Synthesis and Cyclization Reaction of cis,cis-1,3,5-Triformyl-1,3,5-trimethylcyclohexane

Full text of DOI:10.1021/jo962145n

J . Org. Chem. 1997, 62, 1173-1175
1173
Sch em e 1
Syn th esis a n d Cycliza tion Rea ction of
cis,cis-1,3,5-Tr ifor m yl-1,3,5-tr im eth yl-
cycloh exa n e
Hiroshi Izumi,* Osamu Setokuchi, and Yukio Shimizu
National Institute for Resources and Environment,
16-3 Onogawa, Tsukuba, Ibaraki 305, J apan
Hiromi Tobita and Hiroshi Ogino
Department of Chemistry, Graduate School of Science,
Tohoku University, Sendai 980-77, J apan
Received November 18, 1996
The hydrocarbon wurtzitane (fused boat cyclohexane)
has been known since 1974.1 Nielsen and co-workers
prepared the first wurtzitanes incorporating three het-
eroatoms within the ring. They found the isomerization
of triimines to triazawurtzitanes 1.2 However, they also
reported that trioxawurtzitane 2 was not obtained from
trialdehyde 3. The contrast between these two reactions
Sch em e 2
may be explained in terms of their conformations. We
introduced methyl groups into the ipso positions of 3 and
examined the effect of these groups. Here, we report the
synthesis of cis,cis-1,3,5-triformyl-1,3,5-trimethylcyclo-
hexane (7) and the first cyclization of 7 to 1,7,9-trimethyl-
3,5,12-trioxawurtzitane (1,7,9-trimethyl-3,5,12-trioxa-
tetracyclo[5.3.1.12,6.04,9]dodecane) (9). We also present
the conformational analysis of 7 and a plausible mech-
anism for the cyclization of 7 to 9.
As shown in Scheme 1, 7 was obtained in three steps
in an overall yield of 12% from cis,cis-1,3,5-trimethyl-
cyclohexane-1,3,5-tricarboxylic acid (4).3 Esterification
of 4 with 1-propanol in excess, by refluxing with hydrogen
chloride catalyst, led to triester 5. Reduction of 5 with
lithium aluminum hydride in refluxing THF yielded triol
6 (81% yield from 4). Mayer and co-workers first
prepared triol 6 from methyl ester in 60% yield.4 We
obtained triol 6 in higher yield from n-propyl ester 5 than
from the methyl ester.2
Swern oxidation of triol 6 with TFAA instead of oxalyl
chloride also yielded 9 in 23% yield.
As described above, trioxawurtzitane 2 was not ob-
tained from trialdehyde 3. This conflicting result shows
that conformations of these aldehydes play an important
role in the cyclization. Trialdehydes 3 and 7 can exist
in two chair conformations, respectively; three aldehyde
groups are oriented triaxially (3-a , 7-a ) or triequatorially
(3-e, 7-e). Kemp and co-workers3 studied the conforma-
A modified Swern oxidation of triol 6 with oxalyl
chloride2 yielded 7 in 15% yield as well as 5-formyl-cis,cis-
1,3,5-trimethyl-3-(hydroxymethyl)cyclohexane-1-car-
boxylic acid lactone (8)3 in 14% yield. Trialdehyde 7 is
very unstable, and attempts to recrystallize 7 from
CH2Cl2/hexane led to polymers, which are soluble in
hexane. Standing 7 in CDCl3 at room temperature for 1
day gave trioxawurtzitane 9 in quantitative yield (Scheme
2). Both signals of trioxawurtzitane 9 at 4.85 ppm in the
1H NMR spectrum and at 97.60 ppm in the 13C NMR
(DEPT) spectrum indicate that methine protons exist.
The IR spectrum of 9 does not show any νCO band. A
tion of 4 in detail. They ascribed the large chemical shift
difference observed between the equatorial and axial
methylene proton (∆δ > 1.2 ppm) to the triaxial carbonyl
groups. The chemical shift difference of 7, ∆δ ) 1.43
ppm, supports that 7 takes the conformation 7-a . Fur-
thermore, aldehyde groups of 7-a are closer to equatorial
methylene protons than to axial ones. NOE differential
(1) Cupas, C. A.; Hodakowski, L. J . Am. Chem. Soc. 1974, 96, 4668.
(2) Nielsen, A. T.; Christian, S. L.; Moore, D. W.; Gilardi, R. D.;
George, C. F. J . Org. Chem. 1987, 52, 1656.
(3) Kemp, D. S.; Petrakis, K. S. J . Org. Chem. 1981, 46, 5140.
(4) Mayer, H. A.; Fawzi, R; Steimann, M. Chem. Ber. 1993, 126,
1341.
S0022-3263(96)02145-7 CCC: $14.00 © 1997 American Chemical Society

1174 J . Org. Chem., Vol. 62, No. 4, 1997
Notes
Ta ble 1. Rela tive En er giesa for Con for m er s of
Tr ifor m ylcycloh exa n es a n d Tr ioxa w u r tzita n es
Sch em e 4
structure
rel energy (kcal/mol)
7-e
7-a
9
11.33
0.00 b
-11.04
3-e
3-a
2
-1.38
0.00c
-6.85
a
b
HF/6-31G**. -689.504 945 1 hartree. c -572.393 867 2 har-
tree.
Sch em e 3
spectra indicate that aldehyde groups of 7 do not cor-
relate with axial methylene protons but with equatorial
methylene protons.
The cyclization of trialdehydes to trioxawurtzitanes
proceeds with triaxial conformation. The fact that 7
cyclizes and 3 does not suggests that a large difference
in the stability of both axial forms exists. We carried
out ab initio energy calculations for the conformers of the
trialdehydes and the trioxawurtzitanes (Table 1), where
all geometries were optimized at the HF/6-31G** level.5
The data indicate that the cyclizations of both trialde-
hydes to trioxawurtzitanes are exothermic. However, a
large difference is found for the stability of the conformers
of the trialdehydes. The equatorial form is slightly
preferable to the axial one for 3. In contrast with 3, the
axial form of 7 is more stable than the equatorial one.
The stability of the axial forms has a great influence on
the process that 7 cyclizes and 3 does not.
Allowing 7 to stand in CDCl3 yielded 9 quantitatively,
but 7 in acetone-d6 gave decomposition products of 7 very
slowly. 7 did not give 9 either in DMSO-d6 or in C6D6 at
room temperature for 1 day. These results indicate that
a catalytic action exists in the cyclization process. The
addition of radical scavenger galvinoxyl (G•)6 to a CDCl3
solution of 7 inhibited the formation of 9 (Scheme 3). But,
the reaction of radical initiator 2,2′-azobis(2,4-dimeth-
ylvaleronitrile) with 7 in C6D6 at 70 °C did not give 9 at
all, and heating 7 in C6D6 at 70 °C yielded a small
amount of 9 with many decomposition products. Fur-
thermore, bubbling hydrogen chloride into the CH2Cl2
solution of 7 only gave 9. Scheme 4 shows a plausible
mechanism for the cyclization of 7 to 9 in CDCl3. It is
well-known that chloroform decomposes in the presence
of oxygen and light to yield HCl7 and that trioxane
Sch em e 5
formation from formaldehyde is acid catalyzed.8 This
mechanism explains the experimental facts well. The
result that the addition of galvinoxyl inhibited the
formation of 9 suggests that acids are trapped by gal-
vinoxyl. Coppinger described that galvinoxyl is sensitive
to traces of strong acid in hydroxylic or hydrocarbon
solvents.9 Screttas and co-workers reported the reaction
between galvinoxyl G• and a series of substituted acetic
acids.10 They found that hydrogalvinoxyl GH was pro-
duced and proposed Scheme 5. We confirmed micro
amounts of hydrogalvinoxyl in the first stage (1H NMR
assay). After all hydrogalvinoxyl was lost, 2,6-di-tert-
butyl-1,4-benzoquinone6 and 3,5-di-tert-butyl-4-hydroxy-
benzaldehyde6 were mainly produced. The fact that
heating 7 in C6D6 at 70 °C yielded a small amount of 9
(5) Gaussian 94: Frisch, M. J .; Trucks, G. W.; Schlegel, H. B.; Gill,
P. M. W.; J ohnson, B. G.; Robb, M. A.; Cheeseman, J . R.; Keith, T.;
Petersson, G. A.; Montgomery, J . A.; Raghavachari, K.; Al-Laham, M.
A.; Zakrzewski, V. G.; Ortiz, J . V.; Foresman, J . B.; Peng, C. Y.; Ayala,
P. Y.; Chen, W.; Wong, M. W.; Andres, J . L.; Replogle, E. S.; Gomperts,
R.; Martin, R. L.; Fox, D. J .; Binkley, J . S.; Defrees, D. J .; Baker, J .;
Stewart, J . P.; Head-Gordon, M.; Gonzalez, C.; Pople, J . A. Gaussian,
Inc., Pittsburgh, PA, 1995.
(7) Huntress, E. H. Organic Chlorine Compounds; Wiley: New York,
1948; pp 544-560.
(8) (a) Walker, J . F.; Carlisle, P. J . Chem. Eng. News 1943, 21, 1250.
(b) Walling, C.; Augurt, T. A. J . Am. Chem. Soc. 1966, 88, 4163.
(9) Coppinger, G. M. J . Am. Chem. Soc. 1957, 79, 501.
(10) Screttas, C. G.; Heropoulos, G. A.; Karayannis, M. I. Tetrahe-
dron 1984, 40, 5275.
(6) (a) Adam, W.; Haas, W.; Lohray, B. B. J . Am. Chem. Soc. 1991,
113, 6202. (b) Kharasch, M. S.; J oshi, B. S. J . Org. Chem. 1957, 22,
1435. (c) Greene, F. D.; Adam, W. J . Org. Chem. 1963, 28, 3550.

Notes
J . Org. Chem., Vol. 62, No. 4, 1997 1175
hexane led to polymers, which are soluble in hexane and show
indicates that 7 cyclizes to 9 at 70 °C thermally without
catalysts such as HCl.
many complicated broad signals in the 1H NMR spectrum.
7
was purified by washing with hexane. 7: colorless solid; 1H
2
NMR (CDCl3, 500 MHz) δ 0.98 (3H, d, J HH ) 14.5 Hz, CHaHe),
Exp er im en ta l Section
1.01 (9H, s, CH3), 2.41 (3H, d, 2J HH ) 14.5 Hz, CHaHe), 9.35 (3H,
s, CHO); 13C NMR (CDCl3, 125.7 MHz) δ 25.62 (CH3), 38.09
(CCH2C), 45.98 [(CH2)2CCH3], 202.84 (CHO); IR (KBr) νCO 1730
cm-1; MS (EI) m/ z 210 (13, M+ ), 182 (17, M+ - CO); HRMS
calcd for C12H18O3 210.1256, found 210.1287.
Gen er a l Meth od s. All reactions were performed in oven-
dried glassware equipped with a magnetic stirring bar under
argon atmosphere using standard syringe techniques. DMSO
and CH2Cl2 were distilled from CaH2 and stored over molecular
sieves. All other solvents were of dehydrated grade. Reagents
were of commercial grade. 1H (500 MHz) and 13C (125.7 MHz)
NMR spectra were recorded in CDCl3, acetone-d6, C6D6, or
DMSO-d6.
1,7,9-Tr im eth yl-3,5,12-tr ioxawu r tzitan e (1,7,9-Tr im eth yl-
3,5,12-tr ioxatetr acyclo[5.3.1.12,6.04,9]dodecan e) (9). The same
manner that was employed in the preparation of 7 was used
with TFAA instead of (COCl)2 to yield 9. Trioxawurtzitane 9
(0.22 g, 1.1 mmol, 23%) was obtained from 6 (1.02 g, 4.7 mmol).
9 sublimates at 90 °C under atmospheric pressure. 9: colorless
cis,cis-1,3,5-Tr is(h yd r oxym et h yl)-1,3,5-t r im et h ylcyclo-
h exa n e(6).4 Hydrogen chloride was bubbled into a suspension
of cis,cis-1,3,5-trimethyl-1,3,5-cyclohexanetricarboxylic acid (4)
(6.88 g, 26.6 mmol) in 100 mL of dry 1-propanol for 2 h. The
suspension was heated under reflux for 1 h and then cooled to
room temperature. Volatiles were removed under reduced
pressure, and the residue was treated with 50 mL of ice-water.
The oily product was extracted with two 50-mL portions of ether.
After being dried with MgSO4, followed by filtration, the filtrate
was concentrated to remove volatiles, leaving 10.86 g of a
colorless oil containing cis,cis-tri-n-propyl 1,3,5-trimethyl-1,3,5-
cyclohexanetricarboxylate (5) (1H NMR assay). A solution of the
residual oil in THF (35 mL) was then added dropwise to an ice-
cooled suspension of lithium aluminum hydride (4.33 g, 114
mmol) in THF (80 mL). After the reaction mixture was refluxed
for 20 h and cooled to 0 °C, the salts were decomposed with
water/ethanol (10 mL:20 mL). The suspension was filtered, and
the solids were extracted with three 30-mL portions of hot
ethanol. The solvent was removed under reduced pressure, and
the residue was extracted with four 50-mL portions of hot 1,4-
dioxane. The extracts were concentrated to 50 mL and cooled
to yield 6 4 (total yield 4.66 g, 21.5 mmol, 81%). 5: colorless oil;
2
crystals; 1H NMR (CDCl3, 500 MHz) δ 0.85 (3H, d, J HH ) 12.2
2
Hz, CHaHe), 1.04 (9H, s, CH3), 1.34 (3H, d, J HH ) 12.2 Hz,
CHaHe), 4.85 (3H, s, OCHO); 13C NMR (CDCl3, 125.7 MHz) δ
27.86 (CH3), 35.12 [(CH2)2CCH3], 41.59 (CCH2C), 97.60 (OCHO);
MS (EI) m/ z 210 (41, M+); HRMS calcd for C12H18O3 210.1256,
found 210.1244.
Cycliza tion of 7 to 9. Trialdehyde 7 (2 mg, 0.01 mmol) was
added to CDCl3 (0.6 mL) in an NMR tube. Allowing the solution
to stand at room temperature for 1 day gave trioxawurtzitane 9
in quantitative yield (1H NMR assay).
Trialdehyde 7 (4.5 mg, 0.02 mmol) was added to a solution of
galvinoxyl6 (2 mg, 0.004 mmol) in CDCl3 (0.6 mL). After the
solution was allowed to stand at room temperature for 40 h, the
1H NMR signals of 7 did not change.
Trialdehyde 7 (3 mg, 0.01 mmol) was added to a solution of
2,2′-azobis(2,4-dimethylvaleronitrile) (5 mg, 0.02 mmol) in C6D6
(0.6 mL). Heating the solution at 70 °C for 1 day did not give 9
at all (1H NMR assay). Heating 7 (3 mg, 0.01 mmol) in C6D6
(0.6 mL) at 70 °C for 1 day yielded a small amount of 9 with
many other decomposition products.
Hydrogen chloride was bubbled into a solution of 7 (15.6 mg,
0.074 mmol) in CH2Cl2 (10 mL) for 2 h. After the mixture was
stirred at room temperature for 1 day, volatiles were removed
under reduced pressure. 9 (7.6 mg, 0.036 mmol) was yielded in
50% yield, and other products were not confirmed.
Com p u ta tion a l Meth od s. All geometry optimizations and
conformer searches were performed using the Gaussian 94
program.5
3
1H NMR (CDCl3, 500 MHz) δ 0.91 (9H, t, J HH ) 7.5 Hz, OCH2-
CH2CH3), 1.01 (3H, d, 2J HH ) 14.6 Hz, CHaHe), 1.21 (9H, s, CH3),
1.63 (6H, sextet, 3J HH ) 7.3 Hz, OCH2CH2CH3), 2.73 (3H, d, 2J HH
3
) 14.6 Hz, CHaHe), 3.98 (6H, t, J HH ) 6.7 Hz, OCH2CH2CH3);
IR (KBr) νCO 1740 cm-1; MS (EI) m/ z 325 (78, M+ - OCH2CH2-
CH3).
cis,cis-1,3,5-Tr ifor m yl-1,3,5-tr im eth ylcycloh exa n e (7). A
mixture of dry DMSO (4 mL) and dry CH2Cl2 (10 mL) was added
to a solution of (COCl)2 (4.42 g, 34.8 mmol) in 40 mL of dry
CH2Cl2 at -55 °C. After the mixture was stirred for 30 min, a
solution of 1.02 g (4.7 mmol) of 6 in dry CH2Cl2 (25 mL)/DMSO
(20 mL) was added at -55 °C. The solution was allowed to stand
at -55 °C for 2 h. NEt3 (16 mL) was then added, and the
reaction mixture was allowed to warm to 25 °C. Water (50 mL)
was then added, and the CH2Cl2 layer was separated. Concen-
trated HCl (3 mL) was added to the aqueous part, followed by
extraction with two 50-mL portions of CH2Cl2. The combined
CH2Cl2 solutions were washed once with 50 mL of 1 N HCl and
once with 50 mL of an aqueous solution saturated with NaCl.
After being dried with MgSO4 for 30 min, followed by filtration,
the filtrate was concentrated to remove volatiles. The residual
yellow solid was subjected to a silica gel column and eluted with
CH2Cl2/acetone (50/1). 7 (0.15 g, 0.71 mmol, 15%) and 83 (0.14
g, 0.65 mmol, 14%) were obtained as the first and second
fractions, respectively. Attempts to recrystallize 7 from CH2Cl2/
Ack n ow led gm en t. The present calculations were
carried out using the Research Information Processing
System of the Agency of Industrial Science and Tech-
nology (Ministry of International Trade and Industry).
Su p p or t in g In for m a t ion Ava ila b le: 1H NMR and 13C
NMR spectra of compounds 7 and 9, the spectral changes upon
the cyclization of 7 to 9 and upon the addition of galvinoxyl to
the solution of 7, and NOE differential spectra of 7 (7 pages).
This material is contained in libraries on microfiche, im-
mediately follows this article in the microfilm version of the
journal, and can be ordered from the ACS; see any current
masthead page for ordering information.
J O962145N