Structure and Thermal Decomposition of Some 5-(Cyclohexyloxy)thianthreniumyl Perchlorates

Article abstract of DOI:10.1021/jo970893m

Five sets of cis- and trans-substituted cyclohexanols were used for reaction with thianthrene cation radical perchlorate, namely, cis- and trans-cyclohexane-1,2-diol, cis- and trans-2-methyl-, 3-methyl-, and 4-methylcyclohexanol, and cis,cis- and trans,trans-3,5-dimethylcyclohexanol. Reaction in CH₂-Cl₂ solution and precipitation with ether gave the corresponding crystalline 5-(cyclohexyloxy)-thianthreniumyl perchlorate salts. The configuration of each salt in CDCl₃ was shown by ¹H and ¹³C NMR spectroscopy to correspond with the configuration of the cyclohexanol from which it was made. X-ray Ortep diagrams of four of the salts confirmed the structure deduced from NMR spectroscopy. In the NMR, inequivalence of the ¹H and ¹³C signals from the thianthreniumyl 4-and 6-, 1- and 9-, 2- and 8-, and 3- and 7-positions was found when the 1′-position of the cyclohexyl ring was stereogenic. In the four Ortep diagrams, the orientation of the S-O bond was psuedoaxial. Thermal decomposition of the salts made from the monosubstituted cyclohexanols at 100°C in CH₃CN solution gave products consistent with the assigned structures.

Full text of DOI:10.1021/jo970893m

J . Org. Chem. 1997, 62, 8693-8701
8693
Str u ctu r e a n d Th er m a l Decom p osition of Som e
5-(Cycloh exyloxy)th ia n th r en iu m yl P er ch lor a tes
Wenyi Zhao, Henry J . Shine,* and Bruce R. Whittlesey
Department of Chemistry and Biochemistry, Texas Tech University, Lubbock, Texas 79409
Received May 20, 1997X
Five sets of cis- and trans-substituted cyclohexanols were used for reaction with thianthrene cation
radical perchlorate, namely, cis- and trans-cyclohexane-1,2-diol, cis- and trans-2-methyl-, 3-methyl-,
and 4-methylcyclohexanol, and cis,cis- and trans,trans-3,5-dimethylcyclohexanol. Reaction in CH2-
Cl2 solution and precipitation with ether gave the corresponding crystalline 5-(cyclohexyloxy)-
1
thianthreniumyl perchlorate salts. The configuration of each salt in CDCl3 was shown by H and
13C NMR spectroscopy to correspond with the configuration of the cyclohexanol from which it was
made. X-ray Ortep diagrams of four of the salts confirmed the structure deduced from NMR
1
spectroscopy. In the NMR, inequivalence of the H and 13C signals from the thianthreniumyl 4-
and 6-, 1- and 9-, 2- and 8-, and 3- and 7-positions was found when the 1′-position of the cyclohexyl
ring was stereogenic. In the four Ortep diagrams, the orientation of the S-O bond was psuedoaxial.
Thermal decomposition of the salts made from the monosubstituted cyclohexanols at 100 °C in
CH3CN solution gave products consistent with the assigned structures.
In tr od u ction
tions gave N-alkylsulfiliminium- (3) and N,N-dialkylsul-
filiminium perchlorates (4), as shown in abbreviated form
One of the simplest and best-known reactions of
organosulfur cation radicals is that with water, which
gives equal amounts of the parent organosulfur com-
pound and its oxide. The reaction has been studied in
much mechanistic detail with the thianthrene cation
radical (Th•+), shown stoichiometrically in eq 1. Studies
of this reaction, particularly, and of the similar phenothi-
azine cation radical reaction, have a very long history.1
It is evident that, regardless of mechanistic nuances, the
oxygen atom of the water molecule becomes the oxygen
atom of the sulfoxide.
in eqs 2 and 3.4
In contrast with these reactions, particularly of the
organosulfur cation radicals with alkylamines, no studies
were reported, in the earlier years of cation radical
chemistry, of reactions of organosulfur cation radicals
with alcohols. It was known in our own laboratory that
Th•+, for example, was unstable in solutions containing
methanol and ethanol. Reactions at the root of this
instability were known to be relatively slow, because
methanol was used routinely to trap alkyl cations formed
An analogous reaction of Th•+ 2 and of the phenoxathiin
cation radical3 with ammonia was discovered about 20
years ago. That reaction is illustrated in Scheme 1 with
phenoxathiin cation radical, from which both the sulfil-
iminium perchlorate (1) and sulfilimine (2) were isolated.3
In that period also, reactions of Th•+ and of N-methyl-
and N-phenylphenothiazine cation radicals with primary
and secondary alkylamines were reported. Those reac-
in some of our (faster) oxidations of azoalkanes by Th•+ 5
.
Furthermore, we were aware some 20 years ago that
5-(alkyloxy)thianthreniumyl perchlorates may be formed
in reactions of Th•+ with alcohols, but were not able to
pursue the work.6
Recently, we began investigations of the reactions of
Th•+ with alcohols7 and diols8 in CH3CN solutions.
Reaction with alcohols gave alkenes, ethers, and, in some
cases, N-alkylacetamides. Initially, it was thought that
X Abstract published in Advance ACS Abstracts, November 1, 1997.
(1) (a) Shine, H. J . In Organosulfur Chemistry; J anssen, M. J ., Ed.;
Interscience Publishers: New York, 1967; pp 93-117. (b) Shine, H. J .
In Organic Free Radicals; ACS Symposium Series No. 69; Pryor, W.
A., Ed.; American Chemical Society: Washington, D.C., 1978; pp 359-
375. (c) Bard, A. J .; Ledwith, A.; Shine, H. J . Adv. Phys. Org. Chem.
1976, 13, 229. (d) Shine, H. J . In The Chemistry of the Sulfonium
Group; Stirling, C. J . M., Patai, S., Eds.; Wiley: New York, 1981; pp
523-570. (e) Hammerich, O.; Parker, V. D. Adv. Phys. Org. Chem.
1984, 20, pp 55-157.
(2) Shine, H. J .; Silber, J . J . J . Am. Chem. Soc. 1972, 94, 1026.
(3) Mani, S. R.; Shine, H. J . J . Org. Chem. 1975, 40, 2756.
(4) (a) Kim, K.; Shine, H. J . Tetrahedron Lett. 1974, 99. J . Org.
Chem. 1974, 39, 2537. (b) Bandlish, B. K.; Padilla, A. G.; Shine, H. J .
J . Org. Chem. 1975, 40, 2590.
(5) Engel, P. S.; Robertson, D. T.; Scholz, J . N.; Shine, H. J . J . Org.
Chem. 1992, 57, 6178.
(6) (a) Unpublished work of Dr. K. Kim. (b) Bard, A. J .; Ledwith,
A.; Shine, H. J . In Advances Physical Organic Chemistry; Gold, V.,
Bethell, D., Eds.; Academic Press: London, 1976; Vol. 13, p 232.
(7) (a) Shine, H. J .; Yueh, W. Tetrahedron Lett. 1992, 33, 6583. (b)
Shine, H. J .; Yueh, W. J . Org. Chem. 1994, 59, 9, 3553.
(8) Han, D. S.; Shine, H. J . J . Org. Chem. 1996, 61, 1, 3977.
S0022-3263(97)00893-1 CCC: $14.00 © 1997 American Chemical Society

8694 J . Org. Chem., Vol. 62, No. 25, 1997
Zhao et al.
Sch em e 1
Sch em e 2
these products were formed in solution from reactions of
the primary but labile products, 5-(alkyloxy)thianthre-
niumyl perchlorates.7 In time, however, it was found that
these primary products were quite stable in CH3CN and
that the secondary products that had been characterized
and assayed by gas chromatography (GC) were being
formed, in fact, for the most part by thermal decomposi-
tion on the GC column. Thus, 5-(cyclohexyloxy)thian-
threniumyl perchlorate (5a ) was isolated from the reac-
tion of Th•+ClO4 with cyclohexanol in CH2Cl2 solution
-
and was shown to decompose not only on a GC column
but also in CH3CN at 100 °C to cyclohexene and ThO.9
We have now isolated thianthreniumyloxy perchlorates
5b-l from reactions of Th•+ ClO4 with cis-and trans-
-
cyclohexane-1,2-diol, cis- and trans-2-methylcyclohexanol,
cis- and trans-3-methylcyclohexanol, and cis- and trans-
4-methylcyclohexanol, cis,cis- and trans,trans-3,5-di-
methylcyclohexanol, and methanol. We report here the
1H NMR characterization of and structure of 10 of these
11 perchlorate salts, the 13C NMR spectra of 7 of them
(among which the 5-methoxy salt, 5l, served as a model),
and the structure of 4 of them as determined by X-ray
crystallography (Ortep diagrams). We report also the
products of decomposition of salts 5b-i in CH3CN at 100
°C.
For H-1′a we have J -1′a (2′a,6′a) 10.0 Hz, and J -1′a (6′e)
4.91 Hz. For H-2′a we have J -2′a (1′a,3′a) 10.1 Hz and
J -2′a (3′e) 4.87 Hz. In contrast, in the isomer 5c the 1′-H
atom is equatorial and gave a broad and partly resolved
singlet NMR signal diagnostic of an equatorial H atom.
The 2′-H atom is axial and showed coupling with another
axial H (3′-H) and two inequivalent equatorial H atoms
at positions 1′ and 3′. That is, J -2′a (3′a) was 11.2 Hz,
J -2′a (1′e) 2.37 Hz, and J -2′a (3′e) 4.40 Hz. We deduce
that J -2′a (1′e) is the smaller coupling from the general
rule that when an electron-attracting group (here the
thianthreniumyloxy) is antiperiplanar to one of the
coupling H atoms, J is diminished by 1-2 Hz.10
The orientations of the thianthreniumyloxy and OH
groups in 5b and 5c, deduced from NMR spectra, are
supported by the X-ray crystal patterns of these two salts
(Figures 1a and 2a in the Supporting Information). The
Ortep diagrams show clearly that the large thianthre-
niumyloxy group is axial in 5c, and the NMR data show
that this orientation is maintained in CDCl3 solution.
Correspondingly, the Ortep diagram for 5b shows clearly
that each substituent is in the equatorial position.
The 1′-H signals in the NMR spectra for 5d and 5e
are equally diagnostic. In the trans isomer 5d , the axial
proton H-1′a gave rise to a triplet of doublets, J -1′a
(2′a,6′a) 10.1 Hz and J -1′a (6′e) 4.61 Hz. This compares
well with the NMR of the corresponding axial proton in
trans-2-methylcyclohexanol, J ) 9.71 and 4.27 Hz. The
chemical shifts (Table 1) for these axial protons reflect
also the nearby positive charge in 5d (δ ) 4.186) as
compared with δ ) 3.121 in the alcohol. Correspond-
ingly, the NMR for H-1′e in 5e is a doublet of triplets,
with J -1′e (6′e) 4.90 Hz and J -1′e (2′a,6′a) 2.44 Hz. These
data and assignments also comply with the general rule
about the influence of an attached electronegative atom.10
A similar multiplet is seen in cis-2-methylcyclohexanol,
Resu lts a n d Discu ssion
Str u ctu r e of 5b-k . The structure we refer to here
means the position (axial or equatorial) occupied by the
thianthreniumyloxy group on the cyclohexyl ring. Ini-
tially, we felt that in all of these compounds the large
thianthreniumyloxy group would have an equatorial
orientation. 1H NMR spectroscopy and X-ray crystal-
lography (of four of the perchlorate salts) show, however,
that this group has the same orientation in the cyclohexyl
ring as the OH group had in the cyclohexanol from which
the salt was prepared. This can be deduced from the
splitting pattern in the NMR signal for the 1′-H atom of
the cyclohexyl ring. That is, where the thianthrenium-
yloxy group is equatorial the 1′-H atom must be axial
and show the splitting pattern of an axial H coupling with
nearby protons. Correspondingly, where the thianthre-
niumyloxy group is in the axial position, the 1′-H atom
must be equatorial and show the splitting pattern of an
equatorial H. Our NMR results show that 5b,d ,f,h ,j
have (predominantly) an axial 1′-H atom whereas
5c,e,g,i,k have (predominantly) an equatorial 1′-H atom
(Scheme 2). The NMR data are particularly instructive
in the cases of 5b and 5c, because in these cases the
splitting pattern of not only the 1′-H but also of the 2′-H
atom was discernible. In 5b both of these H atoms are
axial and their NMR signals were triplets of doublets.
(9) Zhao, W.; Shine, H. J . Tetrahedron Lett. 1996, 37, 1749. The
numbering of protons (H1 and H4) in this report is incorrect and should
be interchanged.
(10) Eliel, E. E.; Wilen, S. H.; Mander, L. N. Stereochemistry of
Organic Compounds; Wiley-Interscience: New York, 1994; pp 712, 713.

5-(Cyclohexyloxy)thianthreniumyl Perchlorates
J . Org. Chem., Vol. 62, No. 25, 1997 8695
Ta ble 1. 1H Ch em ica l Sh ifts (δ)a for 5b-k a n d Th ia n th r en e 5-Oxid e (Th O)
δ
Hb
5b
5c
5d
5e
5f
5g
5h
5i
5j
5k
Th Oc
7.91
7.60
7.53
7.40
4
6
1
9
2
8
3
7
8.737
8.697
7.921
7.911
7.834
7.822
7.738
7.720
4.581
3.288
8.727
8.630
7.928
7.920
7.820
7.813
7.732
7.708
5.146
3.953
8.766
8.698
7.920
7.913
7.840
7.832
7.744
7.735
4.186
8.735
8.699
8.719
8.711
8.719
8.710
8.677
7.923
7.847
8.693
7.939
7.850
8.703
7.914
7.836
8.713
7.929
7.836
7.933
7.845
7.907
7.928
7.830
7.826
7.743
7.738
4.646
7.837
7.832
7.742
7.736
5.063
7.743
4.773
7.730
4.579
7.740
4.933
7.740
4.674
7.734
5.107
1′
2′
a
b
In CDCl3 solution with TMS as reference for 5b-k . Listed in order of decreasing δ. c In CDCl3 solution with CHCl3 as reference at
7.26 ppm, ref 13.
J ) 5.06 and 2.51 Hz, while the chemical shifts of the
two related protons are δ ) 4.773 for 5e and 3.78 for the
alcohol, again reflecting the positive charge in 5e. 5f,
prepared from cis-3-methylcyclohexanol, has an axial 1′-
H, whose NMR signal was a triplet of triplets, δ ) 4.646.
The alcohol, itself, had a similar NMR multiplet, δ )
3.563. Isomeric 5g and the corresponding trans-3-meth-
ylcyclohexanol had equatorial 1′-H, broad and poorly
resolved multiplets, centered at δ ) 5.063 and 4.061,
respectively. 5h , prepared from trans-4-methylcyclohex-
anol, has an axial 1′-H whose NMR signal was a triplet
of triplets with J -1′a (2′a,6′a) 10.9 Hz and J -1′a (2′e,6′e)
4.51 Hz. The NMR of the related proton in trans-4-
methylcyclohexanol itself was also a triplet of triplets
with J ) 10.7 and 4.36 Hz, with δ ) 3.544 as compared
with 4.58 for 5h . In the cis isomer 5i the axial orienta-
tion of the thianthreniumyloxy group is deducible from
the NMR spectrum of H-1′e and supported by the Ortep
diagram from X-ray crystallography (Figure 3a in the
Supporting Information). The NMR spectrum of H-1′e
was a poorly resolved multiplet, satisfiable with J ap-
proximately 4.7 and 2.9 Hz. In the case of 5k , the axial
orientation of the thianthreniumyloxy group is deduced
from the equatorial orientation of H-1′, a broad, unre-
solved singlet, δ ) 5.107. The isomer 5j had, in contrast,
for H-1′, a triplet of triplets, J ) 11.2 and 4.65 Hz,
centered at δ ) 4.674, diagnostic of an equatorial thian-
threniumyloxy group, and this is seen clearly, also, in
the Ortep projection of Figure 4a in the Supporting
Information. Similar patterns for H-1′ were found in the
corresponding alcohols.
Insofar as structure is concerned, then, the cis-cyclo-
hexanols gave cis salts (5c,e,f,i,j) and the trans-cyclo-
hexanols gave trans salts (5b,d ,g,h ,k ). The cis salts did
not undergo the ring flipping in solution that would have
placed the large thianthreniumyloxy group predomi-
nantly in the equatorial position. This was a surprise
to us. Brown, in a similar finding years ago, found that
group size and conformational disposition were not
necessarily related and invoked the (opposing) contribu-
tions of van der Waals and covalent radii of the group.11
He noted, for example, the report of Berlin and J ensen12
that, contrary to expectation (based on group size), the
amount of axial conformer increased in the series chloro-,
bromo-, and iodocyclohexane. It seems, analogously, that
in 5c,e,g,i,k the large thianthreniumyl group must be
far enough away from the ring (Brown’s covalent radius)
as to not cause the 1,3-diaxial interactions that would
result in ring flipping. Ring flipping, which would place
the large group in the equatorial position would then, in
fact, place a group with, presumably, a larger effective
covalent radius (Me, OH) in the less sterically suitable
axial position. It is remarkable, in fact, that when cis-
cyclohexane-1,2-diol reacts with Th•+, the axial OH group
becomes thianthreniumyloxy (5c). In 5b, ring flipping,
which is proposed (later) to account for the thermal
decomposition of 5b, is inhibited at room temperature
because that would place the remaining OH group in the
axial position, too. It is notable that only one of the OH
groups in these cyclohexane diols reacted with Th•+
.
Insofar as a preferred configuration of the thianthre-
niumyloxy group is concerned, it is, in fact, equatorial in
5a , just as is the hydroxyl group in cyclohexanol. This
1
is deduced again from the H NMR patterns for the 1′-
hydrogen. In 5a , this was an overlapping triplet of
triplets with J -1′a (2′a,6′a) 8.62 and J -1′a (2′e,6′e) 4.31
Hz, while in cyclohexanol the corresponding J were 9.03
and 4.45 Hz. The relevant chemical shifts were 4.691
and 3.611 ppm.
The NMR spectra of compounds 5b-k also had well-
defined signals from the aromatic ring protons. The data
on chemical shifts and coupling constants are given in
Tables 1 and 2. We can relate the NMR spectra of 5b-k
to the spectrum of thianthrene 5-oxide (ThO). ThO is a
symmetrical molecule with four sets of equivalent pro-
tons, H-1,9; H-2,8; H-3,7; and H-4,6. Chemical shifts for
ThO have been reported by Lam,13 in Ternay’s laboratory,
and go from low to high field in the order 4,6; 1,9; 2,8;
3,7. The same order in chemical shifts occurs in 5b-k ,
as deduced from comparison with ThO, from splitting
patterns and from COSY spectra of 5i. 5i was chosen
for COSY spectroscopy because of the simplicity of its
aromatic-region NMR spectrum. That is, the signals
were for four sets of equivalent protons, as in ThO, and
were not complicated with overlapping. The spectrum
consisted of doublets of doublets for H-4,6 and H-1,9 and
triplets of doublets for H-2,8 and H-3,7. The spectrum
from the trans isomer 5h also had the simplicity of
splitting patterns with the exception that a small cou-
pling (J ) 0.637) was found in the spectrum of H-1,9 and
may have been caused by para coupling J -1(4),9(6).
In contrast with the simplicity of the aromatic 1H NMR
spectra of 5h ,i caused by sets of equivalent protons, the
spectra of 5b-g were more complex, because of the
inequivalency of the protons in each set. The spectra of
5b,c exhibited this complexity most clearly. The mul-
tiplet attributable to each of the eight aromatic protons
(13) Lam, W. W. The Synthesis and Multinuclear Magnetic Reso-
nance Studies of Pharmacologically Important Sulfur-Containing
Compounds. Dissertation; University of Texas at Arlington, 1989, p
234.
(11) Brown, H. C.; Klimisch, R. L. J . Am. Chem. Soc. 1966, 88, 1425.
(12) Berlin, A. J .; J ensen, F. R. Chem. Ind. (London) 1960, 998.

8696 J . Org. Chem., Vol. 62, No. 25, 1997
Zhao et al.
Ta ble 2. Cou p lin g Con sta n ts (J ) for Ar om a tic H in 5b-k a n d Th ia n th r en e 5-Oxid e (Th O)
J b
5g
couplinga
5b
5c
5d
5e
5f
5h
7.86
1.36
5i
5j
7.61
f
5k
ThO
7.66
1.39
4(3)
6(7)
4(2)
6(8)
4(1)
6(9)
1(2)
9(8)
1(3)
9(7)
1(4)
9(6)
2(1,3)
8(7,9)
2(4)
8(6)
3(2,4)
7(6,8)
3(1)
7(9)
7.73
c
1.66
1.57
0.638
e
7.62
7.75
1.65
1.55
0.632
e
7.83
c
1.56
1.50
e
7.08d
7.81
7.70
2.03
1.74
e
7.59
7.73
7.88
1.40
7.78
1.44
7.26d
1.50
e
e
e
1.84
e
e
e
e
e
e
e
e
e
e
e
e
e
0.758
7.21
7.89
1.64
1.51
0.638
e
7.53
7.39
1.69
1.56
7.42
7.48
1.72
1.68
7.84
7.68
1.43
1.65
0.730
e
7.51
7.85
1.56
1.39
7.43
7.49
1.63
1.54
7.83
7.84
1.68
1.28
e
7.82
1.55
7.83
1.89
7.64
1.79
7.98
1.76
0.636
7.49
1.48
7.73
1.61
7.98
1.52
e
7.74
1.79
e
7.82
1.55
e
7.56
1.26
e
e
e
e
e
e
e
e
7.62
7.63
1.71
1.74
7.22
7.37
1.74
1.82
7.62
7.62
1.77
1.78
7.32
7.45
1.64
1.84
7.66
7.29
1.46
1.51
7.54
7.51
1.11d
1.44
7.51
1.55
7.48
1.63
7.67
1.42
7.62
1.36
7.46
1.52
7.43
1.85
7.54
1.57
7.49
1.62
7.55
1.31
7.47
1.47
a
Protons are listed in order of decreasing δ. For example, H-4 (before H-6) coupling with H-3, a doublet; and H-2 (before H-8) coupling
b
with H-8 and H-3, a triplet. In Hz; coupling constants are averages when more than one value was measured. For example (5b), the
coupling pattern for H-4 was a dd, so that coupling with H-2 gave two doublets with averaged J ) 1.66 Hz. c Overlap made measurement
uncertain. Low value attributed to poor resolution. e Coupling not evident. f Coupling unresolved.
d
Ta ble 3. Ar om a tic 13C Ch em ica l Sh ifts (δ)a
compound and δ
position
5a
5d
5f
5i
5j
5k
5lb
4
6
4a
5a
1
9
2
8
3
135.744
135.664
135.541
135.508
134.950
134.486
129.539
129.277
128.928
128.863
122.255
121.453
135.824
135.660
135.403
129.352
129.097
121.741
135.779
135.668
135.517
129.394
129.063
135.815
135.645
134.493
129.381
129.123
121.856
135.809
135.643
134.501
129.402
129.079
121.920
135.984
135.709
134.339
129.334
129.264
121.756
136.427
135.715
135.527
129.446
128.947
119.241
7
9a
10a
121.961
121.886
a
b
In ppm with respect to TMS as internal standard. 5-(Methoxy)thianthreniumyl perchlorate.
was readily assignable, but an arbitrary choice had to
be made as to which proton of each pair would be
assigned the downfield shift. The choice was made, as
seen in Table 1, to assign the lower-field shift to the lower
numbered proton, e.g., H-4 rather than H-6. Coupling
constants for most of the proton-proton interactions were
measurable. Only in a few cases were splittings not
distinct enough for measurement or too small to be seen.
The data are given in Table 2.
Inequivalency of protons in a pair of protons was seen
fairly well in 5d ,f,g and to a small extent in 5e. The
chemical shifts and coupling constants are listed in
Tables 1 and 2. The inequivalency of protons in the rings
of 5b-g is attributable to the presence of a stereogenic
center at position C-1′ in the cyclohexyl ring. That is,
regardless of the conformation of the cyclohexyl ring in
5b-g, and regardless of the cyclohexyloxy’s spatial
orientation (whether pseudoaxial or equatorial at the
sulfur atom) and the possiblity of rotation about the S-O
and O-cyclohexyl axes, the protons on the separated
aromatic rings of thianthrenium apparently experience
different environments and respond with different chemi-
cal shifts.14a This effect is absent in 5h ,i.14b
The possibility must be addressed that inequivalency
1
in H signals may have been caused not by the presence
of a stereogenic center in 5b-g but by restriction to
rotation of the cyclohexyl moiety about its S-O bond. To
test this possibility, 5j and 5k were prepared, the premise
being that if rotation were restricted in, say, 5f and 5g,
the 3-methylcyclohexyloxy salts, it should be even more
restricted in the 3,5-dimethylcyclohexyloxy salts, 5j and
5k , and give rise in them to inequivalent NMR signals.
In the event, each of the two astereogenic salts, 5j and
5k , gave aromatic NMR spectra with no sign of inequiva-
lent chemical shifts, supporting the interpretation that
inequivalency in 5b-g is caused by the presence of a
stereogenic center in them.
1
The inequivalency in δ seen in the H NMR spectra
was also seen in the aromatic 13C spectra. Thus, 5a
showed the expected six aromatic 13C peaks of a sym-
metrical 5-(alkyloxy)thianthreniumyl ion (Table 3). The
same complement of peaks was obtained with 5i, 5j, and
5k , and with 5-methoxythianthreniumyl perchlorate, 5l.
On the other hand, 5d showed 12 aromatic 13C peaks,
(14) (a) For a discussion of this effect, see Sanders, J . K. M.; Hunter,
B. K. Modern NMR Spectroscopy; Oxford University Press: Oxford,
1987; pp 299-302. (b) The same effect is seen in simpler 5-(alkyl-
oxy)thianthreniumyl perchlorates in which the alkyl group has a
stereogenic center attached to oxygen, such as 2-pentyl and 2-hexyl.
The effect is absent where the alkyl group is, for example, methyl,
ethyl, isopropyl. Unpublished work of W. Zhao.

5-(Cyclohexyloxy)thianthreniumyl Perchlorates
J . Org. Chem., Vol. 62, No. 25, 1997 8697
1
concordant with its full set of inequivalent H peaks. The
equilibrated conformers at 25 °C to the extent, calculated
from ∆G, of 71%. In contrast with 5a , no change in the
unresolved singlet NMR signal from H-1′ in 5i, the salt
made from cis-4-methylcyclohexanol, could be detected
in going stepwise down to -90 °C. Thus, 5i appeared to
be in one conformation only. Low-temperature NMR
spectroscopy was not carried out with any other examples
of 5 or corresponding alcohols.
Our using NMR data to specify H-1′ as being either
dominantly equatorial or axial appears to be justified.
The aromatic 1H NMR spectra of 5a and 5i at low
temperatures showed no sign of inequivalence. As the
temperature was lowered stepwise to -90 °C, the aro-
matic signals retained their dd (H-4,6 and H-1,9) and td
(H-2,8 and H-3,7) character until at about -45 °C, when
the smaller splittings began to disappear. At -90 °C,
for example, the aromatic signals were doublets for H-4,6
and H-1,9, and triplets for H-2,8 and H-3,7. Our conclu-
sion from these data is that the presence of two cyclohexyl
conformers, as witnessed with 5a , could not be the cause
of the inequivalence in the aromatic NMR signals that
we have noted in some of 5.
12 peaks are listed in Table 3 with assignments of δ
based on the assignments made to the H signals of 5d .
1
On the basis of inequivalency in 1H shifts in 5f, we
expected to find the analogous inequivalency in its 13C
spectrum. However, only seven 13C peaks were found.
We deduce that overlapping of 13C peaks occurred for all
signals except those from C-9a and C-10a. The difference
in results between 5d and 5f can be accommodated by
1
examining the δ data in their H spectra. The differences
(∆δ, ppm) in δ in 5d for H-4 and H-6 (0.068), H-1 and
H-9 (0.007), H-2 and H-8 (0.008), and H-3 and H-7 (0.009)
are larger than for the corresponding shifts in 5f. Those
∆δ, in the same sequence, are 0.008, 0.000, 0.004, and
0.005, indicating that the effect of the unsymmetrical
alkyl group in 5d is larger than in 5f. This difference in
effect shows up not only in the 1H but also in the 13C
spectra.
Last, the effect of the positive charge on chemical shifts
in 5b-k can be seen by comparison with those in ThO
(Table 1). The effect is most marked on the H-4,6
protons, with ∆δ approximately 0.7 ppm, and falls off for
the other sets of protons to approximately 0.3 ppm. The
effect of the positive charge is apparently transmitted
also to the cyclohexyl ring’s H-1′, as seen by comparing
the data for 5d -i with those of the corresponding
alcohols. That is, δ for H-1′ in trans-2-methyl-, cis-2-
methyl-, trans-3-methyl-, cis-3-methyl-, trans-4-methyl-,
and cis-4-methylcyclohexanol was 3.121, 3.779, 4.037,
3.563, 3.544, and 3.947, respectively, while for 5d ,e,f,
g,h ,i, δ was 4.186, 4.773, 4.646, 5.063, 4.579, and 4.933,
respectively. Similarly, δ for H-1′ in 5j and 5k was,
respectively, 4.674 and 5.107, as compared with 3.612
and 4.130 in the corresponding alcohols. The difference
in δ between a compound 5 and its corresponding alcohol
was thus approximately 1 ppm. It is notable that the
equatorial H-1′ had a larger downfield shift than the axial
H-1′ in each pair of salts and alcohols, in agreement with
the generality that axial H generally resonates upfield
of equatorial H.10
A remaining feature of the structure of the salts 5
pertains to the orientation of the S-O bond, that is, as
to whether it is pseudoaxial or pseudoequatorial. An
analogous situation has been presented nicely with
thianthreniumyl bis(ethoxycarbonyl)methylide by Ter-
nay, and this ylide is in the pseudoequatorial conforma-
tion in the solid state.16 In the four cases for which we
have X-ray crystallographic data, 5b, 5c, 5i, and 5j, the
orientation is clearly pseudoaxial, as can be seen in the
projections of Figures 1b-4b in the Supporting Informa-
tion. The orientation has some relevance to our analysis
of NMR spectra in that it could be suggested that the
1
inequivalency found in some of the aromatic H and 13C
signals could be caused in solution by inversion of
configuration at sulfur, leading to a mixture of conform-
ers with different chemical shifts. However, we rule out
this possibility in that it is not seen, for example, with
5i, whose crystallographic orientation is no different from
that of 5b, 5c, and 5j.
In the discussions above, the structures 5a -k have
been treated as if each cyclohexyl group had one identifi-
able conformation. That is, the H-1′ proton has been
characterized as being axial in some cases and equatorial
in others. The corresponding cyclohexanols have been
treated similarly. We recognize that in each of the
alcohols, however, we are identifying the major conformer
of a two-conformer equilibrium. In that respect, Neikam
and Dailey,15a using low-temperature NMR spectroscopy,
found the equilibrium of conformers of cyclohexanol in
CS2 solution to contain 7.1% of the axial-OH conformer.
Also, Bassindale15b has recorded that in cis-4-methylcy-
clohexanol, the conformer with axial-OH and equatorial-
CH3 is preferred to the extent of (our calculation) 88%.
We were unable to detect the separated conformers of
these alcohols in CDCl3 at -50 °C. To go to lower
temperatures with examples of 5 we turned to solvent
CD2Cl2, mp -97 °C. With that, two unresolved but
separate H-1′ peaks were detected in 5a at -90 °C, at δ
) 4.46 (axial-H, 81%) and 4.95 (equatorial-H, 19%). At
25 °C in this solvent one partly resolved multiplet was
observed at δ ) 4.63. These data show that the equato-
rial component of 5a (i.e., with axial H-1′) dominates the
P r od u cts of Th er m a l Decom p osition . Products
were obtained by heating 5 in CH3CN solution at 100 °C
for 30-40 min, and were identified and assayed with gas
chromatography (GC). The product from 5b (trans-
isomer) was cyclohexene oxide (6, 96%), whereas 5c (cis)
gave mainly cyclohexanone (7, 96%) and only a small
amount of 6. From 5d (trans-isomer) were obtained
1-methyl- (8, 45%) and 3-methylcyclohexene (9, 34%) and
some N-(2-methylcyclohexyl)acetamide (10, 14%). The
cis-isomer 5e gave 93% of 8 and 2.7% of 9; none of 10
was obtained. Decomposition of both cis (5f) and trans
(5g) isomers gave mixtures of 3-methyl- (9) and 4-meth-
ylcyclohexene (11), which could not be separated on our
GC columns. They are reported as a mixture (Table 6).
Small amounts of 1-methylcyclohexene (8) were also
obtained from 5f and 5g. 4-Methylcyclohexene (11) was
obtained from the decomposition of both (trans) 5h (87%)
and (cis) 5i (81%) isomers of the 4-methylcyclohexyloxy
salt (Table 7).
In the first reports of reactions of Th•+ClO4 with
-
alcohols it was mistakenly thought that the primary
products, 5-(alkyloxy)thianthreniumyl perchlorates, were
short-lived in solutions containing DTBMP, and that the
(15) (a) Neikam, W. C.; Dailey, B. P. J . Chem. Phys. 1963, 38, 445.
(b) Bassindale. A. The Third Dimension in Organic Chemistry; J ohn
Wiley & Sons: New York, 1984; p 90.
(16) Ternay, A. L., J r.; Baack, J .; Lam, W. W. Phosphorus, Sulfur
Silicon 1992, 70, 19, and references therein.

8698 J . Org. Chem., Vol. 62, No. 25, 1997
Zhao et al.
Ta ble 4. P r od u cts of Decom p osition of 5b a n d 5c in
CH3CN Solu tion Con ta in in g DTBMP a
was based on an earlier finding that 5a reacted very
readily with aqueous K2CO3 to give ThO and cyclohex-
anol.9 Next, a new solution of 5 was heated at 100 °C
and after cooling it to room temperature secondary
products were assayed by GC. Aqueous K2CO3 was next
added to the solution and GC assay was repeated,
showing this time that only the secondary products were
detectable. That is, that 5 had not survived the heating
in solution, for had it done so it would have been
hydrolyzed into ThO and the appropriate alcohol, as
shown schematically in eq 4. Data for this protocol are
given in Tables 4-7.
ratio:e
products/
ThO
products,c mmol × 102
reactant,
run procb mmol × 102
6
7
diold Th ThO
1
a
b
c
a
d
b
c
a
d
b
c
5b, 16.2
10.3 2.0
10.8 1.9
0
8.4 1.4
8.8 0.25
9.2
9.2
1.2 8.6
0.77 9.5
0.68 9.0
0.65 9.5
3.8 4.4 11.6
4.4 4.0 22.9
16.4 0.43 15.0
0
1.09
2
5b, 9.6
0
0
0
0
0
0
0
0
1.5
8.0
0.50 9.0
0.20 9.2
0.19 9.7
1.00
1.00
0.95
0
0
3
5c, 9.9
1.5
8.0
0.60 9.0
0.51 9.1
0.45 9.3
1.14
1.06
1.09
a
Each solution contained DTBMP in small excess over the
amount of 5. b (a) Unheated solution stirred for 75 min and injected
onto column. (b) Preceding solution stirred with 0.4 mL of 2 M
K2CO3 solution for 5 min and again injected onto column. (c)
Preceding solution containing K2CO3 stirred for 18 h (run 1), 20 h
(run 2), 23 h (run 3) and injected onto column. (d) Preceding
solution heated at 100 oC for 30 min, cooled, and injected onto
column. c Columns used were C for 6, 7, and diol, and A for Th
An example of data from this protocol can be seen most
easily in Table 7 with the decomposition of 5h . Entry
8a shows that injection of a solution of 5h onto the GC
column gave 4-methylcyclohexene (11, 99%). However,
addition of 0.3 mL of 2 M aqueous K2CO3 to the unheated
solution followed by GC assay gave, entry 8b, none of 11
but only trans-4-methylcyclohexanol (99%). These two
experiments show that 5h decomposed on the GC col-
umn. In contrast, run 9, Table 7, shows that when
heated in solution at 100 °C for 30 min, 5h gave 11 (entry
c) and that, after heating, none of 5h survived for later
conversion into 4-methylcyclohexanol with K2CO3, entries
9b and 9d. Similar data for 5i are found in Table 7 and
for the other compounds 5b-g in Tables 4-6.
When decomposition of 5 occurs and gives one of the
products 6-11, or when 5 is hydrolyzed to the corre-
sponding alcohol, an equal amount of ThO must be
formed. Therefore, the amount of ThO that was formed
in the various reactions was assayed and is listed in each
entry in the tables. In principle, the amount of ThO
should equal the sum of the amounts of products in each
reaction. That is, the ratio of products/ThO should be
1.00. This ratio is listed in the last column of each table
and is seen for the most part to be close to unity.
Thianthrene (Th) was also formed in every reaction,
usually in small amounts which are listed in the tables.
We do not know the way(s) in which Th was formed.
As an example of the pumping protocol, an unheated
solution of 5h was evaporated to dryness at room
temperature, and the volatile material, collected in liquid
N2, was found to contain only a small amount of 11. The
residue was redissolved in CH3CN and that solution was
heated at 100 °C for 30 min, and again evaporated to
dryness at room temperature. The volatile material now
gave 76% of the 11 available in 5h . The residue from
the second evaporation was dissolved in CH3CN, and
after treatment with K2CO3 the solution was injected into
the GC column, giving a trace of 11 and 18% of N-(4-
methylcyclohexyl)acetamide. These results show that 5h
decomposed when heated in solution almost completely
into 11 and the amide.
d
and ThO. cis- (run 3) or trans- (runs 1, 2) cyclohexane-1,2-diol.
e (Sum of 6, 7 and diol)/ThO.
Ta ble 5. P r od u cts of Decom p osition of 5d a n d 5e in
CH3CN Solu tion Con ta in in g DTBMP a
ratio:f
products,c mmol × 102
reactant,
products/
run procb mmol × 102
8
9
amided ROHe Th ThO
ThO
4
a
b
a
c
b
a
b
d
a
c
b
d
5d , 10.1 5.2 4.3
0
0
0
0.65 10.4
0.93
0.96
1.01
1.07
0.96
0.94
0.99
0.98
0.99
0.90
0.98
0.87
0
0
9.4 0.35 9.8
5
5d , 10.5 5.8 4.6
4.8 3.6
0
0
0
0
0.88 10.3
0.68 10.1
0.40 9.8
0.65 10.2
2.4
1.4
0
0
0
0
0
0
0
4.6 3.4
6
7
5e, 9.8
9.0 0.56
2.8 0
6.9 0.47 9.8
7.2 0.61 10.0
0
0
0
0
2.6 0
5e, 10.1
9.3 0.57
9.2 0.29
9.4 0.27
8.5 0.27
0.72 10.0
0.65 10.6
0.58 9.9
0.49 10.1
a
Each solution contained DTBMP in a small excess over the
b
amount of 5. (a) Unheated solution stirred for 40 min (runs 5,
6), 70 min (run 4), 85 min (run 7), and injected onto column. (b)
Preceding solution stirred with 0.4 mL of 2 M K2CO3 solution for
15 min (runs 4, 5), 5 min (run 6), 10 min (run 7), and again injected
onto column. (c) Preceding solution heated at 100 oC for 30 min
(run 7), 40 min (run 5), cooled, and injected onto column. (d)
Preceding solution containing K2CO3 stirred for 24 h and injected
onto column. c Columns used were B for 8 and 9 and C for all other
d
products. N-(2-Methylcyclohexyl)acetamide. e cis- (runs 6, 7) or
trans- (runs 4, 5) 2-Methylcyclohexanol. f (Sum of 8, 9, amide, and
ROH)/ThO.
secondary products, e.g., alkenes, were formed in solution
at room temperature.7 Only later was it recognized that
the primary products were quite stable at room temper-
ature and that the secondary products had been formed
by thermal decomposition of the primary either in the
GC inlet or on the GC column. Consequently, in the
present work we adopted an experimental procedure to
ensure that the products obtained were indeed formed
by heating compounds 5 in solution and were not formed
from 5 during GC assay. Two experimental protocols
were used. In the first, with each of 5b-i, the solution
of 5 in CH3CN was first shown to decompose into
secondary products on the GC column, i.e., as in our
earlier report.7 The unheated solution was then treated,
usually briefly (20-30 min), with a small amount of
aqueous K2CO3 and shown to give on GC assay none of
the secondary products but only ThO and the alcohol
from which 5 had been made. This part of the protocol
The products of thermal decomposition are related to
the structures of 5b-i. Thus 5b (trans) gave mainly
cyclohexene oxide (6) whereas 5c (cis) gave mainly
cyclohexanone (7). We attribute formation of 6 to the
inversion of configuration of 5b, placing the substituents
in the diaxial configuration, eq 5. Inversions of this kind
prior to product formation have been proposed by others,
for example, to account for the formation of 3-methylcy-
(17) Brown, H. C.; Klimisch, R. L. J . Am. Chem. Soc. 1966, 88, 1430.

5-(Cyclohexyloxy)thianthreniumyl Perchlorates
J . Org. Chem., Vol. 62, No. 25, 1997 8699
Ta ble 6. P r od u cts of Decom p osition of 5f a n d 5g in CH3CN Solu tion Con ta in in g DTBMP a
products,c mmol × 102
ratio:f
products/ThO
run
1
procb
reactant, mmol × 102
5f, 10.1
8
(9 + 11)d
amide
ROHe
Th
ThO
a
b
c
a
d
b
a
b
a
d
b
0.26
0
6.64
0.18
0.12
6.6
8.3
8.1
9.4
0.3
9.73
9.6
9.3
0
0
0
1.37
10.2
9.6
0.91
0.19
0.17
0.77
0.19
0.19
0.68
0.15
0.82
0.15
0.16
9.13
9.9
9.9
9.3
10.0
9.9
8.9
9.5
9.6
10.2
10.3
0.91
1.05
0.99
0.91
1.06
0.93
1.13
1.08
1.10
1.03
0.99
0
2
5f, 10.0
0.44
1.1
1.1
0
0
1.42
0
0
0.63
9.93
0.83
0
1.24
tr
0
0
0
3
4
5g, 10.0
5g, 10.5
0
0
0.94
0.94
tr
tr
a
b
Each solution contained DTBMP in small excess over the amount of 5. (a) Unheated solution stirred for 5 h and injected onto column.
(b) Preceding solution stirred with 0.4 mL of 2 M K2CO3 10 min (run 1), 60 min (run 2), 90 min (run 3), 120 min (run 4), and again
injected onto column. (c) Preceding solution containing K2CO3 stirred for 90 min, and injected onto column. (d) Preceding solution heated
at 100 °C for 30 min, cooled and injected onto column. c Columns used were B for 8, 9, and 11; and A for other products. 9 and 11 could
d
not be separated. e cis- (from 5f) or trans-3-Methylcyclohexanol (from 5g). f (Sum of 8, 9, 11, amide, and ROH)/ThO.
Ta ble 7. P r od u cts of Decom p osition of 5h a n d 5i in
CH3CN Solu tion Con ta in in g DTBMP a
their tosylates that are reported in the literature. Under
purely E2-like conditions, the sulfurane from trans-2-
methylcyclohexanol21 and trans-2-methylcyclohexyl to-
sylate gave 99-100% of 3-methylcyclohexene (9).17,18,22
Thermal decomposition of the trans-tosylate in dimethyl
sulfoxide (DMSO), however, gave 70% of 1-methylcyclo-
hexene (8) and only 30% of 9, and the decomposition was
said to have, therefore, carbonium-ion-pair character.19
The reaction of trans-2-methylcyclohexanol with
Th•+ClO4-, reported earlier,7 gave 8 and 9 in the ratio
66:34, that is, akin to the ratio reported by Hatch for the
tosylate.19 We know now that our earlier results were
from the (then unrecognized) decomposition of 5d on GC
column. That corresponds with our present finding that
5d decomposed at 100 °C in solution to give 8 and 9 in
the ratio 57:43. Thus, decomposition of 5d is thermal
and E1- rather than E2-like in character. In contrast
with 5d , the cis-isomer 5e gave mainly 8 and a small
amount of 9 (run 7a, Table 5), a result that conforms with
either E1- or E2-like elimination. The decompositions
of cis (5f) and trans (5g) salts, from cis- and trans-3-
methylcyclohexanol, could not be distinguished from each
other. Each gave an inseparable mixture of 3- and
4-methylcyclohexene (9 and 11), and also each gave a
small amount of 1-methylcyclohexene (8), Table 6. The
formation of 8 suggests the participation of a rearranging
carbonium ion in the decomposition and would be con-
sistent with an E1-like formation of 9 and 11. Analo-
gously (Table 7), decomposition of 5i (cis) and 5h (trans)
gave only 4-methylcyclohexene (11), accompanied, in the
case of 5i, by a small amount of amide, a result indicative
of carbonium ion involvement.
ratio:
products/
ThO
products,c mmol × 102
reactant,
run procb mmol × 102 11 amided ROHe Th ThO
8
a
b
a
c
b
d
a
b
d
a
c
b
d
5h , 10.0
9.9
0
0
0
0
2.6
1.6
1.6
0
0
0
0
f
f
f
0
0.32 10.2
0.97
1.01
0.96
1.20
1.16
1.03
0.94
0.87
0.89
0.96
0.95
0.85
0.84
9.9 0.12 9.8
9
5h , 10.4
9.8
9.6
9.1
9.0
9.6
0.21
tr
9.8
9.6
8.2
8.8
0
0
0
0
0
0.58 10.2
0.16 10.1
0.09 9.2
0.10 10.3
0.24 10.2
10
11
5i, 9.84
5i, 10.5
8.7 0.85 10.2
8.9
0
0
0
tr
9.9
0.45 10.2
0.15 10.1
0.08 9.7
0
tr
10.5
a
Each solution contained DTBMP in small excess over the
b
amount of 5. (a) Unheated solution stirred for 45 min (runs 8,
10), 70 min (runs 9, 11), and injected onto column. (b) Preceding
solution stirred with 0.3 mL of 2 M K2CO3 solution for 15 min
(run 8), 20 min (runs 9-11), and again injected onto column. (c)
Preceding solution heated at 100 °C for 30 min, cooled, and injected
onto column. (d) Preceding solution containing K2CO3 stirred for
23 h (run 9), 27 h (run 10), 28 h (run 11), and injected onto column.
c Columns used were B for 11 and A for all other products. N-
d
(4-Methylcyclohexyl)acetamide. e cis- (runs 10, 11) or trans- (runs
8, 9) 4-methylcyclohexanol. f GC peaks were small and broad and
could not be integrated.
clohexene in the relatively slow elimination reactions of
trans-2-methylcyclohexyl tosylate.17-19 The formation of
trans-2-(ethyloxy)cyclohexanol from the solvolysis of
trans-2-hydroxycyclohexyl tosylate has been attributed,
also, to neighboring-group participation (i.e., backside
attack) of the hydroxyl group in the tosylate.20 The
preferred formation of 7 from 5c (eq 6) is also under-
standable.
In contrast with 5b-i, and with 5a reported earlier,7
cyclohexyl tosylate decomposed only to a small extent
when heated in CH3CN at 100 °C for 30 min. It did not
react in CH3CN with K2CO3 at room temperature.
Cyclohexyl tosylate did decompose to cyclohexene when
injected onto the GC column, however. These conclusions
are based on experimental sequences similar to those
described in Tables 4-7, and on a pumping protocol, the
results of which are in the Experimental Section. Hatch
and co-workers heated the tosylate under reflux in DMS0
(18) Beltrame, P.; Biale, G.; Lloyd, D. J .; Parker, A. J .; Ruane, M.;
Winstein, S. J . Am. Chem. Soc. 1972, 94, 2240.
(19) Hatch, L. F.; Matar, M. S.; Garvisch, A. Indian J . Chem. 1974,
12, 583.
(20) Roberts, D. D. J . Org. Chem. 1968, 33, 168.
(21) Arhart, R. J .; Martin, J . C. J . Am. Chem. Soc. 1972, 94, 5003.
(22) Froemsdorf, D. H.; McCain, M. E. J . Am. Chem. Soc. 1965, 87,
3983.
Our work with 5d (trans ) and 5e (cis) can be compared
with reactions of cis- and trans-2-methylcyclohexanol and

8700 J . Org. Chem., Vol. 62, No. 25, 1997
Zhao et al.
14 d) either at room temperature or in the freezer until crystals
appeared in the vial, and these were filtered and washed with
ether. In the case of 5j, 10 mL of pentane was used instead
of ether, and the bottle was kept in the freezer.
(95-100 °C) for 3 h in order to complete the formation of
8 and 9. The difference between cyclohexyl tosylate and
5a -i lies in the remarkably good leaving group (ThO) in
the 5 compounds.
Th er m a l Decom p osition of 5b-i. The technique used
is illustrated with 5i. Decom p osition of 5-(cis-4-m eth ylcy-
cloh exyloxy)th ia n th r en iu m yl P er ch lor a te (5i) in MeCN.
A. With ou t Hea tin g. See run 10, Table 7. A solution of 42.2
mg (0.0984 mmol) of 5i and 34.6 mg (0.168 mmol) of DTBMP
in 5 mL of MeCN containing both naphthalene and 2-butanone
as GC standards was stirred for 45 min. GC assay with
columns A and B gave (a) 4-methylcyclohexene (11, 0.096
mmol, 97.6%), DTBMP (0.156 mmol, 92.9%), and thianthrene
5-oxide (ThO, 0.102 mmol, 103%); 4-methylcyclohexanol was
not detected. After 80 min of stirring, 0.3 mL of 2 M K2CO3
solution was injected through the septum. The mixture was
sampled for GC assay after periods of stirring, giving, (b) after
20 min, 0.0021 mmol (2.13%) of 11, 0.087 mmol (88.4%) of cis-
4-methylcyclohexanol, 0.0085 mmol (8.64%) of Th, and 0.102
mmol (104%) of ThO. After (d) 27 h, assay gave only a trace
of 11, 0.089 mmol (90.4%) of cis-4-methylcyclohexanol, a trace
of Th, and 0.099 mmol (101%) of ThO.
B. With Hea tin g. See run 11, Table 7. In a similar way,
a solution of 45.0 mg (0.105 mmol) of 5i and 34.9 mg (0.170
mmol) of DTBMP was (a) sampled for GC assay after 70 min
of stirrring. The solution was then (c) heated at 100 °C for 30
min, cooled, and again sampled for GC assay. To the cooled
solution was added 0.3 mL of K2CO3 solution, and samples
for GC assay were taken after (b) 20 min and (d) 28 h of
stirring. The GC assays gave, in mmol, for assays a, c, b, d:
11, 0.0984 (93.7%), 0.096 (91.4%), 0.082 (78.1%), and 0.088
(83.8%); DTBMP, 0.157 (92.4%), 0.164 (96.5%), 0.161 (94.7%),
and 0.174 (102%); Th, 0.0045 (4.3%), 0.0015 (1.43%), 0.0008
(0.8%), and a trace; ThO, 0.102 (97.1%), 0.101 (96.2%), 0.097
(92.4%), and 0.105 (100%). cis-4-methylcyclohexanol was not
observed in any of the assays. Small amounts of N-(4-
methylcyclohexyl)acetamide were found in the heated solution,
but were not assayed.
C. With P u m p in g Off Vola tile P r od u cts. These data
are not listed in Table 7. A solution of 48.2 mg (0.113 mmol)
of 5i and 35.4 mg (0.172 mmol) of DTBMP, containing only
2-butanone as standard, was stirred for 50 min and a sample
(a) was withdrawn for GC assay. The solution was then
evaporated at room temperature, the volatile materials being
trapped in a receiver cooled in liquid N2. The solution (b) of
volatile materials, to which naphthalene standard was added,
was used for GC assay. The solid residue left after evaporation
was dissolved in 5 mL of MeCN containing 2-butanone as
standard and (c) sampled for assay. The solution was then
heated for 30 min at 100 °C, cooled and (d) again sampled for
assay. The solution was evaporated as before and the collected
distillate (e) was used for assay after addition of naphthalene
as standard. The residue from this second evaporation was
dissolved in 5 mL of MeCN containing both naphthalene and
2-butanone for (f) GC assay. To the solution was added 0.3
mL of K2CO3 solution, and after 10 min of stirring, GC assay
(g) was carried out. DTBMP, Th, and ThO were not measured
in assays a, c, and d. Assays a-d gave for (a), 0.112 mmol
(99.1%) of 11; (b) 0.0069 mmol (6.1%) of 11, 0.039 mmol (22.7%)
of DTBMP, no Th, and a trace of ThO; (c) 0.104 mmol (92.0%),
and (d) 0.092 mmol (81.4%) of 11. Assay e gave 0.064 mmol
(56.6%) of 11 and 0.0086 mmol (5.0%) of DTBMP; (f) gave
0.0016 mmol of 11, 0.005 mmol (4.42%) of N-(4-methylcyclo-
hexyl)acetamide, 0.097 mmol (85.8%) of DTBMP, 0.0012 mmol
(1.06%) of Th, and 0.110 mmol (97.3%) of ThO. The final assay
(g) showed no 11, the same amounts of the amide and Th,
0.096 mmol (85%) of DTBMP, and 0.111 mmol (98.2%) of ThO.
A control experiment was carried out with the evaporation
of a solution of 0.0997 mmol of 11 and 0.224 mmol of
2-butanone in 5 mL of MeCN. GC assay (a) before and (b)
after evaporation gave (a) 0.0973 mmol (97.6%) and (b) 0.088
mmol (88.3%) of 11.
Exp er im en ta l Section
Acetonitrile (Eastman, HPLC grade) was dried by distilla-
tion from CaH2 and again from P2O5, each under N2. Dichlo-
romethane was dried by distillation from P2O5. cis- and trans-
Cyclohexane-1,2-diol and all authentic products except
3-methylcyclohexene (from Lancaster Syntheses, Inc.) were
from Aldrich Chemical Co. cis- and trans-3-Methylcyclohex-
anol, cis,cis- and trans,trans-3,5-dimethylcyclohexanol were
from ChemSampCo (earlier, Wiley Organics),23 while cis- and
trans-4-methylcyclohexanol were from TCI America. Quan-
titative gas chromatographic (GC) assays were made with a
Varian Associates Model 3700 instrument attached to a
Spectra-Physics Model 4290 recorder-integrator. Three col-
umns were used: A, 10% OV-101 on 80-100 mesh Chrom-
WHP, 4 ft × 1/8 in. stainless steel (ss); B, 10% Carbowax 20M
on Chrom-WHP, 6 ft × 1/8 in. ss; and C, 10% OV-17 on 80-
100 mesh Chrom-Q11, 6 ft × 1/8 in. ss. Assays were made
with the use of predetermined concentration factors for
authentic compounds, and internal standards: either naph-
thalene or biphenyl with column A, 2-butanone with column
B, and biphenyl with column C. Columns were heated and
ramped as follows: A and C, injector at 250 °C, detector at
300 °C, oven at 50 °C for 2 min and then 12 °C/min to 250 °C;
B, initially similar settings, but ramped only to 100 °C.
P r ep a r a t ion of 5b -l. Approximately 0.80 mmol of
-
Th•+ClO4 was weighed into a 50-mL flask containing a
magnetic stirrer, and to this was added 15 mL of CH2Cl2. The
flask was capped with a septum through which was added,
after 5 min of stirring, an excess (1.2-3 mmol) of the
appropriate alcohol in 5 mL of CH2Cl2. Stirring was continued
at room temperature until the color of Th•+ had disappeared,
the time for which depended on the amount of alcohol that
was used, namely 2-3 h for approximately 3 mmol and 30 h
for 1.2 mmol. The (usually yellow) solution was concentrated
to about 5 mL in a rotary evaporator and was diluted with 25
mL of dry ether. The white precipitate of 5 was filtered and
washed three times with ether. In most cases, preparation
was also successful if the reaction solution was stirred in an
ice bath rather than at room temperature. In the preparation
of 5g, all procedures were conducted in an ice bath prior to
concentration and dilution with ether, because very little of
5g could be obtained from reaction at room temperature.
Reaction of Th•+ClO4- with cis,cis-3,5-dimethylcyclohexanol in
an ice bath was complete within 140 min. However, similar
reaction with trans,trans-3,5-dimethylcyclohexanol was in-
complete even after 4 d and could be completed only after
stirring at room tempereature for 5 h. Yields (%) and mp (°C,
dec): 5b, 90, 89-90; 5c, 73, 82-85; 5d , 80, 94-96; 5e, 37,
68-70; 5f, 68, 93-96; 5g, 37, 81-82; 5h , 80, 104-107; 5i, 74,
90-91; 5j, 60, 98-100; 5k , 48, 102-104.
For obtaining single crystals for X-ray crystallography
(5b,c,i), the solid was dissolved in 2 mL of CH2Cl2 in a 5-mL
vial. The vial was placed in a larger bottle containing 10 mL
of ether, and the bottle was capped. The bottle was left (2-
(23) There is some disagreement in the literature over the identity
of these 3,5-dimethylcyclohexanols. When supplied to us, the solid
alcohol was named as all cis and the liquid alcohol as cis,trans,trans.
These assignments were consistent with data in earlier literature24 in
which the all cis isomer is said to have mp 40.0-40.3 °C, and the
trans,trans isomer to be liquid at room temperature. However, our
NMR data did not agree with the assigned configurations, in that the
solid alcohol had δ ) 4.12 (equatorial H) and the liquid alcohol had δ
) 3.61 (axial H) for the C-1 H atom. That is, the solid alcohol had to
be the trans-1,3-cis-3,5-trans-1,5 isomer (called trans,trans in the text),
and the liquid alcohol had to be the all cis isomer (called cis,cis in the
text). Our NMR data agree with those of Li and Allinger, namely δ )
4.02 for the trans,trans and 3.48 for the all cis isomer.25 Our
assignments were confirmed, also, by the NMR data for 5j,k and by
Ortep data not only for 5j, but also for the solid, trans,trans alcohol
itself (see Figure 5 in the Supporting Information).
P r ep a r a tion of Am id es. N-Substituted amides were
made by reaction of each of the methylcyclohexanols with an
excess of concentrated sulfuric acid in MeCN. It was recog-
nized that this procedure would probably give a mixture of

5-(Cyclohexyloxy)thianthreniumyl Perchlorates
J . Org. Chem., Vol. 62, No. 25, 1997 8701
cis- and trans-isomers in each case, but this was deemed
sufficient for assaying the small amounts of amide that were
formed in decomposition of some of the 5 in MeCN. The low
melting points of our products indicated indeed that they
contained a mixture of isomers, but separation of isomers was
not observed in the GC traces. Although cis- and trans-N-(2-
methyl-,26 cis- and trans-N-(3-methyl-,27 and cis- and trans-
N-(4-methylcyclohexyl)acetamide28 are known, we did not
attempt to use the tedious methods of their preparation.
Cr ysta l Str u ctu r e Deter m in a tion s. Suitable crystals
were attached to glass fibers. Data collections were performed
at -100 °C for 5b,c,i, and at ambient temperature for 5j on a
Siemens Model P4 automated diffractometer using Mo KR
radiation. Unit cell parameters were determined and refined
by a least-squares fit of 24 high angle relections. Data were
corrected for Lorentz and polarization effects, and semiem-
pirical absorption corrections based upon Ψ-scans were applied
for 5b,c,j. The space group determinations were based upon
a check of Laue symmetry and systematic absences present
and were confirmed by structure solution. The structures were
determined by direct methods followed by successive cycles of
full-matrix least squares refinement and difference Fourier
analysis using the SHELXTL-IRIS software package provided
by Siemens Analytical X-ray Instrument Inc. The parameters
refined included the atomic coordinates and anisotropic ther-
mal parameters for all non-hydrogen atoms, with the exception
of 5i and 5j, for which the thianthrene ring carbon atoms were
refined with isotropic thermal parameters. Hydrogen atoms
were placed in calculated positions but were not refined unless
otherwise specified. ORTEP29 drawings are shown with 30%
probability ellipsoids. Full details for the structure determi-
nations are available in the Supporting Information. The data
for 5i,j were weak and limited in number because of the poorer
quality of the crystals, which did not allow for anisotropic
refinement for all atoms. Consequently, the carbon atoms of
the rigid thianthrene ring were refined isotropically and a
number of ellipsoids for the remaining atoms are distorted.
Ack n ow led gm en t. We thank the Robert A. Welch
Foundation for support (Grant D-028), and Profs. An-
drew L. Ternay and J ohn N. Marx for helpful discus-
sions.
Su p p or tin g In for m a tion Ava ila ble: Figures 1-4, Ortep
diagrams of compounds 5b (Figures 1a,b), 5c (Figures 2a,b),
5i (Figures 3a,b), and 5j (Figures 4a,b). Figure 5, Ortep
diagram of trans,trans-3,5-dimethylcyclohexanol; legends to
the figures (9 pages). This material is contained in libraries
on microfiche, immediately 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.
(24) Mahmoud, B. H. J . Indian Chem. Soc. 1968, 45, 303.
(25) Li, S.; Allinger, N. L. Tetrahedron 1988, 44, 1339. The origin
of the alcohols and the NMR solvent are not given.
(26) Kornprobst, J .-M.; Laurent, A.; Laurent-Dieuzeide, L. Bull. Soc.
Chim. Fr. 1970, 1490.
(27) Hueckel, W.; Thomas, K. D. J ustus Liebig’s Ann. 1961, 645,
177.
(28) Tichy, M.; J onas, J .; Sicher, J . Collect. Czech. Chem. Commun.
1959, 24, 3434.
(29) J ohnson, C. K. ORTEPII. Report ORNL-5138; 1976. Oak Ridge
National Laboratory, TN.
J O970893M