Linear free energy relationships of half-wave reduction potentials of (E)-4-aryl-4-oxo-2-butenoic acids

Article abstract of DOI:10.1016/j.tetlet.2009.11.129

Half-wave reduction potentials of a set of twelve (E)-4-aryl-4-oxo-2-butenoic acids obtained by direct current polarography in methanol are reported. E_1/2 is correlated with Hammett sigma values as well as with values of frontier molecular orbitals calculated at a semiempirical molecular orbital level. Constant potential electrolysis of a representative compound shows that the first polarographic wave corresponds to one-electron reduction. The isolation of the product proves reduction of the activated double bond.

Full text of DOI:10.1016/j.tetlet.2009.11.129

Tetrahedron Letters 51 (2010) 734–738
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Linear free energy relationships of half-wave reduction potentials
of (E)-4-aryl-4-oxo-2-butenoic acids
a
b,*
Ferenc T. Pastor , Branko J. Drakulic´
a
Faculty of Chemistry, University of Belgrade, Studentski Trg 12-16, 11000 Belgrade, Serbia
Department of Chemistry-IChTM, University of Belgrade, Njegoševa 12, 11000 Belgrade, Serbia
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Half-wave reduction potentials of a set of twelve (E)-4-aryl-4-oxo-2-butenoic acids obtained by direct
current polarography in methanol are reported. E1/2 is correlated with Hammett sigma values as well
as with values of frontier molecular orbitals calculated at a semiempirical molecular orbital level. Con-
stant potential electrolysis of a representative compound shows that the first polarographic wave corre-
sponds to one-electron reduction. The isolation of the product proves reduction of the activated double
bond.
Received 22 September 2009
Revised 22 November 2009
Accepted 27 November 2009
Available online 2 December 2009
Keywords:
Ó 2009 Elsevier Ltd. All rights reserved.
Aroylacrylic acids
Half-wave potentials
Linear free energy relationship
Hammett sigma
Frontier orbitals
Many compounds with activated double bonds exert biological
activities. The antiproliferative potency of naturally occurring
flavonoids and related chalcones is well documented.1 A long-
standing interest of our group has been (E)-4-aryl-4-oxo-2-bute-
noic acids (aroylacrylic acids) and their derivatives. It has been
shown that the substitution pattern on the aroyl moiety influences
results in deviation from a straight line. To explore different reac-
tivities amongst more potent compounds, our chosen set includes
several alkyl-substituted derivatives.
Twelve compounds (1–12, Table 1) were prepared as described
8
1
13
previously, and were characterized by H and C NMR spectros-
copy, and by HR-MS (ESI). All the compounds had >98% purity.
2
9
both the antiproliferative activity of the compounds and the reac-
Half-wave reduction potentials of 1–12 were measured by direct
tivity of the distal position of the molecules. So far, acidity con-
3
4
stants and reactivities of the –COOH group have been reported.
Alkyl-substituted aroylacrylic acids are significantly more po-
tent towards human dedifferentiated cell lines than halogen or
methoxy derivatives, whereas MeO– and HO–phenyl-substituted
chalcones exert more potent antiproliferative action.5 Recently,
an extensive study on the correlation between reduction potentials
Table 1
Half-wave reduction potentials for substrates 1–12 and the Hammett
in Equation 1
p
r values used
6
and electron affinities of a set of chalcones was reported. The elec-
trochemical reduction of the activated double bonds in many
molecular classes is well documented;7 to the best of our knowl-
edge, the electrochemical reduction of 4-aryl-4-oxo-2-butenoic
acids has not been described in the literature.
To establish a set for linear free energy relationships (LFERs), it
is common to choose compounds having well-balanced electron-
donating and electron-accepting groups, which span a broad range
of Hammett sigma values. The introduction of compounds bearing
substituents for which Hammett sigma values span a narrow
range, for example, a group of alkyl substituted compounds, often
Substrate
R
ꢀE1/2 (V)
r
p
1
2
3
4
5
6
7
8
9
H
4-Me
4-Et
4-i-Pr
4-n-Bu
4-sec-Bu
4-t-Bu
4-n-Dodecyl
4-F
0.527
0.543
0.536
0.539
0.532
0.543
0.548
0.528
0.532
0.498
0.474
0.569
0
ꢀ0.17
ꢀ0.15
ꢀ0.16
ꢀ0.12
ꢀ0.12
ꢀ0.20
ꢀ0.16
ꢀ0.07
0.11
10
11
12
4-Cl
4-Br
4-MeO
0.23
ꢀ0.27
*
Corresponding author. Tel.: +381 11 3336738; fax: +381 11 2636061.
E-mail address: bdrakuli@chem.bg.ac.rs (B.J. Drakuli c´ ).
0
040-4039/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2009.11.129

F. T. Pastor, B. J. Drakuli ´c / Tetrahedron Letters 51 (2010) 734–738
735
current (DC) polarography in 0.1 M NaCl in MeOH, at 25 ± 0.2 °C
versus a 3.0 M Ag/AgCl reference electrode. On all the polarograms
reported are the mean of three subsequent DC polarograms giving
half-wave potentials within 0.006 V.
(
ꢀ
of 1–12) two waves occurred (Fig. 1a). Waves appeared at about
0.5 V and ꢀ1.1 V versus the 3.0 M Ag/AgCl reference electrode.
Along with two triplets positioned between d 2.8 and d 3.4 in
the H NMR spectra of the products, which belong to the reduced
1
We focused our attention on the electrochemical reaction that cor-
responded to the first wave. In order to determine whether this
reaction was due to the reduction of the double bond or the aroyl
keto group, constant potential electrolysis of 2 was performed (see
2 2
bond (–CH@CH– ? –CH –CH –), an ABX pattern at d 2.5 to d 4.5,
characteristic of a –CH –CHR– moiety was also found (Fig. 2a). At
2
1
the same time, the H NMR spectra clearly showed that the double
bond had been reduced, due to the absence of doublets originating
from –CH@CH– at ꢂd 8 and d 6.9. The aroyl keto group was not re-
experimental), with the addition of CF
press di- or polymerization.
3
COOH in an attempt to sup-
1
3
duced, as was evident from the signal at d 198 in the C NMR spec-
0
0
Substrate 2 was chosen because the AA BB aryl pattern of the
trum (Fig. 2b).
1
signals in its H NMR spectrum allowed the unambiguous detec-
Half-wave reduction potentials were correlated with Hammett
+
tion of the doublets due to the ketovinyl double bond. Moreover,
sigma values (
r
p
for alkyl substituents, MeO and Br, and
r
p
for F
one Me resonance in the aliphatic region of both the 1H and
13
C
and Cl). Hammett sigma values were obtained from the litera-
1
0
NMR spectra could easily be distinguished from the signals of the
products of polymerization.
To find the optimal potential for electrolysis, linear sweep
polarograms were recorded. Figure 1b shows linear sweep polaro-
grams of deaerated 0.1 M LiCl in methanol (i), after the addition of
ture. Equation 1 and derived statistical values were obtained
1
1
using the BILIN program.
A linear correlation having reasonable statistics was obtained,
Equation 1:
ꢀE1/2 = ꢀ0.165 (±0.035)
n = 12, r = 0.958, sd = 0.007, F = 110.263, Q = 0.882, SPRESS
0.009
For the n-dodecyl-derivative 8, a Hammett
p
r + 0.516 (±0.006)
ꢀ3
2
2
.0 ꢁ 10 mol of trifluoroacetic acid (ii), and the same solution
=
ꢀ4
with 1.0 ꢁ 10 mol of 2 (iii). At ꢀ0.7 V, reduction of 2 occurred,
while the reduction of hydrogen was still negligible, so this poten-
tial was chosen as optimal. On the basis of net charge during con-
stant potential electrolysis, we found that the electrochemical
reaction that corresponded to the first wave was a single-electron
reaction; more exactly, the electrolysis required 1.06–1.16 equiv of
electricity per mole of substrate 2. After a few attempts, using dif-
p
r value could not
be found in the literature, and so the value for an n-octyl-group
was used. The exclusion of compound 8 gave a better correlation
2
(r = 0.976, F = 179.483, Q = 0.929).
A graphical presentation of ꢀE1/2 versus
p
r is given in Figure 3.
As can be seen from Equation 1 and the corresponding graph in
Figure 3, reasonable linear dependences exist between E1/2 and
Hammett sigma values. This indicates that substituents influence
ferent molar equivalents of CF
quantitative yield of nonpolymerized product. All the E1/2 values
3
COOH, we were unable to obtain a
ꢀ
3
Figure 1. (a) Polarogram of 5.0 ꢁ 10 M 2 in 0.10 M NaCl in MeOH in the presence of 0.002% Triton X-100; (b) determination of the optimal potential for the constant
ꢀ
3
potential electrolysis: linear sweep polarograms of deaerated 0.1 M LiCl in MeOH (i), after addition of 2.0 ꢁ 10 mol of trifluoroacetic acid (ii), and the same solution with
.0 ꢁ 10ꢀ mol of 2 (iii).
4
1
Figure 2. (a) Aliphatic 1H NMR signals and (b) aromatic 13C NMR signals of the product of electrolysis of substrate 2.

7
36
F. T. Pastor, B. J. Drakuli ´c / Tetrahedron Letters 51 (2010) 734–738
Figure 4.
a SOMO of compound 2 in the radical anion form, obtained using the
semiempirical MO PM6 method.
All the relevant data were collected from MOPAC outputs for
substrates 1–12, including the charges of atoms on which reduc-
tion occurred and the superdelocalizabilities of these atoms, the
highest occupied molecular orbital (HOMO) and the lowest unoc-
cupied molecular orbital (LUMO) energies for the neutral forms,
and the alpha singly occupied molecular orbital (SOMO) and the
corresponding LUMO for radical ions. As a consequence of the
selection of substituents on the compounds included in the set,
perfect statistics cannot be obtained, as alkyl-substituted com-
pounds always deviate from a straight line. However, the intercor-
relation matrix of frontier orbital energy values derived from
semiempirical calculations versus E1/2 should be indicative. In Ta-
ble 2 an intercorrelation matrix of PM6 and RM1 calculated HOMO
and LUMO values for neutral forms, as well as alpha SOMO and the
corresponding LUMO energies for radical anions, and differences
between HOMO–LUMO and SOMO–LUMO versus E1/2 are given.
The results summarized in Table 2 suggest that the RM1 calcu-
lated LUMOs of molecules in their neutral form is better correlated
with E1/2, than those calculated using PM6. In contrast the RM1 cal-
culated (alpha) LUMOs give fair correlation for the radical ion
forms of molecules in an implicit solvent or in a vacuum. Only
Figure 3. Graphical presentation of ꢀE1/2 versus
p
r .
the electronic properties of the activated double bond of the sub-
strates in a regular manner. Similar correlations for chalcones
and other ketovinyl-containing compounds are known in the liter-
ature. To the best of our knowledge, for (E)-4-aryl-4-oxo-2-bute-
noic acids, this is the first report. From Figure 3, it can be seen
that the linearity holds for the alkyl-substituted compounds (2–
), a situation rarely found in the literature.
Having in hand a fair correlation between E1/2 and Hammett
sigma values, we tried to obtain statistically good correlations
with descriptors derived from semiempirical calculations. The
initial structures of 1–12 were constructed by OMEGA using
8
12
1
3
MMFF94s. In this way, the structure obtained was in good agree-
ment with the crystal structure of unsubstituted derivative 1.
1
4
The structures, in their neutral and radical anion forms, were
further optimized by RM115 and PM6 semiempirical molecular
orbital methods. Neutral and ionized forms were treated both in
vacuum and in the implicit solvent MeOH, in which half-wave
potentials were determined. Optimization constraints were im-
posed on the atoms of the –C(O)–CH@ fragment in all molecules
in their neutral forms. Full optimization of structures 1–12 by MO-
16
PM6 calculated
a SOMO energies of molecules obtained both in
implicit solvent and in vacuum exert good correlations with the
experimentally obtained E1/2 values. SOMO–LUMO differences for
radical anions calculated by both PM6 and RM1 semiempirical
molecular orbital methods in vacuum or in an implicit solvent gave
reasonable correlations with experimentally obtained E1/2 values,
while RM1 was superior when an implicit solvent is used. It should
be emphasized that the aroylacrylic acids studied contain a –COOH
group close to the reduction center. This group is prone to hydro-
gen bonding. It is well known that the use of an implicit solvent
does not influence output geometries of optimized structures to
a significant extent, but influences atom charges, and consequently
MO energies. Stabilization of radical anions in a solvent can be seen
from HOMO (neutral forms) and SOMO (radical anions) energies in
the solvent and in vacuum (Table 3).
1
7
PAC2009 unrealistically distorted the –C(O)–CH@ bond. We did
not observe a similar problem for radical anions. Improved geom-
etries could not be assessed at higher levels of calculation; at the
*
*
DFT level (B3LYP/6-311G ) after full optimization, the molecules
were in unrealistic perfectly flat conformations. Similar situations
can be found in the literature for chalcones and related com-
pounds. As the geometries of the radical anions optimized by both
RM1 and PM6 semiempirical MO methods were reasonable, we
considered the reported results to be confined with respect to
the calculation protocols used.18
It should be also mentioned that the optimization algorithms
used decide the final geometries, rather than the level of calcula-
Half-wave reduction potentials (E1/2) experimentally obtained
in MeOH for the set of 12 para-phenyl-substituted (E)-4-phenyl-
4-oxo-2-butenoic acids have been reported. The E1/2 values ob-
tained display good correlations with Hammett sigma values. At-
tempts to obtain good correlations using frontier orbitals of the
molecules optimized at the semiempirical MO level in their molec-
ular and radical anion forms was not successful, contrary to the lit-
erature reports for similar types of compounds. Trials with
optimizations of molecules on different levels of theory (i.e., semi-
empirical and DFT) using different software showed that attention
must be paid to the resulting geometries, prior to the use of
descriptors derived from such calculations. It is a good compromise
to use semiempirical optimized radical anions for single point DFT
21
22
tion. Applying semiempirical MO AM1 or PM3 methods in
2
3
GAUSSIAN03, gave results with perfectly planar output geometries,
*
*
the same as B3LYP/6-311G . This situation is independent of the
choice of starting optimization geometries. For MOPAC2009 and
GAUSSIAN03 calculations, eigenvector and eigenvalue following opti-
mizers, respectively, were used.
Also, both PM6 and RM1 semiempirical MO methods, as applied
in MOPAC2009, distinguished well moieties where radical anions
should form, as is illustrated by the radical anion
ꢀ3.348 eV) of substrate 2, optimized in MeOH, as an implicit sol-
vent (Fig. 4).
a SOMO
(

F. T. Pastor, B. J. Drakuli ´c / Tetrahedron Letters 51 (2010) 734–738
737
Table 2
Intercorrelation (r values) of E1/2 for substrates 1–12 versus frontier molecular orbital energies calculated for the neutral and radical anion forms of the studied compounds
Method
PM6 versus E1/2a
RM1 versus E1/2a
Neutral (in vacuum)
Neutral (in MeOH)
HOMO
LUMO
0.555
0.819
0.607
0.792
HOMO
LUMO
0.531
0.785
0.636
0.846
Radical anion in vacuum
Radical anion in MeOH
a
a
SOMO
LUMO
0.835
0.824
0.499
0.827
a
a
SOMO
LUMO
0.792
0.664
0.356
0.851
Neutral (in vacuum)
Neutral (in MeOH)
HOMO–LUMO
HOMO–LUMO
SOMO–LUMO
SOMO–LUMO
0.084
ꢀ0.329
0.770
0.106
ꢀ0.418
0.861
Radical anion in MeOH
Radical anion in vacuum
0.784
0.787
The numerical values of HOMO, LUMO, and SOMO are in eV.
a
Relevant intercorrelation coefficients are given in italic.
Table 3
A. J. Electrochim. Acta 2004, 49, 455–459; (e) Annapoorna, S. R.; Prasada Rao, M.;
Sethuram, B. J. Electroanal. Chem. 2000, 490, 93–97.
Papa, D.; Schwenk, E.; Villani, F.; Klingsberg, E. J. Am. Chem. Soc. 1948, 70, 3356–
HOMO (neutral form of molecules) and SOMO (radical anions) energies of substrates
1
–12, obtained using the semiempirical MO PM6 method
8
9
.
.
3360.
Substrate
Solvent (MeOH)
Vacuum
Apparatus: DC polarograms were recorded on a VA 797 Computrace, Metrohm,
Switzerland. A three-electrode cell consisting of Metrohm MME in dropping
mercury electrode mode as the working electrode, a platinum rod as the
auxiliary, and aqueous 3.0 M Ag/AgCl as the reference electrode was used. The
desired temperature of the solution was held constant using an Ultra-
Thermostat, type U 10, Medingen, Dresden. Constant potential electrolysis
experiments (and the corresponding linear sweep polarography, used for the
determination of the optimal potential of electrolysis) were performed using a
HOMO
SOMOa
HOMO
SOMOa
1
2
3
4
5
6
7
8
9
ꢀ10.143
ꢀ9.918
ꢀ9.959
ꢀ9.977
ꢀ9.946
ꢀ9.970
ꢀ9.948
ꢀ9.948
ꢀ10.342
ꢀ10.028
ꢀ10.096
ꢀ9.605
ꢀ7.609
ꢀ7.561
ꢀ7.566
ꢀ7.567
ꢀ7.613
ꢀ7.567
ꢀ7.567
ꢀ7.569
ꢀ7.612
ꢀ7.616
ꢀ7.625
ꢀ7.547
ꢀ10.227
ꢀ9.920
ꢀ9.942
ꢀ9.942
ꢀ9.922
ꢀ9.934
ꢀ9.900
ꢀ9.926
ꢀ10.419
ꢀ10.078
ꢀ10.119
ꢀ9.532
ꢀ3.404
ꢀ3.348
ꢀ3.359
ꢀ3.364
ꢀ3.369
ꢀ3.372
ꢀ3.361
ꢀ3.379
ꢀ3.533
ꢀ3.595
ꢀ3.627
ꢀ3.368
CHI 760B Electrochemical Workstation, CH Instruments, USA.
A two-
compartment cell with a 12 mm diameter mercury pool electrode and a
3.5 M aqueous Ag/AgCl reference electrode in the first and a platinum auxiliary
electrode in the second compartment was used. The compartments were
separated with sintered glass. Measurement of half-wave reduction potentials:
DC polarography: The concentrations of the substrates, whose half-wave
10
11
12
ꢀ
3
potentials were determined in 0.10 M NaCl in methanol, were 5.0 ꢁ 10 M.
Triton X-100 was used as a maximum suppressor at a concentration of 0.002%.
All reported E1/2 values are the mean of three subsequent measurements giving
values within 0.006 V. Before the first run, the solution was thermostated over
a
a
SOMO reported.
10 min, and during the last 5 min, deaerated with nitrogen. Between successive
runs, the solutions were deaerated with nitrogen for 30 s. The scan rate was
ꢀ
1
1
0 mV s . The temperature of solutions was 25 ± 0.2 °C. Half-wave potentials
were determined by smoothing using the wave evaluation function of
Metrodata 797 PC-Software 1.0 (smoothing factor 6). Constant potential
electrolysis: The compartments of the electrolytic cell were filled with 0.1 M
LiCl in methanol to equal levels. The volume of the solution in the main
compartment was approximately 16 cm . The solution was deaerated with
nitrogen over 5 min and the linear sweep polarogram was recorded. In the
calculations. This method will be used for a larger set of congeners
in the future.
3
Acknowledgments
ꢀ
4
ꢀ3
main compartment between 5 ꢁ 10 and 2.0 ꢁ 10 mol of trifluoroacetic acid
was added, the solution deaerated again, and the second linear sweep
The authors gratefully acknowledge Professor Jimmy Stewart
for useful suggestions on the definition of radical anions in MOPAC.
All conclusions given in the text are the sole responsibility of the
authors. The Ministry of Science and the Technological Develop-
ment of Serbia support this work, Grants 142010 and 142037.
ꢀ
4
polarogram was recorded. After the addition of 1.0 ꢁ 10 mol of 2 in the
main compartment and deaeration, a third linear sweep polarogram was
recorded. From the recorded linear sweep polarograms, depending on the
number of moles of trifluoroacetic acid added, potentials from ꢀ0.7 to ꢀ0.8 V
versus the 3.5 M aqueous Ag/AgCl reference electrode were found to be optimal
for electrolysis. During electrolysis, the solution in the main compartment was
stirred and purged with nitrogen. The product of electrolysis was extracted
with CHCl
Na SO , and evaporated to dryness. H and C NMR spectra were recorded on
Bruker Avance instruments at 500/125 MHz in CDCl . Chemical shifts (d) are
referred to in ppm, downfield from TMS as the internal standard.
0. Hansch, C.; Leo, A.; Hoekman, D.. Exploring QSAR. In Hydrophobic, Electronic and
Steric Constants; American Chemical Society: Washington, DC, 1995. and
references cited therein.
1. Kubinyi, H. J. Med. Chem. 1977, 20, 625–629. http://www.kubinyi.de/bilin-
program.html.
2. (a) Boström, J. J. Comput. Aided Mol. Des. 2001, 15, 1137–1152; (b) Boström, J.;
Greenwood, J. R.; Gottfries, J. J. Mol. Graphics Modell. 2003, 21, 449–462.
www.eyesopen.com.
3
. The organic layer was washed with H O, dried over anhydrous
2
References and notes
1 13
2
4
3
1
.
.
Patia, H. N.; Dasa, U.; Sharmab, R. K.; Dimmock, J. R. Mini-Rev. Med. Chem. 2007,
, 131–139.
(a) Jurani c´ , Z.; Stevovi c´ , Lj.; Drakuli c´ , B.; Stanojkovi c´ , T.; Radulovi c´ , S.; Jurani c´ , I.
7
1
2
O. J. Serb. Chem. Soc. 1999, 64, 505–512; (b) Drakuli c´ , B. J.; Stanojkovi c´ , T. P.;
Zˇ i zˇ ak, Zˇ . P. unpublished results.
1
1
3.
4.
5.
Pfendt, L.; Popovi c´ , G.; Kosti c´ , M.; Stevovi c´ , Lj.; Drakuli c´ , B.; Jurani c´ , I. J. Chem.
Res. (S) 2000, 6, 260–261.
Stevovi c´ , Lj. S.; Drakuli c´ , B. J.; Jurani c´ , I. O.; Drmani c´ , S. Zˇ .; Jovanovi c´ , B. Zˇ . J. Serb.
Chem. Soc. 1998, 63, 359–365.
Boumendjel, A.; Boccard, J.; Carrupt, P.-A.; Nicolle, E.; Blanc, M.; Geze, A.;
Choisnard, L.; Wouessidjewe, D.; Matera, E.-L.; Dumontet, C. J. Med. Chem. 2008,
1
1
3. Halgren, T. A. J. Comput. Chem. 1999, 720–729.
4. Drakuli c´ , B. J.; Bogdanovi c´ , G. A.; Novakovi c´ , S. B.; Jurani c´ , I. O.: Crystal
structure of (E)-4-phenyl-4-oxo-2-butenoic acid, 41st Meeting of the Serbian
Chemical Society, Belgrade 2003. Book of abstracts, p 191.
5. Rocha, G. B.; Freire, R. O.; Simas, A. M.; Stewart, J. J. P. J. Comput. Chem. 2006, 27,
51, 2307–2310.
6
7
.
.
Hicks, L. D.; Fry, A. J.; Kurzweil, V. C. Electrochim. Acta 2004, 50, 1039–1047.
(a) Zimmer, J. P.; Richards, J. A.; Turner, J. C.; Evans, D. H. Anal. Chem. 1971, 43,
1
1
1
000–1006; (b) Jaeger, D. A.; Bolikal, D.; Nath, B. J. Org. Chem. 1987, 52, 276–
1101–1111.
2
78; (c) Quintana-Espinoza, P.; Yanez, C.; Escobar, C. A.; Sicker, D.; Araya-
6. Stewart, J. J. P. J. Mol. Mod. 2007, 13, 1173–1213.
Maturana, R.; Squella, J. A. Electroanal. 2006, 18, 521–525; (d) Alston, J. Y.; Fry,

7
1
38
F. T. Pastor, B. J. Drakuli ´c / Tetrahedron Letters 51 (2010) 734–738
7. Stewart, J. J. P. J. Comput. Aided Mol. Des. 1990, 4, 1–105; MOPAC2009, Stewart,
J. J. P. Stewart Computational Chemistry, Colorado Springs, CO, USA, 2009,
http://OpenMOPAC.net.
19. Klamt, A.; Schüürmann, G. J. Chem. Soc., Perkin Trans. 2 1993, 799–805.
20. Jmol: an open-source Java viewer for chemical structures in 3D. http://
www.jmol.org/.
1
8. Computational protocols: Initial 3D structures of substrates 1–12 were
generated by OMEGA from SMILES notation. MMFF94s force field was used.
All structures were optimized by semiempirical MO PM6 and RM1 methods,
using MOPAC2009. Constraint was imposed on atoms of the –C(O)–CH@
fragment in molecular forms of compounds, as described in the main text.
Eigenvector following optimizer was used and the structures were optimized
to the root mean square gradient below 0.1 kcal/mol. An implicit solvent was
treated by the COSMO19 formalism, as applied in MOPAC2009. Molecules in
their ground state were treated by configuration interactions (C.I. = 4,4). A
more realistic description of the frontier orbitals should be obtained in this
way. Radical anions were defined by using the MOPAC keyword CHARGE = ꢀ1;
implying the odd-electron systems. Several compounds in their neutral forms
21. Dewar, M. J. S.; Zoebisch, E. G.; Healy, E. F.; Stewart, J. J. P. J. Am. Chem. Soc.
1985, 107, 3902–3909.
22. Stewart, J. J. P. J. Comput. Chem. 1989, 10, 209–220.
23. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.;
Cheeseman, J. R.; Montgomery, J. A.; Vreven, Jr., T.; Kudin, K. N.; Burant, J. C.;
Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.;
Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.;
Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao,
O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Adamo,
C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi,
R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.;
Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.;
Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.;
Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.;
Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.;
Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe,
M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A.,
GAUSSIAN 03, Revision C.02, Gaussian, Inc., Pittsburgh, PA, 2003.
(
1, 4, 9 and 12) were optimized by semiempirical MO AM1 and PM3 methods
*
*
and on the DFT level (B3LYP/6-311G ), as applied in GAUSSIAN03, by
specification that the force constants are computed at every point. Cut-offs
on forces and step size to determine convergence were tightened using the
‘
Tight’ option. Two-electron integrals and derivatives were used (Gaussian
20
keyword Int = UltraFine). The alpha SOMO of 2 was visualized by Jmol, from
the MOPAC output. All computations were performed using a dual core 5 GHz
AMD processor in the Linux or Windows environment.