Reactions of pinacols with one-electron oxidants

Article abstract of DOI:10.1021/jo952217v

Oxidation of the tetraarylpinacols (Ar₂COH)₂, 1a-e, in which Ar = C₆H₅ (1a), 4-ClC₆H₄ (1b), 4-MeC₆H₄ (1c), 4-MeOC₆H₄ (1d) and 4-Me₂NC₆H₄ (1e), by thianthrene cation radical (Th^?+) in CH₃-CN and in CH₂Cl₂ led quantitatively to the corresponding diaryl ketones Ar₂C=O (2a-e), provided a sufficient amount of base, 2,6-di-tert-butyl-4-methylpyridine (DTBMP), was present to prevent presumed acid-catalyzed rearrangement. In the case of 1e, continued oxidation of 2e was also observed. Oxidation of 1a by (4-BrC₆H₄)₃N^?+SbCl ₆^- and (4-BrC₆H₄)₃N^?+SbF ₆^- (Ar₃N^?+) occurred analogously. Evidence for the catalytic, cation-radical rearrangement of 1a by Ar₃N^?+ (reported in earlier literature) and by Th^?+ could not be found. Quantitative oxidation of 1a to 2a and of 1d to 2d was obtained also with NOBF₄, again provided that sufficient DTBMP was present to prevent acid-catalyzed rearrangement. Catalytic, oxidative rearrangement of 1d at room temperature and (as reported in earlier literature) at -5 °C was not observed. Oxidation was also observed of 2,3-diphenyl-2,3-butanediol (3) to acetophenone (9) and of 1,1-dimethyl-2,2-diphenylethanediol (4) to 2a and acetone by Th^?+. Oxidation of 2,3-dimethyl-2,3-butanediol (5) by Th^?+ was not observed. Instead, even in the presence of DTBMP, pinacolone (10) and tetramethyloxirane (11) were formed, through, it is proposed, a mechanism involving complexation with Th^?+.

Full text of DOI:10.1021/jo952217v

J . Org. Chem. 1996, 61, 3977-3982
3977
Rea ction s of P in a cols w ith On e-Electr on Oxid a n ts
Dong Sul Han† and Henry J . Shine*
Department of Chemistry and Biochemistry, Texas Tech University, Lubbock, Texas 79409-1601
Received December 15, 1995X
Oxidation of the tetraarylpinacols (Ar2COH)2, 1a -e, in which Ar ) C6H5 (1a ), 4-ClC6H4 (1b),
4-MeC6H4 (1c), 4-MeOC6H4 (1d ) and 4-Me2NC6H4 (1e), by thianthrene cation radical (Th•+) in CH3-
CN and in CH2Cl2 led quantitatively to the corresponding diaryl ketones Ar2CdO (2a -e), provided
a sufficient amount of base, 2,6-di-tert-butyl-4-methylpyridine (DTBMP), was present to prevent
presumed acid-catalyzed rearrangement. In the case of 1e, continued oxidation of 2e was also
observed. Oxidation of 1a by (4-BrC6H4)3N•+SbCl6 and (4-BrC6H4)3N•+SbF6 (Ar3N•+) occurred
analogously. Evidence for the catalytic, cation-radical rearrangement of 1a by Ar3N•+ (reported in
earlier literature) and by Th•+ could not be found. Quantitative oxidation of 1a to 2a and of 1d to
2d was obtained also with NOBF4, again provided that sufficient DTBMP was present to prevent
acid-catalyzed rearrangement. Catalytic, oxidative rearrangement of 1d at room temperature and
(as reported in earlier literature) at -5 °C was not observed. Oxidation was also observed of 2,3-
diphenyl-2,3-butanediol (3) to acetophenone (9) and of 1,1-dimethyl-2,2-diphenylethanediol (4) to
2a and acetone by Th•+. Oxidation of 2,3-dimethyl-2,3-butanediol (5) by Th•+ was not observed.
Instead, even in the presence of DTBMP, pinacolone (10) and tetramethyloxirane (11) were formed,
-
-
through, it is proposed, a mechanism involving complexation with Th•+
.
In tr od u ction
carried out with equimolar amounts of Th•+ClO4 and
DTBMP. It appears now that this recipe may have
contained an insufficient amount of base, so that reac-
tions attributed to oxygen-transfer to Th•+ may have
been, instead, initiated in part by acid catalysis. In
-
The oxidation of aromatic pinacols [(Ar2COH)2, 1] to
diaryl ketones (Ar2CdO, 2) by one-electron transfer
oxidants has been the subject of a number of reports in
recent years.1-5 Prominent among these reports are
those of Kochi, in which the extremely short lifetimes of
1•+ to scission have been recorded,1 and of Penn, which
describe the rates of electron transfer for and the
characteristics of oxidation of a number of 1 by iron tris-
(phenanthroline) complexes and by dichlorodicyanoben-
zoquinone (DDQ).2 Penn and co-workers have shown,
particularly, that unless precaution with the addition of
a sufficient excess of organic base is taken, the acid
liberated in the oxidation (eq 1) will catalyze the typical
rearrangement of the pinacols.2b,c
particular, reactions of some pinacols with Th•+ClO4
-
were studied in which both rearrangements and oxida-
tions were found, and the cause of the rearrangements
observed then appears now to be uncertain.7 New work
has now been carried out on reactions of Th•+ClO4 and
-
(to some extent) Th•+BF4 with the pinacols 1a (Ar )
-
C6H5), 1b (Ar ) 4-ClC6H4), 1c (Ar ) 4-MeC6H4), 1d (Ar
) 4-MeOC6H4), and 1e (Ar ) 4-Me2NC6H4). Similar work
has been carried out on reactions of Th•+ with 2,3-
diphenyl-2,3-butanediol (3), 1,1-dimethyl-2,2-diphenyl-
ethanediol (4), and 2,3-dimethyl-2,3-butanediol (5). These
works are the subject of this report.
(Ar2COH)2 - 2e- f 2Ar2CdO + 2H+
(1)
Work somewhat analogous to ours has been reported
by Lopez and co-workers8 and by Arce de Sanabia and
Carrio´n,9 that also raises the issue of oxidation versus
rearrangement. For example, reaction of benzopinacol
(1a ) with catalytic amounts (10 mol %) of tris(4-bromo-
phenyl)aminium (Ar3N•+) or tris(2,4-dibromophenyl)-
aminium (Ar′3N•+) hexachloroantimoate in CH2Cl2 is
reported to have caused the quantitative rearrangement
of 1a to phenyl triphenylmethyl ketone (6a ). The rear-
rangement was found to be retarded but not inhibited
by the presence of 10 mol % of 2,6-di-tert-butylpyridine
(DTBP). When reaction of 1a with Ar′3N•+ (50 mol %)
was carried out in CH3CN, both 2a (70%) and 6a (30%)
were formed, a result said to appear to be at variance
with observations by Penn and co-workers2c in that it was
Studies have been made in these laboratories, recently,
of the reactions of alcohols and diols with the thianthrene
cation radical (Th•+ClO4-).6,7 It was recognized in those
studies that the presence of base (2,6-di-tert-butyl-4-
methylpyridine, DTBMP) was essential to prevent acid-
catalyzed reactions. Accordingly, all reactions were
† Present address: Department of Chemistry, Mokpo National
University, Muan-Gun, Chonnam 534-729, Korea.
X Abstract published in Advance ACS Abstracts, J une 1, 1996.
(1) (a) Perrier, S.; Sankararaman, S.; Kochi, J . K. J . Chem. Soc.,
Perkin Trans. 2 1993, 825 and references therein. (b) Sankararaman,
S.; Perrier, S.; Kochi, J . K. J . Am. Chem. Soc. 1989, 111, 6448.
(2) (a) Penn, J . H.; Duncan, J . H. J . Org. Chem. 1993, 58, 2003. (b)
Penn, J . H.; Lin, Z.; Deng, D.-L. J . Am. Chem. Soc. 1991, 113, 1001.
(c) Penn, J . H.; Deng, D.-L.; Chai, K.-J . Tetrahedron Lett. 1988, 29,
3635.
thought that both 2a and 6a might arise from 1a •+ 8
In the work of Arce de Sanabia and Carrio´n,9
.
(3) Ci, X.; Whitten, D. G. J . Am. Chem. Soc. 1989, 111, 3459.
(4) Albini, A.; Mella, M. Tetrahedron 1986, 42, 6219.
(5) (a) Davis, H. F.; Das, P. K.; Reichel, L. W.; Griffin, G. W. J . Am.
Chem. Soc. 1984, 106, 6968. (b) Reichel, L. W.; Griffin, G. W.; Muller,
A. J .; Das, P. K.; Ege, S. N. Can. J . Chem. 1984, 62, 424.
(6) (a) Shine, H. J .; Yueh, W. Tetrahedron Lett. 1992, 33, 6583. (b)
Shine, H. J .; Yueh, W. J . Org. Chem. 1994, 59, 3553.
a
catalytic amount of NOBF4 (20 mol %) was found to cause
quantitative rearrangement of 1d to 6d in CH3CN at -5
(7) (a) Yueh, W. Ph.D. Dissertation, Texas Tech University, 1993.
(b) Shine, H. J . Phorphorus Sulfur Silicon 1994, 95, 429. Abstract of
a paper at the 16th International Symposium on the Organic Chem-
istry of Sulfur.
(8) Lopez, L.; Mele, G.; Mazzeo, C. J . Chem. Soc., Perkin Trans. 1
1994, 779.
(9) Arce de Sanabia, J .; Carrio´n, A. E. Tetrahedron Lett. 1993, 34,
7837.
S0022-3263(95)02217-1 CCC: $12.00 © 1996 American Chemical Society

3978 J . Org. Chem., Vol. 61, No. 12, 1996
Ta ble 1. P r od u cts (m m ol) of Rea ction s of Ben zop in a col (1a )
Han and Shine
product balanced
reactants, mmol × 102
products,c mmol × 102
run
1a
10
9.9
60
Th•+X- or other
DTBMPa
solventb
2a
6a
8a
Th
ThO
1
2
3
40e
40
40
60
61
A
B
A
B
A
B
B
A
A
B
A
B
B
B
B
B
A
A
20
20
15
30
8.2
1.7
tr
30
9.3
11
12
4.7
14
20
0.6
2.7
9.2
16
0.96
1.05
1.07
0.99
0.82
0.85
40e
30
26
35
24
22
40e
tr
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
40
60
40
20
20
10
40
20
40
75
75
75
75
40
40
40e
40e
55
37
18
41e
tr
tr
2.4e
21f
1.7
30
21
15
3.5
7.5
18
14
23
0.98
0.73
1.07
24f
8.1
33
20
tr
tr
40f
33g
40h
7.5j
7.5j
7.5k
7.5k
21.5l
20.4m
i
15
15
4.8
tr
73
tr
73
tr
39
tr
5.3
tr
i
42
10
tr
a
b
2,6-Di-tert-butyl-4-methylpyridine. A ) CH3CN, B ) CH2Cl2. c In runs 3, 4, 8, 10, 13, 15, and 17, 2a (benzophenone) was separated
by TLC and assayed by GC. In runs 12 and 17, 1a was unchanged. In other runs, 1a was consumed completely. DTBMP, when used,
was recovered (GC assay) almost completely. Ratio mmol of 2a /(mmol of Th - mmol of ThO). e X- ) ClO4
.
f X- ) BF4
.
HClO4.
d
-
- g
Bu4N+ClO4
.
All of the 1a was converted into 2a on the GC column, but only 1a (no 2a ) was found with TLC. (BrC6H4)3N•+SbCl6
.
h
k
-
i
j
-
(BrC6H4)3N•+SbF6
.
DTBMPH+ClO4-, which was added in increments over a period of time. See procedure B in the Experimental
-
l
Section. DTBMPH+ClO4-, which was added in one increment. See procedure A in the Experimental Section.
m
°C, and the result was not changed by the presence of
32 mol % of DTBP. Consequently, a catalytic chain
process was proposed in which 1d •+ was formed by
reaction of 1d with NOBF4, and lost a molecule of H2O,
to give the cation radical (7d •+) that next rearranged to
6d •+ (eq 2, An ) p-anisyl). The chain of electron transfer
reactions in which either 1a or the cation radical was
used in excess, and we must first clarify the consequences
of that difference. The products of reaction were identi-
fied and assayed by gas chromatography (GC). 1a (and
its analogs, considered later) decomposed on the GC
column that was used, to 2a . Therefore, when 1a was
used in excess, the formation of 2a from reaction with a
cation radical could not be assessed by direct GC mea-
surement. Instead, the 2a was first isolated with quan-
titative thin layer chromatography (TLC) and next
assayed with GC. When an excess of cation radical was
used, none of 1a remained for decomposition on the
column, and, therefore, direct GC measurements of 2a
could be made. In some cases of excessive amounts of
1a (e.g., runs 5 and 6 in absence of DTBMP), the 1a also
reacted completely, mainly to 6a , as shown by TLC, and,
again, direct GC analysis could be carried out.
was continued by oxidation of 1d by 6d •+ 9
The same
.
behavior was reported for 1e. This report attracted our
interest, since we had found, and we report here, that
both Th•+ClO4- and Th•+BF4- oxidized 1d cleanly to 2d .
Th•+BF4- is prepared by the oxidation of thianthrene (Th)
with NOBF4,10 so that we were surprised that NOBF4
had not also oxidized 1d to 2d cleanly. Further, the half-
life of 1d •+ is reported by Kochi and co-workers to be 66
ps, so that the likelihood that 1d •+ could competitively
and exclusively have time to lose a molecule of water and
become 7d •+ in the chain rearrangement process seemed
questionable.
Reaction of 1a with Th•+ in the presence of sufficient
DTBMP led only to benzophenone (2a ). This is seen in
runs 1-4 (Th•+ClO4 in CH3CN and CH2Cl2) and run 8
-
(Th•+BF4- in CH3CN). If DTBMP was not used (runs 5,
6, and 9) or if an insufficient amount of DTBMP was used
(run 10), all of 1a was consumed and converted mainly
into rearrangement product, phenyl triphenylmethyl
ketone (6a ). These results are analogous to those re-
ported by Penn and co-workers in the reactions of 1a with
tris(phenanthroline)iron complexes [Fe(III)L3], for
example,2b and in reactions of 1d with Fe(III)L3 and
DDQ.2a The quantitative conversion of 1a into 6a by
HClO4 is shown in run 11, while the inertness of 1a to
We report here, then, the reactions of 1a with
Ar3N•+SbCl6 and with Ar3N•+SbF6-, and of 1a and 1d
-
with NOBF4, in the absence and presence of DTBMP.
Most reactions were carried out at room temperature (25
°C) but some with NOBF4 were at -5 °C.
-
oxidation by ClO4 is reported in run 12. It is evident,
then, that unless enough base (DTBMP in Table 1) was
present, acid generated either by hydrolysis of Th•+ by
water11 in the solvent or in the early stage of oxidation
of 1a was the probable cause of rearrangement of 1a .
Penn and co-workers have reported that 2,6-di-tert-
butylpyridinium ion caused rearrangement of 1d ,2a and
Resu lts a n d Discu ssion
Rea ction s of Ben zop in a col (1a ). The results of
these reactions are summarized in Tables 1 and 3. We
will discuss the results in Table 1 first, and their
emphasis is on the reactions of 1a with cation radicals
Th•+ and Ar3N•+ in the presence and absence of the base
DTBMP. The data in Table 1 are derived also from
we found, analogously, that DTBMPH+ClO4 caused
-
rearrangement of 1a unless an excess of DTBMP was also
present (compare run 17 with run 18). We found, also,
(10) Boduszek, B.; Shine, H. J . J . Org. Chem. 1988, 53, 5142.
(11) Murata, Y.; Shine, H. J . J . Org. Chem. 1969, 34, 3368.

Reactions of Pinacols with One-Electron Oxidants
J . Org. Chem., Vol. 61, No. 12, 1996 3979
Ta ble 2. P r od u cts of Rea ction s of P in a cols 1b-e, 3, a n d 4
reactants, mmol × 102
products,c mmol × 102
run
pinacol
Th•+X- or other
DTBMPa
solventb
ketone
pinacol, recovd
Th
ThO
product balanced
1
2
3
4
5
6
7
8
9
10
11
12
13
1b, 10
1c, 9.7
1d , 10
1e, 10
1e, 10
3, 60
3, 41
3, 40
3, 40
3, 40
40e
40e
40e
40e
40e
41e
40e
40g
40g
41h
21i
19j
40g
40
40
50
50
50
60
62
60
60
21
30
20
60
A
A
B
B
A
A
A
A
B
A
A
A
A
2b, 20
2c, 19
2d , 20
2e, 12f
2e, 11f
9, 35
9, 30
9, 26
9, 17
30
29
30
37
40
37
34
32
29
11
10
10
trf
trf
3.1
4.4
9.0
12
1.05
1.00
1.00
44
23
26
30
40
17
21
30
1.03
1.01
1.13
1.00
3, 20
3, 21
4, 40
9, tr
9, tr
2a , 8.5
28
9.9
0.94
a
b
2,6-Di-tert-butyl-4-methylpyridine. A ) CH3CN, B ) CH2Cl2. c Assayed by GC in all runs except run 13 in which 2a was separated
d
by TLC before GC assay. In run 1, column was OV-17; in all other runs, OV-101. Ratio of mmol of ketone/(mmol of Th - mmol of ThO)
in runs 1-3 and 6-9. In run 13, acetone was found by GC but not assayed; the ratio, therefore, is (2 × mmol of 2a )/(mmol of Th - mmol
of ThO). e X- ) ClO4
.
f 2e was found to be oxidized also by Th•+ClO4- (see Experimental Section); hence all Th•+ was converted into Th.
-
X- ) BF4
.
Bu4N+ClO4
.
Bu4N+BF4
.
DTBMPH+ClO4
.
g
-
h
-
i
-
j
-
Ta ble 3. P r od u cts of Rea ction s of 1a a n d 1d w ith NOBF 4
in CH3CNa
that even if DTBMP was present and in excess of the
amount of DTBMPH+ClO4-, small amounts of rearrange-
ment to 6a and 8a , as well as decomposition to 2a
occurred upon injection onto the GC column. If DTB-
MPH+ was neutralized with aqueous K2CO3 before injec-
tion onto the column, the formation of 6a and 8a on the
column diminished and almost quantitative formation of
2a occurred. The data in runs 17 and 18 in Table 1 were,
in fact, obtained by GC after neutralizing the DTBMPH+
with aqueous K2CO3. The distinction between these runs
is that in run 17, having an excess of DTBMP, 1a
remained in the neutralized solution and decomposed to
2a thermally on the column. In contrast, in run 18,
having insufficient DTBMP, 1a was converted by acid
catalysts entirely into 6a before neutralization and no
1a remained in the neutralized solution for decomposition
on the column. Runs 7 and 13-16 are concerned with
“catalytic” amounts of oxidants and are related to the
report of Lopez and co-workers.8 The use of 12 mol % of
reactants, mmol × 102
temp, time,
product,b
run
°C
h
1
NOBF4 DTBMPc mmol × 102
1
2
3
4
5
6
7
8
9
25
25
25
25
25
25
25
-5
-5
25
24
24
24
24
24
24
24
6
a , 40
a , 10
a , 40
d , 40
d , 10
d , 40
d , 100
d , 40
d , 40
d , 40
40
40
8
40
41
8
21
8
8
60
60
2a , 14.5
2a , 18.5
6a , 39.5
2d , 19
2d , 17
6d , 40
2d , 17
2d , trd
6d e
60
60
32
12
6
2
10
f
6d , 40
a
b
25 mL. Because 1a and 1d decomposed to 2a and 2d ,
respectively, on the GC column, 2a and 2d were separated first
by TLC in runs 1, 4, 7, and 8 and next assayed (runs 1, 4, 7) by
GC. In runs 2 and 5, only a trace of 1 remained (TLC) in the
presence of excess of NOBF4, so assay of 2 was made by GC
d
directly. c 2,6-Di-tert-butyl-4-methylpyridine. Most of the 1d re-
mained unchanged. e The major product, by TLC. Small amounts
of 1d and 2d were found, also. f One drop of 70% HClO4; no
NOBF4.
Th•+BF4 (run 7) and 10 mol % of Ar3N•+ (runs 14 and
-
16) without DTBMP caused quantitative rearrangement
of 1a to 6a . In the presence of sufficient DTBMP,
however, conversion into 6a was prevented and was
replaced largely by oxidation (2a ), runs 13 and 15. Thus,
our results show that catalytic amounts (10 mol %) of
one-electron oxidant caused rearrangement (6a ) only if
base was not present. In the presence of base (20 mol
%, runs 13 and 15) rearrangement was suppressed and
oxidation (2a ) prevailed. The results suggest that rear-
rangement was acid-catalyzed rather than originating
from a catalytic, cation-radical reaction, induced by
Rea ction s of Tetr a a r ylp in a cols (1b-e) w ith Th •+
ClO4
-
.
An excess of Th•+ClO4 was used in order to
-
-
obviate TLC separation of products. Oxidations of 1b-d
to 2b-d took place cleanly. The data, including the
ratios 2/(Th-ThO), are given in Table 2. 1e engendered
more extensive oxidation than anticipated. Thus, only
approximately 50% of the expected 2e was obtained. All
of the Th•+ that was used was reduced to Th (runs 4 and
5). The data suggest that 2e itself was oxidized by 2
equiv of Th•+, and this was confirmed with control
experiments (see Experimental Section). Further inves-
tigation of the product(s) of oxidation of 2e was not made.
Rea ction s of 1a a n d 1d w ith NOBF 4. Reactions of
these pinacols with NOBF4 in the presence of a sufficient
amount of DTBMP ended only in oxidation to 2a and 2d ,
respectively (runs 1, 2, 4, and 5, Table 3). Whether or
not TLC separation of 2 was necessary depended on the
amount of 1 used, as explained earlier for 1a . Run 3 with
1a and run 6 with 1d are analogs, but at room temper-
ature, of experiments with 1d and 20 mol % of NOBF4
at -5 °C reported by Arce de Sanabia and Carrio´n,9 while
run 9 at -5 °C replicates their experiment with 1d . In
each case, as reported for 1d ,9 rearrangement was
quantitative. Run 7, at room temperature, and run 8 at
-5 °C with 1d correspond with the reported use of 20
Ar3N•+ 8
.
The last column in Table 1 expresses the quantitative
nature of the oxidations of 1a into 2a , best seen in runs
1-4, 8, and 10. Oxidation of 1a by Th•+ (eq 3) gives 2
1a + 2 Th•+ f 2 2a + 2 Th + 2 H+
(3)
equiv of thianthrene (Th). Reaction of Th•+ with water,
either present in the dried solvent or added in workup
as 2 M K2CO3, gives equal amounts of Th and thian-
threne 5-oxide (ThO),11 eq 4. The ratio 2a /(Th-ThO)
2Th•+ + H2O f Th + ThO + 2 H+
(4)
expresses, therefore, the stoichiometric result of eq 3, and
ideally should be 1.0.

3980 J . Org. Chem., Vol. 61, No. 12, 1996
Han and Shine
Ta ble 4. P r od u cts of Rea ction s of Tetr a p h en yloxir a n e
Ta ble 5. P r od u cts of Rea ction s of
(8a ) w ith Th •+ClO4 in CH3CNa
2,3-Dim eth yl-2,3-bu ta n ed iol (5) w ith Th •+ClO4 a n d
-
-
Th •+BF 4
-
reactants, mmol × 102
products,b mmol × 102
reactants, mmol × 102
products,b mmol × 102
run 8a Th•+ClO4
DTBMP 6a recovd 8a Th ThO
40 10 20 20
10 9.8
-
run
5
Th•+X- DTBMP solventa recovd 5 10f 11g Th ThO
1
2
10
20
40
20
1
2
3
4
63
61
52
61
40c
41c
40d
40c
60
61
60
61
A
B
54
42
39
59
3
2
2
3
23 17
20
11 23 19
14 21 19
19 20
a
b
25 mL. Assayed by GC, OV-101 column.
B
Be
mol % of NOBF4 and 32 mol % of base,9 but whereas the
earlier authors obtained 6d in these conditions we
obtained only 2d . Last, run 10 reports the quantitative
rearrangement of 1d at room temperature with a small
amount (unmeasured) of HClO4, a result that is similar
to that with 1a , run 11, Table 1. Arce de Sanabia and
Carrio´n found 1d to be inert to 20 mol % of H2SO4 under
their conditions of use. The difference between our
a
b
10 mL; A ) CH3CN, B ) CH2Cl2. Assayed by GC on SE-54
capillary column. c X- ) ClO4
.
X- ) BF4 e Saturated with
.
-
d
-
water. f Pinacolone. Tetramethyloxirane.
g
Sch em e 1
9
results with HClO4 and the earlier results with H2SO4
may have occurred because of the difference in temper-
atures (25 °C vs -5 °C) that were used. The variations
in efficiencies of proton-catalyzed pinacol rearrangements
have been noted elsewhere, too.12 Our results indicate
that 1a and 1d are oxidized but not rearranged by
reaction with NOBF4. Bearing in mind that Th is readily
oxidized to Th•+ by NOBF4 and that Th•+ oxidized 1a
10
and 1d to 2a and 2d , it is not surprising to find that these
pinacols are oxidized by NOBF4. The rearrangements
that have been reported here (in the absence of base) and
earlier appear to be acid catalyzed, the acid being
generated, perhaps, by small amounts of oxidation, eq
1.
Rea ction s of P in a cols 3 a n d 4. 2,3-Diphenyl-2,3-
butanediol (3) did not decompose on our OV-101 column
so that GC assay of products could be used directly.
Oxidations by Th•+ in the presence of DTBMP were
quantitative (runs 6-9, Table 2). The pinacol was
ThO, Scheme 1. Because equal amounts of Th and ThO
are formed also from reaction of Th•+ with water, the
yields of Th and ThO in Table 5 should be the same. This
is seen in runs 3 and 4 but there are discrepancies in
runs 1 and 2. The balance between yields of 10 and 11
and yields of Th and ThO is reasonable in runs 2 and 3,
in which CH2Cl2 was the solvent, but is not at all close
in run 1, with CH3CN, the reason for which is not known.
The pinacols we have used have high oxidation poten-
tials2 so that oxidation by Th•+, Ar3N•+, and NO+ (as well
affected little or not at all by Bu4N+ClO4-, Bu4N+BF4
,
-
and DTBMPH+ClO4 in the presence of DTBMP (runs
10-12). Thus 3 behaved similarly to 1a -1d . Pinacol 4
was also oxidized cleanly by Th•+BF4- in the presence of
DTBMP (run 13, Table 2).
-
Table 1 shows that small amounts (traces, tr) of
tetraphenyloxirane (8a ) were obtained in some runs
alongside small amounts of the rearrangement product
6a . We were interested to know if 8a was reactive
toward Th•+, and found (Table 4) that it was unreactive
if base was present. In the absence of base, 8a was
converted into 6a , presumably by acid catalysis.
Rea ction s of 2,3-Dim eth yl-2,3-bu ta n ed iol (5). In
the presence of a large amount of DTBMP, 5 reacted with
Th•+ in dry CH3CN and dry CH2Cl2 to give pinacolone
(10) and tetramethyloxirane (11), Table 5. In CH2Cl2
that had been saturated with water, and, again, in the
presence of DTBMP, 5 was recovered, and only the
products of reaction of Th•+ with water, Th and ThO, were
found. The results in Table 5 suggest, then, that 10 and
11 (runs 1-3) were not formed by acid catalysis, but by
direct reaction with Th•+. A proposal for that reaction is
shown in Scheme 1. The supposition here is that 5
cannot be oxidized by Th•+ and behaves, instead, in the
way that we have proposed for alcohols which also have
high oxidation potentials. In principle, the sum of the
yields of 10 and 11 should equal each of those of Th and
as by Fe(III)L3 ) would not have been expected. Penn2a,b
2
has attributed ease of oxidation to interaction of aryl
rings in both the 1,1- and 1,2-configuration, and the
generality of these oxidations supports that idea. Insofar
as 5 is concerned, oxidation is not promoted, and bonding
at sulfur (12) is an alternative that leads to 10 and 11.
Exp er im en ta l Section
Acetonitrile was refluxed over and distilled from P2O5 and
again from CaH2, each under N2; with dichloromethane, only
P2O5 was used. 2,6-Di-tert-butyl-4-methylpyridine (DTBMP),
from Aldrich, was dried under vacuum. Thin layer chromato-
graphic (TLC) separations for later quantitative assays (by GC)
were made with Analtech Uniplate plates. Quantitative gas
chromatographic (GC) assays were made with
a Varian
Associates Model 3700 instrument attached to a Spectra
Physics Model 4270 or 4290 integrator. The columns used
were 10% OV-101 on 80-100 mesh Chrom WHP, 4 ft × 1/8 in
stainless steel (ss), with injector at 250 °C, detector at 300 °C,
and oven temperature held at 50 °C for 2 min and ramped at
10 °C/min to 250 °C; a 10% OV-17 on 80-100 mesh Chrom Q
II, 6 ft × 1/8 in. ss, used similarly; and an SE-54 capillary, 30
m × 0.25 mm, with injector and detector temperatures at 250
(12) Collins, C. J . Quart. Rev. 1960, 14, 357. See also Collins, C.
J .; Eastham, J . F. In The Chemistry of Functional Groups: The
Carbonyl Group; Patai, S., Ed.; Wiley Interscience: New York, 1966;
Vol. 1, Ch. 15, pp 763, 764.

Reactions of Pinacols with One-Electron Oxidants
J . Org. Chem., Vol. 61, No. 12, 1996 3981
°C, and oven temperature held at 36 °C for 7 min and ramped
12 °C/min to 250 °C. Column use is noted in the tables.
Thianthrene (Th), from Fluka, was purified by column
chromatography, petroleum ether eluent, and was crystallized
from acetone. Thianthrene cation radical perchlorate
procedure was followed with some modification where needed
with other reactants listed in the tables. 1a (220 mg, 0.60
mmol), Th•+ClO4- (126 mg, 0.40 mmol), and DTBMP (123 mg,
0.60 mmol) were placed in a 25 mL volumetric flask, containing
a stirrer magnet bar, and the flask was purged with dry N2
after capping with a septum. Dry CH3CN was introduced to
the mark by syringe, and the mixture was stirred for 3 h, by
which time the color of Th•+ has disappeared and the solution
was pale yellow. Next, 0.2 mL of aqueous 2 M K2CO3 was
injected into the flask and stirring was continued for 1 h. GC
analysis (OV-101), after adding naphthalene as internal
standard, showed Th, ThO, and only 2a as products, and that
no 1a remained. The Th and ThO were assayed (GC). A 10
mL portion of the solution was concentrated under vacuum to
small volume, and this was streaked on a preparative TLC
plate. Development with hexane/ethyl acetate (15:1) showed
that both 1a and 2a were present. The band containing 2a
was removed, and the 2a was extracted with CH2Cl2 for GC
analysis, with naphthalene as an internal standard. The
result is listed in Table 1. When an excess of oxidant was used
in the presence of DTBMP, none of the pinacol remained (as
shown by TLC spotting) for later decomposition into 2a on the
GC column (e.g., runs 1 and 2, Table 1). Therefore, GC
analysis was carried out without intervention of TLC separa-
tion. When an oxidant was used in the absence of DTBMP
(e.g., runs 5-7, 9, Table 1) much of the pinacol underwent acid-
catalyzed reaction and little or none remained (TLC spotting)
for later decomposition on the column. Therefore, GC analysis
was again carried out without prior use of TLC separation.
Reactions with other solid oxidants and with other pinacols
were carried out in much the same way. When HClO4 was
used with 1a (run 11, Table 1), 0.08 mL of 70% HClO4 was
injected into the volumetric flask after the solvent had been
added. Monitoring by GC showed the formation of 91% of 6a
after 10 min, 96% after 1 h, and complete conversion into 6a
after 24 h. When (BrC6H4)3N•+ was used as oxidant, workup
with 0.5 mL (instead of 0.2 mL) of 2 M K2CO3 was used.
Initially, reactant solutions were stirred overnight (24 h), e.g.,
runs 1, 2, 5, 9, 10 of Table 1. Later, shorter times (1-6 h) of
stirring before workup were used, as judged by the disappear-
ance of cation radical color. Solutions were stirred for ap-
proximately 1 h after adding aqueous K2CO3 and before GC
analysis was carried out. Most reactions were repeated two
(Th•+ClO4-, warning)11 and tetrafluoroborate (Th•+BF4- 10 were
)
prepared and assayed as described earlier. Tris(4-bromophen-
-
13
yl)aminium hexafluoroantimonate (TBPA•+SbF6
)
and
-
14
hexachloroantimonate (TBPA•+SbCl6
)
were prepared as
described from TBPA.15 2,3-Dimethylbutane-2,3-diol (pinacol,
5) was dried as described.16 2,3-Dimethyl-2-butene, benzophe-
none (2a ), 4,4′-dichloro- (2b), 4,4′-dimethoxy- (2d ), and 4,4′-
bis(dimethylamino)benzophenone (2e) were from Aldrich. 4,4′-
Dimethylbenzophenone (2c) was from TCI America. Benzo-
pinacol (1,1,2,2-tetraphenyl-1,2-ethanediol, 1a ) was prepared
by irradiation of 2a in isopropyl alcohol,2b mp 188-190 °C
(crystallized from benzene/90-100 °C ligroin); lit.2b mp 185-
189 °C. 1,1,2,2-Tetr a k is(4-ch lor op h en yl)-1,2-eth a n ed iol
(1b) was prepared by reduction of 2b with Zn/ZnCl2,17 mp 173-
175 °C (crystallized from 90-100 °C ligroin); lit.18 mp 173-
174 °C. 1,1,2,2-Tetr a k is(4-m eth ylp h en yl)-1,2-eth a n ed iol
(1c) was prepared from 2c as with 1b, mp 179-181 °C
(crystallized from chloroform/ethanol); lit.2b mp 174-175 °C.
1,1,2,2-Tetr akis(4-m eth oxyph en yl)-1,2-eth an ediol (1d) was
prepared from 2d as with 1a , mp 180-182 °C (crystallized
from hexane/ethylacetate); lit.18 mp 183 °C. 1,1,2,2-Tetr a k is-
[4-(d im eth yla m in o)p h en yl]-1,2-eth a n ed iol (1e) was pre-
pared by reduction of 2e with Mg/I2,19 mp 195-197 °C
(crystallized from chloroform/ethanol and benzene/ethanol);
lit.19 mp 196-197 °C. m eso-2,3-Dip h en yl-2,3-b u t a n ed iol
(3), mp 116-117 °C (crystallized from hexane/ethanol) was
prepared as described;20 lit.20 mp 116.2-117.8 °C; 1,1-d i-
m et h yl-2,2-d ip h en yl-1,2-et h a n ed iol (4) was prepared as
described by Parry.21 The product was not, however, steam
distilled to remove bromobenzene and biphenyl. Instead, the
ether solution from the Grignard reaction was washed with
cold 0.1 M H2SO4 and with water several times and was dried
over MgSO4. The dried solution was concentrated under
reduced pressure, and to the concentrate was added hexane
to precipitate crude 4, the bromobenzene and biphenyl re-
maining in solution. The 4 was crystallized several times from
benzene/petroleum ether, mp 90-90.5 °C. Literature21 mp
89-89.5 °C. P h en yl tr ip h en ylm eth yl k eton e (benzopina-
colone, 6a ),22 mp 185-187 °C (crystallized from benzene/
ligroin), lit.22 mp 179-180 °C; and pinacolone,23 bp 107-108
°C were prepared as described in the literature. Tetr a p h en -
yloxir a n e (8a ) was prepared by oxidation of tetraphen-
ylethene,24 which itself was prepared by reduction and rear-
rangement of 6a 25 and had mp 211-213 °C (crystallized from
ethyl acetate); lit.24 mp 209.7-210.2 °C. Tetr a m eth ylox-
ir a n e (11) was prepared from tetramethylethene as de-
scribed.26
or three times, and the results tabulated are averages.
Rea ction of Ben zop in a col (1a ) w ith DTBMP H+ClO4
-
.
A: A solution of 147 mg (0.402 mmol) of 1a , 20.5 mg (0.100
mmol) of DTBMP, and 62.3 mg (0.204 mmol) of DTBM-
-
PH+ClO4 in 25 mL of CH3CN was stirred for 50 min. GC
analysis gave 0.015 mmol of 2a , 0.346 mmol of 6a , and 0.040
mmol of 8a . After a total of 4 h the solution was neutralized
with 0.2 M K2CO3. GC analysis gave 0.387 mmol of 6a (96%);
2a and 8a were not found.
Rea ction s of Ben zop in a col (1a ). A typical procedure
B: A solution of 147 mg (0.402 mmol) of 1a in 25 mL of CH3-
CN, containing naphthalene as an internal standard, was kept
for 40 min, and an aliquot was injected onto the OV-101
column. Almost all of the 1a was converted into 2a (0.776
mmol, 97%). Small amounts of 6a (0.008 mmol) and 8a (0.007
mmol) were obtained, too. The analysis was repeated after 3
h with a similar result. Then, 85 mg (0.415 mmol) of DTBMP
was added to the solution and GC analysis was repeated,
giving 92% of 2a , 0.006 mmol of 6a , and 0.007 mmol of 8a . To
the solution was next added four increments of 15-16 mg
(e.g., run 3, Table I) for reaction with Th•+ClO4- is given. The
(13) Engel, P. S.; Robertson, D. T.; Scholz, J . N.; Shine, H. J . J . Org.
Chem. 1992, 57, 6178.
(14) Bell, F. A.; Ledwith, A.; Sherrington, D. C. J . Chem. Soc. 1969,
2719.
(15) Baker, T. N., III; Doherty, W. P., J r.; Kelley, W. S.; Newmeyer,
W.; Rogers, J . E., J r.; Spalding, R. E.; Walter, R. I. J . Org. Chem. 1965,
30, 3714.
(16) Furmiss, B. S.; Hannaford, A. J .; Rogers, V.; Smith, P. W. G.;
Tatchell, A. R. Vogel’s Textbook of Practical Organic Chemistry; 4th
ed.; Longman: London, 1978; p 359.
-
(approximately 0.05 mmol) of DTBMPH+ClO4 over a period
of 72 h and after each addition GC analysis was carried out.
The amount of 2a decreased with each incremental addition
(to 0.640 mmol, 80%) while the amounts of 6a and 8a increased
to, finally, 0.034 mmol (8.5%) and 0.064 mmol (16%), respec-
tively. After 75 h, 0.2 mL of 2 M K2CO3 was added to the
solution, and GC analysis gave 0.820 mmol (102%) of 2a , with
0.003 mmol of 6a and 0.004 mmol of 8a . The summation of
(17) Tanaka, K.; Kishigami, S.; Toda, F. J . Org. Chem. 1990, 55,
2981.
(18) Depovere, P.; Devis, R. Bull. Soc. Chim. Fr. 1968, 2470.
(19) Gomberg, M.; Bachmann, W. E. J . Am. Chem. Soc. 1927, 49,
236.
(20) Cram, D. J .; Kopecky, K. R. J . Am. Chem. Soc. 1959, 81, 2748.
(21) Parry, W. J . Chem. Soc. 1911, 99, 1169.
(22) Bachman, W. E. In Organic Syntheses; Blatt, A. H., Ed.;
Wiley: New York, 1943; Coll. Vol. II, p 73.
-
additions of DTBMPH+ClO4 and the analysis after workup
(23) Swinehart, J . S. Organic Chemistry, An Experimental Approach;
Appleton Century Crofts: New York, 1969; p 450.
(24) Cope, A. C.; Trumbull, P. A.; Trumbull, E. R. J . Am. Chem.
Soc. 1958, 80, 2844.
(25) Bachman, W. E. J . Am. Chem. Soc. 1934, 56, 449.
(26) Price, C. C.; Carmelite, D. D. J . Am. Chem. Soc. 1966, 88, 4039.
with K2CO3 is given as run 17, Table 1.
-
Rea ction s of Tetr a p h en yloxir a n e (8a ) w ith Th •+ClO4
.
8a (70 mg, 0.20 mmol) and Th•+ClO4- (63 mg, 0.20 mmol) were
placed with 25 mL of CH3CN in a septum-capped volumetric
flask as described. After 30 min, 0.2 mL of 2 M K2CO3 solution

3982 J . Org. Chem., Vol. 61, No. 12, 1996
Han and Shine
-
was added which discharged the color of Th•+ immediately.
One hour later GC analysis showed quantitative conversion
into 6a . The result in Table 4, run 2, is an average of three
experiments. The reaction was repeated with 10 mmol of 8a ,
40 mmol of Th•+ClO4-, and 40 mmol of DTBMP. Stirring was
continued for 24 h, by which time the color of Th•+ had
disappeared. One hour after addition of 0.2 mL of 2 M K2-
CO3, GC analysis showed none of 6a ; only 8a remained, run
1, Table 4 (two experiments). When this reaction was repeated
but with a lesser amount of DTBMP, periodic GC monitoring
showed the gradual conversion of 8a into 6a , while the color
of Th•+ persisted. The conversion into 6a was complete after
5 h.
Rea ction s of 1b-e w ith Th •+ClO4
.
In each case (runs
1-5, Table 2), the reactants and DTBMP were placed in a 25
mL volumetric flask, containing a stirrer bar, under N2, and
the solvent was added by syringe through the septum. Stirring
was continued for 34 h (1b), 24 h (1c,e), 20 h (1d ), before
adding 0.2 mL of 2 M K2CO3, after which stirring was
continued for 1 h. With 1b,c, the color of Th•+ faded only
slowly, becoming pale yellow on addition of K2CO3. With 1d ,
the color of Th•+ disappeared within 2 h. With 1e, the color
of the solution became deep blue (CH2Cl2) and blue-green (CH3-
CN) immediately, and the color persisted until K2CO3 was
added, whereupon it changed to reddish brown.
Rea ction of 2e w ith Th •+ClO4- in CH3CN. A solution of
Further control tests showed that when a solution of 26.5
mg (0.076 mmol) of 8a , 83.2 mg (0.406 mmol) of DTBMP, and
-
54 mg (0.20 mmol) of 2e, 126 mg (0.40 mmol) of Th•+ClO4
,
-
63 mg (0.206 mmol) of DTBMPH+ClO4 in 25 mL of CH3CN
and 103 mg (0.50 mmol) of DTBMP was prepared in the usual
way. The purple color of the solution (Th•+) changed to reddish
brown overnight, and this color persisted after adding K2CO3
solution. GC analysis gave (average of three experiments) 11.8
mmol of 2e, 38.4 mmol of Th, and a trace of ThO. The GC
trace contained also unidentifiable peaks.
was kept for 2 h and used for GC analysis, 0.68 mmol of 8a
and 0.007 mmol of 6a were found. Yet, if 0.2 mL of 2 M K2-
CO3 was injected 1 h prior to GC analysis, 8a (0.076 mmol)
was fully recovered.
Rea ction s of 1a a n d 1d w ith NOBF 4. A solution of 146
mg (0.40 mmol) of 1a , 48 mg (0.40 mmol) of NOBF4, and 123
mg (0.60 mmol) of DTBMP in 25 mL of CH3CN was prepared
in the usual way. Within a few minutes the color of the
solution became pale green, but was colorless after a few hours.
The solution was kept for 24 h, 0.2 mL of 2 M K2CO3 was added
and an aliquot of 10 mL was concentrated under reduced
pressure and subjected to TLC separation. Only 2a was found
and assayed, run 1, Table 3. This reaction took place at room
temperature and was performed twice.
Rea ction s of 3 a n d 4. The results of these reactions are
listed in Table 2. For the most part, reactions of 3 were carried
out in the standard way and in 25 mL of solvent. Reactions
with Th•+ (runs 6-9) were rapid, the color of Th•+ disappearing
within minutes. These runs were worked up (K2CO3) within
0.5-6.0 h. In run 8, a solution of Th•+BF4- in 5 mL of CH3CN
was added slowly, dropwise to a solution of 3 and DTBMP in
20 mL of CH3CN. In runs 11 and 12 only 10 mL of solvent
was used. Stirring in runs 10-12 was continued for several
hours before addition of K2CO3. GC analysis before addition
of K2CO3 (run 11) gave 0.154 mmol of 3, 0.023 mmol of 3,3-
diphenylbutanone, and traces of acetophenone (9) and oxirane.
The result in Table 2 is for GC after addition of K2CO3. In
A similar reaction was performed (twice) with an excess of
NOBF4 (run 2, Table 3). TLC showed that only a trace of 1a
remained and, therefore, GC analysis was carried out without
TLC separation. No 6a and 8a were found.
In run 3, with no DTBMP (two expts), Table 3, a “catalytic”
amount (20 mol %) of NOBF4 was used, the ratio of 1a /NOBF4
being that used by Arce de Sanabia and Carrio´n, but here at
25 °C rather than -5 °C.9 GC monitoring after 2 h showed
6a and a small amount of 8a , but after 6 h only 6a was found.
Analysis by GC after 24 h gave a quantitative yield of 6a .
The same series of experiments was performed with 1d ,
runs 4-6, Table 3, with similar results. In run 7, the ratio of
reactants is that used by Arce de Sanabia and Carrio´n, but
on a larger scale, and at room temperature rather than at -5
°C. Runs 8 and 9, analogous to 6 and 7, but at -5 °C, were
carried out by cooling a solution of 1d with (run 8) or without
(run 9) DTBMP in 20 mL of CH3CN, and injecting into it 4.2
mL (0.08 mmol) of a precooled, 19 mM solution of NOBF4.
-
-
reaction of 4 with Th•+BF4 (run 13), a solution of Th•+BF4
in 10 mL of CH3CN was added slowly, dropwise, through the
septum to the stirred solution of 4 and DTBMP in 15 mL of
CH3CN. Control experiments showed that rapid addition of
-
Th•+BF4 solution caused some rearrangement of 4.
Ack n ow led gm en t. We thank the Robert A. Welch
Foundation for support, Grant D-028. D.S.H. thanks
Mokpo National University for leave of absence and
support.
J O952217V