An expedient synthesis of cis/trans-1,3-disubstituted cyclobutanols

Article abstract of DOI:10.1016/j.tetlet.2011.06.027

A two-step procedure is described to access 3-alkoxycyclobutanones from chloroacetyl chloride utilizing a step-wise [2+2] ketene cycloaddition followed by catalytic hydrogenation to reduce the α-chlorine in a single reaction sequence. The resulting cyclobutanones can be readily converted into a variety of cis or trans-1,3-disubstituted aminocyclobutanols and cyclobutanediols.

Full text of DOI:10.1016/j.tetlet.2011.06.027

Tetrahedron Letters 52 (2011) 4207–4210
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
An expedient synthesis of cis/trans-1,3-disubstituted cyclobutanols
Stacie Calad a, Douglas Mans a, , Justin-Alexander Morin b, Stacy O’Neill-Slawecki a, Joseph Sisko a
⇑
a GlaxoSmithKline, 709 Swedeland Rd, King of Prussia 19406, USA
b Ontario Institute for Cancer Research, MaRS Center, South Tower, 101 College St, Suite 800, Toronto, ON, Canada M5G 0A3
a r t i c l e i n f o
a b s t r a c t
Article history:
A two-step procedure is described to access 3-alkoxycyclobutanones from chloroacetyl chloride utilizing
Received 10 May 2011
Revised 2 June 2011
Accepted 6 June 2011
Available online 14 June 2011
a step-wise [2+2] ketene cycloaddition followed by catalytic hydrogenation to reduce the a-chlorine in a
single reaction sequence. The resulting cyclobutanones can be readily converted into a variety of cis or
trans-1,3-disubstituted aminocyclobutanols and cyclobutanediols.
Ó 2011 Elsevier Ltd. All rights reserved.
Keywords:
[2+2]
Ketene
Chloroacetyl chloride
Cyclobutanone
Hydrogenolysis
Cyclobutanol
1,3-Disubstituted
The cyclobutanone and corresponding cyclobutanol rings are
important constituents in many natural products, often possessing
novel and/or important bioactivities.1 As such, the synthesis of
these highly strained compounds has been a focus of the synthetic
community for many years. Numerous methods for the construc-
tion of the cyclobutanone framework have been developed over
the years2 including but not limited to: photolytic3 and thermo-
lytic4 [2+2] cycloaddition processes, ring contraction reactions,5
ring expansion of cyclopropanes,6 and radical processes.7
Figure 1.
sometimes conflicting reports in the literature as to the reactivity
of ketenophiles with chloroketene, we undertook a systematic
screen to better map out their reactivity profile (Scheme 1). Entries
a–d were the most appealing substrates as either the alcohol
(entries a and b) or amine (entries c and d) functionality could
be installed directly during the cycloaddition and the reagents
were cheap, easily handled materials. Unfortunately, the lower
reactivity of the vinyl ester compounds failed to give any reaction
at all (entries a, b),9 and the vinyl amides provided only the acyclic
acylated products 4c/d (entries c and d).10 Benzyl vinyl ether (entry
e) was also an attractive substrate due to the mild conditions
needed to reveal the alcohol functionality, but it provided poor iso-
lated yields and a messy reaction profile. The reaction of chloroke-
tene with the extremely reactive dimethyl ketene acetal provided
only the product of C-acylation (entry f). Finally, despite the antic-
ipated difficulty in the cleavage of alkyl ethers, cycloadditions with
commercially available alkyl vinyl ethers (entries g–i) were studied
and the use of i-butyl or t-butyl vinyl ether (i-BVE and t-BVE,
respectively) were found to be more successful than n-butyl vinyl
ether (entry g vs h and i). The greater selectivity and higher yields
During work on a drug development project we required a rapid
and cost effective means of accessing N-Boc-protected trans-3-
aminocyclobutanol 1a ( Fig. 1). Previous work relied on
a
cyclodialkylation of tosylmethyl isocyanide (TOSMIC) with epi-
chlorohydrin to assemble the cyclobutane framework. Unfortu-
nately several issues including overall length, intermediate
stability, and cost of goods made the route untenable.
A survey of the literature, taking into account overall route effi-
ciency/simplicity and material costs, led us to investigate the ther-
mal [2+2] cycloaddition with chloroketene.8 A [2+2] cycloaddition
of chloroketene and electron-rich olefins would assemble the
cyclobutanone scaffold 5 rapidly and efficiently, while also provid-
ing functional group handles for further manipulation (Scheme 1).
Our strategy was to convert the carbonyl of cyclobutanone 5a–i
into either the amine or alcohol of 1 and, therefore, the X-substitu-
ent could be either the alcohol or amine. Despite the disjointed and
⇑
Corresponding author. Tel.: +1 610 270 6298; fax: +1 610 270 4829.
E-mail address: douglas.m.mans@gsk.com (D. Mans).
0040-4039/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2011.06.027

4208
S. Calad et al. / Tetrahedron Letters 52 (2011) 4207–4210
Scheme 1. Ketenophile screen of the 2+2 cycloaddition. aTypical conditions: NMM (1.5 eq) added via syringe pump over 30 min to a solution of enol ether/enamide (1 eq) and
chloroacetyl chloride (1.05 eq) in TBME (7 vol) at 50 °c. bYield referes to the desired cyclobutanone. cReaction performed under ‘‘typical condition’’ as well as inverse addition
in which a solution of chloroacetyl chloride in TBME was added to a solution of enamide and NMM giving identical results. dReaction performed under inverse addition of
chloroacetyl chloride to a solution of enol ether and NMM. Under ‘‘typical conditions’’ acylation of the enol ether by the acid chloride was observed.
(of 5 vs 4) observed with increased steric bulk on the vinyl ethers
may be rationalized by the greater cation stability (via an enhance-
ment in oxonium stability through hyperconjugation) in going
from a 1° to 3° alkyl ether.11 In general, the crude cycloadducts (
5g–i), formed as predominantly the trans isomer,11,12 and could
be carried on without further purification.13 However, in order
for the [2+2] strategy to remain a viable route, two challenges re-
mained: an efficient means of chlorine removal and the ability to
cleave the alkyl ether selectively.
Removal of the C2-chlorine of cylcobutanone 5g–i became our
first of the challenges. Dechlorination of 2-chlorocyclobutanones
possessing a C3-leaving group with Zn°/AcOH has been reported,
however, mediocre yields are typically obtained.14 The poor yields
most likely arise from the competing Boord reaction,15 resulting in
formation of the volatile and unstable cyclobutenone 11. Addition-
ally, the use of stoichiometric Zn° made this process less desirable
on scale. In our hands, treatment of 5h with Zn°/AcOH provided
only a ꢀ50% yield of desired cyclobutanone 7h with large amounts
of i-butanol observed by GC (Scheme 2). Instead, attention was
turned to catalytic hydrogenolysis of 5h.16 A preliminary solvent
and base screen was performed and the optimal conditions
(EtOAc/H2O (2:1, 20 vol), 3 equiv pyridine, 10 wt % Pd/C at 40 psi
H2) provided an 80% yield of cyclobutanone 7h on 30 g scale
(Scheme 2). Pyridine and its derivatives provided the best yields
and conversion rates, while inorganic bases, triethylamine, and
other basic 3° alkyl amines provided inferior results. Other solvents
gave poor to moderate conversions. The need for water in the mix-
ture is unclear at this point.
With conditions for effecting the removal of the chlorine in
hand, we focused next on the alkyl ether cleavage. A variety of clas-
sical ether cleavage conditions17 were examined, but a complete
lack of regioselectivity (cyclobutane C–O bond vs i-butyl C–O bond
cleavage) plagued both compounds 5h and 7h. The desired i-butyl
ether C–O cleavage was eventually realized by treating N,N-
Scheme 2. Catalytic hydrogenolysis and acid promoted ether cleavage.

S. Calad et al. / Tetrahedron Letters 52 (2011) 4207–4210
4209
Scheme 3. Synthesis of cyclobutylamine 9 from t-BVE.
oped to deliver a suitably protected cis-3-aminocyclobutanol in
high yields. The intermediates could be readily transformed into
the desired trans-1a through inversion with KOAc/DMF/130 °C,
debenzylation, and Boc protection. In addition, cyclobutanone 7i
serves as an excellent starting point for elaboration into a variety
of 1,3-disubstituted cyclobutanol structures which find use
throughout natural product synthesis and pharmaceutical devel-
opment (Fig. 2).
In summary, 3-t-butoxycyclobutanone 7i can be obtained in
high yields and purity from t-butyl vinyl ether and chloroacetyl
chloride in a one-pot, two step sequence which is amenable to fur-
ther manipulation to access the trans-1a isomer as well. The cyclo-
butanone core with all the necessary heteratoms present for
further elaboration into cis/trans-3-aminocyclobutanols as well as
a variety of other 1,3-disubstituted cyclobutanes is readily avail-
able on gram to kilogram scale from inexpensive starting materials.
Figure 2. Functional group elaboration of cyclobutanone 7i.
Supplementary data
dibenzylamine 8h, formed by reductive amination of 7h with
dibenzylamine, with strongly Lewis acidic reagents to give amino
cyclobutanol 9 with no decrease in the cis/trans ratio (Scheme 2).
Despite the success utilizing i-BVE, the cleaner and higher yield-
ing reaction profile for the cycloaddition of t-BVE (Scheme 1) and the
more easily cleaved t-butyl group led us to pursue its use instead.
Although the original patent literature8b claimed a need for slow
addition of NMM at 50 °C over 2 h, followed by a 4 h hold, in situ
monitoring with ReactIR showed that the [2+2] cycloaddition be-
tween 2i and 3 was reagent addition controlled and upon complete
addition of the NMM the reaction was complete.18 Ultimately, it was
found that the addition time of base could be shortened to 0.5 h
without detriment to yield. Additionally, the patent claimed8b that
the amine base must have a pKa = 6–8 (NMM has a pKa = 7.4), but a
base screen was undertaken that ultimately showed i-Pr2NEt
(pKa = 10.5) gave 5i in yields as good if not better than NMM (typi-
cally 90–93% crude yield with 95% purity) (Scheme 3).19 Preliminary
scoping studies using Design of Experiment also found that elevated
temperatures are not required for high yields. In fact, the reaction
could be run at 0 °C with no decrease in yield.20 Following the cyclo-
addition, 5i was subjected to the hydrogenation conditions devel-
oped for cyclobutanone 5h and provided the des-chloro 7i in 80–
85% overall yield from t-BVE (Scheme 3). Conversion of 7i into the
N,N-dibenzylcyclobutylamine 8i via reductive amination followed
by treatment with 6 N aq HCl efficiently led to the isolation of the
3-N,N-dibenzylaminocyclobutanol 9 in excellent yields (Scheme 3).
At this point, a four-step sequence in which all of the route
could be telescoped and/or utilize crude material had been devel-
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.tetlet.2011.06.027.
References and notes
1. (a) Thomas, A. F.; Bassiere, Y. In The Total Synthesis of Natural Products;
ApSimon, J. W., Ed.; Wiley: New York, 1988; Vol. 7, pp 275–454; (b) Devon, T.
K.; Scott, A. I. In Handbook of Naturally Occurring Compounds Terpenes;
Academic Press: New York, 1972; Vol. 2,
2. (a) Salaün, J. Sci. Synth. 2005, 26, 557–606; (b) Trost, B. M. In Strain and
Reactivity: Partner for Selective Synthesis in Small Ring Compounds in Organic
Synthesis I; de Meijere, A., Ed.; Springer: Berlin, 1986; pp 3–82.
3. For examples see: (a) Eaton, P. E. J. Am. Chem. Soc. 1962, 84, 2454–2455; (b)
Corey, E. J.; Bass, J. D.; LeMahieu, R.; Mitra, R. B. J. Am. Chem. Soc. 1964, 86,
5570–5583; (c) Corey, E. J.; Mitra, R. B.; Uda, H. J. Am. Chem. Soc. 1964, 86, 485–
492; (d) Alibés, R.; Bourdelande, J. L.; Font, J.; Gregori, A.; Parella, T. Tetrahedron
1996, 52, 1267–1278; (e) Wagner, P. J.; McMahon, K. J. Am. Chem. Soc. 1994,
116, 10827–10828; (f) Tanaka, K.; Toda, F. J. Chem. Soc., Perkin Trans. I 1992,
943–944.
4. For examples with ketene, see: (a) Danheiser, R. L.; Casebier, D. S.; Loebach, J. L.
Tetrahedron Lett. 1992, 33, 1149–1152; (b) Pittol, C. A.; Roberts, S. M.; Sutton, P.
W.; Williams, J. O. J. Chem. Soc., Chem. Commun. 1994, 803–804; (c) Ganz, I.;
Kunz, H. Synthesis 1994, 1353–1358; (d) de Azevedo, M. B. M.; Greene, A. E. J.
Org. Chem. 1995, 60, 4940–4942; For cycloadditions promoted with Lewis
acids, see: (e) Ahmad, S. Tetrahedron Lett. 1991, 32, 6997–7000; (f) Narasaka, K.;
Hayashi, Y.; Shimadzu, H.; Niihata, S. J. Am. Chem. Soc. 1992, 114, 8869–8885;
(g) Narasaka, K.; Hayashi, K.; Hayashi, Y. Tetrahedron 1994, 50, 4529–4542; (h)
Seebach, D.; Dahinden, R.; Marti, R. E.; Beck, A. K.; Plattner, D. A.; Kühnle, F. N.
M. J. Org. Chem. 1995, 60, 1788–1799; For examples with keteniminium ions,
see: (i) O’Brien, J. M.; Kingsbury, J. S. J. Org. Chem. 2011, 76, 1662–1672; (j)
Marchand-Brynaert, J.; Ghosez, L. J. Am. Chem. Soc. 1972, 94, 2870–2872; For a
review on keteniminium 2+2 reactions see: (k) Snider, B. B. Chem. Rev. 1988, 88,
793–811.

4210
S. Calad et al. / Tetrahedron Letters 52 (2011) 4207–4210
5. (a) Malanga, C.; Menicagli, R.; Pecunioso, A.; Lardicci, L. Gazz. Chim. Ital. 1991,
121, 17–20; (b) Ito, H.; Taguchi, T.; Hanzawa, Y. Tetrahedron Lett. 1993, 34,
7639–7640.
A.; Lachowicz, J. E.; Morgan, C. A.; Varty, G. B.; Shih, N.-Y. Bioorg. Med. Chem.
Lett. 2006, 16, 3859–3863; (c) Trost, B. M.; Xie, J. J. Am. Chem. Soc. 2006, 128,
6044–6045; (d) Verniest, G.; Colpaert, J.; Toernroos, K. W.; De Kimpe, N. J. Org.
Chem. 2005, 70, 4549–4552; (e) Neudeck, H.; Brinker, U. H. Tetrahedron Lett.
2005, 46, 1893–1895; For examples of dechlorination with cyclobutanones
possessing a C3 leaving group providing high yields see: (f) Darses, B.; Greene,
A. E.; Coote, S. C.; Poisson, J.-F. Org. Lett. 2008, 10, 821–824; (g) Ferguson, A. C.;
Adlington, R. M.; Martyres, D. H.; Rutledge, P. J.; Cowley, A.; Baldwin, J. E.
Tetrahedron 2003, 59, 8233–8243; (h) Ambrosio, J. C. L.; Santos, R. H. De A.;
Correia, C. R. D. J. Braz. Chem. Soc. 2003, 14, 27–38; For examples of low yielding
dechlorination of cyclobutanones possessing a C3 leaving group see: (i) Wang,
C.; Zhen, Z.; Zhao, J.; Dowd, P. Synth. Commun. 1999, 29, 631–643; (j) Dehmlow,
E. V.; Bueker, S. Chem. Ber. 1993, 126, 2759–2763; (k) Redlich, H.; Lenfers, J. B.;
Kopf, J. Angew. Chem. 1989, 101, 764–765.
6. Nemoto, H.; Fukumoto, K. Synlett 1997, 863–875.
7. (a) Paquette, L. A.; Sugimura, T. J. Am. Chem. Soc. 1986, 108, 3841–3842; (b)
Kocovsky, P.; Srogl, J. J. Org. Chem. 1992, 57, 4565–4567; (c) Weignes, K.;
Schmidbauer, S. B.; Schick, H. Chem. Ber. 1994, 127, 1305–1309; (d) Kocozsky,
P.; Srogl, J.; Pour, M.; Gogoll, A. J. Am. Chem. Soc. 1994, 116, 186–197.
8. (a) Tidwell, T. T. Eur. J. Org. Chem. 2006, 563–576. and references cited therein;
(b) Matsushita, S.; Shimizu, I.; Toyoda, H.; Tanaka, A. US Patent No. US6331653
B1, 2001.
9. The successful use of vinyl acetate in the [2+2] cycloaddition with
trichloroacetyl chloride and Zn°/Cu°-couple has previously been
demonstrated, see: Danheiser, R. L.; Savariar, S.; Cha, D. Org. Synth. 1993, Coll.
Vol. 8, 82–87. however the large amounts of Zn° required for the cycloaddition
and concomitant dechlorination reactions in addition to very low yields for the
dechlorination made this route undesirable.
10. The b-keto N-acyliminium intermediates formed in entries c and d rapidly
deprotonate to give acylated products 4c/d. These results coupled with the
results from entries e–i, Table 1 indicate that a step-wise mechanism is most
likely operative. Separate experiments in which isolated 5 was resubjected to
the reaction conditions failed to provide any of the acyclic acylated material 4.
11. Sieja, J. B. J. Am. Chem. Soc. 1971, 93, 130–136.
15. Swallen, L. C.; Boord, C. E. J. Am. Chem. Soc. 1930, 52, 651–660.
16. See for example: Takuma, Y.; Kasuga, Y.; Watanabe, Naoyuki; Lin, Shu-Su
(Mitsubishi Chemical Corp., Japan) Jpn. Patent JP 2002249454 A, 2002.
17. (a) Bhar, S.; Ranu, B. C. Org. Prep. Proced. Int. 1996, 28, 371–409; (b) Harada, T.;
Kita, K.; Oku, A. Tetrahedron Lett. 1982, 23, 681–684; (c) Ganem, B.; Small, V. R.,
Jr. J. Org. Chem. 1974, 39, 3728–3730; (d) Bhar, S.; Ranu, B. C. J. Org. Chem. 1995,
60, 745–747; (e) Gupta, M. K.; Reddy, B. V. S.; Reddy, K.; Yadav, J. S. Tetrahedron
Lett. 2005, 46, 8493–8495.
18. Results were general for any reaction utilizing chloroacetyl chloride, an amine
base, and an alkyl vinyl ether.
12. Stirring the reaction mixture with a slight excess of base leads to complete
epimerization to the trans isomer in <1 h; storing the cycloadducts neat leads
to complete epimerization over time (ꢀ1 week).
19. Yields based on solution assays. Reaction profile by GC and 1H NMR showed
slightly higher impurity profiles compared to reactions using NMM as the base.
Other bases screened include Et3N, N,N,N0,N0-tetramethyl-1,3-propanediamine,
pyridine, Et3N/pyridine, DABCO, (ꢁ)-Sparteine, and DBU.
13. Acyclic enone 4e–g was purged following chlorine removal (vida infra) by
washing with 3 N HCl.
14. For examples of 2,2-dichlorocyclobutanones with no C3 leaving group see: (a)
Baldwin, J.; Leber, P. A.; Powers, D. C. J. Am. Chem. Soc. 2006, 128, 10020–10021;
(b) Wrobleski, M. L.; Reichard, G. A.; Paliwal, S.; Shah, S.; Tsui, H.-D.; Duffy, R.
20. Yields were found to be comparable, but the cis/trans ratio favored the kinetic
product by ꢀ60:40. See Ref. 4.