An aluminum ate base: Its design, structure, function, and reaction mechanism

Article abstract of DOI:10.1021/ja064601n

An aluminum ate base, i-Bu₃Al(TMP)Li, has been designed and developed for regio- and chemoselective direct generation of functionalized aromatic aluminum compounds. Direct alumination followed by electrophilic trapping with I₂, Cu/Pd-catalyzed C-C bond formation, or direct oxidation with molecular O₂ proved to be a powerful tool for the preparation of 1,2- or 1,2,3-multisubstituted aromatic compounds. This deprotonative alumination using i-Bu₃Al(TMP)Li was found to be effective in aliphatic chemistry as well, enabling regio- and chemoselective addition of functionalized allylic ethers and carbamates to aliphatic and aromatic aldehydes. A combined multinuclear NMR spectroscopy, X-ray crystallography, and theoretical study showed that the aluminum ate base is a Li/Al bimetallic complex bridged by the nitrogen atom of TMP and the α-carbon of an i-Bu ligand and that the Li exclusively serves as a recognition point for electronegative functional groups or coordinative solvents. The mechanism of directed ortho alumination reaction of functionalized aromatic compounds has been studied by NMR and in situ FT-IR spectroscopy, X-ray analysis, and DFT calculation. It has been found that the reaction proceeds with facile formation of an initial adduct of the base and aromatic, followed by deprotonative formation of the functionalized aromatic aluminum compound. Deprotonation by the TMP ligand rather than the isobutyl ligand was suggested and reasoned by means of spectroscopic and theoretical study. The remarkable regioselectivity of the ortho alumination reaction was explained by a coordinative approximation effect between the functional groups and the counter Li⁺ ion, enabling stable initial complex formation and creation of a less strained transition state structure.

Full text of DOI:10.1021/ja064601n

Published on Web 01/31/2007
An Aluminum Ate Base: Its Design, Structure, Function,
and Reaction Mechanism
Hiroshi Naka,*,†,§,¶ Masanobu Uchiyama,*,‡,§,¶ Yotaro Matsumoto,§,¶
Andrew E. H. Wheatley,*, Mary McPartlin, James V. Morey,
and Yoshinori Kondo†
Contribution from the Graduate School of Pharmaceutical Sciences, Tohoku UniVersity,
Aobayama, Aoba-ku, Sendai 980-8578, Japan, AdVanced Elements Chemistry Laboratory,
The Institute of Physical and Chemical Research, RIKEN, 2-1 Hirosawa, Wako-shi, Saitama
351-0198, Japan, Department of Chemistry, UniVersity of Cambridge, Lensfield Road,
Cambridge, CB2 1EW, U.K., “Synthesis and Control,” PRESTO, Japan Science and Technology
Agency (JST), Japan, and Graduate School of Pharmaceutical Sciences,
The UniVersity of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan
Received June 29, 2006; E-mail: naka@mail.pharm.tohoku.ac.jp; uchiyama@mol.f.u-tokyo.ac.jp; aehw2@cam.ac.uk
Abstract: An aluminum ate base, i-Bu3Al(TMP)Li, has been designed and developed for regio- and
chemoselective direct generation of functionalized aromatic aluminum compounds. Direct alumination
followed by electrophilic trapping with I2, Cu/Pd-catalyzed C-C bond formation, or direct oxidation with
molecular O2 proved to be a powerful tool for the preparation of 1,2- or 1,2,3-multisubstituted aromatic
compounds. This deprotonative alumination using i-Bu3Al(TMP)Li was found to be effective in aliphatic
chemistry as well, enabling regio- and chemoselective addition of functionalized allylic ethers and carbamates
to aliphatic and aromatic aldehydes. A combined multinuclear NMR spectroscopy, X-ray crystallography,
and theoretical study showed that the aluminum ate base is a Li/Al bimetallic complex bridged by the
nitrogen atom of TMP and the R-carbon of an i-Bu ligand and that the Li exclusively serves as a recognition
point for electronegative functional groups or coordinative solvents. The mechanism of directed ortho
alumination reaction of functionalized aromatic compounds has been studied by NMR and in situ FT-IR
spectroscopy, X-ray analysis, and DFT calculation. It has been found that the reaction proceeds with facile
formation of an initial adduct of the base and aromatic, followed by deprotonative formation of the
functionalized aromatic aluminum compound. Deprotonation by the TMP ligand rather than the isobutyl
ligand was suggested and reasoned by means of spectroscopic and theoretical study. The remarkable
regioselectivity of the ortho alumination reaction was explained by a coordinative approximation effect
between the functional groups and the counter Li+ ion, enabling stable initial complex formation and creation
of a less strained transition state structure.
Introduction
portant synthetic intermediates in the formation of carbon-
carbon and carbon-heteroatom bonds, especially in aliphatic
Organoaluminum compounds have been widely used both
in industrial and laboratory synthetic chemistry,1 serving as
polymer synthesis catalysts,2 Lewis acid reagents,3 and organic
synthetic building blocks.4 As demonstrated by recent develop-
ments,5 organoaluminum species have become extremely im-
chemistry. Therefore, aromatic aluminum compounds should
be potentially attractive as functional materials and synthetic
building blocks.6 However, aromatic aluminum chemistry has
not been well developed, simply because of the poor synthetic
availability of these systems. A conventional preparative method
for aromatic aluminum compounds has been the transmetalation
of aryllithium or aryl Grignard reagents.7 This method, however,
suffers from the limited compatibility of functional groups on
† Tohoku University.
‡ RIKEN.
§ PRESTO, JST.
University of Cambridge.
¶ The University of Tokyo.
(4) (a) Nagata, W.; Yoshioka, M.; Hirai, S. J. Am. Chem. Soc. 1972, 94, 4635-
4643. (b) Chen, E. Y.-X.; Cooney, M. J. J. Am. Chem. Soc. 2003, 125,
7150-7151. (c) Maruoka, K.; Sano, H.; Shinoda, K.; Yamamoto, H. Chem.
Lett. 1987, 73-76. (d) Kabalka, G. W.; Newton, R. J., Jr. J. Organomet.
Chem. 1978, 156, 65-69.
(5) (a) Liang, B.; Novak, T.; Tan, Z.; Negishi, E.-i. J. Am. Chem. Soc. 2006,
128, 2770-2771. (b) Novak, T.; Tan, Z.; Liang, B.; Negishi, E.-i. J. Am.
Chem. Soc. 2005, 127, 2838-2839.
(6) Ishikawa, T.; Ogawa, A.; Hirao, T. J. Am. Chem. Soc. 1998, 120, 5124-
5125.
(1) (a) Mole, T.; Jeffrey, E. A. Organoaluminum Compounds; Elsevier:
Amsterdam, 1972. (b) Hashimoto, S.; Kitagawa, Y.; Iemura, S.; Yamamoto,
H.; Nozaki, H. Tetrahedron Lett. 1976, 30, 2615-2616. (c) Negishi, E.
J. Organomet. Chem. Libr. 1976, 1, 93-125.
(2) Boor, J. Ziegler Natta Catalysts and Polymerizations; Academic Press: New
York, 1979.
(3) (a) Saito, S. Aluminum in Organic Synthesis. In Main Group Metals in
Organic Synthesis; Yamamoto, H., Oshima, K., Eds.; Wiley-VCH: Wein-
heim, Germany, 2004; Vol 1, Chapter 6 and references therein. (b) Taylor,
M. S.; Zalatan, D. N.; Lerchner, A. M.; Jacobsen, E. N. J. Am. Chem. Soc.
2005, 127, 1313-1317. (c) Wieland, L. C.; Deng, H.; Snapper, M. L.;
Hoveyda, A. H. J. Am. Chem. Soc. 2005, 127, 15453-15456.
(7) Eisch, J. J. In ComprehensiVe Organometallic Chemistry; Wilkinson, G.,
Stone, F. G. A., Abel, E. W., Eds.; Pergamon Press: Oxford, 1982; Vol 6,
Chapter 6 and references therein.
9
10.1021/ja064601n CCC: $37.00 © 2007 American Chemical Society
J. AM. CHEM. SOC. 2007, 129, 1921-1930
1921

A R T I C L E S
Naka et al.
Table 1. Ligand Screenings of the Aluminum Ate Base
aromatic rings with intermediary ArLi or ArMgX species, or
their precursors (alkyllithiums or alkyl Grignard reagents), which
are too highly reactive toward various electronegative functional
groups, such as halogen, amide, and cyano groups and π-de-
ficient heterocycles.8 Hydro- or carbo-alumination, which is
known to be a powerful preparative method in aliphatic
chemistry,9 is ineffective for aromatics because of the structural
limitations of benzene rings.10
R3Al(TMP)Li for Direct Generation of Functionalized Aromatic
Aluminum Compoundsa
To overcome the preparative limitation of aromatic alumi-
nums discussed above, we recently reported that the direct regio-
and chemoselective generation of functionalized aromatic
aluminum compounds was achieved by developing a novel
aluminum ate base i-Bu3Al(TMP)Li (1).11 In this article, we
will give a full account of the design, structure, function, and
reaction mechanism of this aluminum ate base. First we detail
its development and discuss its scope and limitations for the
generation of functionalized aromatic and aliphatic compounds.
Next we characterize the structure of the aluminum ate base
using spectroscopic and theoretical methods. Finally we provide
a comprehensive mechanistic discussion of the directed ortho
alumination of functionalized aromatics by means of X-ray,
NMR, IR, and DFT studies.
entry
Ph−FG
reagentsb
product
yield (%)c
1
2
3
4
5
6
7
2a (PhOMe)
Me3Al(TMP)Li (1a)
Et3Al(TMP)Li
i-Bu3Al(TMP)Li (1)
Me3Al(TMP)Li (1a)
Et3Al(TMP)Li
4a
4a
4a
4b
4b
4b
4b
12
87
88
0
0
0
2a
2a
2b (PhCN)
2b
2b
2b
t-Bu3Al(TMP)Li
i-Bu3Al(TMP)Li (1)
100
a Reaction temperatures for the first step of the reaction are room
temperature for 2a and -78 °C for 2b. b TMP ) 2,2,6,6-tetramethylpip-
eridido. c Isolated yield.
To our delight, deprotonation with Me3Al(TMP)Li (1a) of
2a gave the corresponding iodinated compound 4a in 12% yield
after the iodine quench (entry 1). Use of Et3Al(TMP)Li (entry
2) and i-Bu3Al(TMP)Li (1) (entry 3) improved the system
dramatically, furnishing 4a in high yields. In contrast, depro-
tonative alumination of 2b suffered from significant decomposi-
tion of the CN group when R3Al(TMP)Li (R ) Me, Et, t-Bu)
was used (entries 4-6). Surprisingly, the only exception to this
rule was i-Bu3Al(TMP)Li (1), which gave 2-iodobenzonitrile
(4b) in excellent yield (entry 7). In the deprotonative alumination
of 2b, modification of the i-Bu group in 1 to alkoxides or
amides, of TMP to N(i-Pr)2 or N(TMS)2, or of Li to K, MgCl,
Mg(t-Bu), or Zn(t-Bu) also gave a complex mixture of products.
Solvent effects have also been tested for the alumination of 2b
with 1, and use of noncoordinative solvents such as hexane,
toluene or CH2Cl2, or lower-coordinative ethers resulted in
decreased yields. As we found 1 in THF to be a promising
chemoselective base, we next explored its reactivity with
variously functionalized aromatic compounds (Table 2). In the
following discussions, we define i-Bu3Al(TMP)Li (1) as alu-
minum ate base.
The aluminum ate base (1) was found to be an effective and
regioselective alumination reagent for a variety of (fused)
aromatic compounds bearing electron-donating groups such as
OMe and electron-withdrawing groups such as CN, amide, Cl,
and I (entries 1-12 in Table 2). Notably, deprotonative
alumination occurred with suppression of nucleophilic addition
to carbonyl and CN groups (entries 2, 3, and 5) or benzyne
formation with halogens (entries 4-6, 11, 12) and halogen-
metal exchange reaction at iodine (entries 4-6). Such chemose-
lectivity is considered to be unique to this aluminum ate base,
because neither conventional metal bases (such as RLi) nor even
TMP zincates can coexist with the aryl iodide.18 Heteroaromatics
such as pyridine, indole, benzofuran, and benzoxazole rings were
similarly applicable substrates (entries 13-16). Trifluorometh-
ylbenzene and ferrocenyl esters also tolerated alumination.
Results and Discussion
1. Development of the Aluminum Ate Base. To find ideal
alumination reagents with broad functional group compat-
ibilities, we first tested the halogen-aluminum exchange
reactions of aryl halides using various kinds of organoaluminum
reagents.12 While related alkyllithiums,13 alkylmagnesiums,14
and zincates15 have been known to effect halogen-metal
exchange reactions smoothly, all attempts using organoalumi-
nums were unsuccessful, in spite of extensive investigation. We
then employed a different strategy; the deprotonative alumina-
tion (i.e., hydrogen-aluminum exchange reaction) of function-
alized benzenes. Starting from cleaving an aromatic C-H bond,
this reaction would be more advantageous from the viewpoint
of the availability of aromatic precursors. Pioneering work on
the deprotonation of aliphatic C-H groups using tricoordinated
dialkylamidoaluminums has been reported.16 These reagents,
however, were found to be ineffective for our purpose.
Therefore, we designed a tetracoordinated aluminum ate base,
which should have higher reactivity over conventional tricoor-
dinated aluminums.17 We first screened ligand and counter
cation candidates. Anisole (2a) and benzonitrile (2b) were
selected as model aromatic compounds with electron-donating
and electron-withdrawing groups, respectively (Table 1).
(8) Upton, C. J.; Beak, P. J. Org. Chem. 1975, 40, 1094-1098.
(9) (a) Van Horn, D. E.; Negishi, E.-i. J. Am. Chem. Soc. 1972, 94, 4654-
4672. (b) Eisch, J. J. In ComprehensiVe Organic Synthesis; Trost, B. M.,
Ed.; Pergamon Press: Oxford, 1991; Vol 8, Chapter 3.
(10) Hart, H. In The Chemistry of Triple-Bonded Functional Groups, Supplement
C2; Patai, S., Ed.; John Wiley & Sons Ltd: Chichester, U.K., 1994; Chapter
18.
(11) Uchiyama, M.; Naka, H.; Matsumoto, Y.; Ohwada, T. J. Am. Chem. Soc.
2004, 126, 10526-10527.
(12) For an exceptional example of halogen-aluminum exchange reactions,
see: Maruoka, M.; Fukutani, Y.; Yamamoto, H. J. Org. Chem. 1985, 50,
4412-4414.
(13) Clayden, J. Organolithiums: SelectiVity for Synthesis; Pergamon: Elsevier
Science, Oxford, 2002.
(14) Boudier, A.; Bromm, L. O.; Lotz, M.; Knochel, P. Angew. Chem., Int. Ed.
2000, 39, 4414-4435.
(15) (a) Uchiyama, M.; Koike, M.; Kameda, M.; Kondo, Y.; Sakamoto, T.
J. Am. Chem. Soc. 1996, 118, 8733-8734. (b) Uchiyama, M.; Kameda,
M.; Mishima, O.; Yokoyama, N.; Koikee, M.; Kondo, Y.; Sakamoto, T.
J. Am. Chem. Soc. 1998, 120, 4934-4946.
(16) Yasuda, A.; Tanaka, S.; Oshima, K.; Yamamoto, H.; Nozaki, H. J. Am.
Chem. Soc. 1974, 96, 6513-6514.
(17) (a) Boireau, G.; Abenhaim, D.; Bernardon, C.; Henry-Basch, E.; Sabourault,
B. Tetrahedron Lett. 1975, 16, 2521-2524. (b) Yamamoto, Y.; Yatagai,
H.; Maruyama, K. J. Am. Chem. Soc. 1981, 103, 1969-1975. (c) Sasaki,
M.; Tanino, K.; Miyashita, M. Org. Lett. 2001, 3, 1765-1767.
9
1922 J. AM. CHEM. SOC. VOL. 129, NO. 7, 2007

Aluminum Ate Base
A R T I C L E S
Table 2. Deprotonative Alumination of Functionalized Aromatic Ringsa
a Unless otherwise noted, the deprotonative alumination was carried out using i-Bu3Al(TMP)Li (2.2 equiv) and substrate (1.0 equiv) in THF. b Isolated
yield. Items in parentheses are conditions of metalation. c Value in parentheses is the yield of the 2-deuterate (quenched with D2O).
Scheme 1. Electrophilic Trapping of the Functionalized Aryl
Aluminate Intermediate (3c)
Scheme 2. Thermally Controlled Generation and Suppression of
3-Functionalized Benzyne
convenient one-pot synthesis of functionalized phenols.19 In-
termediate 3c also undergoes copper- and palladium-catalyzed
C-C bond-forming reactions such as allylation, phenylation,
and benzoylation in high yields and with high chemo- and
regioselectivities.
Recently, we reported that the chemo- and regioselective
zincation of meta-functionalized haloaromatics and the genera-
tion of 3-substituted benzynes could be controlled by utilizing
the drastic ligand effects seen in these zincates.20 In the case of
the aluminum ate base, generation of benzynes could be
controlled by changing the reaction temperature (Scheme 2).
The intermediate 3s, generated by the deprotonative alumi-
nation of N,N-diisopropyl-3-bromo-2-iodobenzamide (2s), could
be trapped with an electrophile (I2) at low temperature, while
the generation of a 3-functionalized benzyne proceeded smoothly
at room temperature, and this reacted with 1,3-diphenylisoben-
zofuran to give the corresponding Diels-Alder adduct in
quantitative yield.
Having established a general preparative method for func-
tionalized aromatic aluminum compounds, we next demonstrated
that the functionalized aryl aluminate intermediate 3c (as a
typical intermediate) can be utilized as an aryl anion equivalent
(Scheme 1).
Intermediate 3c, generated by the deprotonative alumination
of 2c using i-Bu3Al(TMP)Li (1), was treated with D2O to give
the corresponding o-deuterated product in quantitative yield.
When the intermediate 3c was exposed to molecular oxygen in
the presence of 0.5 equiv of ZnCl2, the corresponding phenol
was obtained in 56% yield, realizing the regioselective introduc-
tion of an OH group. Since regio- and chemoselective direct
introduction of a hydroxyl group onto an aromatic ring is
generally difficult, the present procedure should provide a new,
As the aluminum ate base 1 proved to be an excellent direct
alumination reagent for functionalized aromatics, we next
(18) (a) Kondo, Y.; Shilai, M.; Uchiyama, M.; Sakamoto, T. J. Am. Chem. Soc.
1999, 121, 3539-3540. (b) Xu, C.; Yamada, H.; Wakamiya, A.; Yamagu-
chi, S.; Tamao, K. Macromolecules 2004, 37, 8978-8983. (c) Uchiyama,
M.; Matsumoto, Y.; Usui, S.; Hashimoto, Y.; Morokuma, K. Angew. Chem.,
Int. Ed. 2007, 46, 926-929. (d) Mulvey, R. E.; Mongin, F.; Uchiyama,
M.; Kondo, Y. Angew. Chem., Int. Ed. 2007, 46, in press (DOI: 10.1002/
anie.200604369).
(19) (a) Maleczka, R. E., Jr.; Shi, F.; Holmes, D.; Smith, M. R., III. J. Am.
Chem. Soc. 2003, 125, 7792-7793. (b) Ishiyama, T.; Takagi, J.; Ishida,
K.; Miyaura, N.; Anastasi, N. R.; Hartwig, J. F. J. Am. Chem. Soc. 2002,
124, 390-391.
(20) Uchiyama, M.; Miyoshi, T.; Kajihara, Y.; Sakamoto, T.; Otani, Y.; Ohwada,
T.; Kondo, Y. J. Am. Chem. Soc. 2002, 124, 8514-8515.
9
J. AM. CHEM. SOC. VOL. 129, NO. 7, 2007 1923

A R T I C L E S
Naka et al.
Table 3. Deprotonative Alumination of Functionalized Allylic
Compounds
ethers and carbamates under mild conditions. When MOM ether
was used, high regioselectivity was accomplished to furnish 6a
(quenched with n-BuCHO) and 7a (quenched with PhCHO) in
good yields (entries 1 and 2). Compound 1 also reacted
chemoselectively with allylcarbamates to give R-substituted
allylic compounds in moderate to excellent yields (entries 3-5).
Sequential alumination and electrophilic trapping of allyl-(2-
methoxyphenyl)ether 5d was found to be highly regio- and
stereoselective (entry 6). As this selectivity was not observed
in unsubstituted allylphenylether, it is possible that the methoxy
group works as an optional directing group for the counter Li+
cation, fixing the relative position and direction of the aldehyde
with respect to the allylaluminate.
Summarizing this section, aluminum ate base 1 has been
found to effect the deprotonative alumination of functionalized
aromatic, heteroaromatic, and allylic compounds with excep-
tionally high chemo- and regioselectivity. In the coming sections,
aiming to obtain suggestions for improving reactivity and
selectivity of the current system and to achieve the logical design
of more efficient reagents, a structural study of i-Bu3Al(TMP)-
Li (1) and a mechanistic investigation of its alumination
reactions are discussed in detail.
2. Structural Study of the Aluminum Ate Base. The
developed aluminum ate base i-Bu3Al(TMP)Li (1) can be
prepared simply by mixing i-Bu3Al and LTMP in a 1:1 ratio in
THF, and one would expect the lithium aluminum ate structure
of 1 to be that shown in Scheme 3.26
a Isolated yield. Diastereomeric ratio of R-products (erythro/threo); entry
1 (87:13), entry 2 (56:44), entry 3 (not determined), entry 4 (82:17), entry
5 (58:42), entry 6 (99:1). E/Z ratios of γ-products were not determined.
extended its scope to functionalized aliphatic compounds.
Various oxygen-substituted allylic species were chosen for
substrates because of their easy accessibility and their potential
transformability.21 For instance, allylcarbamates are attractive
substrates as they can be easily converted to π-allylpalladium
species.22 A conventional direct preparative method of an allylic
anion intermediate requires deprotonation using reactive alkali
metal species such as alkyllithiums and involves low functional
group compatibilities and inefficient R/γ regioselectivity.23
Transmetalation of these allylic alkali metals with Et3Al to give
allylic aluminates, or with Bu3SnCl to give allylic tin com-
pounds, improved the regioselectivity dramatically, reaction with
aldehydes occurring at the R position.24 However, this stepwise
strategy needs strictly controlled reaction conditions in the first
deprotonation step. Moreover, it has been difficult to apply to
allylic compounds with reactive functional groups, such as
carbonyl or phenoxy groups, because the highly reactive
alkyllithium reagents or intermediary allylmetal species cause
SN2 or SN2′ decomposition of allylic substrates or products.25
Therefore, a direct preparative route to a functionalized allylic
anion equivalent under mild conditions via successive regiose-
lective transformations is highly desirable. We found that this
could be achieved by direct generation of the functionalized
allylic aluminum compounds using 1 as base, realizing the one-
pot regioselective transformation under mild conditions
(Table 3).
Scheme 3
We first studied THF solutions of 1, i-Bu3Al and LTMP by
1H, 13C, 7Li, 15N, and 27Al NMR spectroscopy at -50 °C
(Table 4).
The aluminum ate base 1 has been observed as a novel single
species in THF solution. No signals corresponding to i-Bu3Al
or LTMP were detected.32 The 27Al chemical shift (153.4 ppm)
of 1 showed clean formation of a tetrahedrally disposed
aluminum center, indicating strong coordination of the nitrogen
of the TMP ligand to the aluminum. These results confirmed
clean and complete formation of the “ate” complex 1 as shown
in Scheme 3. However, since the three isobutyl groups of 1
were identical on the NMR time scale, the issue left for complete
structural determination was to understand the relative positions
of Li+ ion and the other fragments (Figure 1).
The deprotonative alumination of oxygen-substituted allylic
compounds by 1 was found to be regioselective and tolerant of
(21) Yamamoto, Y.; Asao, N. Chem. ReV. 1993, 93, 2207-2293.
(22) (a) Guibe, F. Tetrahedron 1998, 54, 2967-3042. (b) Hoppe, D. Angew.
Chem., Int. Ed. 1984, 23, 932-948.
(23) Still, W. C.; Macdonald, T. L. J. Am. Chem. Soc. 1974, 96, 5562-5563.
(24) (a) Yamamoto, Y.; Yatagi, H.; Saito, Y.; Maruyama, K. J. Org. Chem.
1984, 49, 1096-1104. (b) Yamamoto, Y. Acc. Chem. Res. 1987, 20, 243-
249. (c) Kadota, I.; Yamamoto, Y. Acc. Chem. Res. 2005, 38, 423-432.
(25) 25. In situ generation of allylmetal species from allylhalides represents an
alternative approach, though the accessibility of functionalized precursors
is a concern. See: Kurosu, M.; Lin, M.-H.; Kishi, Y. J. Am. Chem. Soc.
2004, 126, 12248-12249.
Figure 1. Two possible isomers of the aluminum ate base 1.
(26) Wittig, G. Q. ReV. 1996, 191-210.
9
1924 J. AM. CHEM. SOC. VOL. 129, NO. 7, 2007

Aluminum Ate Base
A R T I C L E S
Table 4. 1H, 13C, 7Li, 15N, and 27Al NMR Spectroscopic Data for THF Solutions of LTMP, i-Bu3Al, and i-Bu3Al(TMP)Li (1)a
1H
13C
7Li
15N
27Al
i-Bu3Al
-0.16 (d, 6H), 1.72 (m, 3H)
0.85 (m, 18H)
1.06 (s, 12H), 1.15 (4H)
1.56 (m, 2H)
0.98 (s, 12H), 1.20 (4H)
1.54 (m, 2H)
23.4 (R), 27.8 (â)
29.1 (γ)
174.4
LTMP
20.9 (γ), 36.2 (Me)
43.3 (â), 53.1 (R)
19.3 (TMPγ), 32.3 (TMPMe
39.0 (TMPâ), 50.1 (TMPR)
24.4 (i-BuR), 30.2 (i-Buâ)
29.4 (i-Buγ)
0.61 (monomer)
1.26 (dimer)
-0.75
79.9
77.0
i-Bu3Al(TMP)Li
)
153.4
-0.47 (6H), 1.69 (m, 3H)
0.77 (m, 18H)
a Reported NMR data for i-Bu3Al (δ(ppm) R-H (0.30), â-H (1.94), γ-H (0.98) in benzene-d6 (ref 27), 27Al (276) in toluene-d8 (ref 28)) and LTMP
7
(δ(ppm) R-C (53.1), â-C (43.1), γ-C (20.9), Me (36.3), Li (0.7 (monomer)) and (1.3 (dimer)), 15N (79.7) in THF-d8) (refs 29-31).
Figure 2. Stationary points of carbon-bridged structure A and nitrogen-bridged structure B for non-, mono-, and disolvated states at the B3LYP/6-31+G*
level with zero-point corrections and entropy term corrections. Free energies given are in kcal/mol and are relative to the monosolvated structure of B.
Two structural isomers are possible: Li+ connected to two
alkyl groups (A) or to TMP and one alkyl group (B). To decide
which is more likely, we compared the relative energies of these
isomers using computational methods (Figure 2).
We first used a theoretical model of Me3Al(Me2N)Li for 1
and Me2O for solvent THF. Non-, mono-, and disolvated states
of A and B were considered. The structure B was found to be
more favorable than A by 10-13 kcal/mol at the B3LYP/6-
31+G* level of theory, irrespective of the solvation numbers,
with monosolvated B found to be of lowest in energy.33
Table 5. 1H, 13C, 7Li, 15N, and 27Al NMR GIAO-Predicted Data of
i-Bu3Al(TMP)Li (1)a
1H
13C
7Li
15N
27Al
1.16 (12H), 1.27 (4H) 18.5 (TMPγ), 30.5 (TMPMe
1.61 (2H)
45.1 (TMPâ), 56.3 (TMPR)
-0.15 (6H), 1.87 (3H) 28.7 (i-BuR), 30.8 (i-Buâ)
0.90 (18H)
25.7 (i-Buγ)
) -2.30 80.5 141.0
a Mono-(Me2O)-i-Bu3Al(TMP)Li has been optimized at the B3LYP/6-
31+G* level of theory. GIAO values were calculated at the B3LYP/6-
311+G* level of theory and referenced to THF (1H, 13C), LiBr (7Li), aniline
(15N), and Al (NO3)3 (27Al). Values of equivalent protons and carbons were
averaged.
(27) Sen, B.; White, G. L. J. Inorg. Nucl. Chem. 1973, 35, 2207-2215.
(28) Benn, R.; Rufinska, A.; Lehmkuhl, H.; Janssen, E.; Kturhrt, C. Angew.
Chem., Int. Ed. 1983, 22, 779-780.
(29) Hall, P. L.; Gilchrist, J. H.; Harrison, A. T.; Fuller, D. J.; Collum, D. B.
J. Am. Chem. Soc. 1991, 113, 9575-9585.
Next, we conducted full optimizations of i-Bu3Al(TMP)Li
with 0, 1, or 2 Me2O solvent molecules using the same level of
theory. In harmony with the Me3Al(Me2N)Li model, the
monosolvated structure has been found to be most stable, and
the GIAO-predicted chemical shifts of this monosolvated i-Bu3-
Al(TMP)Li showed good agreement with the experimental
values (Tables 4 and 5).
(30) Renaud, P.; Fox, M. A. J. Am. Chem. Soc. 1988, 110, 5702-5705.
(31) Romesberg, F. E.; Gilchrist, J. H.; Harrison, A. T.; Fuller, D. J.; Collum,
D. B. J. Am. Chem. Soc. 1991, 113, 95751-5757.
(32) See the SI for each NMR chart.
(33) 27Al NMR GIAO calculation for these structures also supported this view.
27Al NMR GIAO calculation at the level of B3LYP/6-311+G*//B3LYP/
6-31+G* (values in ppm and referenced to Al(NO3)3): A, 135.15; A-Me2O,
134.67; A-2Me2O, 133.33; B, 155.97; B-Me2O, 154.22; B-2Me2O, 153.71.
Experimental value: 153.4 ppm.
9
J. AM. CHEM. SOC. VOL. 129, NO. 7, 2007 1925

A R T I C L E S
Naka et al.
and the R-carbon of a single i-Bu group, with the counter Li+
ion exclusively serving as a recognition point for coordinative
solvents and/or electronegative functional groups.
3. Mechanism of Directed ortho Alumination of Func-
tionalized Aromatic Compounds. Having obtained the struc-
ture of the aluminum ate base that is likely to be experimentally
relevant, we next investigated the reaction mechanism of
directed ortho alumination in detail.
3.1. Characterization of Aluminated Aromatic Com-
pounds. To obtain direct spectroscopic evidence for the
formation of the aluminated arene intermediates, we monitored
reactions using 13C NMR spectroscopy (Figure 4).
Anisole (2a) was chosen as aromatic substrate, and we first
monitored its stepwise reaction with t-BuLi to form ortho-
lithiated anisole (3a′) and then with i-Bu3Al to give ortho-
aluminated compound 3a. In the first step, the chemical shift
for the lithiated ortho carbon moved significantly to low field
(from δ ) 114.3 ppm (spectrum a) to 163.7 ppm (spectrum
b)), revealing the formation of 3a′. Addition of i-Bu3Al gave
ortho-aluminated 3a with the chemical shift of the ortho carbon
at 152.9 ppm (spectrum c). Next we investigated the direct
reaction of 2a with 1. In spectrum d, each signal for the newly
emerged intermediate was found to be identical with that of
stepwise-prepared 3a, with no signals for 3a′ being found. These
results clearly showed that the aluminum ate base reacts with
2a to directly generate ortho-aluminated aryl species 3a.
3.2. Regioselectivity in Deprotonative Alumination. A
considerable number of examples of directed ortho metalation
have been recorded,36 and their regioselectivities have been
generally explained by a complex-induced approximate effect
and/or an acidifying inductive effect of the aromatic substi-
tutents.37 However, the mechanistic details of directed metalation
Figure 3. Molecular structure of 1-THF. Hydrogen atoms are omitted
for clarity. Atoms are plotted at 40% probability.
Finally, we have obtained X-ray crystallographic evidence
for the structure of i-Bu3Al(TMP)Li (1)-THF complex
(Figure 3).
The sequential treatment of HTMP with n-BuLi and i-Bu3Al
in THF afforded crystalline 1-THF. Its structure reveals that
the lithium ion acts as an exclusive recognition point for
electron-rich donors. Notably, the essential structure observed
revealed a core Li-N-Al-C metallocycle34 with a bridging
TMP and single bridging i-Bu ligand (Li1-N1 1.985(4), Li1-
C14 2.258(4), Al1-N1 1.9900(16), Al1-C14 2.067(2) Å).35
The similarity borne by the solid-state structure of 1-THF to
the calculated structure of monosolvated B is striking, and
strongly suggests the validity of our NMR spectroscopic and
computational analysis.
On the basis of this experimental and theoretical evidence,
we have concluded that the aluminum ate base is likely to be a
Li/Al bimetallic complex, bridged by the nitrogen atom of TMP
Figure 4. 13C NMR spectra of (a) anisole (2a) in THF, (b) 2a + t-BuLi, (c) 2a + t-BuLi, then added i-Bu3Al, (d) 2a + i-Bu3Al(TMP)Li. Signals for
2-lithioanisole (3a′ in b) and (2-MeOC5H4)Al(i-Bu)3Li (3a in c, d) are marked with dots.
9
1926 J. AM. CHEM. SOC. VOL. 129, NO. 7, 2007

Aluminum Ate Base
A R T I C L E S
Figure 5. Stationary and transition points during possible deprotonative aluminations of 2a with Me3Al(Me2N)Li-OMe2 at the B3LYP/6-31+G* level of
theory. Energy changes at B3LYP/6-31+G* level during deprotonation with NMe2 ligand or Me ligand are shown in kcal/mol.
with ate complexes seem complicated and unclear.38 To shed
light on the mechanism of directed ortho alumination, we
investigated the possible pathways using DFT calculations
(Figure 5).
in situ FT-IR spectroscopy to monitor the alumination of 2c,
for it possess a CdO group with a strong stretching vibration
(Figure 6).
Addition of 2c to a THF solution of 1 at -25 °C gave a
transient peak at 1585 cm-1 and, along with the disappearance
of this peak, a new signal at 1600 cm-1. Elevating the
temperature to room temperature did not cause any significant
spectral change, and the reaction mixture was quenched with
iodine to give 2-iodo-N,N-diisopropylbenzamide 4c. Throughout
this process no lithiated aromatic compounds (1578 cm-1 in
THF)18a were detected. These results suggested that the signal
at 1600 cm-1 was attributable to the carbonyl group of ortho-
aluminated N,N-diisopropylbenzamide (3c) and that the signal
at 1585 cm-1 originated from an initial complex between 1 and
2c, in accordance with theoretical study. As these frequencies
are lower than that of the substrate (1636 cm-1), a direct
interaction between the carbonyl functional group and the Li+
ion in these intermediates can be strongly implied.
To our delight, a snap shot of the type of initial complex
indicated by in situ FT-IR spectroscopy was obtained by X-ray
crystallographic analysis (Figure 7) of the product of treatment
of preisolated 1a (methyl analogue of 1) with 2c in THF.
Two crystallographically independent rotormers are observed
in the asymmetric unit of the 1:1 electrostatic adduct 1a-2c.
In either case the lithium ion is in contact with the oxygen atom
of the carboxylic amide functional group.41 The 1a component
possesses a heterobimetallic core, retaining the essential struc-
tural features noted for 1-THF. The major difference between
the two adducts lies in the torsional angles observed between
carbon-oxygen and lithium-TMP interactions; at C1-O1-
Li1-N2 53.0(10)° and C26-O2-Li2-N4 40.6(13)° these
suggest some flexibility in the precise arrangement of the two
components of either independent adduct. These structural
results are in harmony with data obtained from NMR spectros-
copy, calculation, and in situ FT-IR observation.
We chose 2a as a model aromatic substrate and Me3Al-
(Me2N)Li-OMe2 (monosolvated B in Figure 3) as the aluminum
ate base on the dual grounds of its relative calculational
simplicity and its successful structural characterization.39 Al-
though the choice of this simplified model system may lead to
an underestimation of the steric effects of bulky groups (e.g.,
i-Bu), the essence of the actual reaction nature should still be
observable using this model system.40
Reaction coordinates started with formation of a relatively
stable initial complex (IM1), with no other pathways (that
omitted formation of this intermediate) being found. This
theoretical result indicated that the regioselectivity of the ortho
alumination reaction can be explained by a coordinative
approximation effect between functional group and Li+ ion,
enabling initial complex formation and orienting the ate base
ligand exclusively toward aromatic ortho hydrogen. To observe
an interaction between Li+ and the directing functional group
on the aromatic substrate in the initial complex, we then used
(34) (a) Niemeyer, M.; Power, P. P. Organometallics 1995, 14, 5488-5489.
(b) Armstrong, D. R.; Craig, F. J.; Kennedy, A. R.; Mulvey, R. E. Chem.
Ber. 2001, 129, 1293-1300. (c) Linton, D. J.; Schooler, P.; Wheatley,
A. E. H. Coord. Chem. ReV. 2001, 223, 53-115. (d) Cui, C.; Schmidt,
J. A. R.; Arnold, J. Dalton Trans. 2002, 2992-2994.
(35) Rutherford, D.; Atwood, D. A. J. Am. Chem. Soc. 1996, 118, 11535-
11540.
(36) Hartung, C. G.; Snieckus, V. The Directed ortho Metalation Reaction. A
Point of Departure for New Synthetic Aromatic Chemistry. In Modern
Arene Chemistry; Astruc, D., Ed.; Wiley-VCH: New York, 2002.
(37) (a) Beak, P.; Meyers, A. I. Acc. Chem. Res. 1986, 19, 356-363. (b) Bauer,
W.; Schleyer, P. v. R. J. Am. Chem. Soc. 1989, 111, 7191-7198. (c)
Whisler, M. C.; MacNeil, S.; Snieckus, V.; Beak, P. Angew. Chem., Int.
Ed. 2004, 43, 2206-2225. (d) Rennels, R. A.; Maliakal, A. J.; Collum,
D. B. J. Am. Chem. Soc. 1998, 120, 421-422. (e) Chadwick, S. T.; Rennels,
R. A.; Rutherford, J. L.; Collum, D. B. J. Am. Chem. Soc. 2000, 122, 8640-
8647.
(38) (a) Clegg, W.; Dale, S. H.; Hevia, E.; Harrington, R. W.; Hevia, E.;
Honeyman, G. W.; Mulvey, R. E. Angew. Chem., Int. Ed. 2006, 45, 2370-
2374. (b) Clegg, W.; Dale, S. H.; Harrington, R. W.; Hecia, E.; Honeyman,
G. W.; Mulvey, R. E. Angew. Chem., Int. Ed. 2006, 45, 2374-2377.
(39) A model system without a solvent molecule was also a possible choice,
and we confirmed that it provided similar energy values to the presently
discussed solvated system.
(40) Both Me3Al(TMP)Li and Me3Al(TMP)Li-THF adopt structures analogous
to that of 1-THF. See the SI for details.
(41) In contrast to recent work on the N,N-diisopropylbenzamide adduct of
(t-Bu)2Zn(TMP)Li, both rotormers of 1a-(N,N-diisopropylbenzamide) place
the TMP ligand in proximity to the target aryl. See ref 38a.
9
J. AM. CHEM. SOC. VOL. 129, NO. 7, 2007 1927

A R T I C L E S
Naka et al.
Figure 6. In situ FT-IR monitoring of the deprotonative alumination of 2c with 1.
Figure 7. Molecular structure of the two crystallographically independent rotormers of 1a-2c complex. Hydrogen atoms are omitted for clarity, and atoms
are plotted at 40% probability.
Scheme 4. TMP-Mediated Deprotonation versus
Isobutyl-Mediated Deprotonation
Overall, it was concluded that the remarkable regioselectivity
of the ortho alumination reaction was explicable by there being
a coordinative approximation effect between functional group
and counter Li+ ion, enabling stable initial complex formation
and a less strained transition structure.
3.3. Ligand Selectivity in Deprotonative Alumination.
Without, as yet, any concrete picture of the reaction pathway
in respect of the ligand selectivity shown by the aluminum ate
base, we extended our investigation to consider transitions states
in deprotonation of a functionalized aromatic by the TMP
nitrogen ligand and by the isobutyl carbon ligand (Scheme 4).
Because it is well-known that nitrogen ligands are generally
Of several possible TSs for the deprotonation of anisole using
more reactive than carbon ligands in mixed ligand aluminum
the Me and N ligand on the model aluminate, we identified
compounds,42 TMP-mediated deprotonation is expected to be
only two TSs, TS1 and TS2 (Figures 5 and 8), that were
energetically plausible. From IM1, both amido-mediated and
more plausible than isobutyl-mediated deprotonation. In the case
of related TMP-zincate base, however, this ligand selectivity
has been reported to be variable depending on the specifics of
(43) (a) Clegg, W.; Dale, S. H.; Drummond, A. M.; Hevia, E.; Honeyman,
the complexes themselves and the reaction conditions.43
G. W.; Mulvey, R. E. J. Am. Chem. Soc. 2006, 128, 7434-7435. (b)
Armstrong, D. R.; Clegg, W.; Dale, S. H.; Hevia, E.; Hogg, L. M.;
Honeyman, G. W.; Mulvey, R. E. Angew. Chem., Int. Ed. 2006, 45, 3775-
(42) (a) Lipton, M. F.; Basha, A.; Weinreb, S. M. Organic Synthesis; Wiley:
New York, 1988; Collect Vol VI, pp 492-495. (b) Maruoka, K.; Oishi,
M.; Yamamoto, H. J. Org. Chem. 1993, 58, 7638-7639.
3778. (c) Mulvey, R. E. Organometallics 2006, 25, 1060-1075. (d)
Uchiyama, M.; Matsumoto, Y.; Nobuto, D.; Furuyama, T.; Yamaguchi,
K.; Morokuma, K. J. Am. Chem. Soc. 2006, 128, 8748-8750.
9
1928 J. AM. CHEM. SOC. VOL. 129, NO. 7, 2007

Aluminum Ate Base
A R T I C L E S
Figure 8. Stationary and transition structures during possible deprotonative aluminations of 2a at the B3LYP/6-31+G* level of theory. Energies are given
in kcal/mol and are relative to structure IM1. Deprotonation by the nitrogen-centered ligand: (a) IM1, (b) TS1, (c) IM2. Deprotonation by the carbon-
centered ligand: (d) IM3, (e) TS2, (f) IM4. Both TS1 and TS2 are denoted as “open form” TS, whereas no “closed form” TS had been found in either
system (ref 44).
alkyl-mediated paths are found. In the NMe2-mediated path,
dissociation of the nitrogen atom from lithium leads to a
transition state structure (TS1) with an activation barrier of 29.8
kcal/mol, producing an aluminated aromatic compound, IM2.
Deprotonation by the Me ligand on Al (IM3-TS2-IM4) was
found to be kinetically unfavorable, requiring a much higher
activation energy of 47.9 kcal/mol (18.6 kcal/mol higher than
NMe2). These results strongly suggest that deprotonative alu-
mination is affected by the TMP ligand rather than by the i-Bu
ligand, which is reasonably consistent with recent reported work
Figure 9. Kohn-Sham highest occupied molecular orbital (HOMO) of
TS1 and TS2 (ref 45).
on TMP zincate.43d
Analysis of local structure of the mediating ligand in either
TS (Figure 8) and of the orbital(s) involved in Cortho-H scission
(Figure 9) provided a key to understanding this remarkable
ligand transfer selectivity.
angle, 155.3°). Here, the unoccupied aromatic σC-H* orbital
needs to interact with the contracted and ill directed sp3-like
orbital of the carbon atom, resulting in a late transition state
(Cortho-H, 1.59 Å) with a higher activation barrier.
In TS1, the nitrogen atom in the NMe2 ligand has a relaxed
tetrahedral structure, allowing optimal interaction of the nitrogen
lone pair with an ortho-hydrogen (Al-N-H angle, 110.2°;
N-H-Cortho angle, 170.5°). In the orbital analysis, therefore,
an efficient linear orbital overlap between the lone pair of
nitrogen and the unoccupied aromatic σC-H* orbital is observed,
allowing an early transition state (Cortho-H, 1.48 Å) with lower
activation energy for the NMe2-mediated deprotonation pathway.
In the case of deprotonation by the Me ligand (TS2), however,
the bridging carbon atom reveals an unfavorable five-coordinate
distorted geometry (Al-CMe-H angle, 73.9°; CMe-H-Cortho
In accordance with the direct 13C NMR spectroscopic
observation of TMP-mediated deprotonation product 3a, we
concluded, with theoretical justification, that the deprotonation
of functionalized arenes with aluminum ate base takes place
by virtue of the TMP ligand rather than the isobutyl ligand.
3.4. Summary of Mechanism in Directed ortho Alumina-
tion. On the basis of experimental and theoretical evidence, it
can be suggested that aluminum ate base 1 deprotonates
functionalized aromatic compounds directly, to generate func-
tionalized aromatic aluminum compounds. Observation of the
initial 1:1 electrostatic complex, in which Lewis acidic Li+
interacts with the functional group of the arene, provided a
mechanistic insight into regioselectivity of this novel directed
ortho alumination. Concerning ligand transfer selectivity of the
aluminum ate base, deprotonation by TMP rather than by an
isobutyl group is preferred.
(44) Excellent ideas and methods to generate the “open form” TS in organo-
lithium (lithium homobimetal) chemistry have been reported, for ex-
ample: (a) Zhao, P.; Collum, D. B. J. Am. Chem. Soc. 2003, 125, 4008-
4009. (b) Zhao, P.; Collum, D. B. J. Am. Chem. Soc. 2003, 125, 14411-
14424. (c) Zhao, P.; Condo, A.; Keresztes, I.; Collum, D. B. J. Am. Chem.
Soc. 2004, 126, 3113-3118. (d) Qu, B.; Collum, D. B. J. Am. Chem. Soc.
2005, 127, 10820-10821.
(45) Kohn, W.; Sham, L. J. Phys. ReV. 1965, 140, A1133-A1138.
9
J. AM. CHEM. SOC. VOL. 129, NO. 7, 2007 1929

A R T I C L E S
Naka et al.
Conclusion
new chemoselective base reagents are in progress in our
laboratory. Furthermore, in conjunction with our recent inves-
tigations,46 complete separation and controlled switching of
reactivities in nucleophilic addition, deprotonation, halogen-
metal exchange, redox, and transmetalation reactions are envis-
aged.
The aluminum ate base, i-Bu3Al(TMP)Li, has been developed
as a novel base reagent, realizing highly chemo- and regiose-
lective deprotonative aluminations of functionalized aromatic
and heteroaromatic compounds, as well as functionalized allylic
compounds. This simple and straightforward method will
provide direct and efficient access to valuable synthetic inter-
mediates.
Full structural characterization of i-Bu3Al(TMP)Li and a
detailed mechanistic investigation of its role in directed ortho
alumination have been conducted by means of NMR spectros-
copy, in situ FT-IR spectroscopic monitoring, and X-ray
crystallographic and theoretical methods, and the origins of the
remarkable regioselectivity and ligand transfer selectivity
observed have been elucidated. With the comprehensive struc-
tural and mechanistic knowledge acquired during this research,
improvement of the reaction system and the logical design of
Acknowledgment. This research was partly supported by the
Astellas Foundation, The Sumitomo Foundation, and a Grant-
in-Aid for Young Scientists (A), HOUGA, and Priority Area
(No. 452) from the Ministry of Education, Culture, Sports,
Science and Technology, Japan (to M.U.). This research was
also supported by the Exploratory Research Program for Young
Scientists (ERYS) (to H.N. and Y.M). We also gratefully
acknowledge the U.K. EPSRC for financial support (M.McP.
and J.V.M.). We thank Dr. Robert P. Davies (Imperial College
London) for aid with crystallography and Professor T. Ohwada
(The University of Tokyo) for his valuable comments.
Supporting Information Available: Details of the experi-
mental procedures, spectral and crystallographic data. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(46) (a) Uchiyama, M.; Matsumoto, Y.; Nakamura, S.; Ohwada, T.; Kobayashi,
N.; Yamashita, N.; Matsumiya, A.; Sakamoto, T. J. Am. Chem. Soc. 2004,
126, 8755-8759. (b) Kobayashi, M.; Matsumoto, Y.; Uchiyama, M.;
Ohwada, T. Macromolecules 2004, 37, 4339-4341. (c) Uchiyama, M.;
Furuyama, T.; Kobayashi, M.; Matsumoto, Y.; Tanaka, K. J. Am. Chem.
Soc. 2006, 128, 8404-8405. (d) Nakamura, S.; Uchiyama, M. J. Am. Chem.
Soc. 2007, 129, 28-29.
JA064601N
9
1930 J. AM. CHEM. SOC. VOL. 129, NO. 7, 2007