Unprecedented dual reactivity of anhydrous potassium hydroxide in cascade cyclopropannelation/Haller-Bauer-scission/Grob-fragmentation reactions

Article abstract of DOI:10.1016/j.tetlet.2010.06.007

We report an unprecedented type of reactivity of 'anhydrous potassium hydroxide' ('APH') in which it plays, over a large variety of related educts, sequentially the role of base and nucleophile. Some insight into the structure of reactive species as well as comparative reactivity of related reagents prepared by fusion of commercially available potassium hydroxide or by adding stoichiometric amount of water to potassium hydride is provided.

Full text of DOI:10.1016/j.tetlet.2010.06.007

Tetrahedron Letters 51 (2010) 4306–4309
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Unprecedented dual reactivity of anhydrous potassium hydroxide in cascade
cyclopropannelation/Haller–Bauer-scission/Grob-fragmentation reactions
*
Alain Krief , Adrian Kremer
Department of Chemistry, Facultés Universitaires N.-D. de la Paix, 61 Rue de Bruxelles, Namur, B-5000, Belgium
a r t i c l e i n f o
a b s t r a c t
Article history:
We report an unprecedented type of reactivity of ‘anhydrous potassium hydroxide’ (‘APH’) in which it
plays, over a large variety of related educts, sequentially the role of base and nucleophile. Some insight
into the structure of reactive species as well as comparative reactivity of related reagents prepared by
fusion of commercially available potassium hydroxide or by adding stoichiometric amount of water to
potassium hydride is provided.
Received 19 April 2010
Revised 27 May 2010
Accepted 1 June 2010
Available online 15 June 2010
Ó 2010 Elsevier Ltd. All rights reserved.
Dedicated to Professor Carmen Najera on
the occasion of her 60th birthday with great
appreciation
Keywords:
Fragmentation
Cyclopropannelation
Hydroxide ion
Wet potassium hydroxide
Anhydrous potassium hydroxide
Vinyl cyclopropane carboxylic acid
Cyclopentanones fused to a cyclopropane in a0,b0-positions and
bearing a gem-dimethyl group in -position and an exo-halogen in
We rationalized1 those results by assuming that potassium
hydroxide acts primarily with 30 as a base (Scheme 2, entry a) or
as a nucleophile (Scheme 2, entry b) to produce 1 or 4, respectively.
We did not however explain why the absence of one or two
methyl groups at C-5 affects the behavior of potassium hydroxide
toward 30. We did not also provide a general solution to transform
all the members of 30 to 1. This is in fact the purpose of this Letter.
In a subsequent work, we found that not only the number of
substituents at C-5 but also the stereochemistry there affect the
nature of the products formed on reaction with potassium hydrox-
ide (Condition A). Thus, whereas 3c0Br produces lactones (Table 1,
entry 4), its epimer at C-5 3bB0 r leads under similar conditions to
the same mixture of the lactones 4c/4b (87/13) and a single stereo-
isomer of 1b in which the latter prevails (Scheme 1, Table 1, Con-
dition A, entry 3 compare to entry 4).
a
b-position 2 (X = Cl or Br) proved to be valuable precursors of vinyl
cyclopropane carboxylic acids 1, especially cis-chrysanthemic acid
1a, on reaction with potassium hydroxide (KOH, aq DMSO, 70 °C,
4 h, Condition A;1,2 t-BuOK and H2O (3–1), 20 °C, <1 h, THF, Condi-
tion B;1b,2 DMSO, Condition C;1b,2 Scheme 1, Table 1). Conditions B
and C which involve ‘anhydrous potassium hydroxide’ proved to be
the most efficient owing to the particularly high nucleophilicity of
this reagent.3
Bicyclo [3.1.0] cyclopentanones 2 have been in turn prepared
from 3,4-dihalogenocyclohexanones 30 (X, Y = Cl, Br)1 or the related
3-halogeno-4-mesyloxy cyclohexanones 300 (X = halogen, Y = mesyl-
oxy)4 and a base such as LDA1,5 or potassium hydroxide (Condition
A).1 The latter conditions proved to be successful with 3a0 and
directly lead to cis-chrysanthemic acid 1a after acid work-up.1 It
involves an exceptional tandem cyclopropannelation/Haller–Bau-
er-scission/Grob-fragmentation reaction which unfortunately
proved to be limited to 3a0 since the related 3c0 and 3d0 instead lead
to the vinyl lactones 4c/4b (87/13 mixture of stereoisomers) and
4d (Scheme 1, Table 1, Condition A).1
We were rather surprised to find that the related 3-bromo-4-
4
00
mesyloxy cyclohexanone 3a
does not behave as the related
Br
1
dibromide 3a0Br which also bears the two geminal methyl groups
at C-5. It mainly leads to a mixture of chrysanthemic acid 1a and
its isomeric chrysanthemolactone 4a in which the latter prevails
(4a/1a: 2/1, Scheme 1, Table 1 entry 7, Condition A). We also ob-
served that the percentage of lactone increases substantially by
carrying out the reaction at 20 °C instead of 70 °C (4a/1a: 9/1,
Scheme 1, Table 1 entry 7, Condition A). Furthermore, whereas 3-
* Corresponding author. Tel.: +32 474 318181; fax: +32 81 724 536.
E-mail address: alain.krief@fundp.ac.be (A. Krief).
4
iodo-4-mesyloxy cyclohexanone 3a0I0 nearly exclusively generates
0040-4039/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2010.06.007

A. Krief, A. Kremer / Tetrahedron Letters 51 (2010) 4306–4309
4307
Scheme 1.
Scheme 2.
chrysanthemolactone 4a (Scheme 1, Table 1 entry 8 Condition A),
(i) Metalation of 3 is best achieved on the more acidic C-6 Ha. It
should provide, under kinetic control, the corresponding
enolate Ba by constant overlap of the orbitals involved
(Scheme 3, entry a).6 Metalation of He, however, the less
acidic of the two hydrogens attached to C-6, achieves the
antiperiplanar arrangement of atoms and bonds suitable
for the cyclization leading to cyclopropane ring (Scheme 3,
entry b) and therefore successful cyclization is expected to
involve an equilibrium.7
4
00
its chloro analogue 3a exclusively produces chrysanthemic acid
Cl
1a (Scheme 1, Table 1 entry 6, Condition A, compare to entries 7
and 8). In the latter case, the bicylic [3.1.0] b-chloroketone1b inter-
mediate 2aCl has been trapped if the reaction is carried out at 20 °C
instead of 70 °C (20 °C, 1 h, 70%).
We envisage that the nature of the products 1 or/and 4 formed
(Scheme 2) could be related to the subtile stereoelectronic effects
as well as to the nature of the leaving groups at C-3 and C-4 on
3. We have ascertained using X-ray crystallography that 3a0Br and
3c0Br possess a chair conformation with the halogens and mesyloxy
group lying in equatorial position as disclosed in 0Figure 1.
(ii) Addition of the hydroxide ion onto the carbonyl carbon8
from the top-face generates the intermediate Ea whose alk-
oxide ion is adequately positioned to favor a fragmentation
reaction in which the atoms and bonds are localized in an
antiperiplanar conformation9,10 leading to the open chain
intermediate F precursor of the vinyl lactones 4 (Scheme 4,
00
Similar results have also been observed for 3dBr, 3a and 3a0I0.
Br
The attack of hydroxyl anion can take place on the C-6 hydro-
gens or on the carbonyl carbon of 3. We assume that:
Table 1
Reactivity of 2,2-dimethylcyclohexanones 3 toward KOH and t-BuOK–H2O
Entry
N
R1
R2
X
Y
Condition A: KOH, aq DMSO 1/4 yield % Condition B: t-BuOK–H2O, THF 1/4 (yield Condition C: t-BuOK–H2O, DMSO 1/4
(1/4 ratio) [T (°C), t (h)]a
%) (1/4 ratio) [T (°C), t (h)]a
(yield %) (1/4 ratio) [T (°C), t (h)]a
1
2
3
4
5
6
7
3a0Cl
Me Me Cl Cl
64/0 (100/0) [70, 4]1
86/0 (100/0) [70, 2]1
65/7 (76/24) [70, 1]
0/81 (0/100) [70, 2]1
0/65 (0/100) [70, 2]1
80/0 (100/0) [20, 1]
94/0 (100/0) [20, 0.4]
94/0 (100/0) [20, 0.4]
92/0 (100/0) [20, 0.4]
86/0 (100/0) [20, 0.4]
72/0 (100/0) [20, 0.6]c
89/0 (100/0) [20, 0.6]
3a0Br Me Me Br Br
65/0 (100/0) [20, 0.5]
3b0Br
H
Me Br Br
3c0Br Me
H
H
Br Br
Br Br
3d0Br
H
Me Me Cl Ms 57/0 (100/0) [70, 6]b
Me Me Br Ms 26/53 (33/67) [70, 1]
7/61 (10/90) [20, 8]
75/0 (100/0) [20, 2]
76d/5 (94/6) [20, 2]
00
3a
Cl
00
3a
Br
8
3a0I0
Me Me
I
Ms 7/66 (10/90) [70, 1]
4/67 (5/95) [20, 3]
89/4 (96/4) [20, 0.6]
36e/5 (88/12) [20, 2]
a
b
c
The yields disclosed refer to that of isolated, analytically pure compounds as single stereoisomer for 1 and as a mixture of stereoisomers for 4 unless otherwise stated.
6% of b-chloroketone 2aCl also recovered.
18% of b-chloroketone 2aCl also recovered.
Chrysanthemic acid 1a proved to be a cis/trans-mixture of isomers 83/17.
Chrysanthemic acid 1a proved to be a cis/trans-mixture of isomers 30/70.
d
e

4308
A. Krief, A. Kremer / Tetrahedron Letters 51 (2010) 4306–4309
entry a). This approach could be hampered with a bulky R2
group (1,3-diaxal interaction). Addition from the bottom-
face however provides instead Ee whose fragmentation is
much less favored (Scheme 4, entry b).
However, the electronic factors are not the only ones which
could play an important role in the process. It was found that the
relative leaving group ability of X and Y on 3 plays also a crucial
role since the better leaving group favors the process in which it
is involved favoring thus the formation of either 1 or 4.
Figure 1.
Scheme 3.
Scheme 4.
Scheme 5.

A. Krief, A. Kremer / Tetrahedron Letters 51 (2010) 4306–4309
4309
Thus vinyl cyclopropane carboxylic acids 1 are obtained from
compounds 3aC0 l, 3aB0 r, 3bB0 r, 3aC00l which possess an axial methyl
group (R2 = Me) at C-5 and a leaving group at C-4 with similar or
better leaving group aptitude than at C-3 (see Scheme 4, entry a;
Table 1, entries 1–3 and 6; Condition A). The higher percentage
of cyclopropane derivative 1 from compounds 3a0Cl, 3aB0 r, 3aC00l bear-
ing a gem-dimethyl groups at C-5 over 3b0Br missing it may be ratio-
nalized by involving the Thorpe-Ingold effect (Table 1 compare
entry 3 to entries 1 and 2; Condition A).11
(0.5 h). Using dried KOH, the bicyclo [3.1.0] beta-bromo cyclo-
pentanone 2aBr (70%) and the vinyl lactone 4a (30%) are formed
whereas using KOH generated from potassium hydride, the bicyclo
[3.1.0] beta-bromo cyclopentanone 2aBr (19%) and cis-chrysanthe-
mic acid 1a (81%) are instead produced.
Dried KOH in THF behaves as KOH in aq DMSO (Condition C).
KOH from KH behaves as ‘AHP’ in chemoselectivity but not in reac-
tivity. ‘AHP’ therefore possesses an exceptional reactivity which
differentiates it from the two ‘anhydrous potassium hydroxide’
and wet potassium hydroxide reagents tested in this study.
The results reported are far beyond the scope of the multistep
one-pot transformation we have finally successfully achieved using
‘APH’. They suggest that potassium t-butoxide and t-butanol which
are also present in ‘APH’ play a crucial role in its reactivity. They
also extend the scope of action of this exceptional reagent far be-
yond what has been already published.3e
It is interesting to note that the vinyl lactones 4 are mainly pro-
duced from (i) 3c0Br and 3dB0 r which miss the axial methyl group
even if the leaving groups at C-3 and C-4 are the same (Table 1, en-
00
tries 4 and 5; Condition A) or (ii) 3a or 3a0I0 which possess an axial
Br
methyl group at C-5 but bear a better leaving group at C-3 than at
C-4 (Table 1, entries 7 and 8; Condition A).
In order to promote the one-pot synthesis of vinyl cyclopropane
carboxylic acids 1 from the cyclohexanones 3 that we were unable
to achieve using potassium hydroxide under condition A, we
decided to use ‘anhydrous potassium hydroxide’ (‘APH’) instead.2,3
‘APH’ is a 2/1 mixture of potassium t-butoxide and potassium
hydroxide and we expected that the former reagent would act as
a powerful base12 to generate the cyclopropane ring whereas the
naked hydroxide anion would achieve, in a second step, the frag-
mentation reaction producing after acidic treatment the vinyl
cyclopropane carboxylic acid 1. This was quite risky since, except
in one case,13 ‘APH’ has been used as a powerful nucleophile.2,3 It
has been successfully used to transform, in high yields and under
mild conditions (ether, 20 °C, few hours), even sterically hindered
esters,3 tertiary amides,3b and non-enolizable ketones1b,2,3 to the
corresponding potassium carboxylates and potassium alcoholates,
-amides or -carbanions, respectively (Scheme 1), without epimer-
General procedure for the ‘one-pot’ cyclization–fragmentation
reactions
With ‘AHP’ in THF: water (21 mg, 1.15 mmol) is added at 20 °C to
a solution of freshly sublimed potassium tert-butoxide (426 mg,
3.8 mmol) in dry THF (4 mL) and the reaction mixture is stirred
for 0.25 h at that temperature. A solution of 3-bromo-4-mesyl-
00
oxy-2,2,5,5-tetramethyl-cyclohexanone 2a (164 mg, 0.5 mmol)
Br
in dry THF (2 mL) is then added dropwise at 20 °C. The reaction
is monitored by TLC (pentane/diethyl ether 80:20) and the reaction
is quenched by icy water (5 mL) after 0.6 h. Aqueous HCl (10%) is
then added (until pH 2) and the solution is extracted with ether
(4 ꢀ 10 mL). The combined organic extracts are washed with water
(2 ꢀ 3 mL), dried over MgSO4, filtered and evaporated under re-
duced pressure. The crude product was purified by column chro-
matography using pentane/diethyl ether (60:40) as eluent to
furnish 75 mg (89%) of 1a.
izing their a-carbon.
The reaction of ‘APH’ on 3 is best achieved in THF using a reac-
tant/reagent ratio (3/t-BuOK/H2O: 1/7.6/2.3) in which it delivers in
very high yield the cis-vinyl cyclopropane carboxylic acids 1 at
20 °C (Table 1, Condition B). The reaction proved to be stereospe-
cific delivering a different stereoisomer of desmethyl cis-chrysan-
themic acids 1b and 1c from each stereoisomers of 3,4-dibromo-
2,2,5-trimethylcyclohexanones 3b0Br and 3cB0 r used (Table 1, entries
3 and 4; Condition B).
References and notes
1. (a) Krief, A.; Lorvelec, G.; Jeanmart, S. Tetrahedron. Lett. 2000, 41, 3871–3874;
(b) Krief, A.; Kremer, A. Tetrahedron Lett. 2010, 51, 3045–3049.
2. Krief, A.; Kremer, A. Synlett 2007, 607–610.
3. (a) Swan, G. A. J. Chem. Soc. 1948, 1408–1412; (b) Gassman, P. G.; Zalar, F. V.
Tetrahedron Lett. 1964, 5, 3031–3034; (c) Gassman, P. G.; Zalar, F. V.; Lumb, J. T.
J. Am. Chem. Soc. 1967, 89, 946–952; (d) Mehta, G.; Venkateswaran, R. V.
Tetrahedron 2000, 56, 1399–1422; (e) Krief, A.; Kremer, A. Chem. Rev., in press.
doi:10.1021/cr9003506.
4. (a) Prepared by (i) non-regioselective ring opening of 3,4-oxido-2,2,5,5-
tetramethyl cyclohexanone with titanium tetrachloride, tetrabromide or
trimethylsilyliodide, (ii) acid hydrolysis, (iii) mesylation and (iv) separation
of the diastereoisomeric mixture of b-halogenomesylates.
5. (a) Krief, A.; Jeanmart, S.; Kremer, A. Heterocycles 2008, 76, 1075–1079;
(b) Krief, A.; Jeanmart, S.; Kremer, A. J. Org. Chem. 2008, 73, 9795–
9797.
6. (a) Corey, E. J.; Sneen, R. A. J. Am. Chem. Soc. 1956, 78, 6269–6278; (b) Behnam,
S. M.; Behnam, S. E.; Ando, K.; Green, N. S.; Houk, K. N. J. Org. Chem. 2000, 65,
8970–8978.
7. (a) Krief, A.; Hobe, M. Tetrahedron Lett. 1992, 33, 6529–6532; (b) Krief, A.; Hobe,
M. Tetrahedron Lett. 1992, 33, 6527–6528.
8. Bürgi, H. B.; Dunitz, J. D.; Lehn, J.-M.; Wipff, G. Tetrahedron 1974, 30, 1563–
1572.
9. (a) Wharton, P. S.; Hiegel, G. A. J. Org. Chem. 1965, 30, 3254–3257; (b) Zurflüh,
R.; Wall, E. N.; Sidall, J. B.; Edwards, J. A. J. Am. Chem. Soc. 1968, 90, 6224–6225;
(c) Heathcock, C. H.; Badger, R. A. J. Org. Chem. 1972, 37, 234–238.
10. (a) Grob, C. A.; Schiess, P. W. Angew. Chem., Int. Ed. Engl. 1967, 6, 1–15; (b) Grob,
C. A. Angew. Chem., Int. Ed. Engl. 1969, 8, 535–546.
Performing the reaction in DMSO does not offer advantages, it is
often slower and produces diastereoisomeric mixtures of cis- and
00
trans-chrysanthemic acid 1a from 3-bromo- 3a and 3-iodo- 3aI00
Br
4-mesyloxy-2,2,5,5-tetrametylcyclohexanones (Table 1, Condition
00
C, cis-1a/trans-1a: 83/17 from 3a and 30/70 from 3a0I0”, compare
Br
to Conditions A and B).
We have been unable to find the origin of the trans-chrysanthe-
mic acid 1a formed in this process. We have however secured, by
independent reactions, that ‘AHP’ in DMSO (Condition C) is unable
to (i) epimerize potassium cis- to trans-chrysanthemate or (ii) to
transform the lactone 4a to 1a. Another hypothesis which involves
a completely different mechanism is presented in Scheme 5. It im-
plies the metalation of the intermediate Fa00 (resulting from the at-
tack of ‘APH’ on the carbonyl group of 3a00) followed by cyclization
through Ga00.
In order to have a better insight in the real species required for
successful synthesis of chrysanthemic acid 1 we have carried out
00
the reaction of 3-bromo-4-mesyloxy-cyclohexanone 3a with
Br
11. Moore, B. R.; Ingold, C. K.; Thorpe, J. F. J. Chem. Soc., Trans. 1915, 107, 1080–
1106.
potassium hydroxide generated by (i) dehydration, on heating, of
powdered commercial potassium hydroxide or (ii) on reacting stoi-
chiometric amounts of water on potassium hydride (1 equiv KH,
1 equiv H2O, THF, 20 °C).3,14 Those reagents miss potassium t-
butoxide as well as t-butanol.
12. As LDA did.1,5 In those cases however competition with an hydroxide ion
playing the role of nucleophile was missing. t-BuOK would act as a strong base
favoring the attack of Ha (Scheme 3) and will not act as a nucleophile toward
the carbonyl group of 3.
13. Markó, I. E.; Vanherck, J.-C.; Ates, A.; Tinant, B.; Declercq, J.-P. Tetrahedron Lett.
2003, 44, 3333–3336.
14. Erman, W. F.; Wenkert, E.; Jeffs, P. W. J. Org. Chem. 1969, 34, 2196–2203.
We found that both reactions carried out in THF at 20 °C are
slower (18 h and 66 h, respectively) than those involving ‘APH’