A new approach to substituted cyclobutanes: Direct β-deprotonation/magnesiation of cyclobutane carboxamides

Article abstract of DOI:10.1055/s-2003-40331

BuMgN(i-Pr)₂ is shown to be kinetically and thermodynamically efficient for deprotonation/magnesiation cis and β to an activating carboxamido group on a cyclobutane ring. The resulting carboxamido-stabilized amido-Grignards are useful reagents. The method provides an entirely new approach to controlled substitution on cyclobutanes. BuMgN(i-Pr)₂ will also metalate pivaloyl amides.

Full text of DOI:10.1055/s-2003-40331

LETTER
1275
A New Approach to Substituted Cyclobutanes: Direct b-Deprotonation/
Magnesiation of Cyclobutane Carboxamides
a
a
a
a
a
b
N
P
ew
A
pproac
h
h
to Substitut
i
e
d
Cyc
l
lobuta
n
i
es p E. Eaton,* Mao-Xi Zhang, Naruyoshi Komiya, Cai-Guang Yang, Ian Steele, Richard Gilardi
a
Department of Chemistry, The University of Chicago, Chicago, IL 60637, USA
Fax +1(773)7022053; E-mail: eaton@uchicago.edu
b
Laboratory for the Structure of Matter, The Naval Research Laboratory, Washington, D.C. 20375, USA
Received 17 March 2003
In memoriam: Professor Ray Lemieux, gentleman and scholar.
is only about five times more acidic than cyclopentane
and 2,500 times less acidic than cyclopropane. Nonethe-
less, as we show now, cyclobutyl carboxamides are effec-
Abstract: BuMgN(i-Pr) is shown to be kinetically and thermody-
namically efficient for deprotonation/magnesiation cis and b to an
activating carboxamido group on a cyclobutane ring. The resulting
2
carboxamido-stabilized amido-Grignards are useful reagents. The tively and usefully deprotonated/magnesiated by
method provides an entirely new approach to controlled substitution BuMgN(i-Pr)2.
on cyclobutanes. BuMgN(i-Pr) will also metalate pivaloyl amides.
2
1
-Methyl-1-(N,N-diisopropylcarboxamido)cyclobutane (3)
Key words: metalation, diastereoselective, amido-Grignards, cy-
clobutanes, substituent effects
was prepared by a-alkylation of the parent 4, itself made
in standard fashion via the acid chloride from commer-
cially available cyclobutane carboxylic acid (Scheme 2).
Although other bases can be used for a-deprotonation/
Alkyl magnesium amides, e.g. BuMgN(i-Pr) , deproto-
1
2
metalation, we chose BuMgN(i-Pr) . It is simple to make,
2
nate and magnesiate mildly acidic CH groups near, but not
reasonably stable even in refluxing THF, and very effec-
tive for stoichiometric deprotonations (butane being
formed irreversibly). Reaction of the intermediate a-ami-
do-Grignard 5 with methyl iodide gave 3, isolated in
1
necessarily bearing, a carboxamido group. Thus, phenyl,
cyclopropyl 1, and cubyl carboxamides are converted
(
e.g. Scheme 1) into synthetically useful carboxamido-
2
stabilized vicinal amido-Grignards 2.
8
8–95% yield as a colorless solid, mp 55.0–55.5 °C.
Scheme 1
Scheme 2
How acidic must the hydrogen of the CH group be for a
useful rate of deprotonation/magnesiation? Any quantita-
tive answer would require a new scale incorporating the
activating and stabilizing effects of the carboxamido
group. Qualitatively, we expect (and find) a rough corre-
BuMgN(i-Pr) in THF/heptanes will deprotonate/magne-
2
siate the b-position of cyclopropyl amide 1 stoichiometri-
cally at 20 °C. The corresponding cyclobutyl amide 3 is
unreactive under these conditions, but treatment with ex-
3
lation of rate with relative kinetic acidities: benzene, ca.
cess BuMgN(i-Pr) in THF/heptanes at reflux (73 °C)
2
+
8
+4
+4
1
0 ; cubane, 6 × 10 ; cyclopropane, 7 × 10 . On the
gave a good rate and produced what is taken to be the
4
same scale, the kinetic acidities of cyclopentane and cy-
clohexane are much less: 5.7 and 1.0, respectively. Under
b-carboxamido stabilized amido-Grignard 6. Addition of
the cooled reaction mixture to a solution of elemental
iodine in THF (Scheme 3) gave a mixture of two iodides
standard conditions (vide infra), BuMgN(i-Pr) fails to
2
1
5
significantly magnesiate 5- or 6-membered rings b to a
carboxamido group thereupon, even when the a-position
is blocked. Cyclobutane CH (relative kinetic acidity = 28)
in an H NMR ratio of ca. 10:1. We isolated a 77% yield
of the major iodide; its stereochemistry is as shown in 7,
established unambiguously by X-ray crystallography.6
The minor isomer is the epimeric iodide.
a
Synlett 2003, No. 9, Print: 11 07 2003.
Art Id.1437-2096,E;2003,0,09,1275,1278,ftx,en;S02803ST.pdf.
©
Georg Thieme Verlag Stuttgart · New York

1
276
P. E. Eaton et al.
LETTER
Equation 1
We believe that proton removal from activated CH groups
is done by the lone pair of nitrogen in BuMgN(i-Pr) or
2
Mg[N(i-Pr) ] , the latter present in trace amounts. Car-
2
2
bon-bound metals are kinetically poor at deprotonating
CH groups, whereas metal salts of amines are much bet-
Scheme 3
ter. Indeed, we have found that Bu Mg is useless in this re-
2
In the major isomer, the iodine has been introduced trans
to the activating amide group, opposite to the stereospec-
ificity found for deprotonation/magnesiation/iodination
1
gard, whereas Mg[N(i-Pr) ] is effective kinetically (but
2
2
not thermodynamically). For this reason we show 11
Scheme 5) as a precursor of 6 (and perhaps a part precur-
sor for 7, 8, and 9). If Mg[N(i-Pr) ] is the actual proton
(
1
of the cyclopropane analog. This may be due to the lesser
2
2
configurational stability of cyclobutyl radicals and anions
abstractor, 6 would be formed directly. In all cases the
reactions are driven forward by formation of butane.
7
relative to the cyclopropyl species. However, exactly as
in the cyclopropane series, carboxylation of 5 (Scheme 4)
8
led to the cis-acid 8 in 80% isolated yield. The stereo-
chemistry of 8 was assigned unambiguously by X-ray sin-
gle crystal analysis of the corresponding methyl ester.6b
Methylation of 6 with CH I/CuI gave a solid (mp ca.
3
8
4
1 °C; 64% isolated yield); structure 9 was obtained by
6
c
X-ray crystallography. As in the carboxylation, the
major product is formed with retention of configuration at
the Grignard carbon of 6.
Scheme 5
Deprotonation/magnesiation of cyclopropyl carboxam-
ides is essentially a stoichiometric reaction at room tem-
perature requiring just one equivalent of BuMgN(i-Pr)2.1
This is not so with the corresponding cyclobutane amides.
Scheme 4
About two equivalents of BuMgN(i-Pr) are needed for
2
9
good conversion, even at reflux. This is probably due to
To our knowledge, these are the first examples of ‘ortho-
metalation’ on a cyclobutane. We expect that amido-Grig-
nard 6 will have much the same flexibility and utility in
synthesis as its cyclopropane analog in reactions with
the diisopropyl amine liberated in the course of Scheme 5.
The freed amine reacts with BuMgN(i-Pr) , slowly at
2
room temperature (the cyclopropyl case), but rapidly
1
above 50 °C, giving butane and Mg[N(i-Pr) ] . As some
electrophiles.
2 2
BuMgN(i-Pr) is thus destroyed, there is need for an
2
To demonstrate the exceptional potential of the method
for preparing substituted cyclobutanes, we briefly investi-
gated introduction of another methyl group onto 9, now in
the cis b¢-position. It soon became clear that 9 is deproto-
nated/magnesiated more slowly than 3, presumably a re-
sult of statistics as well as steric hindrance. Nonetheless,
excess.
Deprotonation/magnesiation of cyclobutane carboxam-
ides with BuMgN(i-Pr) opens new synthetically useful
2
paths. As we shown with numerous examples in the cyclo-
1
propane series, carboxamido-stabilized amido-Grignards
enable stereospecific introduction of a variety of substitu-
ents. Permutations and combinations of these reactions
provide an entirely new method for the preparation of sub-
stituted cyclobutanes. The use of alkyl magnesium homo-
chiral amides should provide for selection between the
enantiotopic protons cis and b to the carboxamido group
as we have shown elsewhere for cyclopropanes.1
reaction of 9 with excess BuMgN(i-Pr) in THF/heptane
2
at reflux for 48 hours followed by methylation with CH3I/
CuI gave the trimethyl compound 10 in 36% isolated yield
along with 25% of recovered starting material
(
Equation 1). The mirror plane symmetry in the assigned
1
13
structure 10 follows from the H- and C NMR spectra
which show the b- and b¢-methyl groups to be magnetical-
ly equivalent.
0
Synlett 2003, No. 9, 1275–1278 ISSN 1234-567-89 © Thieme Stuttgart · New York

LETTER
New Approach to Substituted Cyclobutanes
1277
In theory, were base and solvent completely stable, and This result gives us hope that we can find a base and reac-
were there no destructive side reactions, any CH-group tion conditions for effective deprotonation/magnesiation
near an appropriately placed carboxamido group would be in still less favorable cases. If, as we suspect, unwanted
deprotonated/magnesiated completely by BuMgDA, the products like those in Scheme 5 arise from the reducing
conversion taken to completion by formation of butane. In properties (in the Meerwein–Ponndorff–Verley sense) of
reality, most compounds we have tried with CH kinetic Mg[N(i-Pr) ] , it should be better for deprotonation/mag-
2
2
acidities less than that of cyclobutane reacted only very nesiation to derive the base instead from a tertiary amine.
slowly with BuMgN(i-Pr) at <100 °C. The 1-N,N-diiso- This and the use of different solvents and higher tempera-
2
propylcarboxamides of 1-methylcyclopentane, 1-methyl- tures are under investigation.
cyclohexane, and adamantane gave little if anything
derived from the desired deprotonation/magnesiation. In-
stead, on prolonged treatment, a complex mixture of prod-
References
ucts was produced, many (and maybe all) from reactions
at the carboxamido substituent. Some N-dealkylated car-
boxamides were found, presumably the result of Hof-
mann-type eliminations. Mg[N(i-Pr) ] , probably formed
(1) (a) Eaton, P. E.; Zhang, M. X. Angew. Chem. Int. Ed.; 2002,
1, 2169. (b) See also: Eaton, P. E.; Lee, C. H.; Xiong, Y. J.
4
Am. Chem. Soc. 1989, 111, 8016. (c) Eaton, P. E.; Xiong,
Y.; Lee, C. H. J. Chin. Chem. Soc. 1991, 38, 303.
2
2
(
2) The activating/stabilizing carboxamido group need not be on
the b-carbon, see: Eaton, P. E.; Lukin, K. J. Am. Chem. Soc.
by disproportionation of BuMgN(i-Pr) , seems responsi-
2
ble for the other major products, for these were formed
1993, 115, 11370.
much more quickly when pure Mg[N(i-Pr) ] was used
(3) Streitwieser, A. Jr.; Caldwell, R. A.; Young, W. R. J. Am.
Chem. Soc. 1969, 91, 529.
2
2
instead of BuMgN(i-Pr) . For example, reaction of the
2
(
4) We assign the liganding interaction shown by analogy to that
seen in a cyclopropyl example for which we have a single
crystal X-ray structure: Eaton, P. E.; Zhang, M. X.; Steele,
I., unpublished results
adamantyl carboxamide 12 with four equivalents of
Mg[N(i-Pr)2]2 in refluxing THF/heptane (Equation 2)
gave, amongst other things, the known carbinol 13 and the
strange ene-amide 14, identified by single crystal X-ray
(
5) Typical Procedure. Dry diisopropylamine (202 mg, 2.0
mmol) was added dropwise to a stirred 1 M solution of
6
d
analysis. Spectroscopically similar ene-amides (oils)
were formed even from the more acidic compounds 3 and
Bu Mg in heptanes (Aldrich, 2.0 mL, 2.0 mmol) at 0 °C. The
mixture was stirred at r.t. for 1.5 h to give a pale yellow
2
1
5 (vide infra) on treatment with Mg[N(i-Pr) ] , but their
2 2
BuMgN(i-Pr) solution. Amide 3 (99 mg, 0.5 mmol) in THF
structures were not identified as conclusively.
2
(
1.5 mL) was added at r.t. to the base solution so made. The
resulting pale yellow solution was taken to reflux (73 °C,
bath at 85 °C) for 4 h. Afterwards, the solution was cooled to
r.t. and added to I (1 g, 4 mmol) in THF (5 mL) at 0 °C. The
2
mixture was stirred at r.t. for 1.5 h then quenched with aq
1
0% HCl (5 mL), extracted with CHCl (3 × 10 mL) and
3
washed with aq sat. (NH ) SO (10 mL). The extract was
4
2
4
washed with an aq 10% Na S O (10 mL), then again with
2
2
3
Equation 2
aq sat. (NH ) SO (2 × 10 mL), dried over Na SO and
4 2 4 2 4
concentrated at r.t. in vacuo. The residual brown liquid was
chromatographed (silica gel, CH Cl ) to give 7 (124 mg,
2
2
The kinetic acidity of a hydrogen on a methyl group is
somewhat greater than that of a methylene hydrogen and
thus lies between a cyclobutane and a cyclopentane.11
This enhancement, even with the statistical boost of 3:2,
proved insufficient for significant deprotonation/magne-
siation of a b-to-the-carboxamide hydrogen on the methyl
group of 1-methyl-1-(N,N-diisopropylcarboxamido)cy-
clopentane; standard conditions favored other reactions
1
77%) as a colorless solid: mp 119–120 °C. H NMR (400
MHz, CDCl ): d = 4.96 (dd, J = 10 and 9 Hz, 1 H), 3.73
3
(sept, J = 7 Hz, 1 H), 3.27 (sept, J = 7 Hz, 1 H), 2.48 (m, 1
H), 2.31–2.42 (m, 2 H), 2.02 (m, 1 H), 1.47 (s, 3 H), 1.39 (d,
J = 7 Hz, 3 H), 1.38 (d, J = 7 Hz, 3 H), 1.17 (d, J = 7 Hz, 3
1
3
H), 1.15 (d, J = 7 Hz, 3 H). C NMR (100 MHz): d = 174.1,
9.7, 47.9, 46.0, 32.6, 28.8, 28.1, 25.7, 20.7, 20.4., Anal.
Calcd for C H INO: C, 44.59; H, 6.86; N, 4.33. Found: C,
4
1
2
22
4
4.97; H, 6.79; N, 4.36. The epimer of 7 was eluted
(
vide supra). On the other hand, it appears that the statis-
1
somewhat later from the column: mp 109–110 °C. H NMR
tical boost provided by the 9 hydrogens in the pivaloyl
carboxamido 15 is sufficient for useful deprotonation/
magnesiation. Reaction of 15 with excess BuMgN(i-Pr)2
in THF/heptane at reflux for 8 hours, followed by carbox-
ylation and diazomethane esterification, gave a 56% iso-
lated yield of the succinic acid derivative 16 (Equation 3).
(400 MHz, CDCl ): d = 4.21 (dd, J = 6.0 and 5.6 Hz, 1 H),
3
3.45 (sept, J = 6.4 Hz, 1 H), 3.29 (sept, J = 6.8 Hz, 1 H), 3.17
(
dd, J = 20.0 and 11.2 Hz, 1 H), 2.85–2.75 (m, 1 H), 2.09 (m,
1
3
H), 1.64–1.57 (m, 1 H), 1.53 (s, 3 H), 1.40 (d, J = 6.8 Hz,
H), 1.38 (d, J = 6.8 Hz, 3 H), 1.35 (d, J = 6.4 Hz, 3 H), 1.18
1
3
(
d, J = 6.4 Hz, 3 H). C NMR (100 MHz): d = 173.2, 52.0,
8.3, 46.1, 30.8, 30.2, 29.2, 22.2, 21.6, 21.1, 20.5, 20.2. MS
ES): 324.0 [M + 1].
4
(
(
6) Crystallographic data have been deposited with the
Cambridge Crystallographic Data Centre, e-mail:
deposit@ccde.cam.ac.uk, as supplementary publications
numbered: (a) 206444; (b) 173571; (c) 209582 (racemate);
(
d) 205769.
Equation 3
Synlett 2003, No. 9, 1275–1278 ISSN 1234-567-89 © Thieme Stuttgart · New York

1
278
P. E. Eaton et al.
LETTER
(
7) Remarkably little is known about what happens at the
(9) Bu Mg as supplied commercially (Aldrich) has a mix of
2
+
magnesiated carbon during Grignard reactions with {E }.
See: Gawley, R. E. In Grignard Reagents. New
n-butyl and sec-butyl groups. Thus, two bases result from
reaction with HN(i-Pr) : n-BuMgN(i-Pr) and sec-
2
2
Developments; Richey, H. G. Jr., Ed.; Wiley: Chichester,
BuMgN(i-Pr) We have made each separately and have
.
2
000, 139.
found only small differences in their ability to deprotonate/
magnesiate 3.
(10) Zhang, M. X.; Komiya, N.; Eaton, P. E. 224th ACS National
Meeting, Boston; 2002, 341-ORGN.
(
8) We estimate that we would have seen in the NMR spectra of
the crudes any other isomer present in >10% yield.
(
11) Best value at present for neopentane:cyclohexane relative
kinetic acidity is 8.2:1 (per hydrogen) obtained using
CsCHA at 50 °C. Private Communication, Prof. A.
Streitwieser, March 4, 2003.
Synlett 2003, No. 9, 1275–1278 ISSN 1234-567-89 © Thieme Stuttgart · New York