Relationship between molecular structure and electron targets in the electroreduction of benzocarbazolediones and anilinenaphthoquinones. Experimental and theoretical study.

Article abstract of DOI:10.1021/jo011083k

We report the synthesis and voltamperometric reduction of 5H-benzo[b]carbazole-6,11-dione (BCD) and its 2-R-substituted derivatives (R = -OMe, -Me, -COMe, -CF(3)). The electrochemical behavior of BCDs was compared to that of the 2-[(R-phenyl)amine]-1,4-naphthalenediones (PANs) previously studied. Like PANs, BCDs exhibit two reduction waves in acetonitrile. The first reduction step for the BCDs represents formation of the radical anion, and the half-wave potential (E(1/2)) values for this step are less negative than for that of the PANs. The second reduction wave, corresponding to the formation of dianion hydroquinone, has E(1/2) values that shift to more negative potentials. A good linear Hammett-Zuman (E(1/2) vs sigma(p)) relationship, similar to that for the PAN series, was also obtained for the BCDs. However, unlike the PANs, in the BCDs, the first reduction wave was more susceptible to the effect of the substituent groups than was the second wave, suggesting that the ordering of the two successive one-electron reductions in BCDs is opposite that in PANs. This is explained by the fact that the electron delocalizations in the two systems are different; in the case of BCDs there is an extra aromatic indole ring, which resists loss of its aromatic character. The electronic structures of BCD compounds were, therefore, investigated within the framework of the density functional theory, using the B3LYP hybrid functional with a double zeta split valence basis set. Our theoretical calculations show that the O(1).H-N hydrogen bond, analogous to that previously described for the PAN series, is not observed in the BCDs. Laplacians of the critical points (nabla(2)rho) and the natural charges for the C-O bonds indicate that the first reduction wave for the BCDs corresponds to the C(4)-O(2) carbonyl, while in the PAN series the first one-electron transfer occurred at the C(1)-O(1) carbonyl. Natural bond orbital analysis showed that, in all the BCDs, the lowest unoccupied molecular orbital (LUMO) is located at C(4), whereas for the PANs, the LUMO is found at C(1). The good correlation between the LUMO energy values and the E(1/2) potentials (wave I) established that the first one-electron addition takes place at the LUMO. Analysis of the molecular geometry confirmed that, in both series of compounds, the effect of the substituent groups is mainly on the C(4)-O(2) carbonyl. These results explain the fact that reduction of the C(4)-O(2) carbonyl (voltammetric wave II in the PANs and voltammetric wave I in the BCDs) is more susceptible to the effect of the substituent groups than is reduction of the C(1)-O(1) carbonyl (wave I in the PANs and wave II in the BCDs).

Full text of DOI:10.1021/jo011083k

J . Org. Chem. 2002, 67, 3673-3681
3673
Rela tion sh ip betw een Molecu la r Str u ctu r e a n d Electr on Ta r gets
in th e Electr or ed u ction of Ben zoca r ba zoled ion es a n d
An ilin en a p h th oqu in on es. Exp er im en ta l a n d Th eor etica l Stu d y
N. Mac´ıas-Ruvalcaba,† G. Cuevas,† I. Gonza´lez,‡ and M. Aguilar-Mart´ınez*,†
Instituto de Quı´mica, Universidad Nacional Auto´noma de Me´xico, Ciudad Universitaria,
04510 Me´xico D.F., Mexico, and Departamento de Quı´mica, Universidad Auto´noma
Metropolitana-Iztapalapa, Apartado Postal 55-534, 09340 Me´xico D.F., Mexico
marthaa@servidor.unam.mx
Received November 19, 2001
We report the synthesis and voltamperometric reduction of 5H-benzo[b]carbazole-6,11-dione (BCD)
and its 2-R-substituted derivatives (R ) -OMe, -Me, -COMe, -CF3). The electrochemical behavior
of BCDs was compared to that of the 2-[(R-phenyl)amine]-1,4-naphthalenediones (PANs) previously
studied. Like PANs, BCDs exhibit two reduction waves in acetonitrile. The first reduction step for
the BCDs represents formation of the radical anion, and the half-wave potential (E1/2) values for
this step are less negative than for that of the PANs. The second reduction wave, corresponding to
the formation of dianion hydroquinone, has E1/2 values that shift to more negative potentials. A
good linear Hammett-Zuman (E1/2 vs σp) relationship, similar to that for the PAN series, was also
obtained for the BCDs. However, unlike the PANs, in the BCDs, the first reduction wave was more
susceptible to the effect of the substituent groups than was the second wave, suggesting that the
ordering of the two successive one-electron reductions in BCDs is opposite that in PANs. This is
explained by the fact that the electron delocalizations in the two systems are different; in the case
of BCDs there is an extra aromatic indole ring, which resists loss of its aromatic character. The
electronic structures of BCD compounds were, therefore, investigated within the framework of the
density functional theory, using the B3LYP hybrid functional with a double ú split valence basis
set. Our theoretical calculations show that the O1‚‚‚H-N hydrogen bond, analogous to that
previously described for the PAN series, is not observed in the BCDs. Laplacians of the critical
2
points (∇ F) and the natural charges for the C-O bonds indicate that the first reduction wave for
the BCDs corresponds to the C4-O2 carbonyl, while in the PAN series the first one-electron transfer
occurred at the C1-O1 carbonyl. Natural bond orbital analysis showed that, in all the BCDs, the
lowest unoccupied molecular orbital (LUMO) is located at C4, whereas for the PANs, the LUMO is
found at C1. The good correlation between the LUMO energy values and the E1/2 potentials (wave
I) established that the first one-electron addition takes place at the LUMO. Analysis of the molecular
geometry confirmed that, in both series of compounds, the effect of the substituent groups is mainly
on the C4-O2 carbonyl. These results explain the fact that reduction of the C4-O2 carbonyl
(voltammetric wave II in the PANs and voltammetric wave I in the BCDs) is more susceptible to
the effect of the substituent groups than is reduction of the C1-O1 carbonyl (wave I in the PANs
and wave II in the BCDs).
In tr od u ction
cides7 and fungicides.8 The biological activity of the
quinoid compounds is reportedly due to the redox chem-
istry of the quinone system.1,9 Quinones are capable of
accepting one or two electrons to form the corresponding
radical anion (Q•-) and hydroquinone dianion (Q2-).
These species interact with crucial cellular molecules
such as oxygen, DNA, and proteins, altering and possibly
controlling their biological activity.1,9,10 To effect such
Organic molecules containing a quinone moiety are
important in a wide variety of biological activities.1 Many
such molecules are used pharmaceutically as agents
against tuberculosis,2 malaria,3 bacterial infections,4 and
tumor growth,5 in water management and agriculture to
combat larvae,6 in mollusk infestations,6 and as herbi-
* To whom correspondence should be addressed at the Departa-
mento de Fisicoqu´ımica, Facultad de Qu´ımica, Universidad Nacional
Auto´noma de Me´xico, Ciudad Universitaria, 04510 Me´xico D.F.,
Mexico.
(6) Lopez, J . N. C.; J ohnson, A. W.; Grove, J . F.; Bulhoes, M. S. Cienc.
Cult. (Sao Paulo) 1977, 29, 1145-1149.
(7) U.S. Rubber Co. British Patent 862 489, 1959. Takeda Chemical
Industry Co. Ltd. J apanese Patent 18 520, 1963. Ube Industries Ltd.
J apanese Patent 126 725, 1979. Shell Internationale Research
Maatschappij B.V. British Patent 1 314 881, 1973.
(8) Clark, N. G. Pestic. Sci. 1985, 16, 23-32.
(9) (a) Monks, T. J .; Hanzlik, R. P.; Cohen, G. M.; Ross, D.; Graham,
D. G. Toxicol. Appl. Pharmacol. 1992, 112, 2-16. (b) Wolstenholm, G.
E. W., O’Conner, C. M., Eds. Quinones in Electron Transport;
Churchill: London, 1961.
(10) Brunmark, A.; Cadenas, E. Free Radical Biol. Med. 1988, 7,
435-477.
† Universidad Nacional Auto´noma de Me´xico.
‡ Universidad Auto´noma Metropolitana-Iztapalapa.
(1) Morton, R. A., Ed. Biochemistry of Quinones; Academic Press:
New York, 1965.
(2) Oeriu, I.; Benesch, H. Bull. Soc. Chim. Biol. 1962, 44, 91-100.
Oeriu, I. Biokhimiya 1963, 28, 380-383.
(3) Prescott, B. J . Med. Chem. 1969, 12, 181-182.
(4) Silver, R. F.; Holmes, H. L. Can. J . Chem. 1968, 46, 1859-1864.
(5) Hodnett, E. M.; Wongwiechintana, C.; Dunn, W. J .; Marrs, P. J .
Med. Chem. 1983, 26, 570-574.
10.1021/jo011083k CCC: $22.00 © 2002 American Chemical Society
Published on Web 05/08/2002

3674 J . Org. Chem., Vol. 67, No. 11, 2002
Mac´ıas-Ruvalcaba et al.
Ta ble 1. Str u ctu r es a n d Releva n t Sp ectr oscop ic Da ta
for th e BCDs a n d for th e P ANs Stu d ied in Th is Wor k
fine-tuned manipulations, the molecules not only must
be capable of accepting electrons at some site with
appropriate speed and energy, but also should accept
electrons in the correct site.11
The chemical and biological importance of quinone
redox systems has motivated a great deal of research
evaluating the electrochemical behavior of the quinone-
hydroquinone systems.12-24 Most investigations focus on
the reduction kinetics and mechanisms in protic and
aprotic media,12-14 determination of the relationship
between chemical structure and redox potentials,15-19 and
how adding proton donor species influences the electro-
reduction mechanisms.20-24
In this paper, we concentrated on determining the
favored site of electron acceptance within the asymmetri-
cally substituted p-naphthoquinones. We previously used
electrochemical and theoretical methods to show that, for
the 2-[(R-phenyl)amine]-1,4-naphthalenediones (PANs),
the first electron transfer corresponds to the reduction
of the R-carbonyl to the amino group (carbonyl, C1-O1,
Table 1). Here, we extended the investigation to another
series of quinones, the 2-R-5H-benzo[b]carbazole-6,11-
diones (BCDs) (Table 1). The purpose of this study was
to determine the precise electron acceptance site of the
BCD molecules and to establish a relationship between
molecular structure and the reduction target of the
electrons for the two quinone series PANs and BCDs.
We synthesized five BCD derivatives on which volta-
mmetric studies in acetonitrile and analyses of their
molecular geometry using computational chemistry were
performed. The hybrid B3LYP functional was employed
to optimize the molecular geometry for the BCDs studied.
We determined the Wiberg bond indexes (WBIs), natural
charges, densities at critical bond points, Laplacians, and
ellipticities for each of the derivatives. The correlation
between the electrochemical and theoretical data, for both
the PAN19 and the BCD compounds, showed that the
presence of intramolecular hydrogen bonds (O1‚‚‚H-N)
plays an important role in determining the acceptance
a
The UV-vis spectra were determined using EtOH as solvent,
and the λmax values are in nanometers. ꢀmax values are in L mol-1
b
cm-1 c These compounds have not been previously described; see
.
the Experimental Section.
site of the first electron in the electrochemical reduction
of the quinone system. The use of acetonitrile as solvent
in the electrolytic medium was particularly important
since the quinones and their corresponding anions are
solvated far less efficiently25 than in water, and conse-
quently, the intramolecular hydrogen bonds are more
important.26 Moreover, aprotic solvents mimic nonpolar
environments of the cell where much of the biological
electron transfer occurs.27
The biological activity of quinoid compounds depends
not only on the ability of the quinone system to accept
electrons, but also on the site of electron acceptance
within the quinone itself.11 Determining acceptance sites
could provide valuable information on the mechanisms
of biological activity and aid in designing new molecules
with greater and more efficient biological activity than
those currently available.
Resu lts a n d Discu ssion
(11) Baum, R. M. Chem. Eng. News 1987, 19-21.
(12) (a) Chambers, J . Q. In The Chemistry of the Quinoid Com-
pounds; Patai, S., Rapport, Z., Eds.; Wiley: New York, 1988; Vol. II,
Chapter 12, pp 719-757; 1974; Vol. I, Chapter 14, pp 737-791. (b)
Evans, D. H. Carbonyl Compounds. Encyclopedia of Electrochemistry
of the Elements; Marcel Dekker: New York, 1978; Vol. XII, Chapter
1, p 3. (c) Laviron, E. J . Electroanal. Chem. 1986, 208, 357-372.
(13) Peover, M. E. In Electroanalytical Chemistry; Bard, A. J ., Ed.;
Dekker: New York, 1967; pp 1-51.
Sp ectr oscop y. In ethanol, all the PANs and BCDs
studied produced very similar electronic absorption spec-
tra with a band (x) in the 266-302 nm region (log ꢀ )
4.32-4.65) and a very broad, low energy band (y) in the
visible region centered between 452 and 480 nm (log ꢀ )
3.67-3.78) for PANs and 348 and 430 nm (log ꢀ ) 3.69-
3.93) for BCDs (Table 1). These bands are due to π-π*
electronic transitions of the aromatic systems in the
molecules. This was confirmed by the fact that both bands
underwent a bathochromic shift when the polarity of the
(14) Kolthoff, I. M.; Lingane, J . J . Polarograpy, 2nd ed.; Inter-
science: New York, 1952; Vols. I and II.
(15) (a) Zuman, P. Substituent Effects in Organic Polarography;
Plenum Press: New York, 1967; Chapters I-III and VIII. (b) Zuman,
P. Collect. Czech. Chem. Commun. 1962, 27, 2035-2057.
(16) Li, C.-Y.; Caspar, M. L.; Dixon, D. W. Electrochim. Acta 1980,
25, 1135-1142.
(17) Glezer, V.; Turovska, B.; Stradins, J .; Freimanis, J . Electrochim.
Acta 1990, 35 (11/12), 1933-1940.
(18) Suresh Babu, S.; Subba Reddy, G. V.; Rajendra Kumar Reddy,
P.; J ayarama Reddy, S. U. Scientist Phyl. Sci. 1995, 7 (1), 43-53.
(19) Aguilar-Mart´ınez, M.; Cuevas, G.; J ime´nez-Estrada, M.; Gonza´-
lez, I.; Lotina-Hennsen, B.; Macı´as-Ruvalcaba, N. J . Org. Chem. 1999,
64, 3684-3694.
(20) Given, P. H.; Peover, M. E. J . Chem. Soc. 1960, 385-393.
(21) Eggins, B. R.; Chambers, J . Q. J . Electrochem. Soc. 1970, 117
(2), 186-192.
(25) Although dipolar aprotic solvents ordinarily have large dipole
moments, they are relatively ineffective at solvating the negative ions
by dipole interactions because the positive ends of the dipole are
usually buried in the middle of the molecule.26
(26) Lowry, T. H.; Richardson, K. S. Mechanism and Theory in
Organic Electrochemistry; Harpenter & Row Publishers: New York,
1987; Chapter 2. (b) Kosower, E. M. An Introduction to Physical
Organic Chemistry; J ohn Wiley & Sons: New York, 1986; Part 2.
(27) (a) Li, C.-Y.; J enq, J . Electrochim. Acta 1991, 36, 269-276. (b)
Crawford, P. W.: Gross, J .; Lawson, K.; Cheng, C. C.; Dong, Q.; Liu,
D. f.; Luo, Y. L.; Szczepankiewicz, B. G.; Heathcock, C. H. J . Electro-
chem. Soc. 1997, 144, 3710-3715. (c) J ezziorek, D.; Ossowski, T.; Liwo,
A.; Dyl, D.; Nowacka, M.; Wo’znicki, W. J . Chem. Soc., Perkin Trans.
2 1997, 229-236; (d) Weiss, H.; Friedrich, T.; Hofhaus, G.; Preis, D.
Eur. J . Biochem. 1991, 197, 536. (e) Ashnagar, A.; Bruce, J . M.; Dutton,
P. L.; Prince, R. C. Biochim. Biophys. Acta 1984, 801, 351-359. (f)
Crawford, P. W.; Carlos, E.; Ellegood, J . C.; Cheng, C. C.; Dong, Q.;
Liu, D. F.; Luo, Y. L. Electrochim. Acta 1996, 41, 2399-2403.
(22) (a) Ortiz, J . L.; Delgado, J .; Baeza, A.; Gonza´lez, I.; Sanabria,
R.; Miranda, R. J . Electroanal. Chem. 1996, 411, 103-107. (b)
Gonza´lez, F. J .; Aceves, J . M.; Miranda, R.; Gonza´lez, I. J . Electroanal.
Chem. 1991, 310, 293-303.
(23) Gupta, N.; Linschitz, H. J . Am. Chem. Soc. 1997, 119, 6384-
6391.
(24) Uno, B.; Okumura, N.; Goto M.; Kano, K. J . Org. Chem. 2000,
65, 1448-1455.

Electroreduction of BCDs and PANs
J . Org. Chem., Vol. 67, No. 11, 2002 3675
Ta ble 2. Volta m m etr ic P a r a m eter sa for P ANsb a n d BCDs
solvent was increased (i.e., MeBCD in acetonitrile shows
two bands at 267 and 392 nm, while in ethanol these
bands are located at 271 and 402 nm). Some quinones
show specific absorptions originating from the substitu-
ent; these will not be discussed in this paper.
Table 1 shows that, in the BCDs, molecular closure
gives rise to the aromatic indole system, causing a shift
to shorter wavelengths for the y-band. These results
indicate that, in the PAN series, displacement of the
electronic density between the naphthoquinone ring and
the substituted aniline is more efficient than between the
indole and the naphthoquinone in the BCDs, since the
indole system apparently behaves as an independent
group.
It is well-known that both the electron-accepting and
electron-donating substituents allow the general π-π*
bands contributed from the HOMO-LUMO configuration
to be shifted to longer wavelengths.28 However, we found
that, for both series of compounds, the λmax for the y-band
shows an interesting shift, which depends on the sub-
stituent. The electron-donor groups produced a change
in the absorption toward longer wavelengths, while the
electron-accepting groups caused the opposite effect
(Table 1). These changes seem more important in the
BCD molecules.
I
E1/2 (mV) ((E1/2 + E1/2II)/2)
∆E1/2 (mV)
I
II
compd
wave I
wave II
(E1/2 - E1/2
)
ln Kc
NQ
-1036
-1236
-1216
-1209
-1126
-1147
-1203
-1208
-1188
-1078
-1057
-1495
-1809
-1773
-1685
-1554
-1602
-1788
-1796
-1778
-1738
-1727
459
573
557
476
428
455
585
588
590
660
670
17.9
22.3
21.7
18.5
16.7
17.7
22.8
22.8
22.9
25.7
26.1
p-MeOPAN
p-MePAN
PAN
p-COCH3PAN
p-CF3PAN
MeOBCD
MeBCD
BCD
COCH3BCD
CF3BCD
a
Determined by cyclic voltammetry at 100 mV/s using a glassy
carbon working electrode and Pt counter electrode. The potentials
are given with respect to the Fc+/Fc redox couple. Because of the
insolubility of many of the quinones in the electrolytic medium,
b
not all of them were used at the same concentrations. The E1/2
data for the compounds of the PAN series were obtained from ref
19. c Logarithms of the equilibrium constant of eq 7 producing the
I
radical anion (ln K ) (F/RT)(E1/2 - E1/2II)).
molecule as shown in Table 2. Such behavior indicates
that substitution of the different anilines in the p-
naphthoquinone nucleus (PAN system) or fusion of the
p-naphthoquinone ring with an indole 5-substituted
(BCD system) (Table 1) increases electron density (F) on
the quinone ring. Such an increase is explained by the
fact that, in these molecules, the lone pair of nitrogen in
the R-position to the C1-O1 carbonyl group is displaced
toward the enone system of the quinone as shown in
Figure 1 (structures I and IV). For a single substituent
group, the E1/2 potentials corresponding to the first
reduction wave were less negative for the BCDs than for
the PANs. With the exception of MeOBCD, the E1/2
potentials corresponding to the second electron transfer
(wave II) present the opposite effect (Table 2).
For PAN and its derivatives, as well as for the BCD
series, introduction of electron-donor groups (i.e., -OMe
and -Me) displaced E1/2 potentials for the first reduction
wave toward a more negative region than the E1/2
potentials corresponding to the parent compound (PAN
or BCD). Introduction of electron-accepting substituents
(-COMe and -CF3) changed the E1/2 potentials to a less
negative region (Table 2). Electron-accepting groups
reduced the electron density of the electroactive group,
facilitating the reduction process, while electron-donor
groups have the opposite effect. The degree of positive
or negative change is directly related to the magnitude
of the electronic effect of the substituent group. Thus,
for the BCDs, the molecule with a substituted -CF3
group, which has an electron-accepting effect slightly
greater than that of the -COMe group (σp ) 0.54 and
0.50, respectively),29 results in a greater positive change
for E1/2 of both reduction waves than that observed for
the molecule substituted with a -COMe group. The
analogous situation occurs in PAN derivatives for the
electron-donor groups. The p-MeOPAN molecule under-
goes a greater negative change than p-MePAN, agreeing
with the Hammett σp constants for these substituent
groups (for -OMe σp ) -0.27, and for -Me σp ) -0.17).29
Nonetheless, for the BCD series, when R ) -OMe, both
reduction waves present a less negative change than that
observed in the MeBCD (Table 2). Therefore, in this
Electr och em istr y. Voltammetric studies of all the
PANs and BCDs evaluated here (Table 1) were performed
at 25 °C using a glassy carbon working electrode and 0.1
M tetraethylammonium tetrafluroborate in acetonitrile
as the electrolytic medium. The BCDs were less soluble
in acetonitrile than their PAN counterparts. BCDs
containing electron-accepting substituents were less
soluble than BCDs with electron-donor groups. For all
PAN and BCD compounds, the voltammograms, recorded
in the potential range from +433 to -2233 mV, produced
two reversible waves. The first wave is attributed to the
addition of an electron to the quinone (Q) to produce a
semiquinone radical anion (Q•-) and the second one to
the subsequent addition of an electron to the semiquinone
radical anion, producing a dianion hydroquinone (Q2- 12-14
(eqs 1 and 2).
)
I
Q + e h Q•-
•- + e h Q2-
E1/2
(1)
(2)
II
Q
E1/2
From the voltammetric curves, we evaluated the
following parameters: (a) half-wave potentials, E1/2
)
(Epa +Epc)/2 corresponding to waves I and II, where Epa
and Epc correspond to the anodic and cathodic peak
I
potentials, respectively, (b) the values of ∆E1/2 ) E1/2
-
E1/2II, where E1/2 and E1/2 correspond to half-wave
potentials of waves I and II, respectively, and (c) the
equilibrium constants for the comproportionation reac-
tion (ln K) (Table 2) (see below). Table 2 also includes
the voltammetric parameters previously obtained for
naphthoquinone (NQ) for comparison.19
I
II
Effect of Ch em ica l Str u ctu r e a n d Electr on ic Ef-
fect of th e Su bstitu en t Gr ou p s on E1/2 P oten tia ls.
Both reduction waves, Q to Q•- (wave I) and Q•- to Q2-
(wave II), produced more negative E1/2 potentials for the
PANs and BCDs than for the naphthoquinone (NQ)
(28) Silverstein, R. M.; Clayton Bassler, G.; Morrill, T. C. Spectro-
metric Identification of Organic Compounds, 5th ed.; J ohn Wiley &
Sons: New York, 1991; Chapter 7, pp 289-316.
(29) Hansch, C.; Leo, A.; Taft, R. W. Chem. Rev. 1991, 91, 165-
195.

3676 J . Org. Chem., Vol. 67, No. 11, 2002
Mac´ıas-Ruvalcaba et al.
F igu r e 1. Resonant hybrids for (a) PANs and (b) BCDs, showing displacement of the lone pair of the nitrogen toward the quinone
ring (structures I and IV) and toward the indole ring in the BCDs (structures II and III).
system, the link between the quinone ring and the
substituted aromatic ring (C3-C6′ bond) results in the
inductive effect of the substituents being felt in the
quinone system. Since -OMe is an electron-accepting
group by inductive effect (σm ) 0.12), the magnitude of
the negative change observed indicates that there is
resonance competition between the electron-donor effect
and the inductive electron-acceptor effect. Comparing the
E1/2 for MeOBCD with that of BCD demonstrates that
the electron-donor effect predominates.
To establish a quantitative relationship of the magni-
tude of the effect of substituents on the electrochemical
reduction of the PANs and the BCDs, we correlated the
Hammett σp constants of the different substituent groups
with the ∆E1/2 potentials of waves I and II as presented
in eqs 3-6.15
available, we used the standard Hammett values as an
approximation of the substituent effect, as suggested by
Hammett (σp ) -0.27 (-OMe), -0.17 (-Me), 0 (H), 0.5
(-COMe), 0.54 (-CF3)).29
The sensitivity of the electrochemical reduction to the
electronic effect of the substituents, characterized by the
values of the slope15 in eqs 3-6, indicates that even when
there is no direct conjugation between the substituent
group and the quinone system in either of the two series
of compounds, electrochemical reduction of the quinone
system is sensitive to electronic perturbation of the
substituents. The positive values of the slopes illustrate
that an increase in the electron attraction of the sub-
stituent facilitates the reduction of the quinone. Equa-
tions 3 and 4 show that, for the PAN series, the reduction
of Q to Q•- (wave I) is less susceptible to the effect of the
substituents than is the reduction of Q•- to Q2-(wave II)
(slope values 123 ( 7.6 and 277 ( 6.4 mV, respectively).
In previous studies19 we found that, for this system, the
first electron transfer corresponds to the reduction of
R-carbonyl to the amino group (C1-O1 carbonyl), and
wave II is associated with reduction of the C4-O2
carbonyl. Thus, the greater susceptibility of the second
wave to the electronic effect of the substituent in the PAN
series occurs since, in these molecules, displacement of
the electron pair of the amine nitrogen mainly affects the
electron density of the C4-O2 carbonyl (Figure 1, struc-
ture I). The magnitude of such a displacement has been
shown to be influenced strongly by the type of substituent
group present in the phenyl ring.19 Surprisingly, for
BCDs we observe that the first wave is more susceptible
to the effect of the substituents than the second wave,
slope values 190 ( 8.7 and 80 ( 6.0 mV, respectively (eqs
5 and 6). We explain this behavior in the section below
on theoretical calculations.
for the PAN series
∆E1/2(wave I) ) (123 ( 7.6)σp + 7.5 ( 1.3
(n ) 5, r2 ) 0.9517) (3)
∆E1/2(wave II) ) (277 ( 6.4)σp + 32.9 ( 2.1
(n ) 5, r2 ) 0.9323) (4)
for the BCD series
∆E1/2(wave I) ) (190 ( 8.7)σp + 18.4 ( 2.3
(n ) 5, r2 ) 0.9621) (5)
∆E1/2(wave II) ) (80 ( 6.0)σp + 2.3 ( 0.8
(n ) 5, r2 ) 0.9323) (6)
Here ∆E1/2 is the difference in half-wave potentials
between the substituted quinone and the parent refer-
ence compound PAN or BCD. To assess correctly the
effect of the substituents on the second reduction peak,
Effect of th e Ch em ica l Str u ctu r e a n d Electr on ic
Effect of th e Su bstitu en t Gr ou p s on Ra d ica l An ion
Stability. Two successive well-separated reduction waves,
-
σp• constants for radical anions should be used. Since
-
the σp• values for the studied compounds are not

Electroreduction of BCDs and PANs
J . Org. Chem., Vol. 67, No. 11, 2002 3677
corresponding to one electron charge transfer each, are
associated with the presence of stable radical anions. To
obtain a quantitative measurement of the stability of the
radical anions, we should refer to the equilibrium con-
stants for the disproportionation reaction (inverse of eq
7); however, this type of reaction, probably because of the
extremely low equilibrium constants, is not usually
mentioned in the literature.17,30,31 Instead, the opposite
reaction, comproportionation,32,33 which involves the
reaction of Q2- with Q to yield Q•- (eq 7), is discussed
more frequently. In this work we discuss the stability of
the radical anions (Q•-) in terms of the equilibrium
constants K of the comproportionation reaction. The
values of ln K were calculated from the difference
between the potentials of the first and the second charge
transfers as indicated in eq 8, where the K values are
expressed as shown in eq 9. Thus, a high value of ln K
means that the Q•- is not prone to disproportionation;
hence, it is more stable.
displacement of the radical anion is directly related to
the type of the substituent in R. Electron-donor groups
favor displacement of the free pair of electrons in the
nitrogen toward the quinone system (Figure 2, structure
VI), while electron-accepting groups facilitate displace-
ment of the electron density of the nitrogen toward the
aniline system (Figure 2, structure VII).19 Thereby, if R
is an electron-accepting group, the extra electron, initially
located at C1, can become stabilized only over the
naphthoquinone ring19 (Figure 2, structures VII-X). If
R is an electron-donor group, the radical anion can
become stabilized not only over the naphthoquinone ring,
but also over the aniline ring, as shown in structures XI
and XII of Figure 2. This behavior explains the apparent
inconsistency that, for the PAN molecules, electron-donor
groups increase the radical anion stability.
Th eor etica l Ca lcu la tion s. To explain the experimen-
tal observations described above, we performed a com-
plete optimization of the molecular geometry for the
compounds MeOBCD, MeBCD, BCD, ClBCD, COMe-
BCD, CF3BCD, CNBCD, and NO2BCD as molecular
models corresponding to the BCD series. These studies
were performed within the framework of the density
functional theory34 with the B3LYP hybrid functional,
using a double ú split valence basis set as implemented
in Gaussian 94 (G94) programs.35 The relevant geometric
data are included in Table S1 of the Supporting Informa-
tion. Table S2 (Supporting Information) shows the WBIs,
calculated with the natural bond orbital (NBO) program36
at the B3LYP/6-31G(d,p) level of theory, as well as the
densities of the relevant critical points (F), the Laplacians
Q
2- + Q h 2Q•-
(7)
(8)
F
I
II
ln K )
(E1/2 - E1/2
)
RT
[Q•-]2
[Q][Q2-
K )
(9)
]
Since the E1/2 values for all the quinones were evalu-
ated under the same experimental conditions, the ob-
served changes in the ln K values shown in Table 2 can
only be associated with the structural modifications of
the molecules, such as the joining of the p-naphtho-
quinone ring with an indolic 5-sustituted ring (BCDs),
or with p-substituted aniline (PANs), as well as the
electronic effect of the substituents.
2
(∇ F), and the ellipticities (ꢀ) determined using the AIM
program.37 Table S3 (Supporting Information) includes
natural charges, and Table 3 shows the energies of the
highest occupied molecular orbital (HOMO) and the
lowest unoccupied molecular orbital (LUMO), the hard-
ness (η), and the total energy of the molecules (Etotal).
Numeration for the different atoms in the PAN and BCD
molecules is shown in Table 1.
As shown in Table 2, the BCDs have higher ln K values
than the PANs. This indicates that the radical anion
formed in the BCDs is more stable, toward dispropor-
tionation, than the radical anion obtained for the PAN
derivatives. One would expect the stability constants of
-COMe and -CF3 compounds to be greater than those
of -OMe and -Me compounds because of the strong
electron-accepting capacity of -COMe and -CF3 groups.
However, this behavior was observed only for the BCD
compounds (Table 2). The opposite effect was seen for
the PAN derivatives since, in this case, substitution of
electron-donor groups leads to more stable radical anions
than when R is an electron-accepting group (Table 2).
The greater stability of Q•- for the BCDs compared to
the PANs is due to the fact that more resonant forms
participate in stabilization of the radical anion in the
BCDs than in the PANs. This is explained by the fact
that the extra electron can be displaced over the four
rings of the molecule, regardless of the type of substituent
in R. In the PAN system, however, the degree of
Geometric data included in Table S1 (Supporting
Information) illustrate that none of the BCD molecules
studied deviate from planarity, allowing the nitrogen
atom to maintain
a p-type orbital (determined by
NBO analysis, vide infra) with an electron pair that
allows it to participate in conjugation with the ben-
zene ring and with the quinone rings. The analysis of
2
the bond distances, WBIs, F values, and ∇ F values, for
the C1′-N-C2-C1-O1 fragment (Tables S1 and S2)
indicates that, in the BCD system, the electronic density
of the indole system is also displaced toward the C1-C2
(34) Malkin, V. G.; Malkina, O. L.; Eriksson, L. A.; Salahub, D. R.
In Modern Density Functional Theory. A Tool for Chemistry; Seminar,
J . M., Politzer, P., Eds.; Elsevier: Amsterdam, 1995.
(35) Frisch, M. J .; Trucks, G. W.; Schlegel, H. B.; Gill, P. M. W.;
J ohnson, B. G.; Robb, M. A.; Cheeseman, J . R.; Keith, T.; Petersson,
G. A.; Montgomery, J . A.; Raghavachari, K.; Al-Laham, M. A.;
Zakrzewski, V. G.; Ortiz, J . V.; Foresman, J . B.; Cioslowski, J .;
Stefanov, B. B.; Nanayakkara, A.; Challacombe, M.; Peng, C. Y.; Ayala,
P. Y.; Chen, W.; Wong, M. W.; Andres, J . L.; Replogle, E. S.; Gomperts,
R.; Martin, R. L.; Fox, D. J .; Binkley, J . S.; Defrees, D. J .; Baker, J .;
Stewart, J . P.; Head-Gordon, M.; Gonzalez, C.; Pople, J . A. Gaussian
94, revision D.4; Gaussian, Inc.: Pittsburgh, PA, 1995.
(36) NBO 4.0: Glendening, E. D.; Badenhoop, J . K.; Reed, A. E.;
Carpenter, J . E.; Weinhold, F., Theoretical Chemistry Institute,
University of Wisconsin, Madison, WI, 1994. Reed, E. A.; Weinstock,
R. B.; Weinhold, F. J . J . Chem. Phys. 1985, 83, 735-746. Reed, E. A.;
Weinhold, F. J . J . Chem. Phys. 1985, 83, 1736-1740. Reed, E. A.;
Curtiss, L. A.; Weinhold, F. Chem. Rev. 1988, 88, 899-926.
(37) Biegler-Koenig, F. W.; Bader, R. F. W.; Tang, T. H. J . Comput.
Chem. 1982, 3 (3), 317-328.
(30) J ensen, B. S.; Parker, V. D. J . Am. Chem. Soc. 1975, 97, 5211-
5217.
(31) Leventis, N.; Elder, I. A.; Gao, X.; Bohannan, E. W.; Sotiriou-
Leventis, C.; Rawashdeh, A. M. M.; Overschmidt, T. J .; Gaston, K. R.
J . Phys. Chem. B 2001, 105, 3663-3674.
(32) (a) Kitagawa, T.; Toyoda, J .; Nakasuji, K.; Yamamoto, H.;
Murata, I. Chem. Lett. 1990, 897-900. (b) Illescas, B.; Mart´ın, N.;
Segura, J . L.; Seoane, C. J . Org. Chem. 1995, 60, 5643-5650.
(33) (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.

3678 J . Org. Chem., Vol. 67, No. 11, 2002
Mac´ıas-Ruvalcaba et al.
F igu r e 2. Resonance hybrids showing stabilization of the radical anion of the PANs where R is an electron-donor (ED) group or
an electron-accepting (EA) group.
Ta ble 3. En er gy of th e F r on tier Or bita ls a n d Th eir Neigh bor s in h a r tr ees for th e Qu in on es MeOBCD, MeBCD, BCD,
ClBCD, COMeBCD, CF 3BCD, CNBCD, a n d NO2BCD
R
EHOMO-1
EHOMO
ELUMO
ELUMO+1
η
Etotal
OMe
Me
H
-0.22924
-0.23324
-0.24061
-0.24220
-0.24742
-0.25618
-0.25625
-0.26707
-0.21073
-0.22287
-0.22672
-0.23481
-0.23695
-0.23934
-0.24632
-0.24905
-0.10089
-0.10232
-0.10383
-0.10989
-0.11202
-0.11280
-0.11775
-0.11986
-0.03873
-0.03951
-0.04103
-0.04797
-0.05120
-0.04990
-0.05515
-0.08248
0.05492
0.06028
0.06145
0.06246
0.06247
0.06327
0.06429
0.06460
-934.89465
-859.69225
-820.37233
-1279.96543
-973.02320
-1154.40724
-912.61332
-1024.87154
Cl
COMe
CF3
CN
NO2
bond as shown in the resonant structure III (Figure 1),
and not only toward the carbonyl C4-O2, as in the PANs
(Figure 1, structure I). The natural charges shown in
Table S3 indicate that the electrophilic character of C1
decreases when R is an electron-donor group, confirming
participation of the resonant form III (Figure 1).
of the changes in bond lengths is significantly less than
changes in bond lengths for the PANs.19 In the BCDs,
there is, therefore, less participation of the free electron
pair of the nitrogen toward the quinone system than in
the PAN molecules, as demonstrated by the BCD y-band
transition shifting to shorter wavelengths than in the
PANs. This behavior is due to the fact that, in the BCDs,
the indolic nucleus loses its aromatic character upon
conjugating with the C4-O2 carbonyl (Figure 1, structure
IV). The results in Tables S1 and S2 show that even
when both fragments C1′-N-C2-C3-C4-O2 and C1′-N-
C2-C1-O1 demonstrate sensitivity to the effect of the
substituents, the C1′-N-C2-C1-O1 fragment registers
less significant effects as the substituent group changes.
Electronic density analysis of the critical bond points
showed that, for the BCD, there was no evidence of
Like previous results reported for the PANs,19 in the
BCD series, a bonding pattern also exists for the
C1′-N-C2-C3-C4-O2 fragment. Increased electron-
donor capability in the substituent increases the C1′-N,
C2-C3, and C4-O2 bond lengths, and decreases the N-C2
and C3-C4 bond lengths. This pattern is explained by
participation of the resonant form IV presented in Figure
1. The topological analysis of electron density shown in
Table S2 established the same pattern in terms of the
2
WBIs, F values, and ∇ F values. However, the magnitude

Electroreduction of BCDs and PANs
J . Org. Chem., Vol. 67, No. 11, 2002 3679
O1‚‚‚H-N hydrogen bonding, which had been previously
observed for the PAN series.19 However, the presence of
a weak interaction between the proton at C5′ and the
C4-O2 carbonyl oxygen (O2‚‚‚H-C5′) was revealed (Figure
S1 of the Supporting Information).
PAN derivatives,19 one would expect the BCDs to have
more negative E1/2 potentials for the first wave than the
PAN molecules. However, we observe in Table 2 that this
is not true. This suggests that the LUMO energy level is
not the only property determining the reactivity of the
molecule in accepting electrons. Neither can we attribute
the BCD facility in accepting the first electron to the fact
that the radical anion of these molecules is more stable
that of the PAN derivatives (Table 2). If this were true,
the molecules p-MeOPAN, MeBCD, and MeOBCD, which
have similar ln K values, would also present similar E1/2
values for wave I, and this was not the case (Table 2).
Thus, it is possible to establish that both the LUMO
energies for Q and the stability of the radical anions (Q•-)
are factors that significantly affect the half-wave poten-
tials for the first reduction transfer (in accordance with
the corresponding Nernst equation).
In agreement with this, the energetics of the second
reduction step would be associated with the SOMO
(singly occupied molecular orbital) of the radical anions
and the electronic repulsion in the SOMO, as well as the
stability of the dianion hydroquinones. In our work,
SOMO energy values were not determined, and this is
the reason the relationships with the half-wave potentials
of the second wave are not included.
For the substituted BCDs, NBO analysis establishes
that, unlike the LUMO difference, which is always found
at the C4 atom, the HOMO is located on different atoms
depending on the substituent. For MeOBCD, the HOMO
corresponds to an np-type orbital of the unshared electron
pairs of the -OMe group oxygen. For CF3BCD, the
HOMO corresponds to an np-type orbital of the fluoride
atoms, for CNBCD, to the np-type electron pair of the
-CN group nitrogen, for ClBCD, to an unshared electron
pair of the chlorine atom, and for BCD and MeBCD, to
an np-type orbital of the O1 of the quinone.
We last discuss the molecular orbital interpretation of
the optical absorption properties displayed by quinones
studied in this work. From data in Table 3, we establish
that the energy gap between the LUMO and HOMO is
greater for the BCD compounds than for the PAN
derivatives,19 which is in excellent agreement with the
experimental values of λmax for the y-band. These results
suggest that the lower energy absorption band (y-band)
of BCDs and PANs could result from the HOMO-LUMO
electronic transitions. Experimental observations for the
BCDs, where the λmax values for the y-band are more
susceptible to the effect of the substituents than those
of the PAN series of compounds, are explained by the
fact that, in the BCDs, the LUMO is localized at C4,
whereas for the PANs, it is at C1, and the C4-O2 carbonyl
is more susceptible to the effect of the substituents.
Increases in bond distances and decreases in Laplacian
and electronic density values for the O2‚‚‚H-C5′ interac-
tion, observed in the BCDs compared to the correspond-
ing interaction O1‚‚‚H-N in the PANs,19 indicate that the
O2‚‚‚H-C5′ interaction is much weaker than the hydrogen
bond O1‚‚‚H-N observed in the PANs. It is convenient
to mention that we have no experimental evidence for
the intramolecular O2‚‚‚H-C5′ interaction in the BCDs.
Furthermore, this hydrogen bond is incompatible with
the electronic density delocalization on the indolic system.
According to the delocalization of Figure 1 (structure II),
the carbon atom C5′ accepts a partial negative charge,
which decreases the polarization of the C5′-H bond,
causing the partial positive charge, required for the
hydrogen bond, on that H atom to be diminished. Thus,
this hydrogen bond is only a result of theoretical calcula-
tions obtained under conditions very different from the
experimental conditions. Therefore, in the BCDs the
absence of the O1‚‚‚H-N interaction provokes the pos-
sibility of polarizing the C1-O1 carbonyl to be lost, so
access of the first electron to C1 is hindered. We empha-
size that, for the BCDs, both the natural charge of the
C4 carbon and the Laplacian value for the C4-O2 bond
are more positive than those corresponding to the C1
carbon and the C1-O1 carbonyl, indicating that, in this
system, the C4-O2 carbonyl has a more electrophilic
character than the C1-O1 carbonyl. Thus, the greater
electrophilic character of the C4-O2 carbonyl indicates
that, for the BCD systems, the first electron transfer
corresponds to the reduction of the C4-O2 carbonyl and
not the C1-O2 carbonyl, as is the case for the PANs.19
Analysis of frontier molecular orbitals supports this
scheme since in all the BCDs, the LUMO is located in
the Rydberg orbital located at C4, while for the PANs,
the LUMO is found at C1. In addition, the good linear
relationships between the LUMO energy level (ELUMO
)
and the E1/2 potentials for the first reduction wave (eqs
10 and 11) confirm the fact that, in these molecules, the
first electron addition occurs at the LUMO.
for the BCD series
E1/2(wave I) ) -475ELUMO (eV) - 2518
(n ) 5, r2 ) 0.9801) (10)
for the PAN series
E1/2(wave I) ) -173ELUMO (eV) -1708
(n ) 5, r2 ) 0.9670) (11)
Con clu sion s
The values of the slopes in eqs 10 and 11 also agree
with the susceptibility to the effect of the substituents
previously seen in the linear trends of eqs 3 and 5.
Negative values of the slopes in eqs 10 and 11, on the
other hand, indicate that when the LUMO energy level
increases, the facility of reduction of the quinone molecule
decreases. This suggests that the LUMO energy level
could be considered an index of the facility of the molecule
to accept electrons. On the basis of this fact and consider-
ing that the ELUMO values for a single substituent are
greater for the BCD derivatives (Table 3) than for the
We synthesized five BCDs. In the synthetic strategy,
the PANs were used as precursors for the BCDs. Cyclic
voltammograms of PANs and BCDs exhibited two one-
electron reduction waves to form the corresponding
radical anion and dianion. In this study the combination
of spectroscopic UV-vis data with the molecular orbital
quantum-chemical calculations allowed us to establish
that closure of the ring in the PAN molecules (connection
between C3 and C6′) that gives rise to the BCD system
results in a decrease in the electron density of the
quinone system, as well as loss of the hydrogen bond

3680 J . Org. Chem., Vol. 67, No. 11, 2002
Mac´ıas-Ruvalcaba et al.
260 °C; IR (KBr, cm-1) 3400, 3232, 3070, 2934, 1666, 1592,
1531, 1483, 1433; UV-vis (EtOH; λmax, nm (ꢀ)) 211 (35889),
271 (44790), 401.5 (8556); 1H NMR (CDCl3) δ 12.74 (s, 1H,
NH), 8.13 (dd, 1H, J ) 7.2, 1.7 Hz, H8), 8.016 (dd, 1H, J ) 2.5,
0.6 Hz, H5′), 8.09 (dd, 1H, J ) 7.2, 1.7 Hz, H5), 7.76 (td, 1H, J
) 7.4, 1.8 Hz, H7), 7.71 (td, 1H, J ) 7.4, 1.8 Hz, H6), 7.46 (d,
1H, J ) 8.7 Hz, H2′), 7.20 (dd, 1H, J ) 7.0, 1.5 Hz, H3′), 2.47
(s, 3H, CH3); 13C NMR (CDCl3) δ 180.13 (C1), 177.28 (C4),
134.11 (C1′), 133.34 (C7), 132.91 (C8a), 132.91 (C4a), 132.52 (C4′),
132.24 (C6), 128.31 (C3′), 125.75 (C8), 125.53 (C5), 124.16 (C2),
121.57 (C6′), 121.56 (C5′), 116.98 (C3), 113.06 (C2′), 21.11 (CH3);
MS m/z 261, 260, 232, 204. Anal. Calcd for C17H11NO2: C,
78.15; H, 4.24; N, 5.36. Found: C, 77.88; H, 4.17; N, 5.08.
2-Aceth yl-5H-ben zo[b]ca r ba zole-6,11-d ion e (COMeB-
CD). Recrystallization from glacial acetic acid gave a 12%
yield: mp 265-266 °C; IR (KBr, cm-1) 3430, 3173, 3136, 2926,
2858, 1667, 1639, 1588, 1385; UV-vis (EtOH; λmax, nm (ꢀ)) 205
(11332), 232 (14479), 266 (31135), 368 (4897); 1H NMR (CDCl3)
δ 13.16 (s, 1H, NH), 8.81 (d, 1H, J ) 0.9 Hz, H5′), 8.14 (dd,
1H, J ) 7.0, 1.5 Hz, H8), 8.10 (dd, 1H, J ) 7.5, 1.5 Hz, H5),
7.95 (dd, 1H, J ) 8.7, 1.8 Hz, H3′), 7.81 (td, 1H, J ) 7.5, 1.5
Hz, H6), 7.76 (td, 1H, J ) 7.5, 1.5 Hz, H7), 7.60 (d, 1H, J ) 8.7
Hz, H2′), 2.67 (s, 3H, CH3); 13C NMR (CDCl3) δ 179.94 (C1),
171.15 (C4), 140.28 (C1′), 138.30 (C4′), 133.74 (C6), 132.72 (C7),
132.45 (C4a), 132.35 (C8a), 125.87 (C8), 125.74 (C5), 125.74 (C3′),
124.82 (C6′), 123.97 (C5′), 123.25 (C2), 118.27 (C3), 113.47 (C2′),
26.289 (CH3), 196.47(CdO); MS m/z 289, 274, 264, 190. Anal.
Calcd for C18H11NO3: C, 74.73; H, 3.83; N, 4.84. Found: C,
74.45; H, 4.18; N, 4.53.
2-(Tr iflu or om e t h yl)-5H -b e n zo[b]ca r b a zole -6,11-d i-
on e (CF 3BCD). Recrystallization from nitrobenzene gave a
16% yield: mp 250-252 °C; IR (KBr, cm-1) 3400, 3244, 2920,
1658, 1593, 1475, 1402; UV-vis (EtOH; λmax, nm (ꢀ)) 206
(24943), 226 (18334), 280 (40512), 348 (6360); MS m/z 315, 287,
279, 166, 149. Anal. Calcd for C17H8F3NO2: C, 64.77; H, 2.56;
F, 18.08; N, 4.44. Found: C, 64.52; H, 2.48; F, 17.48; N, 4.16.
Electr och em ica l P r oced u r e. Solven t a n d Su p p or tin g
Electr olyte. Acetonitrile (AN) was dried overnight with CaCl2
and purified by distillation on P2O5 under vacuum.40 Traces
of water in the solvent were eliminated by contact with 3 Å
molecular sieves in the dark. The absence of the characteristic
-OH bands in the IR spectrum confirmed complete elimination
of water traces. Tetraethylammonium tetrafluroborate (Et4-
NBF4) was dried under vacuum at 60 °C.
Electr od es, Ap p a r a tu s, a n d In str u m en ta tion . Cyclic
voltammograms were performed at 25 °C. A solution of 0.1 M
tetraethylammonium tetrafluroborate in acetonitrile was used
as the electrolytic medium. Since the solubility of the amino-
quinones in this medium varied, it was not possible to use all
the compounds at the same concentrations. Depending on their
solubility, the quinone concentrations varied from 0.34 to 0.64
mM. Prior to electrochemical determinations, N2 was bubbled
through all solutions for 60 min. The inert atmosphere was
especially critical in this study since dissolved oxygen in
solution not only can be reduced at potentials close to the
quinone reduction potentials, but also is capable of oxidizing
the radical anion generated in the first stage of reduction, as
occurs with other quinones.41
Voltammetric curves were recorded using an electrochemical
analyzer interfaced with a personal computer. Measurements
were made over a potential range between +433 and -2233
mV with a scan rate of 10-8000 mV/s. Cyclic voltammetry
for all the experiments used a three-electrode cell equipped
with a 7 mm2 glassy carbon working electrode. Prior to
measurements, this electrode was cleaned and polished with
0.05 µm alumina, wiped with a tissue, and sonicated in
distilled water for 2-4 min. The counter electrode consisted
of a platinum wire with a greater area than the working
electrode. A saturated calomel electrode (SCE) served as the
reference electrode.
between the oxygen of the C1-O1 carbonyl and the
nitrogen atom (O1‚‚‚H-N). The decrease in electron
density on the naphthoquinone system explains the
slightly less negative reduction potentials measured for
the first one-electron transfer in the BCDs compared with
PANs.
On the other hand, the absence of the interaction
O1‚‚‚H-N and the delocalization of the electronic density
on the carbonyl C1-O1 are factors that contributed to
diminishing the polarization of the C1-O1 carbonyl,
indicating that the C4-O2 carbonyl must be more elec-
trophilic than the C1-O1 carbonyl. Thus, in the BCDs,
the first wave corresponds to reduction of the C4-O2
carbonyl, as supported by the fact that, in all the BCDs,
the LUMO is at C4, and not at C1, as observed for the
PAN derivatives.19 From the obtained results, it was
possible to establish that the reactivity of the quinones
to accept the first electron depends on both the LUMO
energy of the quinone and the stability of the radical
anion toward the comproportionation reaction.
Greater susceptibility to the effect of the substituents
of the E1/2 potentials of wave I with respect to wave II in
the BCDs is due to the fact that the displacement of the
unshared electrons of the nitrogen, which depends on the
substituent R, mainly modifies the electron density of the
C4-O2 bond, with less effect on the electron density of
the C1-O1 bond.
Exp er im en ta l Section
Syn th esis. 1H and 13C NMR spectra (500 and 75 MHz,
respectively) were recorded in CDCl3 using TMS as a reference.
Signals of multiplicity are described as follows: s, singlet; d,
doublet; t, triplet; dd, doublet of doublets; td, triplet of doublets.
Molecular numbering is given in Table 1. Mass spectra (MS)
were performed by electron impact with a beam energy of 70
eV. IR spectra were recorded in KBr pellets. Elemental
analyses were performed by Galbraith Laboratories, Inc.
Melting points are not corrected.
Gen er a l P r oced u r e for th e P r ep a r a tion of 2-R-5H-
ben zo[b]ca r ba zole-6,11-d ion es (BCDs). The synthesis and
characterization of PAN and its derivatives are described
elswhere.19 The preparation of BCDs (Table 1) was performed
according to Bittner,38 whose method involves oxidative cou-
pling of the corresponding PANs19 in the presence of palla-
dium(II) acetate. Equal amounts of the corresponding PAN,
benzoquinone, and Pd(OAc)2 in glacial acetic acid were placed
in a three-neck, round-bottom flask (equipped with a ther-
mometer and a reflux condenser with a humidity-absorbing
trap). The reaction mixture was stirred at reflux temperature
o
(110 C) under nitrogen for the required reaction time, which
varied for the PAN used. The disappearance of the starting
material and appearance of the product were monitored by
thin-layer chromatography using mixtures of CH2Cl2/EtOH.
Once the reaction time was completed, the mixture was
allowed to cool to room temperature and filtered to eliminate
the Pd(0) formed during the reaction. To ensure elimination
of all the AcOH, the reaction mixture was evaporated to
dryness at reduced pressure. The residual solid was purified
by column chromatography using the mixture of solvents
mentioned above to give the corresponding benzocarbazoledi-
ones. All the compounds obtained by column chromatography
were later recrystallized from the appropriate solvent. The
preparation of BCD and MeOBCD was performed as de-
scribed.38,39
2-Meth yl-5H-ben zo[b]ca r ba zole-6,11-d ion e (MeBCD).
Recrystallization from acetone gave a 35.5% yield: mp 257-
(40) Coetzee, J . F.; Cunningham, D. K.; Mc. Guire, D. K.; Pad-
manabban, A. Anal. Chem. 1962, 34, 1139-1143.
(41) Wardman, P.; Tai-Shun L.; Sartorelli, A. C. J . Med. Chem. 1986,
29, 1381-1384.
(38) Bittner, S.; Krief, P.; Massil, T. Synthesis 1991, 215-216.
(39) Luo, Y.-L.; Chou, T.-C.; Cheng, C. C. J . Heterocycl. Chem. 1996,
33, 113.

Electroreduction of BCDs and PANs
J . Org. Chem., Vol. 67, No. 11, 2002 3681
Redox potentials for all compounds were obtained by
measuring the current i in the working electrode as a function
of the potential E (V) vs the SCE. Contact with the reference
electrode was made using a Luggin tube filled with electrolytic
medium. To establish an internal reference system, ferrocene
was added to the studied solutions, and all the redox potentials
reported in this study refer to the ferrocenium/ferrocene redox
couple (Fc+/Fc) as recommended by IUPAC.42 In this case
the potential for the Fc+/Fc redox couple, determined by
voltamperometric studies, was 443 ( 2 mV.
Com p u ta tion a l Meth od s. Complete geometry optimiza-
tion was performed according to the density functional theory
(DFT) using the B3LYP/6-31G(d,p) level with the Gaussian
94 program (G94).35 The functional hybrid B3LYP defines the
functional exchange as a Hartree-Fock local linear combina-
tion in terms of gradient exchange.43
The exchange functional is combined with a local correlation
function with a corrected gradient. The C*EcLYP+(1-C)-
*ECVWN correlation functional is used, where LYP is the Lee,
Yang, and Parr44 correlation functional, which includes both
local and nonlocal terms, and VWN is the Vosko, Wilk, and
Nusair correlation functional, to provide the local correlation
required in excess, since LYP contains a local term essentially
equivalent to VWN.43 The 6-31G(d,p) orbital base used adds
polarization functions for heavy atoms and hydrogen. NBO
analysis was performed using version 3.1, which is included
in G94.36 Densities computed at B3LYP/6-31G(d,p) from the
G94 output were used with the AIMPAC37 set of programs to
calculate the properties of critical points (cps) in the charge
2
density, F values, ∇ F values, and ꢀ values.
Ack n ow led gm en t. We thank the Consejo Nacional
de Ciencia y Tecnolog´ıa (CONACYT) for the financial
support given via Grant Nos. 28016E and 32420-E, and
for a fellowship to N.M.-R., who is also grateful to DGEP
for a complementary fellowship. We are grateful to the
Direccio´n General de Servicios de Co´mputo Acade´mico
(DGSCA), Universidad Nacional Auto´noma de Me´xico,
and to the Escuela de Ciencias Qu´ımicas, Universidad
La Salle, for their computational support, as well as for
the generous gift of supercomputer CPU time. We really
appreciate the suggestions and recommendations of the
reviewers that improved this work.
Su p p or tin g In for m a tion Ava ila ble: Relevant geometry
data, Wiberg bond indexes, properties of the relevant critical
points, and natural charges for the BCDs (Tables S1-S3 and
Figure S1) and complete computational results in the form of
Z-matrixes with the computed total energies. This material
is available free of charge via the Internet at http://pubs.acs.org.
(42) Gritzner, G.; Ku¨ta, J . Pure Appl. Chem. 1984, 4, 462-466.
(43) Stephens, P. J .; Devlin, F. J .; Chabalowski, C. F.; Frisch, M. J .
J . Phys. Chem. 1994, 98, 11623-11627. Becke, A. D. J . Chem. Phys.
1993, 98, 1372-1377, 5648-5652.
(44) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. 1988, B37, 785-789.
Miehlich, B.; Savin, A. Stoll, H.; Preuss, H. Chem. Phys. Lett. 1989,
157 (3), 200-206.
J O011083K