Self-Exchange Electron Transfer Rate Constants
J. Am. Chem. Soc., Vol. 120, No. 32, 1998 8239
Experimental Section
TTF, TMTSF, and DMP were obtained from Aldrich and used as
received. Solutions of TMTSF+ and TTF+ for kinetic studies were
prepared by 1 equiv oxidation of TMTSF and TTF by NOPF6 in
acetonitrile. Solutions of DMP+ were prepared by 1 equiv oxidation
of DMP by FeCp′2PF6 in acetonitrile. The syntheses of 22/tBuMe,24a
22/iPr2,24b 22/tBuiPr,24b 22/tBuPh,24c iPrMeN)2,24d iPr2NNMe2,24d
N[333]N,17 and their related radical cation salts used in this work have
been described elsewhere. Me2N)2 (Fluka) was purified by GC.
Et2N)2, nPr2N)2, and nHx2N)2 were prepared by a common method;
the details for Et2N)2 are given here. A distinct advantage of this
methods is that it avoids the use of N-nitrosoamines, previously used
in these syntheses, which are carcinogenic and mutagenic. [Caution:
tetraethylhydrazine is extremely volatile!] Bromoethane (15.2 mL, 201
mmol) was slowly added dropwise at room temperature through a
condenser to a stirring solution of hydrazine monohydrate (9.7 mL,
200 mmol) and absolute ethanol (10 mL) during which it refluxed
without additional heat. Heat was then applied to maintain gentle reflux
for an additional hour. The solution was allowed to cool overnight
with stirring. A NaOH solution (10.010 g, 0.25 mol, in 30 mL of H2O)
was added to dissolve the hydrazine salts. The solution was extracted
with pentane (2 × 20 mL), washed with a saturated NaCl solution,
dried over MgSO4, and evaporated to yield a mixture of mono-, di-,
and triethylhydrazine (0.222 g). The aqueous layer was saturated with
solid NaCl, extracted with ether (2 × 20 mL), dried over MgSO4, and
evaporated to yield an additional mixture of mono-, di-, and triethyl-
hydrazine (4.748 g). Acetaldehyde (7.8 mL, 139.7 mmol), while being
kept at a temperature below 0 °C, was added dropwise under nitrogen
to a stirring mixture of mono-, di-, and triethylhydrazine in ether (4.970
g obtained in the previous reaction) and acetonitrile (100 mL). Upon
completion, NaBH3CN (2.941 g, 44.5 mmol) was added and stirred.
Acetic acid (5.4 mL, 93.6 mmol) was added in small amounts over a
period of 45 min, and the solution was stirred overnight under nitrogen
at room temperature. Concentrated HCl (9 mL, 37.5%) was then added
until the solution became acidic and gases no longer evolved.. Upon
evaporation of the acetonitrile, a NaOH solution (9.919 g, 25.0 mmol,
in 30 mL of H2O) was slowly added, while the mixture was cooled in
Figure 5. AM1 ∆Hv(calcd) versus estimated ∆Gqv ) ∆Gq (fit) - ∆Gq
ii
s
for aromatics and hydrazines.17 Three points are plotted for each couple,
those obtained using ∆Gq (TMTSF0/+) ) 0, 1, and their average (shown
v
as the filled circle). The dotted line has a slope of 1.
obtaining useful ∆Gqii values. Figure 1 most clearly shows the
predictive value of the resulting ∆Gq (fit) for the 31 compounds
ii
studied.
The ∆Gq (fit) values for aromatic compounds are smaller than
ii
∆Gs values calculated using Marcus’s familiar dielectric con-
tinuum formula, but when values of ∆Gq scaled to experimental
s
values (using ∆Gq (TMTSF0/+) ) ∆Gq (TMTSF0/+)), and are
s
ii
estimated assuming the r-1 dependence it predicts, a plot of
∆Gq (inter) ) ∆Gq (fit) - ∆Gq (est) versus AM1-calculated
v
ii
s
vertical reorganization enthalpy is linear, suggesting that varia-
a room-temperature water bath, until the solution became basic.
A
tions in V do not very much affect the >4 kcal/mol range in
saturated NaCl solution (30 mL) was then added with stirring to the
basic solution. The reaction mixture was extracted with ether (3 × 20
mL) and dried over MgSO4. The ether was removed by distillation
(760 mmHg, 45 °C) to yield tetraethylhydrazine (4.50 g). A sample
(2.998 g, 20.8 mmol) of crude tetraethylhydrazine was filtered through
100 g of silica gel (2% ether in pentane, Rf ) 0.39) and evaporated in
several fractions to yield tetraethylhydrazine. Traces of the ether were
removed by distillation (760 mmHg, 45 °C) through a vacuum-jacketed
30 cm Vigreux column to yield pure tetraethylhydrazine as a slightly
yellow oil (0.752 g, 5.22 mmol). 1H NMR (300 MHz, CDCl3): δ 2.460
(q, 2H), 1.024 (t, 3H).
∆Gq (fit) for these aromatic compounds. However, because
ii
Marcus’s cross reaction eq 2 assumes constant preexponential
factors, effects of variations of V between the couples must be
incorporated into the kii(fit) values obtained. ∆Gq (fit) values
ii
should be larger than ∆Gq values determined under self-
ii
exchange conditions unless V for the cross reaction is as large
on average as for the related self-exchanges, and ∆G*ii(fit)
becomes increasingly smaller than ∆Gq (fit) as V for the cross
ii
reaction drops. When a larger range of structural variation is
considered, such effects are clearly present. A smaller V appears
principally responsible for the smaller kii(fit) values for fer-
rocenes than for TMPD0/+, and that the 2500-fold higher
reactivity of Me2N)20/+ than Et2N)20/+ originates from a larger
V for Me2N)20/+, resulting from smaller steric interactions
allowing direct NN, π system overlap with an ET partner, which
apparently does not occur significantly even for n-alkylhydra-
The cyclic voltammetry and stopped-flow experiments were con-
ducted as previously described.5 The E°′ values for TMTSF0/+ and
TTF0/+ require special comment because literature values vary so much
with solvent and who reported them. For example, values in methylene
23a
chloride have been reported at 0.54 V (TMTSF0/+
)
and ∆E°′ )
E°′(TMTSF0/+) - E°′(TTF0/+) ) 0.167,23b but more recently at 0.53
and 0.52 V, respectively (∆E°′ ) 0.01),23c and in benzonitrile values
of 0.42 and 0.24 V (∆E°′ ) 0.18),23d and 0.47 and 0.40 V (∆E°′ )
0.07)23c have been reported. E°′ values are strongly affected by ion
pairing in nonpolar solvents, which probably contributes to the problems
in internal consistency for the literature values. We have not found
zines. A plot of ∆Hv(calcd) versus estimated ∆Gq (fit) values
v
(∆Gq (fit) values “corrected” using r values to allow for λs
ii
effects) is shown as Figure 5;22 the dashed line shows a slope
of 1. It is not clear how accurate the ∆Hv values are, but larger
0/+
∆Gqv values for the more hindered hydrazines than Me2N)2
(23) (a) Lerstrup, K.; Toulham, D.; Bloch, A.; Poeler, T. Cowan, D. J.
Chem. Soc., Chem. Commun. 1982, 336. (b) Bechgaard, K.; Cowan, D. O.;
Bloch, A. N. J. Chem. Soc., Chem. Commun. 1974, 937. (c) Zambounis, J.
S.; Christen, E.; Pfeiffer, J.; Rihs, G. J. Am. Chem. Soc. 1994, 116, 925.
(d) Wudl, F.; Aharon-Shalom, E. J. Am. Chem. Soc. 1982, 104, 1154.
(24) (a) Nelsen, S. F.; Wolff, J. J.; Chang, H.; Powell, D. R. J. Am.
Chem. Soc. 1991, 113, 7882. (b) Nelsen, S. F.; Chen, L.-J.; Petillo, P. A.;
Evans, D. H.; Neugebauer, F. A. J. Am. Chem. Soc. 1993, 115, 10611. (c)
Nelsen, S. F.; Ismagilov, R. F.; Powell, D. R. J. Am. Chem. Soc. 1996,
118, 6313. (d) Nelsen, S. F.; Peacock, V.; Weisman, G. R. J. Am. Chem.
Soc. 1976, 98, 5269.
probably result at least partially from smaller V values for the
more hindered compounds.
(22) The bis-N,N′-bicyclic hydrazines are 21/21, 22/u22, and 21/u22,
which are untwisted in both oxidation states. AM1 calculations underesti-
mate λ′v for 22/220/+ relative to bis-N,N′-bicyclic systems which are not
twisted in the neutral form because they incorrectly determine 22/22° to be
untwisted.6 A point for 22/220/+ was therefore omitted. Both the 33)2PD
and 33N)2 calculations are not for the AM1 energy minima but for neutral
compound twist angles which are independently known to be realistic.