The decomposition of diazo-compounds induced by nucleophiles. The decomposition of 9-diazofluorene in the presence of hydroxide or alkoxide ions

Article abstract of DOI:10.1039/b211139j

9-Diazofluorene, on treatment with stoichiometric or substoichiometric amounts of quaternary ammonium hydroxide or methoxide or of potassium t-butoxide in solution in aqueous or alcoholic dimethyl sulfoxide or acetonitrile at 30°C, decomposes with evolution of nitrogen to yield fluorenone azine [bis(fluorenylidene)hydrazine] in almost quantitative yield. Studies are reported of the identity of the minor by-products, together with an examination of the kinetics of the reactions and additional spectroscopic experiments. The general rate equation is v = k[FIN₂]^3/2[Nu^-]^1/2, where Nu^- represents the nucleophile. It is concluded that the oxyanions, most probably after nucleophilic attack on 9-diazofluorene under these basic conditions to generate a new anionic species, are capable of transferring an electron to the diazo-compound. This initiates decomposition of further diazo-molecules in a process of electron-transfer chain catalysis that bears some similarities to, but has also some differences from, that encountered using cathodic initiation. In support of this interpretation it is found that reactions can be accelerated by the introduction of carbon acids (fluorene, 9-phenylfluorene) into the reaction mixtures. The nature of the initiation, propagation and termination steps of the chain mechanism are discussed.

Full text of DOI:10.1039/b211139j

View Article Online / Journal Homepage / Table of Contents for this issue
The decomposition of diazo-compounds induced by nucleophiles.
The decomposition of 9-diazofluorene in the presence of hydroxide
or alkoxide ions
Linda J. McDowell, M. Mehdi Khodaei† and Donald Bethell*
Department of Chemistry, The University of Liverpool, Liverpool, UK L69 3BX
Received 13th November 2002, Accepted 30th January 2003
First published as an Advance Article on the web 25th February 2003
9
-Diazofluorene, on treatment with stoichiometric or substoichiometric amounts of quaternary ammonium
hydroxide or methoxide or of potassium t-butoxide in solution in aqueous or alcoholic dimethyl sulfoxide or
acetonitrile at 30 ЊC, decomposes with evolution of nitrogen to yield fluorenone azine [bis(fluorenylidene)hydrazine]
in almost quantitative yield. Studies are reported of the identity of the minor by-products, together with an
examination of the kinetics of the reactions and additional spectroscopic experiments. The general rate equation is
3/2
Ϫ 1/2
Ϫ
v = k[FlN ] [Nu ] , where Nu represents the nucleophile. It is concluded that the oxyanions, most probably after
2
nucleophilic attack on 9-diazofluorene under these basic conditions to generate a new anionic species, are capable
of transferring an electron to the diazo-compound. This initiates decomposition of further diazo-molecules in a
process of electron-transfer chain catalysis that bears some similarities to, but has also some differences from, that
encountered using cathodic initiation. In support of this interpretation it is found that reactions can be accelerated
by the introduction of carbon acids (fluorene, 9-phenylfluorene) into the reaction mixtures. The nature of the
initiation, propagation and termination steps of the chain mechanism are discussed
Introduction
solutions of quaternary ammonium hydroxide or methoxide
and of potassium t-butoxide usually in dimethyl sulfoxide con-
taining water, methanol or t-butyl alcohol respectively. A few
experiments were also conducted in acetonitrile for comparison
with the earlier electrochemical studies. In the course of the
work it became necessary to re-examine the role of carbanions
as part of a chain initiation process. Accordingly the kinetic
effect of introducing carbon acids such as fluorene and
1
In 1953 Beringer and co-workers reported that 9-diazofluor-
ene, on treatment with fluorenyl lithium in tetrahydrofuran, was
converted into fluorenone azine [bis(fluorenylidene)hydrazine]
which they believed arose by simple nucleophilic attack of the
carbanion on the terminal nitrogen atom of the diazo-com-
pound. This interpretation was later revived by McDonald and
2
Lin. It is well known that phosphorus-centred nucleophiles do
9
-phenylfluorene into the highly basic reaction mixtures was
indeed react in this way, forming stable adducts (phospha-
investigated.
3
zines), but the carbanion adducts are negatively charged and
cannot yield the azine without loss of a negatively charged
entity (e.g. electron, hydride ion). This prompted publication of
an alternative view that the interaction was not one between
nucleophile and electrophile, but rather involved initial electron
Results
Reactions in the presence of tetraethylammonium hydroxide
4
transfer from the carbanion to the diazo-compound. Thus
Treatment in an atmosphere of oxygen-free nitrogen of 9-
diazofluorene in solution in either dimethyl sulfoxide or
acetonitrile with a small amount of a 20% aqueous solution
of tetraethylammonium hydroxide at 30 ЊC led to formation of
a deep purple colouration (λmax = 570 nm) and smooth and
rapid evolution of nitrogen, accompanied by precipitation in
high yield of fluorenone azine. Essentially complete conversion
could be achieved with as little as 0.5 mol% hydroxide with
respect to the diazo-compound; a small amount of the diazo-
compound was trapped unchanged within particles of the azine
during the precipitation. The progress of the reaction could be
followed either by infrared spectrophotometry of the diazo-
complete decomposition of the diazo-compound could be
induced by substoichiometric quantities of the carbanion,
although labelling revealed that the carbanion could be
incorporated to some extent into the azine product. Sub-
5
sequently, Handoo et al. presented additional evidence,
including EPR results that indicated that odd-electron species
were generated when 9-diazofluorene was treated with meth-
anolic KOH in DMSO solution. It had been established that
conversion of 9-diazofluorene into fluorenone azine could also
6
be initiated by cathodic reduction in acetonitrile solution in
a process often referred to as electron transfer chain (ETC)
7
Ϫ1
catalysis. The kinetics of this process in a stirred reactor have
compound at 2050 cm on samples of the reaction mixture, or,
8
been studied; while it proved possible to establish the kinetic
more conveniently by determining the increase in pressure in
the closed reaction vessel arising from the nitrogen evolution;
the results were kinetically equivalent. At a given hydroxide
concentration, reactions in DMSO were almost an order of
magnitude faster than those in acetonitrile. The reaction
products when nitrogen evolution ceased were determined
quantitatively by hplc (Table 1) and indicated that in general
more than 95% of the diazofluorene had been converted
into the azine. Pressure changes in the system were consistent
with this. In addition to the azine, small amounts of fluorene
and a few of its derivatives (9-fluorenone, its hydrazone and
behaviour in terms of ETC catalysis, the identity of the chain
carrier involved in the critical electron transfer could not be
determined unambiguously. Although the transformation of
9
-diazofluorene into the azine is of limited synthetical
importance, the reaction provides a valuable test system
for elucidating the sorts of reactive intermediates that can be
generated from diazoalkanes or diazoketones in the presence of
nucleophiles. Accordingly we report here on kinetic studies
of the decomposition of 9-diazofluorene brought about by
9
-fluorenol) were also present, especially when relatively high
†
On leave of absence from the Department of Chemistry, Razi Uni-
versity, Kermanshah, Iran.
hydroxide concentrations were used.
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 3
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 9 9 5 – 1 0 0 3
995

View Article Online
a
Table 1 Products of the reaction of 9-diazofluorene with tetraethylammonium hydroxide at 30 ЊC
b
b
Solvent
DMSO
1.0
CH CN
3
3
1
0 [Et NOH]/M
2.5
5.0
10.0
2.5
5.0
10.0
4
c
Products (%)
Fl᎐N–N᎐Fl
Fl᎐N–NH2
Fl᎐O
FlH–OH
FlH2
96.1
0.1
0.3
1.0
0.1
95.9
0.3
1.1
0.5
1.3
95.8
0.2
0.8
1.5
0.6
94.2
1.6
3.1
2.0
1.5
99.1
nd
0.9
nd
nd
99.8
97.5
nd
2.3
nd
0.1
99.3
96.4
0.2
2.8
0.3
0.2
FlN Conversion(%)
97.9
97.5
98.9
99.0
99.1
2
a
b
c
Fl = 9-fluorenylidene; nd = not detected. Containing 1.0 M water. Percentage of 9-diazofluorene (initially 0.05 M) converted into each product.
The kinetics of the reaction in both DMSO and in
acetonitrile were followed at 30 ЊC by pressure change in a
stirred, closed system incorporating a pressure transducer.
Some typical pressure time curves are shown in Fig. 1 to
illustrate their form, the dependence of rate on hydroxide con-
centration and the relative rates in DMSO and acetonitrile
solutions. Addition of a second aliquot of 9-diazofluorene to a
reaction mixture in which decomposition of a first aliquot was
complete gave a pressure/time curve that was indistinguishable
from the first provided that the solutions were not exposed to
air. It was also established that increase of the concentration
of water in the reaction mixture at constant hydroxide con-
centration led to a roughly proportionate increase in the time
to 50% reaction. In most experiments the water concentration
was held at 1.0 M.
Fig. 2 Dependence of 3/2-order rate coefficients for the decomposition
of 9-diazofluorene at 30 ЊC on the hydroxide or alkoxide concentration.
concentration of quaternary ammonium hydroxide suggests an
apparent order in hydroxide ion close to unity in DMSO, the
results fit into a broader pattern of results with other bases and
over a wider concentration range that suggest an order closer to
1
/2 (see below).
The results in Table 2 indicate that the rate of reaction is
unaffected by the presence of fluorenone azine (at saturation
level at the start of reaction) or by fluorenone. The presence
of p-dinitrobenzene (at a concentration one half of that of the
hydroxide), however, caused inhibition of the decomposition;
this was total at low hydroxide concentration, but led to a
halving of the observed rate coefficient at a higher base concen-
tration. The presence of diazodiphenylmethane in the reaction
mixture had no effect on the decomposition of 9-diazofluorene;
the additional diazo-compound was unchanged after all the
diazofluorene had been transformed into the azine and the
reaction rate was unchanged. This finding is analogous to our
previous observation on the reaction of diazofluorene and
diazodiphenylmethane with potassium t-butoxide in t-butyl
Fig. 1 Pressure/time curves for the decomposition of 9-diazofluorene
(
50 mM) induced by tetraethylammonium hydroxide in 1 M aqueous
DMSO or CH CN at 30 ЊC. Hydroxide concentrations: ᭿ 5 mM; ꢀ 2.5
3
mM; ᮀ 1 mM in DMSO; ꢀ 2.5 mM in CH CN.
3
Curves of the type shown in Fig. 1 were used to determine
the reaction order in 9-diazofluorene. It was evident by inspec-
tion that the order was higher than one since successive half-
lives of pressure build-up during each experiment were not
constant, and, in separate experiments, values increased with
decreasing initial concentration of the diazo-compound. The
inference was confirmed by the curvature of plots of ln (p Ϫ p )/
4
alcohol/DMSO (9:1 v/v).
Reactions in the presence of benzyltrimethylammonium
methoxide or potassium t-butoxide in DMSO solution
Treatment of 9-diazofluorene in DMSO solution with sub-
stoichiometric amounts of benzyltrimethylammonium meth-
oxide (as a solution in methanol) or potassium t-butoxide
(as a solution in t-butyl alcohol) in an atmosphere of nitrogen
at 30 ЊC led to transformation of the diazo-compound into
fluorenone azine in high (>90%) yield (Table 3). By-products
were 9-fluorenone together, in the case of the methoxide
reactions, with two unidentified products having m/z-values of
224 and 226. When reactions were conducted in the presence of
∞
t
(
p∞ Ϫ p ) = ln [FlN ] /[FlN ] versus time. Linear plots with
0 2 t 2 0
excellent correlation coefficients, typically over 2 to 3 half-lives,
Ϫ1/2
were, however, achieved by plotting [FlN2]
against time, and
this implies that the reaction is of 3/2 order in diazofluorene,
the resultant rate coefficients being independent of the initial
concentration of the diazo-compound. Table 2 is a compilation
of the values of the corresponding rate coefficients. Although a
plot (Fig. 2; main graph) of the rate coefficients against the
9
96
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 9 9 5 – 1 0 0 3

View Article Online
a
Table 2 Observed rate coefficients for the reaction of 9-diazofluorene with tetraethylammonium hydroxide at 30 ЊC
3
3
Ϫ
a
3
Ϫ1/2 Ϫ1
1
0 [FlN ]/M
10 [HO ]/M
Additive (conc/M)
10 k(HO)/M
s
r
2
b
Reactions in DMSO
5
5
00
0
0
0.25
0.50
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
2.50
5.00
4.53 ± 0.04
8.66 ± 0.14
16.99 ± 0.45
18.32 ± 0.30
16.86 ± 0.17
18.38 ± 0.46
17.06 ± 0.31
20.73 ± 0.32c
No reaction
18.36 ± 0.29
40.72 ± 0.89
79.14 ± 6.82
129.0 ± 7.9
58.60 ± 4.47
0.9993
0.9979
0.9969
0.9981
0.9994
0.9946
0.9981
0.9988
1
50
25
10
50
50
50
50
50
50
50
50
Fl᎐N–N᎐Fl (sat.)
Fl᎐O (10 mM)
PDNB (0.5 mM)
Ph CN (50 mM)
0.9986
0.9986
0.9912
0.9907
0.9886
2
2
10.0
10.0
PDNB (5 mM)
b
Reactions in CH CN
3
5
00
5
2
00
5
2
5
0
2.50
5.00
5.00
5.00
6.16 ± 0.20
10.48 ± 0.22
10.98 ± 0.35
9.65 ± 0.50
24.67 ± 0.71
25.61 ± 0.47
24.80 ± 1.63
74.08 ± 6.20
0.9924
0.9975
0.9931
0.9777
0.9975
0.9985
0.9791
0.9897
1
1
0
5
10.0
0
5
0
10.0
10.0
25.0
a
b
Additives: Fl᎐N–N᎐Fl, fluorenone azine; Fl᎐O, 9-fluorenone; PDNB, p-dinitrobenzene; Ph CN , diazodiphenylmethane. Containing 1.0 M water.
2
2
Table 3 Products of the reaction of 9-diazofluorene with benzyltrimethylammonium methoxide in methanolic DMSO and with potassium
a
t-butoxide in DMSO/t-butyl alcohol at 30 ЊC
Products (%)
3
3
Ϫ
b
c
1
0 [FlN ]/M
10 [RO ]/M
Conversion/%
Fl᎐N–N᎐Fl
Fl᎐O
m/z 224
m/z 226
2
Methoxide reactions
2
2
2
1
1
1
5
5
5
2.5
2.5
2.5
25
12.5
5.0
12.5
12.5
12.5
96.4
98.5
97.9
98.0
100
90.9
95.1
96.2
91.3
62.5
48.6
2.8
2.4
0.8
3.0
28.9
39.0
2.2
0.8
0.9
2.0
7.7
8.7
1.2
0.15
0.1
0.7
1.7
3.6
d
e
100
t-Butoxide reactions
2
2
2
5
5
5
25
12.5
5.0
100
100
100
98.2
97.7
96.5
1.7
2.2
3.1
a
c
b
[
ROH] = 1.0 M. Fragmentation: m/z 224 (8%), 207 (21), 195 (22), 181 (52), 180 (80), 178 (79), 165 (20), 152 (39), 82 (52), 57 (80), 44 (100).
d
e
Fragmentation: m/z 226 (6%), 208 (6), 194 (3), 181 (100), 165 (12), 152 (19), 82 (14), 57 (14), 44 (18). In air-saturated solution. In oxygen-saturated
solution.
air or oxygen, the yield of azine was drastically reduced with
corresponding increases mainly in the yield of fluorenone but
also of the two unidentified products.
concentration (Table 4). The addition of a spin trap, phenyl
t-butyl nitrone (PBN) had a negligible effect, as did the long-
lived radical di-t-butyl nitroxide (DTBN).
Reactions were again followed by the increase in pressure as a
function of time. Again the results fitted a 3/2 order law with
respect to the diazo-compound for both alkoxides, and the
order with respect to the alkoxide was close to 1/2 over the
whole range of concentrations examined, as demonstrated
in the inset in Fig. 2. Interestingly, the points in the plot in
Fig. 2 suggest that the rate is dependent on the alkoxide con-
centration but not evidently its identity; hydroxide, methoxide
and t-butoxide at a given concentration give roughly the same
rate coefficient. It would seem that matters such as basicity and
size of the initiating base make little difference to the kinetically
important processes.
The reactions of 9-diazofluorene with sodium borohydride in
DMSO and the influence of carbon acids on the kinetics of
base-induced decomposition
Previous investigations of the transformation of the diazo-
ketone azibenzil (benzoyl phenyl diazomethane) into the corre-
sponding azine showed that the reaction could be initiated
9
cathodically in acetonitrile solution, and by sodium boro-
10
hydride in DMSO solution. An anionic chain mechanism was
proposed for the borohydride reaction in which a key step is the
transfer of a hydride ion to the diazo-carbon atom of the
diazoketone. Products were the corresponding azine plus
deoxybenzoin, the result of conversion of the diazo-function
into a methylene group. In order to establish whether an
As with hydroxide-induced reactions, the addition of p-di-
nitrobenzene had an inhibiting effect, reducing the observed
3
/2-order rate coefficient by some 30–40% depending on the
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 9 9 5 – 1 0 0 3
997

View Article Online
Table 4 a Effects of additives on observed rate coefficients for the reaction of 9-diazofluorene with benzyltrimethylammonium methoxide in DMSO
at 30 ЊC
3
3
. . .
a
3
Ϫ1/2 Ϫ1
1
0 [FlN ]/M
10 [MeO ]/M
Additive (conc/M)
10 k(MeO)/M
s
r
2
25
25
25
25
25
25
25
25
6.25
6.25
6.25
5.00
5.00
5.00
2.50
2.50
89.3
80.4
87.1
69.9
40.8
1.70
44.7
31.6
0.9966
0.9991
0.9962
0.9967
0.9967
0.9996
0.9993
0.9975
Ϫ3
PBN (2.5 × 10 M)
Ϫ3
DTBN (3.25 × 10 M)
Ϫ3
PDNB (2.5 × 10 M)
Ϫ3
MDNB (2.5 × 10 M)
Ϫ3
PDNB (1.25 × 10 M)
1
3
analogous hydride transfer pathway was possible for 9-diazo-
fluorene, similar experiments with sodium borohydride in
DMSO were carried out, but these gave a somewhat different
pattern of behaviour than that seen with azibenzil. The prod-
ucts were fluorenone azine and fluorenone hydrazone, the latter
being formed in an amount roughly corresponding to the mol
percent of sodium borohydride used. For example, with reac-
tion mixtures 0.05 M in diazofluorene, yields of hydrazone were
9-t-butoxyfluorene (pK 21.3; no observable reaction): clearly
there is no simple relation between activity and strength of the
carbon acid. As shown in Fig. 3, the observed 3/2-order rate
coefficients increase linearly with the square root of the concen-
tration of the C-acid, a dependence more strictly adhered to
with fluorene than with 9-phenylfluorene. It is to be noted that
the fluorene line in Fig. 3 extends into a region where the fluor-
ene concentration exceeds the concentration of potassium
t-butoxide. The effect of increasing the (excess) butoxide
concentration at a fixed concentration of carbon acid was
less than proportional and close to a square root dependence,
although the scatter of data points is substantial.
2
3% and 14% when the sodium borohydride concentrations
were 0.01 and 0.005 M respectively. Reaction mixtures
remained red for a short period after mixing the reagents,
during which nitrogen evolution was rather slow, and this was
followed by a colour change to purple (λmax 570 nm), as
observed in the base-induced decompositions, and this was
accompanied by much more rapid gas evolution. The timescale
for completion of the reaction was much shorter than com-
parable reactions of azibenzil, but much longer than observed
for base-induced reaction of diazofluorene at the same reactant
concentrations. Furthermore it was found, using sodium
borodeuteride, that, although the induction period was
extended, once rapid nitrogen evolution started, the rate was
little changed from that observed in reactions using NaBH4.
It has previously been reported that transformation of
azibenzil into the corresponding azine could be induced by
10
carbanions, for example that formed by deprotonation of
deoxybenzoin, and it has been argued that the reaction follows
the same hydride-transfer chain mechanism as that induced by
sodium borohydride. Preliminary experiments had suggested
that diazofluorene decomposition was not promoted by
added carbanions generated from carbon acids by tetraethyl-
ammonium hydroxide in slightly aqueous DMSO. Subsequent
5
study showed that this was not generally true. The present
work is concerned with the effect of carbon acids on reactions
in DMSO in the presence of potassium t-butoxide. The
presence of, for example, fluorene had a dramatic accelerating
effect on the reaction rate. It was necessary to work in a solvent
consisting of a mixture of DMSO and t-butyl alcohol in a ratio
of 1:9 by volume in which solvent the 3/2 order rate coefficient
for the transformation of the diazofluorene (25 mM) into azine
by potassium t-butoxide (12.5 mM) alone was rather slow
Fig.
decomposition on the concentration of carbon acid: ᭿ fluorene;
9-phenylfluorene.
3
Dependence of the rate coefficients for 9-diazofluorene
ᮀ
Electron paramagnetic resonance and uv/visible spectroscopy
During and after completion of decomposition of 9-diazo-
fluorene induced by hydroxide or alkoxide ions in DMSO or by
potassium t-butoxide in conjunction with fluorene in 9:1 t-butyl
alcohol/DMSO, the reaction mixtures displayed strong EPR
signals. In each case the spectrum consisted of a single very
broad signal upon which was superimposed an envelope
spanning about 13 G and comprising some 39 lines with a
uniform hyperfine splitting of 0.31 G. Fig. 4 shows a typical
spectrum from the reaction of the diazo-compound with
benzyltrimethylammonium methoxide. The appearance of the
Ϫ3
Ϫ1/2 Ϫ1
[k(BuO) = 1.0 × 10
M
s
at 30 ЊC]. In this reaction
medium, the reactions in the presence of fluorene were all
accelerated and afforded the azine in more than 98% yield.
9
,9Ј-Bifluorenyl could be identified among the products; for
example, in a reaction of diazofluorene (25 mM) with potas-
sium t-butoxide (12.5 mM) in the presence of fluorene (6.25
mM), 4.5% of the fluorene was converted into bifluorenyl. No
corresponding dimeric product was detected in the reactions in
which the added carbon acid was 9-phenylfluorene. The kinetics
of the reactions followed the 3/2-order dependence on diazo-
fluorene concentration described earlier but the fit to a first
order dependence was almost equally good. The kinetic effects
of carbon acid addition on the observed rate coefficients are
in Table 5. It can be seen that the order of effectiveness of the
C-acids in promoting the formation of the azine is 9-phenyl-
5
39-line spectrum was rather similar to that reported by others
14
and ascribed to the azine anion radical. Such differences as
15
there are can be ascribed to the somewhat higher temperature
(ca. 25 ЊC) at which the present spectra were recorded. In all
cases the radical species had slightly different g-values when
generated using different bases: tetraethylammonium hydrox-
ide, 2.00270; benzyltrimethylammonium methoxide, 2.00277;
potassium t-butoxide, 2.00273. In the diazofluorene reactions,
the radical concentration was estimated roughly to correspond
1
1
11
fluorene (pK in DMSO 17.9) > fluorene (pK 22.6) >> 9-t-
12
13
butylfluorene (pK 24.4) > 9-methoxyfluorene (pK 22.1) >>
9
98
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 9 9 5 – 1 0 0 3

View Article Online
Table 5 Effects of carbon acids on observed rate coefficients for the reaction of 9-diazofluorene (25 mM) with potassium t-butoxide in
DMSO-t-butyl alcohol (1:9 v/v) at 30 ЊC
Ϫ
a
3
Ϫ1/2 Ϫ1
[
BuO ]/mM
Additive
Conc/mM
10 k(FlHR)/M
s
r
1
1
1
1
1
1
1
1
1
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
1.00
24.7
38.8
55.4
78.1
90.0
155
192
214
46.3
101
91.3
98.0
110
0.9376
0.9838
0.9823
0.9828
0.9916
0.9788
0.9858
0.9656
0.9891
0.9889
0.9891
0.9901
0.9738
0.9970
0.9805
0.9929
0.9900
0.9900
0.9993
0.9009
FlH2
FlH2
FlH2
FlH2
FlH2
FlH2
FlH2
FlH2
FlH2
FlH2
FlH2
FlH2
FlH2
FlHPh
FlHPh
FlHPh
FlHPh
0.39
0.78
1.56
3.12
6.25
12.5
19.0
25.0
6.25
6.25
6.25
6.25
6.25
1.56
3.12
4.25
6.25
6.25
6.25
6.25
12.5
6.25
6.25
6.25
6.25
6
.25
1
2
3
4
1
1
1
1
1
1
1
1
9.0
5.0
7.5
0.0
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
98.5
106.4
118
110
FlHCMe3
FlHOMe
FlHOCMe3
FlHPh
FlHPh
FlHPh
0.54
0.05
No reaction
151
94.8
106
0.9929
0.9910
0.9955
0.9954
0.9878
6
.25
1
2
3
9.0
5.0
1.3
FlHPh
FlHPh
119
125
a
Carbon acids: FlH fluorene; FlHPh 9-phenylfluorene; FlHCMe 9-t-butylfluorene; FlOMe 9-methoxyfluorene; FlOCMe 9-t-butoxyfluorene.
2
3
3
DMSO) with an equimolar solution of hydroxide, methoxide or
t-butoxide. The timescale of events was roughly the same as
that of the decomposition of 9-diazofluorene; for example with
equimolar (5 mM) azine and tetraethylammonium hydroxide
the half-life of colour production was some 600 s.
Discussion
The decomposition of 9-diazofluorene induced by quaternary
ammonium hydroxide and methoxide, and by potassium
t-butoxide with or without the addition of carbon acids such
as fluorene affords fluorenone azine which is precipitated in
almost quantitative yield. Evolution of nitrogen occurs at a rate
Fig. 4 EPR spectrum of a reaction mixture containing 9-diazo-
fluorene (5 mM) and benzyltrimethylammonium methoxide (2 mM) in
DMSO solution at ambient temperature.
3/2
Ϫ 1/2
d[N ]/dt ∝ [FlN ] [RO ] . Free radicals are detectable in the
2 2
reaction mixtures by EPR spectroscopy and the EPR signals
and colour of the reaction mixtures are identical to those
produced when 9-diazofluorene or the azine are subjected to
cathodic reduction, or when the azine is treated with the bases
mentioned above. The question then is whether the base-induced
decomposition of the diazo-compound (with or without addi-
tion of a carbon acid) is analogous to the cathodically induced
reaction which is thought to involve electron transfer chain
catalysis, or whether a non-radical, polar mechanism operates,
with the radical formation a side reaction independent of the
diazoalkane decomposition and involving perhaps only the
more readily reduced azine and the base solution. In order to
illustrate the problem, we present possible, but by no means
unique, pathways of these two types in Scheme 1. They are
formulated in terms of a common first step, involving diazo-
fluorene and the nucleophile. Attack of the nucleophile on the
diazo-carbon atom would be expected to generate nitrogen and
a 9-substituted fluorenyl anion; since fluorenyl alkyl ethers in
the presence of potassium t-butoxide do not appear to promote
efficient decomposition of diazofluorene, this pathway seems
unlikely, even though a 9-methoxy-substituent in fluorene is
reported to have a strongly radical-stabilising effect as judged
to about 5% of the initial concentration of the diazo-com-
pound (1.25 to 10 mM). Increasing the concentration of dia-
zofluorene above this range, or using acetonitrile as the solvent
led to much less well resolved spectra. Reactions carried out in
DMSO with addition of PBN gave EPR spectra of slightly
reduced intensity but with no evidence of the change in the
appearance of the spectra that would indicate radical trapping
and formation of nitroxyl species. Reactions in the presence of
DTBN (1 mM) led to the disappearance of some 95% of the
intensity of its characteristic triplet EPR signal; the 39-line sig-
nal observed in reaction mixtures not containing DTBN was
absent.
Within a few seconds of the initiation of all reactions, the
mixtures took on a deep purple colouration, and spectroscopic
investigation showed strong absorbance at 380, 545 (sh) and
5
80 nm, with much weaker bands at 742, 824 and 880 nm. The
colour in all cases was discharged on exposure of reaction
mixtures to the air. It was similar in appearance to that
observed during preparative electrolysis of 9-diazofluorene at a
4
platinum cathode in acetonitrile. Significantly, the spectrum
is identical to that resulting from spectroelectrochemical
16
experiments on the reduction of fluorenone azine in the same
from its effect on the acidity-oxidation potential. Reversible
nucleophilic addition to the terminal nitrogen of the diazo-
compound could, however, afford a carbanion which could
attack a further molecule of 9-diazofluorene either by nucleo-
14
solvent. Indeed we have found that the same colour, and
the same 39-line EPR spectrum as described above, could be
generated by treatment of fluorenone azine (typically 5 mM in
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 9 9 5 – 1 0 0 3
999

View Article Online
Scheme 1
philic addition to the terminal nitrogen atom, followed by loss
of the original nucleophile and a molecule of nitrogen (Pathway
A) or by electron transfer to the second molecule of the diazo-
compound (Pathway B).
On the basis of Pathway A, the azine yield should be high,
and indeed, in all cases, was found to be greater than 90%, even
when the oxyanion concentration was equal to the initial con-
centration of 9-diazofluorene. Using hydroxide ion as the
nucleophile, 9-fluorenol was detected in yields that increased
with the hydroxide concentration, but the amounts were small
and could be attributed to a side reaction in which the nucleo-
phile attacked the diazo-carbon. One of the unidentified prod-
ucts from reactions in the presence of methoxide ions has m/z =
On the basis of the establishment of a steady state for the
concentration of the chain carrier C as formulated in eqn. (1)
and (2),
Ϫ
n
2
d[C]/dt = k
[Nu ][FlN
] Ϫ k
[C] = 0
(1)
I
2
T
whence
Ϫ
n 1/2
[C] = {(k /k )[Nu ][FlN ] }
I T 2
(2)
If the rate of propagation is much larger than the rate of
initiation, then the rate of product formation is given by eqn.
3).
(
2
24 corresponding to FlH–N᎐N–OMe. The other by-products
᎐
d[Fl᎐N–N᎐Fl]/dt = d[N ]/dt = k [C][FlN ] =
cannot be rationalised in terms of the mechanism, and, along
with the observed free radicals, would have to be attributed to
separate side-reactions. Turning to the kinetic form expected for
Pathway A, it is possible that the rate-limiting step would be the
initial attack of the nucleophile on the diazo-compound, the
subsequent reaction with a second diazo-molecule being rapid
and producing an adduct whose fragmentation would be
irreversible. If this were so, then the observed rate law would be
᎐
᎐
2
P
2
1/2
Ϫ 1/2
1 ϩ n/2
k
(k
/k
) [Nu ] [FlN
]
2
(3)
P
I
T
Ϫ
For the reactions induced by hydroxide or alkoxide (Nu
᎐
᎐
Ϫ
Ϫ
HO or RO ), the chain formulation fits the observed kinetic
form if the number of 9-diazofluorene molecules (n) partici-
pating in the rate-limiting step of the initiation process is one.
Note that in the propagation step it has been assumed that the
two diazofluorene molecules which are converted by the chain
carrier into azine are involved in two separate steps, the first of
which determines the overall rate of propagation.
Ϫ
d[N ]/dt ∝ [FlN ][RO ], which is not observed. Rate-limiting
2
2
attack by 9-hydroxydiazenyl- or alkoxydiazenyl-fluorenyl
anion, rapidly formed in low equilibrium concentration, on a
second molecule of diazofluorene or fragmentation of the
resultant adduct cannot be rate-determining since this would
lead to a kinetic law second order in the diazo-compound and
first order in alkoxide ion. These kinetic features have not been
observed, and we believe that the nucleophilic addition/elimin-
ation mechanism (Pathway A) can therefore be ruled out.
The observation of fractional orders of reaction is often
associated with radical chain reactions. In the present instance,
all the reactions show an order in diazofluorene of 3/2 and in
hydroxide/alkoxide of 1/2. In the case of reactions promoted by
fluorene, the order of reaction with respect to the carbon acid is
In the case of reactions promoted by added carbon acids
in the presence of potassium t-butoxide, the nucleophile is
presumed to be the related, rapidly formed carbanion, FlR ,
Ϫ
the concentration of which is given by eqn. (4),
Ϫ
Ϫ
Ϫ
[FlR ] = [FlHR] [BuO ]/([BuO ] ϩ
total
Ϫ
Kequ[BuOH]) = [Nu ] (4)
where [FlHR]total is the total concentration of carbon acid
added to the reaction mixture and Kequ is the equilibrium
constant for the protonation of the carbanion according to:
1
/2 and, at fixed fluorene concentration of 6.25 mM, increasing
Ϫ
Ϫ
the t-butoxide concentration in the range 6.25 to 40 mM led to
an increase in rate that showed a rough dependence on the
square root of the butoxide concentration. These reaction
orders can be fitted to the standard formulation of a chain
reaction, analogous to Pathway B but including a chain-
termination step, viz.,
FlHR ϩ BuO = FlR ϩ BuOH
In such cases, the rate expression becomes:
d[Fl᎐N–N᎐Fl]/dt =
1/2
Ϫ
Ϫ
k (k /k ) {[FlHR] [BuO ]/([BuO ] ϩ
P
I
T
total
1/2
3/2
Kequ[BuOH])} [FlN ]
(5)
2
Ϫ
ؒ
ؒ
Ϫ
Initiation: Nu ϩ n FlN2 C [Nu ϩ FlN2
Propagation: C ϩ FlN2 Intermediate ϩ N2 Rate constant kP
Intermediate ϩ FlN2 Fl᎐N–N᎐Fl ϩ C
Termination: 2 C termination products Rate constant kT
] Rate constant kI
and the apparent order in butoxide concentration will be one
half for weak carbon acids or less for carbon acids that
are substantially deprotonated under the reaction conditions.
᎐
᎐
1
000
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 9 9 5 – 1 0 0 3

View Article Online
Consistent with this, the plot of observed 3/2 order rate co-
initiation process has the additional virtue of rationalising the
formation of the by-product having m/z = 224 observed in
1/2
efficients against [FlHR]total is linear for fluorene, but shows
downward curvature as the butoxide concentration increases, in
the case of 9-phenylfluorene (Fig. 3).
methoxide induced reactions as Fl᎐NNHOMe, but it would
᎐
require that the rate-limiting step of initiation be the formation
of the 9-(methoxydiazenyl)fluorenyl anion rather than the sub-
sequent electron transfer if it is to be consistent with the overall
fractional order in the diazo-compound. It should be noted,
however, that the isomeric adduct formed by attachment of the
Having established that the kinetic form of the reactions fits
the chain model with second order termination in the chain
carrier concentration, we need to consider the identity of the
carrier, its formation and destruction. The observation that the
hydroxide- and methoxide-induced reactions are inhibited by
p-dinitrobenzene, a good one-electron acceptor, points to an
electron-transfer chain mechanism, but it is necessary to
explain why chain termination has a different kinetic form
from that previously observed in electrochemically induced
decomposition of 9-diazofluorene in a stirred, divided reactor.
The simplest formulation of the reaction in the absence of
added carbon acids is that initiation of the reaction involves
direct electron transfer from the hydroxide/alkoxide ion in
DMSO solution to diazofluorene, generating a hydroxy- or
alkoxy-radical and a diazofluorene anion radical, the chain
carrier. In this model, propagation of the chain would be by
attack of the carrier on a second diazofluorene molecule in the
rate-limiting step, generating molecular nitrogen and a mole-
cule of fluorenone azine as its anion radical. The latter would
be involved in an electron-transfer equilibrium with a further
molecule of diazofluorene, regenerating the chain carrier and
giving azine which would precipitate. A significant concen-
tration of the azine anion radical would remain in solution and
would be detectable by EPR spectroscopy. Termination of the
chains would be by dimerisation of the carrier species, a reac-
Ϫ
oxyanion to C-9 of the diazo-compound, Fl(OR)N2 could be
the key intermediate, electron-transfer being facilitated by
Ϫ
Ϫ
ؒ
irreversible fragmentation: Fl(OR)N2 ϩ FlN2
FlN2
ϩ
ؒ
FlOR ϩ N . We cannot discount this possibility, but we
suspect that Fl(OR)N2 may not have a long enough lifetime.
2
Ϫ
Although we cannot unambiguously rule out direct electron
transfer from the oxyanions to the diazo-compound, we believe
that a nucleophile/diazoalkane adduct is responsible. Our
observation that electron transfer also takes place to fluorenone
azine in hydroxide or alkoxide solutions in DMSO when no
9-diazofluorene is present might seem to suggest direct transfer
from the oxyanions, but again the transfer might be mediated
by an azine/oxyanion adduct, although we have no direct
evidence for this.
The identity of the chain carrier as the diazofluorene anion
radical needs further consideration. We have previously shown
that, when generated electrochemically in dimethylformamide
or acetonitrile solution, it rapidly dimerises forming a dianion
which, at room temperature, rapidly loses molecular nitrogen to
22
form fluorenone azine as its strongly basic dianion. We have
8
previously argued that, in cathodically induced conversion of
8a
tion we have previously identified and reported. In order to
account for the by-products; fluorenone and 9-fluorenol, the
oxygen centred radical produced in the initiation step would
have to be trapped irreversibly by diazofluorene generating
molecular nitrogen and 9-hydroxyfluorenyl radicals some
of which might be expected to undergo disproportionation;
N-protonation of diazofluorene anion radical would generate
diazofluorene into fluorenone azine, which follows an electron-
transfer chain catalysis mechanism, the diazofluorene anion
radical is unlikely to be the chain carrier since it would dimerise
in preference to propagating the chain by reaction with further
diazoalkane molecules under the prevailing reaction conditions.
It was suggested that the carrier was probably the s-cis-fluor-
enone azine anion radical. In the present experiments in a
different solvent system, in which there is an appreciable
concentration of water or alcohol to act as a proton source, the
same considerations may not apply. Even though the kinetic
form of the propagation step appears to be the same, that of the
termination step is different, and it has to be a dimerisation or
disproportionation in the base induced reactions. Moreover the
rate of generation of the chain carriers is not under independ-
ent experimental control as is the case in electrolysis. For
simplicity therefore we propose that the carrier is indeed the
9-diazofluorene anion radical, and that propagation involves
attack on a molecule of diazofluorene, yielding after nitrogen
loss the fluorenone azine anion radical (the intermediate in
Scheme 1, Pathway B) which transfers its extra electron to a
further diazo-molecule in a rapidly established electron-transfer
equilibrium. Dimerisation of the carrier could terminate the
chain but only if the dimer is inactivated as an electron transfer
agent; in the present conditions that could be achieved by
rapid protonation either before or after loss of a molecule of
nitrogen, although no corresponding products were detected.
On this basis the observed second-order rate constant estimated
9
-diazenylfluorenyl radical, disproportionation of which would
produce a molecule of fluorenone hydrazone and regenerate
17
one of diazofluorene, as suggested in pulse radiolysis studies.
The addition of fluorene to reaction mixtures containing potas-
sium t-butoxide accelerates the conversion of diazofluorene
into fluorenone azine and this is consistent with the simple
formulation if initiation of the chains is mediated by fluorenyl
anions. Electron transfer from the carbanion would afford
9
9
-fluorenyl radicals, leading to the corresponding radical dimer,
,9Ј-bifluorenyl, as is indeed observed, and to the 1/2-order
kinetic dependence on the fluorene concentration. Reactions in
the presence of 9-phenylfluorene fit less well; the kinetic
dependence on carbon acid concentration is somewhat less well
defined and no products attributable to 9-phenylfluorenyl
radicals were detected.
There are other more serious problems with the simple
formulation, however, particularly in respect of the initiation
process, the identity of the chain carrier so produced, and the
termination process. There has been much discussion over the
possibility of hydroxide and alkoxide ions taking part in outer
sphere electron transfer reactions with organic acceptors.
Ϫ3
Ϫ1 Ϫ1
from the inset in Fig. 2 of 11.4 × 10
M
s
is equal to k (k /
P I
18
19
20
1/2
Eberson, Sawyer and Roberts and Djaasberg et al. in
particular have advanced compelling arguments based on
Marcus theory that this should not be feasible and that altern-
ative pathways involving the generation of better one-electron
donors should be sought. In dipolar aprotic solvents in which
anion solvation energies are much lower than in protic solvents,
the situation should be more favourable, and reports continue
to appear of outer sphere electron transfer from hydroxide ion,
k ) . The value of k from cathodically initiated reactions
T
P
Ϫ1 Ϫ1
is 9.7 M 5
s
and k taken as the dimerisation of the carrier
T
Ϫ1 Ϫ1
is 2 × 10 M s ; hence the value of k would have to be 0.2
I
Ϫ1 Ϫ1
M
s , which seems reasonable in the light of the rate of
generation of the anion radical from fluorenone azine and
tetraethylammonium hydroxide.
The effect of adding carbon acids to diazofluorene solutions
containing potassium t-butoxide is to accelerate the conversion
of diazofluorene into fluorenone azine. The kinetic form of the
reactions indicates that the carbanion formed by deprotonation
of the carbon acid facilitates electron transfer to diazofluorene,
generating the chain carrier, as expected for a species much
21
in some cases assisted by subsequent irreversible steps. An
alternative to direct electron transfer from oxyanions is that
shown in Scheme 1 (Pathway B) in which the anionic diazo-
fluorene/nucleophile adduct initiates the chain by transferring
an electron to a second molecule of diazofluorene. Such an
17,18
more readily oxidised than the oxyanions.
It would seem
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 9 9 5 – 1 0 0 3
1001

View Article Online
that thereafter the behaviour is essentially the same as in
reactions without added carbon acid. Our earlier experiments
have shown that the radicals formed after electron transfer from
carbanions to diazofluorene can dimerise but can themselves
attack diazo-molecules and generate azines.
temperature under oxygen-free nitrogen, the solutions were
mixed and, by inversion of the apparatus, the mixture trans-
ferred to the EPR tube for recording of the spectra. Where
necessary, phenyl t-butyl nitrone or di-t-butyl nitroxide were
introduced; both were commercial samples used as received.
Apart from a few experiments using infrared spectro-
photometry, most kinetic studies were carried out in a thermo-
statted (30 ЊC), two-legged, modified Warburg apparatus that
was equipped with a two way tap to permit degassing and
flushing with oxygen-free nitrogen and a side arm housing a
pressure transducer (R.S. Components Ltd., Type 303–337).
The transducer response (mV), recorded on a chart recorder
(Servoscribe 1S) or using a digital voltmeter and calibrated
against a mercury manometer, was linear over the range
of pressures generated during reaction. For a kinetic experi-
ment, DMSO solutions of the diazo-compound and, where
appropriate, other additives were syringed into one of the legs,
the base solution in DMSO into the other. Using fixed volumes
of reaction solutions, the pressure changes recorded corre-
sponded closely with expectation based on the yield of azine
when reaction was complete.
4
In summary, we have shown that hydroxide and alkoxide
anions can induce the transformation of 9-diazofluorene into
fluorenone azine in high yield. This reaction is accelerated by
the addition of carbon acids such as fluorene, but inhibited
by p-dinitrobenzene, a good one-electron acceptor. Reaction
mixtures exhibit EPR signals. The reaction shows fractional
orders in the diazo-compound, the oxyanions and the carbon
acid. It is argued that this is all consistent with an electron-
transfer chain mechanism with second order termination.
Suggestions have been made concerning the nature of the
initiation step; although direct (outer sphere) electron transfer
from the oxyanion to diazofluorene, generating the diazo-
fluorene anion radical as the chain carrier, provides the simplest
interpretation of the kinetics, redox-potential considerations
make this unlikely. A diazofluorene/nucleophile adduct as the
electron-transfer agent is a likely alternative.
Experimental
Acknowledgements
Materials
The authors are indebted to Professor J. H. P. Utley for valuable
discussions. Financial sponsorship from SERC (LJM) and
the Islamic Republic of Iran (MMK) is also gratefully
acknowledged.
9
-Diazofluorene, mp 94–5 ЊC [from petrol (bp 60–80Њ)], was
23
prepared as previously described;
from hplc analysis,
contamination by fluorenone (the only impurity) was less than
.02%. Fluorenone azine was prepared from fluorenone and
0
24
its hydrazone in the presence of acetic acid, mp 273 ЊC (from
m-xylene). Of the carbon acids used, fluorene, 9-t-butylfluorene
and 9-phenylfluorene were recrystallised commercial samples
while the 9-methoxy- and 9-t-butoxy-fluorenes were prepared
by refluxing 9-bromofluorene with silver nitrate in the
appropriate alcohol as solvent; all had mp in agreement with
literature values. Dimethyl sulfoxide was dried by standing
overnight over calcium hydride, followed by distillation under
reduced pressure, discarding the first 15% of the distillate.
Acetonitrile (hplc grade) was passed through a short column
of activated alumina immediately prior to use. Tetraethyl-
ammonium hydroxide was a commercial 20% solution in water
and benzyltrimethylammonium methoxide was a commercial
References
1
F. M. Beringer, J. A. Farr and S. Sands, J. Am. Chem. Soc., 1953, 75,
984.
3
2
3
R. N. McDonald and K.-W. Lin, J. Am. Chem. Soc., 1978, 100, 8028.
H. Staudinger and J. Meyer, Helv. Chim. Acta., 1919, 2, 619,635;
D. Bethell, S. F. C. Dunn, M. M. Khodaei and A. R. Newall,
J. Chem. Soc., Perkin Trans. 2, 1989, 1829; D. Bethell, M. P. Brown,
M. M. Harding, C. A. Herbert, M. M. Khodaei, M. I. Rios
and K. Woolstencroft, Acta Crystallogr., Sect. B, 1992, 48, 683;
D. Bethell, R. Bourne and M. Kazlan, J. Chem. Soc., Perkin Trans.
25
2
, 1994, 2081.
4
J. M. Bakke, D. Bethell, P. J. Galsworthy, K. L. Handoo and
D. Jackson, J. Chem. Soc., Chem. Commun., 1979, 890.
5 K. L. Handoo, K. Gadru and C. K. Jain, Indian J. Chem., Sect. A,
1983, 22B, 526.
4
0% solution in methanol; both were used as supplied, and
6
R. N. McDonald, K. J. Borhani and M. D. Hawley, J. Am. Chem.
Soc., 1978, 100, 995, but see V. D. Parker and D. Bethell, Acta Chem.
Scand. Ser. B, 1981, B35, 691 for an alternative interpretation of the
electrochemistry.
their concentrations determined by titration with standard
hydrochloric acid. The stock solutions were diluted with water
or methanol as appropriate before mixing with the dipolar
aprotic solvent so as to have a concentration of the hydroxylic
component of 1.0 M in all cases. Solutions of potassium
t-butoxide in t-butyl alcohol were freshly prepared as previously
described and then diluted with the alcohol where necessary
before mixing with DMSO. Except in the experiments involving
the addition of carbon acids, the concentration of t-butyl
alcohol in the reaction mixtures was 1.0 M.
7
See for example, M. Chanon and M. L. Tobe, Angew. Chem. Int. Ed.
Engl., 1982, 21, 1.
8 (a) D. Bethell, L. J. McDowall and V. D. Parker, J. Chem. Soc.,
Perkin Trans. 2, 1984, 1531; (b) D. Bethell, P. J. Galsworthy,
K. L. Handoo and V. D. Parker, J. Chem. Soc., Chem. Commun.,
1
1
980, 534; (c) D. Bethell and V. D. Parker, J. Am. Chem. Soc., 1986,
08, 895.
9
D. Bethell, L. J. McDowall and V. D. Parker, J. Chem. Soc., Chem.
Commun., 1984, 308.
Instrumentation
10 D. Bethell and L. J. McDowall, J. Chem. Soc., Chem. Commun.,
984, 1408.
1
Quantitative infrared measurements were made on a Perkin
Elmer 1320 instrument using calcium fluoride cells. Hplc
analyses were carried out on a Varian 5000 liquid chromato-
graph using a RP18-2776 reverse phase column and 85%
aqueous acetonitrile as eluant; uv detection was at 254 nm.
Mass spectra and GCMS data were obtained on a VG 7070E
instrument using a capillary column with OV1 as the stationary
phase.
1
1 W. S. Matthews, J. E. Bares, J. E. Bartmess, F. G. Bordwell,
F. J. Cornforth, G. E. Drucker, Z. Margolin, R. J. McCallum,
G. J. McCollum and N. R. Vanier, J. Am. Chem. Soc., 1978, 97, 7006.
2 F. G. Bordwell, G. E. Drucker and G. J. McCollum, J. Org. Chem.,
1
1
1
1
982, 47, 2504.
3 F. G. Bordwell, M. Van Der Puy and N. R. Vanier, J. Org. Chem.,
976, 41, 1885.
1
4 See Y. P. Kitaev, V. K. Ivanova, A. S. Mukhtarov and L. N. Orlova,
New Research on the Electrochemistry of Organic Compounds, 8th
All-Union Conference on the Electrochemistry of Organic
Compounds, ed. L. G. Feoktistov, Zinatne, Riga, 1973, p. 168;
Y. P. Kitaev, V. K. Ivanova, A. S. Mukhtarov, L. N. Orlova and
A. Ladygin, Akad. Nauk SSSR, Ser. Khim., 1974, 1, 72.
5 K. M. White, PhD Thesis, Liverpool University, 1982.
6 F. G. Bordwell and M. J. Bausch, J. Am. Chem. Soc., 1986, 108,
1979.
EPR spectra were recorded on a Varian E-4 X-band spec-
trometer using a two-legged glass arrangement fused to the
EPR sample tube. A solution of the diazo-compound in
DMSO was syringed into one leg of the apparatus and a solu-
1
1
Ϫ
tion of the base RO /ROH in DMSO into the other and
degassed by freeze–pump–thaw cycles; after bringing to room
1
002
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 9 9 5 – 1 0 0 3

View Article Online
1
1
7 J. E. Packer and R. L. Willson, J. Chem. Soc., Chem. Commun.,
983, 733; J. E. Packer and R. L. Willson, J. Chem. Soc., Perkin
Trans. 2, 1984, 1415.
8 L. Eberson, Acta Chem. Scand. Ser B, 1982, B36, 533; L. Eberson,
Acta Chem. Scand. Ser. B, 1984, B38, 439; L. Eberson, Adv. Phys.
Org. Chem., 1982, 18, 79; L. Eberson, Electron Transfer Reactions in
Organic Chemistry, Springer Verlag, Heidelberg, 1987.
9 D. T. Sawyer and J. L. Roberts, Acc. Chem. Res., 1988, 21, 469.
0 K. Djaasberg, S. R. Knudsen, K. N. Sonnichsen, A. R. Andrade and
S. U. Pedersen, Acta Chem. Scand., 1999, 53, 938.
6709; W. Y. Zhao and Y. C. Liu, Acta Chim. Sinica., 1991, 49, 1028;
1
R. Bacaloglu, A. Blasko, C. A. Bunton, F. Ortega and C. Zucco,
J. Am. Chem. Soc., 1991, 113, 238; R. Bacaloglu, A. Blasko,
C. A. Bunton, F. Ortega and C. Zucco, J. Am. Chem. Soc., 1992, 114,
7708.
22 For
a review of electrochemically generated bases including
dianions, see J. H. P. Utley, Topics in Current Chemistry, Springer
Verlag, Heidelberg, 1987, vol. 142, p. 133; J. H. P. Utley and
M. F. Nielsen, in Organic Electrochemistry, 4th edn., ed. H. Lund
and O. Hammerich, Marcel Dekker, New York, 2000, p. 1227.
23 H. Staudinger and O. Kupfer, Chem. Ber., 1911, 44, 2197.
24 S. Cohen, F. Cohen and C. Wang, J. Org. Chem., 1963, 28, 1479.
25 A. Kliegl, Chem. Ber., 1929, 62, 1327.
1
2
2
1 See, for example, S. M. Mattar and D. Sutherland, J. Phys. Chem.,
1
991, 95, 5129; J. M. Robinson, E. T. Flynn, T. L. McMahan,
S. L. Simpson, J. C. Trisler and K. B. Conn, J. Org. Chem., 1991, 56,
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 9 9 5 – 1 0 0 3
1003