NMR of alkali metal-substituted benzophenone radical anions
587
to second-degree polynomials with R2 D 0.999 or better.
Regarding the extraction of conclusions from the deviation
of plots from linearity, we should point out that in the
last 20 years or so during which we have been performing
measurements on a wide variety of paramagnetic mate-
rials, such as aromatic radical anions in THF, galvinoxyl
in toluene and Cr(acac)3 in THF, where there is either no
aggregation in solution or, if there is, no dissociation to
smaller aggregates or monomeric species on dilution, we
have consistently observed a perfectly linear shift vs con-
centration relationship and the molar shifts obtained have
been reproduced to within 0.01 ppm molꢀ1. Therefore, we
are convinced that any deviation from linearity, such as
reported here, is indicative of ‘structural’ changes of the
paramagnetic solute caused by dilution (see, for example,
Ref. 10).
deviation of the NMR paramagnetic solvent shift vs con-
centration curve from that of the linear dilution plot.11,12
Some of the possible transformations that may be deduced
by this criterion are (1) determination of stoichiometry,13
(2) changes in the state of ion pairing, i.e. tightening, loos-
ening or separation of the ion pair,12 (3) changes in the
degree of aggregation or in the structure (i.e. whether
tight or loose) of the aggregate,10 (4) disproportionation
into a dianion and the corresponding parent molecule,14
and (5) decomposition.14 Earlier reports from this labora-
tory have dealt with, inter alia, the effect of added good
and poor cosolvents15 or magnesium 2-ethoxyethoxide16 on
the shape of the shift versus concentration relationship of
alkali metal benzophenone and fluorenone radical anions.
In the latter cases, however, the origin of the changes
in the shift vs concentration relationship was not ade-
quately addressed. Relevant theoretical studies have also
recently been carried out concerning the solvation mech-
anism of benzophenone radical anion,17 while another
investigation which also impinges on this paper relates
to the kinetics of the electroreduction of benzophenone at
a mercury ultramicroelectrode in seven different aprotic
solvents.18
Kinetic measurements were carried out on solutions
of the alkali metal monosubstituted benzophenone radical
anion generated in situ by electron transfer from the
corresponding alkali metal naphthalene radical anion and
following the paramagnetic solvent NMR shift with respect
to time.
In this work, we employed 1H NMR spectroscopy in
an attempt to probe the stability of certain alkali metal
aromatic ketone radical anions in tetrahydrofuran (THF)
with respect to the perturbation caused by the addition
of poorly solvating cosolvents such as diethyl ether and
triethylamine. The specific objective was to improve our
understanding of the effect of the substituent, the cation and
the variation of concentration plus the solvent composition
on the structure of monosubstituted benzophenone radical
anions in solution. In addition, we examined the speed with
which alkali metal monosubstituted benzophenone radical
anions attain their ‘equilibrium’ structure.
Specific examples
2-Methylbenzophenone radical anions
Lithium 2-methylbenzophenone radical anion in THF gave,
on dilution with THF, shift versus concentration relation-
ships for both the ˛- and ˇ-proton resonances of THF which
were slightly convex. Dilution with triethylamine led to a
shift versus concentration relationship exhibiting a barely
noticeable concavity. An analogous situation occurred on
dilution with diethyl ether.
Lithium 2-methylbenzophenone radical anion in THF
generated by electron transfer from lithium naphthalene
radical anion in the same solvent gave a paramagnetic shift
invariant with time.
Sodium 2-methylbenzophenone radical anion gave linear
shift versus concentration plots on dilution with THF, tri-
ethylamine or diethyl ether. Sodium 2-methylbenzophenone
radical anion generated by electron transfer exhibited a shift
which did not change with time.
RESULTS
Data collection, general remarks
Data were collected by adding increments of solvent or
cosolvent to a standardized THF solution of the substi-
tuted benzophenone radical anion and recording the shift
of the corresponding ˛- or ˇ-proton resonance of THF,
which in turn was plotted against the formal concentra-
tion of the radical anion. Formal concentration refers to
the concentration of the monomeric species regardless of
whether or not the actual species in solution is an aggre-
gate. In the case of dilution with diethyl ether, owing to
interfering resonance signals, the shift could be referred
to the ˇ-protons of THF only. For comparison reasons, in
dilution experiments with triethylamine, the shift was also
referred to the ˇ-proton resonance signal of THF. From
dilution experiments with THF, the shift vs concentration
lines afforded the relevant molar paramagnetic solvent NMR
shifts, whereas the υ vs concentration relationships from the
dilution experiment with diethyl ether and triethylamine
were used to extract the initial slopes. Molar paramagnetic
solvent shifts and initial slopes (or slopes) are summarized
in Table 1. Concave or convex dilution lines were fitted
Potassium 2-methylbenzophenone radical anion behaved
in a similar way to the sodium analogue.
Sodium 3-methylbenzophenone radical anion generated
by electron transfer from sodium naphthalene radical
anion gave a paramagnetic solution, whose paramagnetism
decayed slowly up to complete disappearance. Addition of
fluorenone caused reappearance of paramagnetism.
Table 1 summarizes the above results together with the
data for the other alkali metal monosubstituted benzophe-
none radical anions. The table also includes information
regarding (1) the solubility of the radical anion in THF,
(2) the magnetic state of the solution, (3) the form of the
υ vs concentration relationship, (4) the molar paramagnetic
solvent shift referred to the ˛-proton resonance of THF in
solution containing THF only, (5) the initial slope or slope of
the υ vs concentration relationship referred to the ˇ-proton
resonance signal of THF and (6) the speed with which the
radical anion attains its equilibrium structure.
Copyright 2001 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2001; 39: 586–592