissolved
in THF solution or treated with slower reacting substrates the
dismutation pathway B will take on more prominence.
1
20.14 ppm. TMPH was detected in the H NMR spectra of the filtrates
after the isolation of crystalline product 3 from THF solutions.
§ Crystal data for 3: C37H69AlLiNO4, Mr = 625.85, triclinic, space group
˚
P1, a = 9.8880(6), b = 10.5709(8), c = 10.8837(8) A, a = 108.652(3), b =
We thank the EPSRC and the Royal Society (University
Research Fellowship to E. H.) for their generous sponsoring of this
research.
3
˚
˚
107.329(4), c = 99.374(4)u, V = 986.15(12) A , Z = 1, l = 0.71073 A, m =
0.086 mm21, T = 123 K; 20423 reflections, 3833 unique, Rint = 0.061; final
refinement to convergence on F2 gave R = 0.0494 (F, 2688 obs. data only)
and Rw = 0.1087 (F2, all unique data), GOF = 1.017. CCDC 660804. For
crystallographic data in CIF or other electronic format see DOI: 10.1039/
b713913f
Notes and references
{ All reactions were carried out under a protective argon blanket.
Synthesis of [{Li?(THF)4}+{Al(iBu)4}2] (2): BuLi (5 mmol, 3.13 ml of a
1.6 M solution in hexane) was added to a mixture of THF (4 mL) and
TMPH (5 mmol, (0.85 mL)) at 278 uC and the mixture was stirred for
10 min at 0 uC. Then, iBu3Al (5 mmol, 5 mL of a 1 M solution in hexane)
was added to the mixture at 278 uC and the mixture was stirred for 30 min
at 0 uC to give a pale yellow solution and a white solid. Heating the solution
to refluxing temperature was needed to form a clear solution. Bench
cooling of this solution afforded colourless crystals of 2 (0.71 g, 23%).
Under these conditions, [THF?Li(m-TMP)(m-iBu)Al(iBu)2], 1, failed to
crystallise despite several attempts (the only product deposited as a solid
was 2). 1H NMR (400.13 MHz, d6-benzene, 293 K): d 3.48 (m, 16H, OCH2
1 M. Uchiyama, Y. Naka, Y. Matsumoto and T. Ohwada, J. Am. Chem.
Soc., 2004, 126, 10526.
2 For a review, see: J. J. Eisch, in Comprehensive Organometallic Chemistry;
ed. G. Wilkinson, F. G. A. Stone and E. W. Abel, Pergamon Press,
Oxford, 1982, vol. 6, ch. 6.
´
3 J. Garc´ıa-Alvarez, E. Hevia, A. R. Kennedy, J. Klett and R. E. Mulvey,
Chem. Commun., 2007, 2402.
4 Y. Naka, M. Uchiyama, Y. Matsumoto, A. E. H. Wheatley,
M. McPartlin, J. V. Morey and Y. Kondo, J. Am. Chem. Soc., 2007,
129, 1921.
5 ‘‘Intermediate’’ in the sense that it is still carbanionically active and has
not been subjected to a subsequent electrophilic quenching step.
6 For full spectroscopic and crystallographic characterization of this
compound see the Electronic Supplementary Information (ESI) asso-
ciated with the paper cited in ref. 3.
7 For a structural study of a reaction carried out in hexane solution where a
TMEDA-stabilised lithium TMP-aluminate functions as a dibasic alkyl-
´
amido base, see: J. Garc´ıa-Alvarez, D. V. Graham, A. R. Kennedy,
R. E. Mulvey and S. Weatherstone, Chem. Commun., 2006, 3208.
8 For a review of such alkali-metal-mediated alumination (and metallation
in general), see: R. E. Mulvey, F. Mongin, M. Uchiyama and Y. Kondo,
Angew. Chem., Int. Ed., 2007, 46, 3802.
9 A search of the Cambridge Structural Database scored 296 hits for
structures containing the cation [Li(THF)4]+, 17 of which have an Al
atom in the counterion. Cambridge Structural Database, Version 5.28
with updates to May 2007. For an example, see: J. Vollet, G. Stosser and
H. Schno¨ckel, Inorg. Chim. Acta, 2007, 360, 1298.
THF), 2.45 (sept, 4H, 3JHH = 6.6 Hz, CH2CH(CH3)2), 1.45 (d, 24H, 3JHH
=
=
3
6.6 Hz, CH2CH(CH3)2), 1.40 (m, 16H, CH2 THF), 0.17 (d, 8H, JHH
6.6 Hz, CH2CH(CH3)2). 13C{H} NMR (100.63 MHz, d6-benzene, 293 K):
d 67.75 (OCH2 THF), 28.96 (CH2CH(CH3)2), 27.45 (CH2CH(CH3)2),
24.75 (CH2 THF), 22.92 (CH2CH(CH3)2). 7Li NMR (155.50 MHz,
d6-benzene, 293 K, reference LiCl in D2O at 0.00 ppm): d 20.95 ppm.
Reaction between [{Li?(THF)4}+{Al(iBu)4}2] (2) and [PhC(LO)NiPr2]: In
a Schlenk tube, isolated 2 was dissolved in neat THF solution to give a pale
yellow solution. A molar equivalent of N,N-diisopropylbenzamide was
introduced, and the mixture was further stirred for 3 h. 1H NMR
spectroscopic analysis of this mixture established that 2 does not metallate
the benzamide.
Synthesis of [(THF)3?Li{O(LC)N(iPr)2(C6H4)}Al(iBu)3] (3): Under Ar
atmosphere, BuLi (5 mmol, 3.13 ml of a 1.6 M solution in hexane) was
added to a mixture of THF (4 mL) and TMPH (5 mmol, (0.85 mL)) at
i
278 uC and the mixture was stirred for 10 min at 0 uC. Then, Bu3Al
(5 mmol, 5 mL of a 1 M solution in hexane) was added to the mixture at
This journal is ß The Royal Society of Chemistry 2007
Chem. Commun., 2007, 5241–5243 | 5243