The redox chemistry of 4-benzoyi-N′-methyipyridinium cations in acetonitrile with and without proton donors: The role of hydrogen bonding

Article abstract of DOI:10.1021/jp0107199

In anhydrous CH3CN, 4-benzoyl-W-methyIpyridinium cations undergo two reversible, well-separated (ΔE_1/2 ～ 0.6 V) one-electron reductions in analogy to quinones and viologens. If the solvent contains weak protic acids, such as water or alcohols, the first cyclic voltammetric wave remains unaffected while the second wave is shifted closer to the first. Both voltammetric and spectroelectrochemical evidence suggest that the positive shift of the second wave is due to hydrogen bonding between the two-electron reduced form of the ketone and the proton donors. While the one-electron reduction product is stable both in the presence and in the absence of the weak-acid proton donors, the two-electron reduction wave is reversible only in the time scale of cyclic voltammetry. Interestingly, at longer times, the hydrogen bonded adduct reacts further giving nonquaternized 4-benzoylpyridine and 4-(a-hydroxybenzyl)pyridine as the two main terminal products. In the presence of stronger acids, such as acetic acid, the second wave merges quickly with the first, producing an irreversible two-electron reduction wave. The only terminal product in this case is the quatemized 4-(α-hydroxybenzyl)-N-methyIpyridinium cation. Experimental evidence points toward a common mechanism for the formation of the nonquaternized products in the presence of weaker acids and the quaternized product in the presence of CH₃CO₂H.

Full text of DOI:10.1021/jp0107199

J. Phys. Chem. B 2001, 105, 3663-3674
3663
The Redox Chemistry of 4-Benzoyl-N-methylpyridinium Cations in Acetonitrile with and
without Proton Donors: The Role of Hydrogen Bonding
Nicholas Leventis,*,† Ian A. Elder, Xuerong Gao, Eric W. Bohannan,
Chariklia Sotiriou-Leventis,*,‡ Abdel Monem M. Rawashdeh, Travis J. Overschmidt, and
Kimberly R. Gaston
Department of Chemistry, The UniVersity of Missouri at Rolla, 142 Schrenk Hall, Rolla, Missouri 65409-0010
ReceiVed: February 26, 2001
In anhydrous CH3CN, 4-benzoyl-N-methylpyridinium cations undergo two reversible, well-separated (∆E1/2
∼ 0.6 V) one-electron reductions in analogy to quinones and viologens. If the solvent contains weak protic
acids, such as water or alcohols, the first cyclic voltammetric wave remains unaffected while the second
wave is shifted closer to the first. Both voltammetric and spectroelectrochemical evidence suggest that the
positive shift of the second wave is due to hydrogen bonding between the two-electron reduced form of the
ketone and the proton donors. While the one-electron reduction product is stable both in the presence and in
the absence of the weak-acid proton donors, the two-electron reduction wave is reversible only in the time
scale of cyclic voltammetry. Interestingly, at longer times, the hydrogen bonded adduct reacts further giving
nonquaternized 4-benzoylpyridine and 4-(R-hydroxybenzyl)pyridine as the two main terminal products. In
the presence of stronger acids, such as acetic acid, the second wave merges quickly with the first, producing
an irreversible two-electron reduction wave. The only terminal product in this case is the quaternized 4-(R-
hydroxybenzyl)-N-methylpyridinium cation. Experimental evidence points toward a common mechanism for
the formation of the nonquaternized products in the presence of weaker acids and the quaternized product in
the presence of CH3CO2H.
1. Introduction
cations.17 For such radicals, possibilities for disproportionation
have not been excluded,18 and although coupling of 4-substituted
In general, chemical or electrochemical reduction of ketones
yields alcohols.1 More specifically, in aprotic media electro-
chemical reduction gives primarily pinacol dianions through
radical-radical, radical-ketone or ion-ketone coupling.2 In the
presence of proton donors such as acetic acid, two-electron (2-
e) reduction also yields carbinols.2g,3 The same reactivity pattern
is demonstrated by pyridine-based aromatic ketones. For
instance, one-electron (1-e) reduction of 4-acetylpyridine is
followed by dimerization leading to the corresponding pinacol,
while 2-e reduction at pH ) 6.5 leads to methyl-4-pyridyl-
methanol.4
In contrast, pyridinium salts demonstrate chemically reversible
electrochemical behavior upon reduction. Notable examples
include the NAD+/NADH couple that functions as an electron-
transfer catalyst from flavoenzymes (iron-containing flavopro-
teins) to ATP in the respiratory chain,5 and the various viologens
(N,N′-diquaternized-4,4-bipyridinium salts).6 The latter have
been employed widely as redox mediators,7 electrochromic
materials,8,9 electron acceptors in photoinduced electron trans-
fers,10 and redox probes in self-assembled monolayers,11 den-
drimers,12 silicates,13 zeolites,14 and semiconductors.15
Significant attention has been focused also on the 1-e
reduction of 4-acyl-, 4-carboxy-, 4-carbomethoxy-, 4-cyano- and
4-benzoylpyridinium cations, partly owing to the stability of
the resulting radicals.16 ESR spectroscopy, dipole moment
considerations and chemical trapping experiments, point toward
a pyridinium-localized radical as the 1-e reduced form of these
pyridinium rings at the 4-position is generally considered
sterically hindered, it has been observed; on the other hand,
lack of coupling at the 2- and 6-positions has been difficult to
reconcile.19 Overall, the chemical stability and the intense color
of the 1-e reduced form of 4-benzoylpyridinium cations has led
to their consideration as inter- and intramolecular photoinduced
electron-transfer acceptors,16,20 and as electrochromic materials.21
Owing to disproportionation,18 however, the 1-e reduced form
is expected to exist in equilibrium with low concentrations of
the 2-e reduced species; therefore, the long-term survivability
of the former depends on the chemical stability of the latter. In
this context, a careful literature survey suggests that the
generation and fate of the 2-e reduced form of this class of
compounds is still lacking.
The present study comprises the first detailed investigation
of the redox chemistry in CH3CN of 1 and its analogues 2 and
* To whom correspondence should be addressed.
† Phone: (573) 341-4391. E-mail: leventis@umr.edu.
‡ Phone: (573) 341-4353. E-mail: cslevent@umr.edu.
10.1021/jp0107199 CCC: $20.00 © 2001 American Chemical Society
Published on Web 04/05/2001

3664 J. Phys. Chem. B, Vol. 105, No. 17, 2001
Leventis et al.
3. The latter two compounds have been synthesized as models
of 1 linked to conjugated systems for applications in molecular
wires.
For compound identification, 1H and 13C NMR spectra were
obtained using a Varian INOVA 400 MHz NMR spectrometer,
and are referenced versus TMS. Infrared spectra were recorded
on a Nicolet Magna-IR model 750 spectrophotometer. GC-
MS was done with a Hewlett-Packard Corp. model 5890
instrument, equipped with a J&W Scientific DB-5 MS column
(3.0 m, O. D.)0.25 mm, i.d. ) 0.25 µm) attached to a Hewlett-
Packard Corp. mass selective detector (MSD) model 5972 (EI,
scan mode), using helium as the carrier gas. UV-vis spectra
were recorded with an Ocean Optics, Inc., model CHEM2000
miniature fiber optic spectrophotometer. Elemental analyses
were performed by Oneida Research Services, Inc., Whiteboro,
NY.
Our data shows that, in CH3CN, 4-benzoyl-N-methylpyri-
dinium salts possess a second reVersible 1-e reduction wave
which moves to more positive potentials in the presence of
proton donors. On the basis of voltammetric and spectropho-
tometric evidence, this behavior, previously unknown for this
class of compounds, is attributed to hydrogen bonding of the
2-e reduced forms with the proton donors, and is reminiscent
of what is demonstrated by quinones in nonaqueous media.22
In contrast to quinones, however, with weaker acids (pKa>15),
and at time scales much longer than the time scale of the cyclic
voltammetry, the hydrogen-bonded assemblages lead to proton
transfer and eventual formation of nonquaternized 4-(R-hy-
droxybenzyl)pyridine and 4-benzoylpyridine. With acetic acid,
proton transfer is faster, taking place within the time-scale of
the cyclic voltammetry and leading to the quaternized 4-(R-
hydroxybenzyl)-N-methylpyridinium cation. From a practical
viewpoint, the important conclusion of this study is that if proton
donors are meticulously excluded, 4-benzoylpyridinium salts
can be used as reversible 1-e acceptors in lieu of either quinones
or viologens.
Electrochemical digital simulations were carried out on an
iMac G3 Macintosh computer (64 MB RAM), using the Absoft
Corporation FORTRAN 77 Compiler 4.4, as described else-
where.23a
Materials. All starting materials and reagents were purchased
from Acros or Aldrich and were used as received unless noted
otherwise. All reactions were carried out under nitrogen in
flame-dried glassware. Bromobenzene was washed consecu-
tively with concentrated H2SO4, 10% NaOH, and H2O followed
by drying over CaCl2 and vacuum distillation from sodium
pellets using a glass helix packed column. Diisopropylamine
was distilled from NaOH. Phenylacetylene was distilled through
a vigreux column. Compound 1 was prepared as described
before.7b,23a
2. Experimental Section
Methods. All electrochemical experiments were carried out
at room temperature (23 ( 1 °C) with an EG&G 263A
potentiostat controlled by the EG&G model 270/250 research
electrochemistry software 4.30. Data were plotted using a
Gateway 2000 computer and the Microcal Origin 4.1 software
package. Cyclic voltammetry was conducted with an Au disk
working electrode (1.6 mm in diameter, 0.0201 cm2), and an
aqueous Ag/AgCl reference electrode, both commercially avail-
able from Bioanalytical Systems, Inc., West Lafayette, IN. A
Au foil (2.5 cm2) was used as counter electrode. All cyclic
voltammograms have been 80% compensated for the solution
resistance. Our bulk electrolysis procedures have been described
elsewhere.23a Here, bulk electrolysis was conducted in conjunc-
tion with NMR spectroscopy for identification of terminal
products; for that purpose NaClO4 was chosen as an NMR-
silent supporting electrolyte. At the end of the electrolysis period,
the contents of the working compartment of the H-cell were
transferred in a pear-bottom flask, the solvent (CH3CN/proton
donor) was removed with a rotary evaporator, the solids were
dried under vacuum, CD3CN was added, and the NMR spectrum
was recorded. Spectroelectrochemistry was conducted with a
dual ITO-electrode thin-layer-cell as described before (cell
thickness∼190 µm).7a All electrolytic solutions were degassed
with N2 or Ar.
All volumetric glassware was rinsed with acetone, washed
with MicroTM cleaning solution in boiling water, rinsed with
copious amounts of distilled water, and dried at 150 °C. The
Au foil electrode was cleaned in a H2O2/c.H2SO4 (1:4 v/v)
solution and oven-dried. All working electrodes were succes-
sively polished with 6, 3, and 1 µm diamond paste, washed
with water and acetone, and then air-dried.
Anhydrous CH3CN was purchased from Aldrich. Tetrabutyl-
ammonium perchlorate (TBAP) was used as a supporting
electrolyte, and was prepared from an aqueous solution of
tetrabutylammonium bromide (Aldrich) and 70% HClO4. The
resulting perchlorate precipitate was washed with copious
amounts of deionized water (∼20 l), freeze-dried, recrystallized
from methanol/ether, freeze-dried again and finally dried under
vacuum at 80 °C.
4-(4-Bromobenzoyl)pyridine (4). Isonicotinic acid (15.012
g, 121.94 mmol) in thionyl chloride (50 mL) was stirred at 100
°C for 1.5 h. Excess SOCl2 was removed under reduced
pressure, and bromobenzene (80 mL) was added to the yellow
residue. Upon cooling in an ice-bath, AlCl3 (58 g, 0.43 mol)
was added and the mixture was allowed to slowly return to room
temperature. The solution was then heated to 90 °C and stirred
for 4 h. The reaction was quenched by slowly pouring the
reaction mixture into 400 mL of a 4.2% w/v HCl/ice-water
solution. A pH ) 4 was reached with the addition of Na2CO3
and then a saturated aqueous NaOH solution was added to attain
pH ∼ 10. The organic layer was extracted with CHCl3 and dried
with CaCl2. The solvents were removed with a rotary evaporator,
and the crude product was loaded on a silica column (packed
with hexane) and was eluted first with CH2Cl2 and then with
CH2Cl2/CH3OH, 49:1 v/v. Solvent removal yielded 4 as a white
1
solid (28.4 g, 89%): mp 124-125 °C; H NMR (400 MHz,
CDCl3) δ 8.83 (dd, J ) 4.4 Hz, J ) 1.6 Hz, 2 H), 7.68 (m, 4
H), 7.56 (dd, J ) 4.4 Hz, J ) 1.6 Hz, 2 H); 13C NMR (100
MHz, CDCl3) δ 193.98, 150.45, 143.79, 134.56, 131.98, 131.52,
128.82, 122.66; IR (KBr pellet) ν 3295 (w), 3100 (w), 3073
(w), 3019 (w), 1656 (s), 1582 (s), 1542 (m), 1420 (m) cm-1
;
GC-MS m/z (rel intensity) 263 (M+, 52), 261 (55), 185 (100),
183 (99), 157 (59), 155 (57), 76 (64), 51 (74).
4-(4-[(Trimethylsilyl)ethynyl]benzoyl)pyridine (5). Tri-
methylsilylacetylene (1.694 g, 17.25 mmol), dissolved in
diisopropylamine (5 mL), was added to a solution of 4-(4-
bromobenzoyl)pyridine (2.997 g, 11.43 mmol), dichlorobis(tri-
phenylphosphine)palladium(II) (0.153 g, 0.218 mmol), and CuI
(0.041 g, 0.22 mmol) in benzene (20 mL) and diisopropylamine
(60 mL). After stirring overnight (14 h), the solvents were
removed by rotary evaporation. The residue was passed through
a column of silica gel (packed with hexane) using CH2Cl2 for
elution. Two recrystallizations from hexane yielded light-brown
transparent crystals: mp 85-87 °C; 1H NMR (400 MHz,
CDCl3) δ 8.92 (br s, 2 H), 7.76 (m, 2 H), 7.58 (m, 4 H), 0.28
(s, 9 H); 13C NMR (100 MHz, CDCl3) δ 194.43, 150.57, 144.25,

4-Benzoyl-N-methylpyridinium Cations
J. Phys. Chem. B, Vol. 105, No. 17, 2001 3665
135.33, 132.27, 130.13, 128.71, 123.05, 103.94, 99.13, -0.20;
IR (KBr pellet) ν 3315 (w), 3066 (w), 2959 (m), 2905 (w),
(m), 1500 (w), 1450 (w), 1400 (s) cm-1; GC-MS m/z (rel
intensity) 283(M+, 68), 205 (100), 176 (84), 150 (41).
4-(4-[(Phenyl)ethynyl]benzoyl)-N-methylpyridinium tet-
2160 (s), 1656 (s), 1595 (s), 1555 (m), 1401 (s), 1280 (s) cm-1
;
-
rafluoroborate (3). (CH3)3O+BF4 (0.409 g, 2.77 mmol)
GC-MS m/z (rel intensity) 279 (M+, 28), 264 (100).
dissolved in 10 mL CH3NO2 was added dropwise to a solution
of 4-(4-[(phenyl)ethynyl]benzoyl)pyridine (0.502 g, 1.77 mmol)
in CH3NO2 (60 mL) at room temperature. The solution was
stirred for 20 min, and at the end of this period, the product
was precipitated by addition of diethyl ether. The crude product
was collected by vacuum filtration and recrystallized from water
4-(4-Ethynylbenzoyl)pyridine (6). The entire amount of 5
from above was dissolved in CH3OH (80 mL) along with a
catalytic amount of NaOH (0.027 g, 0.675 mmol). The solution
was stirred at room-temperature overnight (14 h) and then
poured into H2O (115 mL). Diethyl ether (50 mL) and brine
(100 mL) were added and the layers were separated. The
aqueous layer was further extracted with diethyl ether (3 × 100
mL). The combined organic layers were dried with MgSO4 and
filtered. Rotary evaporation of the solvents left a solid which
was sublimed at 100 °C under vacuum to yield 6 as white
needles (1.91 g, 81% overall yield from 4): mp 119-121 °C;
1H NMR (400 MHz, CD3CN) δ 8.78 (dd, J ) 4.4 Hz, J ) 1.6
Hz, 2 H), 7.8-7.6 (m, 4 H), 7.57 (dd, J ) 4.4 Hz, J ) 1.6 Hz,
2 H), 3.65 (s, 1 H); 13C NMR (100 MHz, CD3CN) δ 195.45,
151.40, 144.90, 136.92, 133.17, 131.10, 127.96, 123.67, 83.27,
82.48; IR (KBr pellet) ν 3300 (w), 3160 (s), 3046 (m), 2100
(s), 1660 (s), 1602 (s), 1548 (s), 1414 (s) cm-1; GC-MS m/z
(rel intensity) 207 (M+, 46), 129 (100), 101 (51), 75 (33), 51
(48). Anal. Calcd for C14H9NO: C, 81.14; H, 4.38; N, 6.76.
Found: C, 81.39; H, 4.48; N, 6.79.
1
to yield 3 as a white solid (0.355 g, 52%): H NMR (400 MHz,
CD3CN) δ 8.82 (d, J ) 6.4 Hz, 2 H), 8.18 (d, J ) 6.4 Hz, 2
H), 7.84 (dd, J ) 6.7 Hz, J ) 1.9 Hz, 2 H), 7.74 (dd, J ) 6.7
Hz, J ) 1.9 Hz, 2 H), 7.60 (m, 2 H), 7.45 (m, 3 H), 4.4 (s, 3
H); 13C NMR (100 MHz, CD3CN) δ 192.02, 152.79, 147.44,
134.72, 132.93, 132.75, 131.51, 130.47, 129.85, 128.28, 123.16,
118.36, 94.64, 89.00, 49.65; IR (KBr pellet) ν 3140 (w), 3110
(w), 3075 (w), 3066 (w), 2210 (w), 1667 (s), 1575 (w), 1480
(m), 1413 (m) cm-1. Anal. Calcd for C21H16NOBF4: C, 65.50;
H, 4.19; N, 3.64. Found: C, 65.34; H, 3.94; N, 3.68.
4-(r-Hydroxybenzyl)pyridine (8).24 Sodium borohydride
(2.4 g, 63 mmol) was added to a solution of 4-benzoyl-
pyridine (9.09 g, 49.6 mmol) in methanol (125 mL). The mixture
was stirred at room temperature under nitrogen for 2 h. Methanol
was then removed under reduced pressure and the solid was
washed with 50 mL of water and dissolved in 200 mL of ethyl
acetate. The water layer was extracted with two 50 mL portions
of ethyl acetate. The ethyl acetate layers were combined and
dried with sodium sulfate. Ethyl acetate was removed under
reduced pressure, and the solid was triturated with hexane. The
solution was filtered and the precipitate was dried in the vacuum
oven at 40 °C overnight to give 8 (8.4 g, 91%): mp 120-122
°C; 1H NMR (400 MHz, CD3CN) δ 8.48 (dd, J ) 1.6 Hz, J )
4.6 Hz, 2 H), 7.32 (m, 7 H), 5.76 (d, J ) 3.6 Hz, 1 H), 4.11 (d,
J ) 3.6 Hz, 1 H); 13C NMR (100 MHz, CD3CN) δ 154.75,
150.61, 144.93, 129.55, 128.64, 127.57, 122.13, 74.92; IR (KBr
pellet) ν 3400 (br, m), 3150 (br, s), 2825 (m), 1608 (s), 1562
(s), 1461 (s), 1414 (m), 1199 (m), 1051 (s), 1004 (s), 790 (m),
770 (s), 705 (s), 660 (s), 605 (s) cm-1. Anal. Calcd for C12H11-
NO: C, 77.81; H, 5.99; N, 7.56. Found: C, 77.88; H, 6.03; N,
7.52.
4-(4-Ethynylbenzoyl)-N-methylpyridinium perchlorate (2).
4-(4-Ethynylbenzoyl) pyridine (1.914 g, 9.24 mmol) and CH3I
(5.327 g, 37.53 mmol) were dissolved in 50 mL CH3CN. The
mixture was stirred under reflux for 4 days while new portions
of CH3I (∼5 g, 0.04 mol) were added daily. At the end of the
four-day period, the reaction mixture was cooled to room
temperature, and diethyl ether was added to complete precipita-
tion of the product which was then filtered and dried to yield
2.918 g, (90%) of crude 4-(4-ethynylbenzoyl)-N-methylpyri-
dinium iodide as a light-orange solid. A portion of the product
(1.451 g, 4.16 mmol) was dissolved in H2O. Excess of 70%
HClO4 was added dropwise. The resultant off-white precipitate
was recrystallized once from CH3OH/Et2O and once from H2O
1
to yield white crystals of 2 (0.901 g, 67%): H NMR (400 MHz,
CD3CN) δ 8.81 (d, J ) 6.4 Hz, 2 H), 8.17 (d, J ) 6.4 Hz, 2
H), 7.80 (dd, J ) 6.7 Hz, J ) 2.0 Hz, 2 H), 7.70 (dd, J ) 6.7
Hz, J ) 2.0 Hz, 2 H), 4.39 (s, 3 H), 3.72 (s, 1 H); 13C NMR
(100 MHz, CD3CN) δ 192.06, 152.45, 147.40, 135.20, 133.46,
131.32, 129.18, 128.19, 83.40, 82.98, 49.62; IR (KBr pellet) ν
3240 (s), 3135 (w), 3066 (m), 2173 (w), 2113 (w), 1675 (s),
1605 (s), 1575 (m), 1480 (m), 1400 (m) cm-1. Anal. Calcd for
C15H12NO5Cl: C, 56.00; H, 3.76; N, 4.35. Found: C, 56.14;
H, 3.63; N, 4.36.
4-(r-Hydroxybenzyl)-N-methylpyridinium tetrafluoro-
borate (10). Compound 10 was prepared following an identical
preparative procedure to the one used for the synthesis of 3:
1H NMR (400 MHz, CD3CN) δ 8.51 (d, J ) 6.6 Hz, 2 H), 7.99
(d, J ) 6.6 Hz, 2 H), 7.40 (m, 5 H), 6.01 (s, 1 H), 4.56 (br s,
1H), 4.22 (s, 3 H); 13C NMR (100 MHz, CD3CN) δ 164.62,
146.00, 142.57, 129.94, 129.50, 127.91, 125.66, 74.26, 48.53.
4-(4-[(Phenyl)ethynyl]benzoyl)pyridine (7). Phenylacetylene
(0.670 g, 6.56 mmol) was added dropwise to a stirred solution
of 4 (1.148 g, 4.380 mmol), dichlorobis(triphenylphosphine)-
palladium(II) (0.058 g, 0.083 mmol), and CuI (0.018 g, 0.094
mmol) in a 100 mL mixture of diisopropylamine and benzene
(4:1, v:v). After stirring for 21 h, solvents were removed by
rotary evaporation. The residue was extracted with CH2Cl2 and
passed through a silica column (packed with hexane) using CH2-
Cl2 for elution. Recrystallization from CHCl3 yielded 7 as a
3. Results
3.1. Synthesis of 1-3. Compounds 1-3 were prepared via
quaternization of the corresponding free bases. 4-Benzoyl-
pyridine, the free base of 1, is available commercially, while 6
and 7 were synthesized via a common intermediate, 4-(4-
bromobenzoyl)pyridine (4) (Scheme 1). Compound 4 was
prepared from isonicotinic acid in two steps through a Friedel-
Crafts acylation of bromobenzene with isonicotinoyl chloride,
following standard procedures. Pd(PPh3)2Cl2/CuI-catalyzed
Heck-type coupling reactions between 4 and trimethylsilylacety-
lene or phenylacetylene gave 5 and 7, respectively.25 Compound
5 was deprotected to form the terminal alkyne, 6, using a
catalytic amount of NaOH in methanol. Methyl iodide is a
commonly used quaternizing agent, but since the iodide salts
of the 4-benzoylpyridinium cations considered here are not
soluble in CH3CN, metathesis to the corresponding perchlorates
1
white solid (0.777 g, 63%): mp 147-148 °C; H NMR (400
MHz, CDCl3) δ 8.83 (dd, J ) 4.4 Hz, J ) 1.7 Hz, 2 H), 7.81
(dd, J ) 6.8 Hz, J ) 1.9 Hz, 2 H), 7.66 (dd, J ) 6.8 Hz, J )
1.9 Hz, 2 H), 7.58 (dd, J ) 4.4 Hz, J ) 1.7 Hz, 2 H), 7.56 (m,
2 H), 7.37 (m, 3 H); 13C NMR (100 MHz, CDCl3) δ 189.99,
146.10, 139.90, 130.59, 127.46, 127.38, 125.78, 124.65, 124.52,
124.14, 118.43, 118.15, 89.07, 84.04; IR (KBr pellet): ν 3040
(w), 3019 (w), 2220 (m), 1649 (s), 1600 (s), 1595 (m), 1548

3666 J. Phys. Chem. B, Vol. 105, No. 17, 2001
Leventis et al.
SCHEME 1: Synthesis of 2 and 3
TABLE 1: Redox Data for Compounds 1-3 Obtained with
Cyclic Voltammetry in Anhydrous Acetonitrilea
∆Ep-p(1)
(mV)
∆Ep-p(2)
(mV)
compound
E1/2(1) (V)
E
1/2(2) (V)
1
-0.626 ( 0.011 69 ( 3b -1.260 ( 0.010 72 ( 4b
70 ( 4c 73 ( 4c
-0.582 ( 0.005 70 ( 5b -1.190 ( 0.010 74 ( 4b
71 ( 4c 73 ( 6c
-0.592 ( 0.003 66 ( 2b -1.188 ( 0.008 71 ( 3b
71 ( 1c 73 ( 1c
2
3
a E1/2’s are reported versus an aq Ag/AgCl reference electrode. b At
100 mV/s. c At 40 mV/s.
electron wave approximately 0.1-0.3 V more negative than the
first,21 it was deemed important to investigate the effect of
different amounts of proton donors in the electrolytic solution.
Figure 1. Electrochemical response at 40 mV/s of 1 (s, 2.82 mM), 2
(- - -; 2.89 mM), and 3 (‚ ‚ ‚; 2.93 mM) in CH3CN/0.5 TBAP using a
Au disk electrode (0.0201 cm2).
Perhaps not for obvious reasons, but intuitively of little
surprise is the fact that the effect of proton donors on the redox
chemistry of 1-3 depends on their chemical identity. The proton
donors used in this study, and their pKa values in water are
presented in Table 2.
At concentrations up to 1.0 M, phenylacetylene does not have
a significant effect on the voltammograms of 1-3. The rest of
the proton donors are separated into two groups: (Group I) tert-
butyl alcohol (t-BuOH), 1-butanol (n-BuOH), H2O, and CH3-
OH; and (Group II) diethyl malonate, phenol, and acetic acid.
or tetrafluoroborates, which are both soluble in CH3CN, is
required. Alternatively, using trimethyloxonium tetrafuoroborate
as the methylating agent yields directly the tetrafluoroborate
salt and metathesis is not necessary.26
3.2. Electrochemical Reduction of 1-3. All initial redox
characterization of compounds 1-3 was conducted in anhydrous
CH3CN, where they all demonstrate two well-separated 1-e
reduction waves (Figure 1). The peak-to-peak separation (∆Ep-p
)
The common effect of the group I proton donors is demon-
strated with 2 and methanol in Figure 2: as the concentration
of CH3OH increases, the first wave remains unaffected, while
the second one moves to more positive potentials. The two
waves retain their relative sizes and their peak-to-peak separa-
tions, meaning that both waves still correspond to 1-e reductions.
However, as Figure 3 demonstrates, the relative effectiveness
in both waves is ∼70 mV and rather insensitive to the potential
sweep rate, at least in the range 40-120 mV/s. Table 1
summarizes the electrochemical data for compounds 1-3.
Since the only other reference to a second electron reduction
for this class of compounds reports that in aqueous electrolytes
N-alkyl-4-benzoylpyridinium salts possess an irreversible second

4-Benzoyl-N-methylpyridinium Cations
J. Phys. Chem. B, Vol. 105, No. 17, 2001 3667
TABLE 2: Proton Donors Used in This Study and Their pKa Values toward Water
Ph-CtCH
23.227
t-butanol
1728a
1-butanol
∼1628a
H2O
15.7428a
CH3OH
15.228a
CH2(CO2Et)2
1328a
phenol
9.929
CH3CO2H
4.7529
Figure 2. Electrochemical response of 2 (2.67 mM) at 40 mV/s in
CH3CN/0.5 M TBAP containing varying amounts of CH3OH. (;)
[CH3OH] ) 0.0 M, (- - -) [CH3OH] ) 0.34 M, (-‚-) [CH3OH] )
1.44 M, (‚ ‚ ‚) [CH3OH] ) 6.83 M.
Figure 4. Spectroelectrochemistry of 1 (10 mM) using a dual ITO-
electrode thin layer cell (∼190 µm thick) in the absence (A-C) and in
the presence (D-F) of 1.5 M of n-butanol.
More light is shed on the effect of group I proton donors on
the redox chemistry of 1-3 by thin-layer-cell (TLC) spectro-
electrochemistry. Figure 4 shows representative results obtained
with 1 in the absence and the presence of a typical group I proton
donor (n-BuOH), while all spectroelectrochemical data are
summarized in Table 4. It is observed that the 1-e reduction of
1 in the absence or presence of group I proton donors yields
practically identical absorption spectra with λmax∼507 nm
(Figure 4A,D). The red species responsible for that absorption
is stable for at least 20 min under inert atmosphere, and upon
reoxidation at -0.3 V vs Ag/AgCl the colorless solution of 1
is recovered completely. Although cyclic voltammetry (Figure
2) suggests that both the one- and the two-electron reduction
products of 1-3 are stable within the time-scale of cyclic
voltammetry, spectroelectrochemistry now indicates that at least
the 1-e reduction product of 1 does not pinacolize or react with
group I proton donors even at much longer periods.
Reduction of 1 beyond its second wave (e.g., at -1.4 to -1.5
V vs Ag/AgCl) both with and without group I proton donors
yields an early accumulation of the 1-e reduction product, which
subsequently retreats in favor of the new absorbance corre-
sponding to the yellow 2-e reduced form (Figure 4B,E). It is
noted further that in the presence of the group I proton donors
the λmax of the 2-e reduction product is blue-shifted by about
32-34 nm relative to its absorption in the pure CH3CN-based
electrolyte. As will be defended below, that blue shift is
consistent with a hydrogen bonding interaction between the 2-e
reduced forms of 1-3 and the proton donors.
In the absence of a proton donor, the absorption spectrum of
the 2-e reduction product does not decay, and the original
colorless state is recovered completely by potentiostating the
TLC at -0.3 V vs Ag/AgCl (Figure 4C). In the presence of
group I proton donors, however, the spectrum of the 2-e reduced
Figure 3. Cumulative results demonstrating the effect of varying the
concentration of group I proton donors (H2O, CH3OH, n-BuOH,
t-BuOH) on the position of the second electron reduction wave of 2.
Correlation coefficients g0.97. (Data obtained from experiments similar
to those shown in Figure 2.)
of the group I proton donors in shifting the second wave closer
to the first one is not the same, with H2O or methanol being
the most effective, and t-BuOH the least effective. All relevant
data concerning 1-3 and group I proton donors are summarized
in Table 3.

3668 J. Phys. Chem. B, Vol. 105, No. 17, 2001
Leventis et al.
TABLE 3: Slopes and Intercepts (both in mV) of the Plots of ∆E1/2 vs log[Proton Donor]a,b
proton donor
compound
H2O
CH3OH
n-BuOH
t-BuOH
slope
intercept
slope
intercept
slope
intercept
slope
intercept
1
2
3
-161 ( 1
-148 ( 1
-185 ( 2
455 ( 2
428 ( 1
442 ( 2
-100 ( 1
-117 ( 1
-106 ( 1
463 ( 2
435 ( 2
445 ( 3
-90 ( 1
-74 ( 1
-71 ( 1
489 ( 2
478 ( 2
467 ( 2
-50 ( 1
-48 ( 1
-51 ( 1
561 ( 1
544 ( 1
531 ( 1
a ∆E1/2 ) E1/2(1) - E1/2(2). ∆E1/2 reports the position of the second wave, E1/2(2), with respect to the fixed position of the first one, E1/2(1). b Data
comprise averages of two experiments at two different sweep rates (40 and 100 mV/s).
TABLE 4: Summary of Visible Absorption Data after One-
and Two-Electron Reductions of 1 in the Presence of Group
I Proton Donorsa
1-e reduction
product
2-e reduction
product
H-donor
none
H2O
CH3OH
n-BuOH
λ
max (nm)
λmax (nm)
506
509
509
507
429
398
395
397
a [proton donor])4.7 M.
species evolves slowly: for example in the presence of n-BuOH
the absorbance with λmax ) 397 nm decays, and a new
absorption feature with λmax ) 584 nm grows gradually (Figure
4F). That new feature increases even under open circuit
conditions, and persists even after the thin layer cell is
potentiostated back at -0.3 V. Nevertheless, that absorption
feature begins to decay eventually, leading to a final pale-yellow
solution with no distinct absorption features in the visible.
Practically identical spectroelectrochemical results have been
obtained in the presence of methanol. In the presence of water
the absorption of the 2-e reduction product also decays slowly,
but the transient absorption at ∼580 nm is not observed.
Generally, in the presence of group I proton donors, the 2-e
reduction product reacts slowly toward terminal products having
no distinct absorption in the visible. Separate bulk electrolyses
of 1 at -1.4 V vs aq Ag/AgCl in the presence of 5 M of each
of the group I proton donors reveal a surprisingly clean
1
reaction: solvent removal and H NMR spectroscopy of the
residue in CD3CN shows clearly two nonquaternized major
terminal products, 8 and 9:
1
Figure 5. A. Typical H NMR spectrum in CD3CN of the product
mixture after a 3 h electrolysis of 1 (3 mM) in CH3CN/0.1 M NaClO4
in the presence of 1-butanol (5 M). By comparison with part B of this
Figure, no unreacted 1 has remained at the end of the electrolysis. B.
1H NMR spectrum of the reaction mixture after 1 (3 mM) was stirred
for 3 h at room temperature in CD3CN in the presence of 3 mM
t-BuO-K+. The signals correspond to the R-protons relative to the
N-atom for each compound.
The structures of 8 and 9 were assigned by comparing their
1H NMR signatures with those of authentic samples (the
preparation of 8 is described in the Experimental Section, while
9 is commercially available). Fortuitously, the R-protons relative
to the aromatic N-atom are more downfield than the other
aromatic protons, and with the resolution of the 400 MHz NMR
spectrometer they appear as a doublet in the quaternary salts
and as a doublet of doublets in the free bases. Figure 5A shows
representative 1H NMR data in the region of the R-protons after
electrolysis of 1 in the presence of 1-butanol. The two signals
can be integrated, and the ratio of 8 to 9 is reported in Table 5.
Figure 5B shows the results of a “control” experiment in which
1 was stirred with t-BuO-K+ in CD3CN at room temperature
(23 ( 1 °C) for 3 h; under these conditions, 1 reacts cleanly
with the base producing 9. Similar results have been obtained
with 1 and CH3ONa.
demonstrated in Figure 6. Here, only 0.5 mol-equivalent amount
of the proton donor (dashed lines) versus 1 is enough to generate
a new wave (marked with arrows) between the first and second
electron reductions in the absence of proton donors (solid lines),
while the original second wave is still present but its size has
been diminished. Acetic acid, the stronger acid in the group,
causes the most positive shift of the new intermediate wave.
When the concentration of the group II proton donor increases
to a 5 molar excess vs 1, the original second-electron wave
disappears completely, and the new wave is shifted to even more
positive potentials (Figure 6, dotted lines). Again, that phenom-
enon is more pronounced with acetic acid than with phenol,
which in turn is more effective than diethyl malonate. Specif-
The electrochemical signature of the group II proton donors
(diethylmalonate, phenol and acetic acid) is different, and is

4-Benzoyl-N-methylpyridinium Cations
J. Phys. Chem. B, Vol. 105, No. 17, 2001 3669
TABLE 5: Terminal Product Ratios Relative to Unreacted
Starting Material after 2-e Electrolytic Reductions of 1 in
the Presence of Group I Proton Donorsa-c
H-donor
[1]
[8]
[9]
H2O
1
0
0
0
0
0.44
1
0.18
1
0.80
1
CH3OH
n-BuOH
t-BuOH
0
a In all experiments, [1])3.0 mM in CH3CN/0.1 M NaClO4, [proton
donor] ) 5 M, electrolysis time )3 h, applied potential)-1.4 V vs
Ag/AgCl. The current decays to less than 5% of the original value in
15-20 min; the working electrode, however, was kept under potential
control for the entire 3-h period. The product ratios were determined
from data similar to those of Figure 5A. b A zero value means that the
compound is undetectable by 1H-NMR. c GC-MS analysis of the
reaction mixture after electrolysis revealed no volatile products other
than 8 and 9.
Figure 7. Data supporting a stepwise 2-e and 2-H transfer mechanism
during the irreversible reduction of 1 in the presence of 5-molar excess
of CH3CO2H. A: cyclic voltammetry of 1 (3.02 mM) at 30 V/s in
CH3CN/0.5 M TBAP in the presence of 15.07 mM of CH3CO2H. B.
Spectroelectrochemical signature of 1 (10.08 mM) using a dual ITO
electrode thin layer cell at -0.8 V vs Ag/AgCl in a CH3CN/0.5 M
TBAP solution containing 50.37 mM CH3CO2H. The spectrum was
taken 3 min after the potential had been applied.
permanent absorption feature above 300 nm. Clearly, any
intermediates react quickly and the terminal reduction product
does not absorb in the visible. A weak transient absorption
feature (λmax∼510 nm, A ∼ 0.05, Figure 7B) corresponds to
1
the 1-e reduction product of 1. By comparison of the H and
13C NMR spectra of the reaction mixture from bulk electrolysis
with the spectra of an authentic sample (see Experimental
Section), 4-(R-hydroxybenzyl)-N-methylpyridinium cation (10)
was identified as the only terminal product from the reduction
of 1 in the presence of a 5 M excess of acetic acid. (After a 3
Figure 6. Electrochemical response of 1 at 100 mV/s in CH3CN/0.5
M TBAP in the presence of of group II proton donors. (s) [HA] )
0.0 M, (- - -) [HA]/[1] ) 0.5, (‚ ‚ ‚): [HA]/[1] ) 5. A: HA ) diethyl
malonate, [1] ) 9.81 mM. B: HA ) phenol, [1] ) 3.02 mM. C: HA
) CH3CO2H, [1] ) 3.02 mM.
1
h electrolysis, H NMR analysis showed only two compounds
present: 1 and 10; [1]/[10] ) 1.06.)
ically, with acetic acid as the proton donor, the new wave is so
positive-shifted that it merges with the first one, which now
represents an irreversible 2-e reduction (the current has doubled
and the reoxidation wave has disappeared, Figure 6C). Increas-
ing the potential sweep rate to 30 V/s, we have been able to
regain some reversibility in the 2-e reduction of 1 in the presence
of 5 molar excess of acetic acid (compare Figure 6C, dotted
line, and Figure 7A).
4. Discussion
TLC-electrolysis of 1 in the presence of 0.5 mol equiv amount
of phenol after the first, second, and third waves produces
absorption features corresponding respectively, (a) to the 1-e
reduction product (λmax)507 nm), (b) both the 1- and the 2-e
reduction products (∼507 nm and ∼395 nm), and (c) a broader
spectrum for the 2-e reduction product (λmax∼417 nm).
TLC-electrolysis of a 10 mM solution of 1 in the presence
of ca. 50 mM of acetic acid does not produce any strong or
Although reports involving benzoylpyridinium salts are
scarce,20,21,30 benzoylpyridines, themselves have been investi-
gated for their potential pharmacological properties,31 and have
been synthesized via reaction of cyanopyridines with phenyl
Grignard reagents,31b-d Friedel-Crafts reactions of nicotinoyl
chlorides with substituted benzenes,31a,d,f,g,32 and by the oxidation
of benzylpyridines or 1-phenyl-1-pyridylmethanols with potas-
sium permanganate or chromic acid.31g,33

3670 J. Phys. Chem. B, Vol. 105, No. 17, 2001
Leventis et al.
The synthetic procedure of Scheme 1 was selected because
both 2 and 3 emerge from the common intermediate 4.
Alternatively, 3 can be prepared via a Pd(PPh3)2Cl2/CuI-
catalyzed coupling of bromobenzene with 6, which in turn can
be synthesized from the reaction of 4-vinylphenylmagnesium
bromide with 4-cyanopyridine followed by acid hydrolysis,28b
bromination of the double bond, and dehydrohalogenation. This
procedure, however, is lengthier and the yield of 6 is very low
(<10%), probably due to polymerization of the Grignard
reagent.34
Within the time-scale of cyclic voltammetry and in the
presence or absence of group I proton donors, 1-3 undergo
two sequential, chemically reversible (ip,c/ip,a≈ 1 ( 0.05 for both
waves), 1-e reductions. The ∆Ep-p values of both waves of all
three compounds are practically independent of the potential
sweep-rate, indicating reversibility. Therefore, in the ensuing
the cyclic voltammograms of 2 and 3 with respect to 1 (Figure
1), the position of the second wave is more sensitive to
substitution than that of the first one. This indicates, on the basis
of the Hammett-type of argumentation,28c that the first electron
goes into the pyridinium ring, while the second electron
generates a negative charge on the carbonyl group, which is
closer to the point of substitution.
Spectroelectrochemistry provides further understanding for
the redox chemistry of 1-3 by showing clear evidence for a
homogeneous comproportionation reaction between 12 and 1
producing 11, according to eq 3. Although the red 1-e reduction
product, 11, is incompatible with the electrode held at potentials
beyond the second reduction wave (e.g., at -1.5 V vs Ag/AgCl,
Figure 4B), nevertheless it is still produced transiently on the
way to yellow 12. Also, the similar spectral evolution of Figures
4B and E suggests that comproportionation reactions proceed
unobstructed, and 11 is formed irrespective of the chemical
identity of the 2-e reduced form in the absence or presence of
group I proton donors.
discussion the cyclic voltammetric half-wave potentials, E1/2
,
of both reduction waves are assumed to correlate with the
concentrations of the involved species via the Nernst equation.
In the absence of proton donors, the redox chemistry of 1-3
(X ) H-, H-CtC-, and Ph-CtC-) is described by eqs 1
and 2. The two possible 1-e reduced products, 11 and 11a, are
not canonical forms: 11 is a nonaromatic neutral species, while
the zwiterionic radical 11a is the typical 1-e reduction product
expected from a ketone.
Group I proton donors shift positive only the second reduction
waves of 1-3, in a way that depends on their chemical identity
and concentration (Figure 3 and Table 3). According to the
literature, analogous behavior has been encountered with
quinones in nonaqueous media, and has been debated for
decades; it was recognized early that an ECE type mechanism
(initial electron transfer, E, followed by a homogeneous
chemical reaction, C, and a second electron transfer, E) had to
be excluded, because it would affect the position and the shape
of both waves.35 Subsequently, attention was focused on an EEC
type of mechanism, but the older literature commonly implied
that a direct proton transfer was involved immediately after the
second electron reduction. If that model was to be adopted for
explaining the positive shift of the second wave of 1-3 in the
presence of group I proton donors, it would involve protonation
of 12 according to eq 4. However, digital simulations show that
eq 4 introduces chemical irreversibility in both waves, and that
has been traced to the limited supply of A-, which is produced
and consumed via eq 4. It is thereby concluded that in order
for eq 4 to keep the supply of 12 at the level demanded by eq
2, a large excess of A- should be present in the electrolytic
solution at all times.
It is evident that in order to preserve the reversibility of both
waves of 1-3 and explain the positive shift of E1/2(2) in the
presence of group I proton donors, the interaction between 12
and the proton donors should not produce the conjugate base
A-. In this context, comparison of the spectroelectrochemical
data of Figures 4B and E suggests development of hydrogen
bonding between 12 and the group I proton donors: the blue-
shifted absorption of the 2-e reduction product in the presence
of the group I proton donors, indicates that the HOMO of 12
has been stabilized by about 5.4 kcal/mol, which is a typical
value for a hydrogen bonding interaction.28d (In contrast, the
absorption spectrum of 11 is not affected by the presence of
group I proton donors: see Table 4 or compare Figures 4A and
D.) From a purely voltammetric perspective, it has been
proposed that hydrogen bonding between proton donors and the
oxygen of a carbonyl group renders the latter more easily
reducible and creates polarographic prewaves in the electro-
chemical reduction of ketones.36 Hydrogen bonding has been
Both isomers 11 and 11a yield the same ionic species 12 by a
second electron reduction. According to the literature, the nature
of the 1-e reduced form of other 4-substituted pyridinium salts
conforms to an 11-type structure.17 It is also noteworthy that in
analogy to 11a the 1-e reduction product of 4-acetylpyridine
leads to pinacolization,4 while the 1-e reduction of the 4-acetyl-
N-methylpyridinium cation leads only to coupling in the
4-position relative to the aromatic nitrogen, obviously via a
radical analogous to 11.19 In this realm, our spectroelectro-
chemical data show that the chemical stability of the 1-e reduced
form of 1 extends well beyond the time-scale of the cyclic
voltammetry, and structure 11 provides an adequate explanation
for both the lack of pinacolization and the lack of coupling due
to steric hindrance. Furthermore, by comparative analysis of

4-Benzoyl-N-methylpyridinium Cations
J. Phys. Chem. B, Vol. 105, No. 17, 2001 3671
considered also between phenol and the 1-e reduction product
of 1-hydroxy-9,10-anthraquinone as a possible explanation for
a new polarographic wave (between the two original ones) in
the presence of a stoichiometric amount of phenol.35d However,
digital simulations show that any interaction of the 1-e reduced
form affects the position and slope of both waves, leaving the
2-e reduced form as the only candidate for developing hydrogen
bonding that would affect the position of only the second wave.
In this regard, the electrochemistry of quinones was revisited
recently by Cupta et al., who introduced hydrogen bonding into
digital simulations in order to explain their voltammetry in
nonaqueous media.22
Accepting that only the 2-e reduced forms of 1-3 develop
hydrogen bonding with the group I proton donors, there are
four possible mechanisms explaining the positive shift of the
second wave. 12--HAm represents the hydrogen-bonded adduct
Figure 8. Digital simulations demonstrating the effect of [HA] and
of “m” on the cyclic voltammetry of a redox-active species (0.001 M)
reacting according to mechanism I. [HA] is in M. The simulations were
conducted using an iteration time interval of 1 ms. The diffusion
coefficients of all three redox states were assumed equal so the cyclic
voltammograms are undistorted by the homogeneous comproportion-
ation reaction between the redox-active species and its 2-e reduced
form.23
On the basis of eqs 5 and 6, the value of E1/2(2) in the presence
of HA is given for mechanisms III and IV by eqs 7 and 8,
respectively.
of 12 to either “m” molecules of HA or a mth member aggregate
of HA (HAm). Mechanisms I and II assume preassociation of
the proton donors in aggregates, and should become important
at higher concentrations of HA. In relatively dilute solutions of
HA, mechanisms III and IV are expected to predominate. For
example, if HA ) H2O, Bertie et al. have reported that when
the mole-fraction of water in acetonitrile is less than ∼0.15 (or
[H2O] < ∼3 M), there are “enough CH3CN molecules to
surround each water molecule, and the water molecules increas-
ingly form just a single hydrogen bond to one of the surrounding
CH3CN molecules.”37 Mechanism III is a limiting case of
mechanism IV for very fast establishment of the equilibrium
between 12 and HA. Digital simulations (Figure 8) provide
supporting evidence for both mechanisms by showing that
mechanism III for example, is fully capable of reproducing the
main facets of Figures 2 and 3, and Table 3; namely that, in
the presence of an associative interaction (e.g., hydrogen
bonding) between 12 and HA, the first wave remains unchanged,
whereas the second one moves to more positive potentials with
(a) increasing concentration of HA and (b) increasing number
of HA molecules entering the association. Now, which one of
the two mechanisms, III or IV, is actually the correct one is
decided as follows.
E1/2(2) ) E1/2°(2) + m 0.059 log[HA]
(7)
E1/2(2) ) E1/2°(2) + 0.059 log K + m 0.059 log[HA] (8)
According to eqs 7 and 8, E1/2(2), irrespective of mechanism,
should vary linearly with log[HA], and the slope should depend
on the number of HA molecules participating in the 12- -HAm
assemblages. The same argument has been advanced by
Ravenga et al. for the redox behavior of anthraquinone in
aqueous media,35f and by Cupta et al. for the electrochemistry
of quinones in nonaqueous media.22 If mechanism III is correct,
the intercept of ∆E1/2 vs log[HA] should be independent of the
identity of the acid, while the opposite should be valid if
mechanism IV is correct. In fact, the subtle difference between
mechanisms III and IV is that in the latter, the shift of the E1/2(2)
depends also on the equilibrium constant, K, of the interaction
of 12 with HA. That equilibrium constant depends on the
tendency of HA to form hydrogen bonding, which in turn
depends on the polarization of the H-A bond, and therefore is
related to the acidity of the proton donor. In conclusion, if
mechanism IV is correct, the acid strength should play an
important role in the positive shift of the second wave.
According to Figure 3 and Table 3, the slope of ∆E1/2 vs the
log[HA] varies with the identity of the group I proton donor,
and for compound 3 for example, the values of the slope are
0.86, 1.2, 1.8 and 3.1 times 0.059 V for tert-butyl alcohol,
n-butanol, CH3OH, and H2O, respectively. Because these
multiples of 0.059 V represent the aVerage number of HA
molecules associated with 12 in 12--HAm, this trend most
probably reflects the relative size of the proton donors around
the O-atom. Structures 12a-c illustrate the hydrogen bonded
associations of the 2-e reduced forms of 1-3, following a first
According to mechanisms III and IV, the relationship between
the electrode potential, E, and the ratio of the electrode-surface
concentrations, [11]/[12--HAm], along the second wave are given
at 25 °C by eqs 5 and 6, respectively, where E1/2°(2) is the
half-wave potential of the second wave in the absence of HA.
E ) E1/2°(2) + m 0.059 log[HA] + 0.059 log[[11]/
[12--HAm]] (5)
E ) E1/2°(2) + 0.059 log K + m 0.059 log[HA] +
0.059 log[[11]/[12--HAm]] (6)

3672 J. Phys. Chem. B, Vol. 105, No. 17, 2001
Leventis et al.
significant figure approximation for the participating number
of HA molecules.38
Figure 9. Linear correlation of the ∆E1/2 [)E1/2(1) - E1/2(2)] values
of 1 vs the pKa values of group II proton donors. E1/2(2) refers to the
E1/2 value of the new wave that appears between the original waves
when [HA]/[1] ) 0.5 (refer to Figure 6).
The intercepts of the plots of ∆E1/2 vs log[H2O] and log[CH3-
OH] are numerically close (Table 3), but the intercepts with n-
and tert-butyl alcohol are progressively higher. The equilibrium
constants for the formation of the various 12--HAm adducts are
calculated from the intercepts (Table 3) and the E1/2°(2) values
(Table 1) via eq 9. For compound 1, for example, it is calculated
electrolysis of 1 at -0.8 V in the presence of a 5-molar excess
of CH3CO2H (Figure 7B), support the fact that the 2-e and the
2-H transfer do not take place concurrently, but sequentially. If
10 was formed via a simultaneous 2-e/2-H transfer, the 2-e
reduction wave of 1 in the presence of 5-molar excess of CH3-
CO2H would not show any signs of chemical reversibility at
any sweep rate, because 10 is not oxidizable in the range
between -1.5 and 0 V vs Ag/AgCl. Furthermore, under 2-e
transfer conditions, 11 can only be produced via the compro-
portionation reaction of 1 with the 2-e reduction product or its
hydrogen-bonded form (eq 3). Therefore, if protons and
electrons were transferred concurrently, 11 should be produced
from the reaction of 1 with 10; but as it was stated, 10 is not
oxidizable between -1.5 and 0 V, so its reaction with 1 is not
favored thermodynamically. Based on the above, the most
probable mechanism for the reduction of 1-3 in the presence
of acetic acid is described by eqs 10-13. Equations 10 and 11
leave the possibility open for (a) a large positive shift of the
second wave by hydrogen bonding so that it merges with the
first (Figure 6C), (b) some electrochemical reversibility at higher
sweep rates (Figure 7A), and (c) a comproportionation reaction
that generates 11 transiently (Figure 7B). In this realm, eq 12
suggests that 12--HAm is capable of comproportionation with 1
(a fact also supported by the spectroelectrochemical data in the
presence of n-BuOHscompare Figure 4B and E), while eq 13
intercept ) E1/2°(2) + 0.059 log K
(9)
that K12-(t-BuOH) ) 17 M-1, K12-(n-BuOH) ) 2.9 × 102 M-1
,
1
1
K12-(CH OH) ) 7.9 × 102 M-2, and K12-(H O) ) 1.08 × 103
2
3
3
2
M-3. These equilibrium constants increase as the acid strength
of the group I proton donor increases (Table 2), supporting
mechanism IV.
A correct mechanism should be also consistent with the data
obtained with group II proton donors. Indeed, being stronger
acids, group II proton donors shift the second wave much more
effectively than their group I counterparts (compare Figures 2
and 6). Therefore, to monitor the effect of all three group II
proton donors it was necessary to work with sub-mol equivalent
amounts relative to 1-3. But then, there is stoichiometrically
not enough group II proton donor for all molecules of 1-3 to
associate with, and therefore the original second waves survive
(albeit reduced in sizesFigure 6, dashed lines) while a new
wave appears between the two old ones. The E1/2’s of the new
intermediate waves correlate well with the pKa’s of the three
group II proton donors (Figure 9), supporting their origin from
a hydrogen bonding interaction of 12, in analogy to the
inferences in conjunction with mechanism IV and the group I
proton donors. The conclusions based on the voltammetric data
are also consistent with the spectroelectrochemical data obtained
with 0.5 mol-equivalent amount of phenol relative to 1, showing
the coexistence of 11 with 12--HAm after the new intermediate
wave, and of 12 with 12--HAm after the original second wave.
Finally, it is noted that with only a moderately higher
concentration of CH3CO2H relative to 1, the second wave
merges with the first one, becoming an irreversible 2-e wave
(Figure 6C, dotted line). Bulk electrolysis under these conditions
at -0.8 V vs Ag/AgCl yields only one product, the colorless
4-(R-hydroxybenzyl)-N-methylpyridinium cation (10). Forma-
tion of 10 requires transfer of two electrons and two protons.
The question whether the protons are transferred concurrently
or sequentially with the electrons, is addressed as follows. Both
the partial reversibility during cyclic voltammetry at 30 V/s
(Figure 7A) and the low transient absorbance of 11 during TLC
m
describes the irreversible formation of 10 from 12--HOCOCH3
via a 2-H transfer.
Finally, the last issue to be addressed is the different product
distribution upon electrolysis of 1-3 in the presence of group
I proton donors vs CH3CO2H. Bulk electrolysis of 1 in the
presence of group I proton donors yields only nonquaternized
terminal products (compounds 8 and 9). No trace of 10 was
ever detected among the electrolysis products in the presence
of group I proton donors, and no trace of either 8 or 9 was ever
detected in the presence of CH3CO2H. The immediate question
is whether 9 is formed from 8 or vice versa. As it was mentioned
in the beginning of this section, alcohols analogous to 8 can be

4-Benzoyl-N-methylpyridinium Cations
J. Phys. Chem. B, Vol. 105, No. 17, 2001 3673
oxidized to the corresponding ketones (compounds analogous
to 9) only with strong oxidizing agents such as KMnO4 or
chromic acid.31g,33 Therefore, 9 cannot be formed from 8 in the
reducing environment during bulk electrolysis. On the other
hand, the first-electron reduction of 9 has E1/2 ) -1.47 V vs
Ag/AgCl,20 but in analogy to 4-acetylpyridine,4 the 1-e reduced
form of 9 is expected to pinacolize, and not to form 8. No trace
product should be an alkoxide (or HO-). Stronger nucleophiles,
such as CH3O- and n-BuO-, react with intermediate 13
according to eq 15 giving 8 (Table 5), while all four conjugate
bases of the group I proton donors react with 1 produced at the
end of the electrolysis by air oxidation of either 13, or unreacted
12- -HAm, to give 9 (eq 16).40
In the presence of CH3CO2H and in analogy to eq 14, 12--
(HOCOCH3)m should first yield 13 and CH3CO2-. Owing not
only to the stronger acidity of CH3CO2H relative to the group
I proton donors, but also to the weaker nucleophilicity of
1
of such pinacol has been detected by H NMR among our
electrolysis products, so the reduced form of 9 was not present
during the electrolytic formation of 8. All in all, 8 and 9 are
formed independently of one another, and 9, being an oxidized
form, is probably formed from unreacted 1 after the three-hour
long electrolytic experiment is ended.
-
CH3CO2 compared to HO- and RO-, 13 follows a different
reaction pathway as shown by eq 17.
Indeed, spectroelectrochemistry shows that in the presence
of group I proton donors the irreversible reaction of 12--HAm
toward terminal products is slow (Figure 4F). As soon as the
reducing environment is removed, the remaining 12--HAm is
expected to be converted back to 1, owing to the reversibility
of the voltammetric second wave in the presence of group I
proton donors (Figure 2). According to the control experiment
of Figure 5B, once 1 is formed, it reacts with alkoxide leading
to dequaternization and formation of 9. (A literature survey
shows that nucleophile-induced dequaternization of pyridinium
salts is known but perhaps the full latitude of this reaction may
not have been appreciated.39) The next question is related to
the origin of the alkoxide in the electrolytic solution. This is
not difficult to reconcile: 8 must be produced from 12--HAm
via a two-step proton transfer (eqs 14 and 15), whose side
5. Conclusions
It has been reported for the first time that in nonaqueous CH3-
CN, 4-benzoyl-N-methylpyridinium cations, such as 1-3,
undergo two successive well-separated 1-e reductions, in
analogy to viologens and quinones. The chemical inertness and
the intense color of the 1-e reduced forms, 11, suggest that 1-3
could be considered for applications as redox mediators or
electrochromic compounds. Although the 2-e reduction of
4-benzoyl-N-methylpyridinium cations has not been studied
before, nevertheless it has been recognized that 12 could exist
in equilibrium with 11, according to the disproportionation
reaction of eq 18.18 Due to the wide separation of the two
reduction waves of 1-3 (Table 1), the room-temperature
equilibrium constant of eq 18 in the absence of a proton donor
is calculated equal to 2 × 10-11. Therefore, the equilibrium
concentration of 12 (and also of 1) in a 3 mM solution of 11 is
1.27 × 10-5 mM. In the presence of water or alcohols, however,
the second wave moves closer to the first one, hence the
equilibrium constant of eq 18 increases significantly. This fact,
in conjunction with the eventual reaction of 12 with the group
I proton donors, might limit applications where the 1/11 redox
couple should be able to survive millions of cycles for very
long periods of time (in the order of years) as in certain
electrochromic devices.42 In this regard, the results presented
here suggest that under rigorous exclusion of proton donors such
extended operational lifetimes may be possible.
Acknowledgment. We gratefully acknowledge support from
The Petroleum Research Fund, (administered by the ACS, Grant
35154-AC5), and from the University of Missouri Research
Board. We thank Dr. D. Hasha of the University of Missouris
Rolla for his assistance with NMR and Dr. M. Komarynsky of
Washington University in St. Louis for his assistance with the
H-cell.

3674 J. Phys. Chem. B, Vol. 105, No. 17, 2001
Leventis et al.
(23) (a) Leventis, N.; Gao, X. J. Electroanal. Chem. 2001, 500, 78. (b)
Amatore, C.; Bonhomme, F.; Bruneel, J.-L.; Servant, L.; Thouin, L. J.
Electroanal. Chem. 2000, 484, 1. (c) Rongfeng, Z.; Evans, D. H. J.
Electroanal. Chem. 1995, 385, 201.
(24) Efange, S. M. N.; Michelson, R. H.; Remmel, R. P.; Boudreau, R.
J.; Dutta, A. K.; Freshler, A. J. Med. Chem. 1990, 33, 3133.
(25) (a) Sonogashira, K.; Tohda, Y.; Hagihara, N. Tetrahedron Lett.
1975, 50, 4467. (b) Takahashi, S.; Kuroyama, Y.; Sonogashira, K.; Hagihara,
N. Synthesis 1980, 627.
References and Notes
(1) Feoktistov, L. G. In Organic Electrochemistry; 2nd ed.; Baizer,
M. M., Lund, H., Eds.; Marcel Dekker: New York, 1983; p 315.
(2) (a) Stocker, J. H.; Jenevein, R. M.; Kern, D. H. J. Org. Chem.
1969, 34, 2810. (b) Rudd, E. J.; Conway, B. E. Trans. Faraday Soc. 1971,
67, 440. (c) Nadjo, L.; Saveant, J. M. J. Electroanal. Chem. 1973, 44, 327.
(d) van Tilborg, W. J. M.; Smit, C. J. J. R. Neth. Chem. Soc. 1979, 98,
532. (e) Swartz, J. E.; Mahachi, T. J.; Kariv-Miller, E. J. Am. Chem. Soc.
1988, 110, 3622. (f) Tanko, J. M.; Drumright, R. E. J. Am. Chem. Soc.
1992, 114, 1844. (g) Mattiello, L.; Rampazzo, L. J. Chem. Soc., Perkin
Trans. 1993, 2243. (h) Liotier, E.; Mousset, G. Mousty, C. Can J. Chem.
1995, 73, 1488.
(3) (a) Elving, P. J.; Leone, J. T. J. Am. Chem. Soc. 1958, 80, 1021.
(b) Wawzonek, S.; Gundersen, A. J. Electrochem. Soc. 1960, 107, 537.
(4) (a) Hermolin, J.; Kopilov, J.; Gileadi, E. J. Electroanal. Chem. 1976,
71, 245. (b) Kopilov, J.; Shatzmiller, S.; Kariv, E. Electrochim. Acta 1976,
21, 535. (c) Kopilov, J.; Kariv, E.; Miller, L. L. J. Am. Chem. Soc. 1977,
99, 3450.
(26) Back, K.; Huenig, S.; Reinold, P. Tetrahedron 1988, 44, 3295.
(27) (a) Streitweiser, A., Jr.; Reuben, D. M. E. J. Am. Chem. Soc. 1971,
93, 1794. (b) Patrick, T.; Disher, J. M.; Probst, W. J. J. Org. Chem. 1972,
37, 4467.
(28) Smith, M. B.; March, J. March’s AdVanced Organic Chemistry.
Reactions, Mechanisms, and Structure, 5th ed.; John Wiley and Sons: New
York, 2001, (a) p 330; (b) p 1217; (c) p 368; (d) p 98.
(29) Graham Solomons, T. W. Solomons Organic Chemistry; 6th ed.;
John Wiley and Sons: New York, 1966; p 98.
(5) Lehninger, A. L. Biochemistry, 2nd ed.; Worth Publishers: New
York, 1978; p 494.
(30) Munavalli, S.; Poziomek, E. J.; Landis, W. G. Heterocycles 1986,
24, 1883.
(6) Monk, P. M. S. The Viologens: Physicochemical Properties,
Synthesis and Applications of the Salts of 4,4′-Bipyridine; John Wiley and
Sons: New York, 1998.
(7) (a) Leventis, N.; Chen, M.; Liapis, A. I.; Johnson, J. W.; Jain, A.
J. Electrochem. Soc. 1998, 145, L55. (b) Leventis, N.; Gao, X. J. Phys.
Chem. B 1999, 103, 5832.
(8) (a) Imahori, H.; Norieda, H.; Nishimura, Y.; Yamazaki, I.; Higuchi,
K.; Kato, N.; Motohiro, T.; Yamada, H.; Tamaki, K.; Arimura, M.; Sakata,
Y. J. Phys. Chem. B 2000, 104, 1253. (b) Fukushima, M.; Tatsumi, K.;
Tanaka, S.; Nakamura, H. EnViron. Sci. Technol. 1998, 32, 3948. (c)
Moretto, L. M.; Ugo, P.; Zanata, M.; Guerriero, P.; Martin, C. R. Anal.
Chem. 1998, 70, 2163. (d) Yuan, R.; Watanabe, S.; Kuwabata, S.;
Yoneyama, H. J. Org. Chem. 1997, 62, 2494. (e) Nakamura, Y.; Kamon,
N.; Hori, T. Chem. Soc. Jpn. 1989, 62, 551.
(31) (a) Cavallini, G.; Milla, E.; Grumelli, E.; Ravenna, F.; Grasso, I.
Farmaco, Ed. Sci. 1958, 12, 853. (b) Breen, M. P.; Bojanowski, E. M.;
Cipolle, R. J.; Dunn, W. J., III.; Frank, E.; Gearien, J. E. J. Pharm. Sci.
1973, 62, 847. (c) McCaustland, D. J.; Chien, P.-L.; Burton, W. H.; Cheng,
C. C. J. Med. Chem. 1974, 17, 993. (d) Ho¨gberg, T.; Ulff, B.; Renyi, A.
L.; Ross, S. B. J. Med. Chem. 1981, 24, 1499. (e) Ba¨ckvall, J.-E.; Nordberg,
R. E.; Nystro¨m, J.-E.; Ho¨gberg, T.; Ulff, B. J. Org. Chem. 1981, 46, 3479.
(f) Earley, J. V.; Gilman, N. W. Synth. Commun. 1985, 15, 1271. (g)
Carvalho, I.; Miller, J. Heterocycl. Commun. 1995, 1, 403. (h) Takemoto,
M.; Yamamoto, Y.; Achiwa, K. Chem. Pharm. Bull. 1996, 44, 853.
(32) (a) Wolffenstein, R.; Hartwich, F. Chem. Ber. 1915, 48, 2043. (b)
Mirek, J. Zesz. Nauk. Uniw. Jagiellon., Pr. Chem. 1965, 10, 61. Chem.
Abstr. 1967, 66, 37125h. (c) Carmellino, M. L.; Pagani, G.; Pregnolato,
M.; Terreni, M. Pestic. Sci. 1995, 45, 227.
(9) Bookbinder, D. C.; Wrighton, M. S. J. Electrochem. Soc. 1983,
130, 1080.
(33) (a) Crook, K. E.; McElvain, S. M. J. Am. Chem. Soc. 1930, 52,
4006. (b) Teague, P. C. J. Am. Chem. Soc. 1947, 69, 714.
(10) (a) Konishi, T.; Fujitsuka, M.; Ito, O.; Toba, Y.; Usui, Y. J. Phys.
Chem. A 1999, 103, 9938. (b) Zahavy, E.; Seiler, M.; Marx-Tibbon, S.;
Joselevich, E.; Willner, I.; Du¨rr, H.; O’Connor, D.; Harriman, A. Angew.
Chem., Int. Ed. Engl. 1995, 34, 1005. (c) Inada, T. N.; Miyazawa, C. S.;
Kikuchi, K.; Yamauchi, M.; Nagata, T.; Takahashi, Y.; Ikeda, H.; Miyahi,
T. J. Am. Chem. Soc. 1999, 121, 7211. (d) Warren, J. T.; Chen, W.; Johnston,
D. H.; Turro, C. Inorg. Chem. 1999, 38, 6187. (e) Borsarelli, C. D.;
Braslavsky, S. E. J. Phys. Chem. A 1999, 103, 1719.
(11) (a) Leventis, N.; Sotiriou-Leventis, C.; Chen, M.; Jain, A. J.
Electrochem. Soc. 1997, 144, L305. (b) Li, J.; Chen, G.; Dong, S.
Electroanalysis 1997, 9, 834. (c) Tang, X.; Schneider, T. W.; Walker, J.
W.; Buttry, D. A. Langmuir 1996, 12, 5921. (d) Tang, X.; Schneider, T.;
Buttry, D. A. Langmuir 1994, 10, 2235. (e) Katz, E.; Itzhak, N.; Willner,
I. Langmuir 1993, 9, 1392. (f) De Long, H. C.; Buttry, D. A. Langmuir
1992, 8, 2491. (g) Creager, S. E.; Collard, D. M.; Fox, M. A. Langmuir
1990, 6, 1617. (h) Lee, K. A. B. Langmuir 1990, 6, 709. (i) De Long, H.
C.; Buttry, D. A. Langmuir 1990, 6, 1319.
(12) (a) Vo¨gtle, F.; Plevoets, M.; Nieger, M.; Azzellini, G. C.; Credi,
A.; De Cola, L.; De Marchis, V.; Venturi, M.; Balzani, V. J. Am. Chem.
Soc. 1999, 121, 6290. (b) Jockusch, S.; Ramirez, J.; Sanghvi, K.; Nociti,
R.; Turro, N. J.; Tomalia, D. A. Macromolecules 1999, 32, 4419. (c) Pollak,
K. W.; Leon, J. W.; Fre´chet, J. M. J.; Maskus, M.; Abrun˜a, H. D. Chem.
Mater. 1998, 10, 30.
(13) (a) Ruetten, S. A.; Thomas, J. K. Langmuir 2000, 16, 234. (b)
Zhang, G.; Mao, Y.; Thomas, J. K. J. Phys. Chem. B 1997, 101, 7100. (c)
Pfennig B. W.; Chen, P.; Meyer, T. J. Inorg. Chem. 1996, 35, 2898.
(14) (a) Vitale, M.; Castagnola, N. B.; Ortins, N. J.; Brooke, J. A.;
Vaidyalingam, A.; Dutta, P. K. J. Phys. Chem. B 1999, 103, 2408. (b)
Castagnola, N. B.; Dutta, P. K. J. Phys. Chem. B 1998, 102, 1696.
(15) (a) Bavykin, D. V.; Savinov, E. N.; Parmon, V. N. Langmuir 1999,
15, 4722. (b) Logunov, S.; Green, T.; Marguet, S.; El-Sayed, M. A. J. Phys.
Chem. A 1998, 102, 5652. (c) Robins, D. S.; Dutta, P. K. Langmuir 1996,
12, 402.
(34) Mirviss, S. B. J. Org. Chem. 1989, 54, 1948.
(35) (a) Wawzonek, S.; Berkey, R.; Blaha, E. W.; Runner, M. E. J.
Electrochem. Soc. 1956, 103, 456. (b) Given, P. H.; Peover, M. E. J. Chem
Soc. 1960, 385. (c) Umemoto, K. Bull. Chem. Soc. Jpn. 1967, 40, 1058.
(d) Piljac, I.; Murray, R. W. J. Electrochem. Soc. 1971, 118, 1758. (e)
Pekmez, K.; Can, M.; Yildiz, A. Electrochim. Acta 1993, 38, 607. (f)
Revenga, J.; Rodriquez, F.; Tijero, J. J. Electrochem. Soc. 1994, 141, 330.
(36) (a) Paduszek-Kwiatek, B.; Kalinowski, M. K. Electrochim. Acta
1984, 29, 1439. (b) Kwiatek, B.; Kalinowski, M. K. Aust. J. Chem. 1988,
41, 1963.
(37) Bertie, J. E.; Lan, Z. J. Phys. Chem. B 1997, 101, 4111.
(38) The concave-down curvatures in Figure 3 are most probably due
to the formation of HA-aggregates as the concentration of HA increases.
An analysis of mechanisms I and II shows that the proton donor aggregation
number should be also reflected into the slope of the ∆E1/2 vs log[HA]
plots, which therefore should curve downwards with increasing [HA], as
observed.
(39) (a) Bailey, T. D.; McGill, C. K. U.S. Patent 4,158,093, 1979. (b)
Hart, L. S.; Killen, C. R. J.; Saunders: K. D. J. Chem. Soc., Chem. Commun.
1979, 24. (c) Kutney, J. P.; Greenhouse, R. Synth. Commun. 1975, 119.
(40) The reaction of 1 with base is most probably completed in the rotary
evaporator (bath temperature ∼50 °C) during solvent removal. Furthermore,
it may be noted in Table 5 that with H2O and t-BuOH as proton donors, no
8 was formed, whereas in the case of t-BuOH no 1 was recovered either.
These results may seem contradictory because HO- and RO-, which are
necessary for the formation of 9, are byproducts of the formation of 8, and
therefore 8 should have been always present among the electrolysis products.
We have noted, however, that even at slightly elevated temperatures (e.g.,
>30 °C) the reaction of 1 with either CH3O- or t-BuO- is more complicated
than a simple dequaternization, yielding along with 9 intractable products,
which, according to the literature, should be attributed to pyridinium ring
1
opening.41 (In this regard, H NMR spectra in CD3CN show a consistent
(16) Jones, G., II.; Mabla, V. J. Org. Chem. 1985, 50, 5776.
(17) (a) Kosover, E. M.; Poziomek, E. J. J. Am. Chem. Soc. 1964, 86,
5515. (b) Grossi, L.; Minisci, F.; Pedulli, G. F. J. Chem. Soc., Perkin II
1977, 943. (c) Grossi, L.; Minisci, F.; Pedulli, G. F. J. Chem. Soc., Perkin
II 1977, 948.
set of two doublets at 5.95 and 4.72 ppm with a coupling constant of 8.15
Hz, and more complicated patterns in the 2.4-4.6 ppm region.) The same
NMR signals are detectable among the electrolysis products in the presence
of CH3OH and t-BuOH, and it is thus suggested that the yields of 8 and 9
are compromized by competing pyridinium ring opening side reactions.
(18) Neta, P.; Patterson, L. K. J. Phys. Chem. 1974, 78, 2211.
(19) Frangopol, M.; Frangopol, P. T.; Trichilo, C. L.; Geiger, F. E.;
Filipescu, N. J. Org. Chem. 1973, 38, 2355.
(20) Shu, C.-F.; Wrighton, M. S. Inorg. Chem. 1988, 27, 4326.
(21) Yoshiike, N.; Kondo, S.; Fukai, M. J. Electrochem. Soc. 1980, 127,
1496.
(41) (a) Kavalek, J.; Lycka, A.; Machacek, V.; Sterba, V. Collect. Czech.
Chem. Commun. 1975, 40, 1166. (b) Yakovlev, M. Yu.; Kadushkin, A. V.;
Solov’eva, N. P.; Anisimova, O. S.; Granik, V. G. Tetrahedron 1998, 54,
5775.
(42) (a) Byker, H. J. Proc. Electrochem. Soc. 1994, 94, 2. (b) Byker,
H. J. U.S. Patent 5,128,799, 1992. (c) Byker, H. J. U.S. Patent 4,902,108,
1990.
(22) Cupta, N.; Linschitz, H. J. Am. Chem. Soc. 1997, 119, 6384.