Experimental and theoretical study on the structure-property relationship of novel 1-aryl-3-methylsuccinimides

Article abstract of DOI:10.1016/j.molstruc.2016.09.086

A series of ten 1-aryl-3-methylsuccinimides was synthesized and their solvatochromic properties were studied in a set of fifteen binary solvent mixtures. The solute-solvent interactions were analyzed on the basis of the linear solvation energy relationshi

Full text of DOI:10.1016/j.molstruc.2016.09.086

Accepted Manuscript
Experimental and theoretical study on the structure-property relationship of novel 1-
aryl-3-methylsuccinimides
Nebojša R. Banjac, Bojan Đ. Božić, Jelena M. Mirković, Vesna D. Vitnik, Željko J.
Vitnik, Nataša V. Valentić, Gordana S. Ušćumlić
PII:
S0022-2860(16)31034-1
10.1016/j.molstruc.2016.09.086
MOLSTR 22998
DOI:
Reference:
To appear in: Journal of Molecular Structure
Received Date: 22 June 2016
Revised Date: 28 September 2016
Accepted Date: 29 September 2016
Please cite this article as: N.R. Banjac, B.E. Božić, J.M. Mirković, V.D. Vitnik, E.J. Vitnik, N.V. Valentić,
G.S. Ušćumlić, Experimental and theoretical study on the structure-property relationship of novel 1-
aryl-3-methylsuccinimides, Journal of Molecular Structure (2016), doi: 10.1016/j.molstruc.2016.09.086.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
Experimental and theoretical study on the structure-property relationship of novel
1-aryl-3-methylsuccinimides
Nebojša R. Banjaca, Bojan Đ. Božićb*, Jelena M. Mirkovićb, Vesna D. Vitnikc, Željko J.
Vitnikc, Nataša V. Valentićb, Gordana S. Ušćumlićb
a
Faculty of Agriculture, Food Technology and Biochemistry, University of Belgrade,
Nemanjina 6, 11080, Belgrade, Zemun, Serbia
b
Department of Organic Chemistry, Faculty of Technology and Metallurgy, University
of Belgrade, Karnegijeva 4, P.O. Box 3503, 11120 Belgrade, Serbia
c
Department of Chemistry, ICTM, University of Belgrade, Studentski trg 12-16, 11000
Belgrade, Serbia
*Corresponding author: e-mail: bbozic@tmf.bg.ac.rs; tel.: +381113303869; fax:
+381113370387.
1

ACCEPTED MANUSCRIPT
Abstract
A series of ten 1-aryl-3-methylsuccinimides was synthesized and their solvatochromic
properties were studied in a set of fifteen binary solvent mixtures. The solute-solvent
interactions were analyzed on the basis of the linear solvation energy relationship (LSER)
concept proposed by Kamlet and Taft. The electronic effect of the substituents on the
UV-Vis absorption and NMR spectra was analyzed using the simple Hammett equation.
Moreover, the B3LYP, CAM-B3LYP, and M06-2X functionals using the 6–311G(d,p)
basic set have been assessed in light of the position of experimental absorption maxima
obtained for these compounds. The integration grid effects have also been evaluated. An
interpretation of the substituent-effect transmission through the molecular skeleton and
the nature of the HOMO and LUMO orbitals based on quantum-chemical calculations is
given. The values of partial atomic charges from the atomic polar tenzors (APT), natural
population analysis (NBO), and charges fit to the electrostatic potential using the B3LYP,
CAM-B3LYP, and M06-2X methods are produced and correlated with different
experimental properties. In order to estimate the chemical activity of the molecule, the
molecular electrostatic potential (MEP) surface map is calculated for the optimized
geometry of 1-phenyl-3-methylsuccinimide.
Keywords: Succinimide; Absorption spectra; Solvent effect; Substituent effect; Quantum
chemical calculation.
2

ACCEPTED MANUSCRIPT
1. Introduction
Five-membered heterocycles with nitrogen in their structure are widely used in
different fields of science such as medicine, biology, and chemistry. Succinimides
(pyrrolidine-2,5-diones) and hydantoins (imidazolidine-2,4-diones) have been promising
candidates for anticonvulsive agents since they contain necessary structural features
responsible for this activity. This comprises a cyclic imide, one or more carbonyl groups,
and a lipophilic part of the molecule consisting of phenyl or alkyl groups related to
heterocyclic rings in certain relative orientations [1–6]. Additionally, it has been
experimentally established that succinimides possess anticonvulsive activity since they
are efficient against the maximal electroshock (MES) and subcutaneous
pentylenetetrazole (scPTZ) screens set standards for drug testing in early phases [7–9].
Correlation studies between chromatographic data and the selected structural
features described by ADMET parameters (refers to absorption, distribution, metabolism,
elimination, and toxicity in pharmacokinetics) of 1-aryl-3-phenyl-, 1-aryl-3,3-diphenyl-
and 1-aryl-3-methylsucccinimides have been reported previously using computational
medicinal chemistry methods [2,10]. According to these results, the investigated
succinimides of the most promising pharmacokinetics have been selected for further
pharmacodynamic drug development. It has been demonstrated that the affinity of a
particular molecule, defined as a potential anticonvulsive drug, to a specific receptor
highly depends on the interactions of the molecule with its surrounding space [11]. A key
feature of the molecules responsible for the transport through biological systems is its
lipophilicity. According to the literature, most active central-nervous system drugs have
optimum lipophilicity, defined as the logarithm of the 1-octanol/water partition
coefficient (log P), close to 2 [12,13]. An ability to transport through biological
membranes by passive diffusion is another important property of the molecules,
necessary for the in biological activity. This process involves a series of partitioning
steps, in combination with diffusion through several regions, i.e., partitioning between
aqueous intra- and extracellular media and phospholipid membranes [2]. Therefore, for a
further study of succinimides as potential anticonvulsant drugs, their solvatochromic
properties need to be taken into consideration.
3

ACCEPTED MANUSCRIPT
The purpose of this research is the application of the solvatochromic comparison
method for quantification and correlation of multiple solvent effects with properties and
interactions of the investigated succinimides which have potential pharmaceutical
applications. To the best of our knowledge, the structural spectroscopic and
solvatochromic properties of 1-aryl-3-methylpyrrolidine-2,5-diones have not been
thoroughly studied. In this work, ten 1-aryl-3-methylpyrrolidine-2,5-diones (Scheme 1)
have been synthesized and their solvatochromism has been studied in a set of fifteen
binary solvent mixtures. The effects of specific and non-specific interactions between
succinimide and solvent molecules on UV absorption maxima have been interpreted by
the linear solvation energy relationship (LSER) concept using the Kamlet-Taft equation
[14]. Moreover, the effects of substituents on the UV absorption frequencies and 13C
NMR chemical shifts have been analyzed by a linear free energy relationship (LFER)
model, using the simple Hammett equation [15]. In order to support structural and
solvatochromic properties of the investigated succinimides, density functional theory
(DFT/6–311G(d,p)) calculations have been carried out. The UV spectroscopic studies, as
well as HOMO and LUMO analysis, have been used to elucidate the intramolecular
charge transfer (ICT). Furthermore, the molecular electrostatic potential (MEP) surface
map has been calculated for the optimized geometry of the investigated succinimides in
order to evaluate the chemical activity of these molecules.
Scheme 1. Molecular structure of the 1-aryl-3-methylpyrrolidine-2,5-diones 1–10 with
the atoms and torsion angle labeled.
2. Results and discussion
2.1. Spectroscopic analysis
2.1.1. UV-spectroscopic measurements
The UV absorption spectra of the 1-aryl-3-methylpyrrolidine-2,5-diones 1–10
have been recorded in three pure solvents and selected binary solvent mixtures in the
range of 200–400 nm. Wavenumbers of the absorption maxima measured in different
solvents and their mixtures are given in Table 1. Table 1 shows that the UV absorption
wavenumbers strongly depend on the nature of the substituent, and that they are less
4

ACCEPTED MANUSCRIPT
affected by solvent properties. All substituents cause bathochromic band shifts relative to
the unsubstituted compound 4.
Table 1. Long-wavelength UV absorption maxima of 1-aryl-3-methylpyrrolidine-2,5-
diones 1–10.
The weakly electron-donating –CH3 group induces the smallest positive
solvatochromism, while the strongly electron-withdrawing –NO2 group exhibits the
largest solvent-induced bathochromic shift. UV absorption band shifts (Table 1) are
ascribed to different π→π* transitions involving the π-electron system of the investigated
compounds (Figure 1). The π-delocalization is mainly a result of the π-electron density
transmission from the phenyl ring, caused by electronic effects of the substituents in 4-
position, to the carbonyl groups of the succinimide unit. The bathochromic band shifts
observed for all compounds support the fact that a π→π* transition is responsible for UV
absorption. Moreover, the planarity of the molecule is the feature which significantly
affects transmission of the resonance effect through the molecule and therefore influences
the contribution of the π-delocalization to the UV absorption of each compound.
The strength of the specific and nonspecific interactions between the succinimide
and solvent molecules changes upon excitation as a consequence of the different
stabilization in the ground and excited states. In all solvents and binary mixtures, for
compounds 1–10, a small bathochromic shifts have been observed as a result of the
solvent-solute interactions. This is characteristic for molecules with low dipolarity in the
ground state, as electronic transitions in these molecules are associated with an
intramolecular electron transfer and where delocalization of electrons is possible between
nitrogen and oxygen, i.e. overlapping of π-electrons of carbonyl groups and a lone pair of
amide nitrogen (Fig. 1).
Figure 1. Mesomeric structures of 1-aryl-3-methylsuccinimides.
5

ACCEPTED MANUSCRIPT
2.1.2. Solvent effects on the electronic spectra (LSER analysis)
In order to explain the solvatochromism of 1–10, their UV absorption data have
been correlated with the simplified Kamlet–Taft solvatochromic equation [16]:
ν = ν0 + sπ* + aα
(1)
where π* is an index of the solvent dipolarity/polarizability, α is a measure of the solvent
hydrogen-bonding donor (HBD) acidity, and ν0 is the absorption value in cyclohexane as
reference solvent [14]. The relative susceptibilities of absorption frequencies to the
corresponding solvent parameters are expressed through the regression coefficients s and
a (Eq. 1). The solvent hydrogen-bond basicity parameter β is excluded from the
correlations due to its insignificant impact. Kamlet-Taft’s solvatochromic comparison
model is widely used for the quantitative interpretation of specific and nonspecific
solvent effects on UV-Vis absorption spectra since this model supposes attractive
interactions between the investigated solute and its surrounding and enables evaluation of
the capability of the solutes to form hydrogen bonds with the solvent. The correlations of
the absorption wavenumbers νmax have been achieved by multiple linear regression
analysis. The results of the correlation analysis are presented in Table 2 and the
percentage contributions of each interaction in overall solvation are given in Table 3. The
independent coefficients ν0, s, and a (Table 2) are fitted at the 95% confidence level.
Table 2. Regression fits to the solvatochromic parameters of Eq. 1.
Table 3. Percentage contribution of the two solvatochromic parameters of Eq. 1.
Тhe absorption wavenumbers (νmax) of the 1-aryl-3-methylsuccinimides 1–10
measured in pure solvents and in selected binary solvent mixtures show a satisfactory
correlation with the solvent parametars π* and α (Table 2). The results presented in Table
2 show that in all cases (excluding compounds 3, 5, and 8) the absolute value of a is
greater than s, which means that the solvent HBD acidity plays a more important role on
their solvatochromism than non-specific (solvent dipolarity/polarizability) interactions.
For the excluded compounds (3, 5, and 8), the solvent dipolarity/polarizability has a
slightly higher significance than HBD, while in all cases the solvent hydrogen-bond
acceptor basicity has no influence on the solvatochromism of 1–10.
6

ACCEPTED MANUSCRIPT
From the analysis of the absorption wavenumbers according to the Kamlet-Taft
equation (1), it has been found that the negative signs of s (excluding 4, 5, 6, and 8) and a
(excluding 4 and 8) for all 1-aryl-3-methylsuccinimides (Table 2) indicate a
bathochromic band shift with increasing solvent dipolarity/polarizability and solvent
hydrogen-bond donor acidity. This suggests better stabilization of the electron in excited
state by solvent molecules relative to the ground state.
Based on these results, it can be concluded that interactions of solvent with the
investigated compounds are very complex and that the nature of the substituents have a
strong influence on the UV absorption maxima. This phenomenon is mainly caused by
the differences in the conjugational or migrating abilities of the electron lone pair on
succinimide nitrogen.
2.1.3. Effects of substituents on the UV and NMR spectra of 1–10 (LFER analysis)
In order to gain insight into factors affecting the position of the UV absorption
maxima and the transmission of the electronic effects, the absorption wavenumbers of the
1-aryl-3-methylsuccinimides 1–10 have been correlated with Hammett substituent
+
constants, σp and σp , according to Hammett Equation [15]:
s = ρ·σ + h
(2)
where s is the substituent-dependent value: absorption wavenumbers (νmax), ρ is the
proportionality constant expressing the sensitivity of νmax to the substituent effects, σ is
the corresponding substituent constant (measuring the electronic effect of the
substituents), and h is the intercept (i.e., describes νmax of the unsubstituted member of the
series, σ = 0).
From Table 1 it can be observed that the UV absorption maxima of the amides
with both electron-donating and electron-accepting substituents are shifted to higher
wavelengths in all pure solvents and their binary mixtures, as compared to the
unsubstituted amide 4.
+
The plots νmax vs. the σp [17] and σp [18] substituent constants show deviations
from the Hammett equation, therefore the effect of the electron-donating and electron-
accepting substituents are taken into consideration separately. Different slopes for
electron-donating substituents (–OH, –OCH3, –CH3), and electron-accepting substituents
7

ACCEPTED MANUSCRIPT
(–Cl, –Br, –I, –COOH, –COCH3, –NO2) are acquired in all solvents and binary mixtures.
+
A better correlation of νmax is obtained with the electrophilic σp substituent constants
than with the σp constants, indicating an extensive electron delocalization in the aryl
+
amide part of the molecule. Figure 2 shows the dependence of νmax on the σp constants
for all succinimides (excluding 7) in pure solvents and binary mixtures.
+
Figure 2. Relationship between νmax and σp for the 1-aryl-3-methylsuccinimides 1–6 and
8–10.
For electron-accepting substituents, linear correlations of the absorption
+
wavenumbers, νmax, and the substituent constants, σp , with high values of the
proportionality constant, ρ, are obtained indicating, a significant influence of the
electronic effect of these groups on the π-electron density transmission according to the
mesomeric structures presented in Fig. 1. Moreover, significantly lower values of the
proportionality constant and their positive sign for electron-donating substituents indicate
larger solvent effects which diminish the substituent effect on the electronic transition
due to their interaction with the amide group of the succinimide ring.
Table 4. The results of the correlation between νmax 10–3/cm–1 and σp for the 1-aryl-3-
+
methylsuccinimides.
13C NMR chemical shifts of the signals for the two carbonyl C atoms in
succinimide moiety (Scheme 1) for the ten of 1-aryl-3-methylsuccinimides are given in
Table 5 as the substituent chemical shifts (SCS) compared to the parent compound 4.
Table 5 shows that electron-donating substituents exhibit downfield shifts, compounding
to a decrease in electron density at both atoms, whereas electron-accepting substituents
induce an opposite effect, related to an increase of electron density at both atoms (upfield
shift).
Table 5. 13C NMR substituent chemical shifts (δ/ppm).
8

ACCEPTED MANUSCRIPT
For a quantitative assessment of the substituent effects on the 13C NMR data of 1–
13
10, analogously to Eq. 2 the LFER Eq. 3 has been applied. A better correlation of C
+
SCSs for C2 and C5 is achieved with the σp substituent constants than with the σp
constants. The electronic effects of the substituents have no influence on the SCS of C3
and C4 of the succinimide ring. From the ρ values given in Table 6 it can be noticed that
there is no significant difference in the influence of electronic effects of substituents on
the C2 and C5 chemical shifts indicating a moderate impact of the substituents on the
phenyl nucleus on the electron density of the investigated atoms. These results are in
complete accordance with the molecular structure of 1-aryl-3-methylpyrrolidine-2,5-
dione as well as with the mechanism of transmission of the substituent electronic effects
throughout the molecule shown in Figure 1.
Table 6. Results of the SCS data correlations according to Eq. (2).
2.2. Computational studies
2.2.1. Theoretical calculation of optimal geometries and spectral data
2.2.1.1. Conformational stability of the investigated succinimides
A conformational analysis was carried out for the succinimides 1–10 in order to
locate all stable conformers. Their geometries are very rigid, so the only conformational
changes are possible in the position of the methyl and phenyl groups. As the rotation of
the methyl group shows a negligible change in the energy of the system, only the
positions of the phenyl group are investigated. For a better insight in changes of energy as
a consequence of the phenyl rotation, a potential energy scan (PES) was done. During the
calculation, all geometrical parameters were simultaneously relaxed while the torsion
angle θ (rotation around N1-C(Ph) bond, see Scheme 1) was varied in steps of 5° from 0–
360° (Fig S1, Supplementary material). Results show that only two conformers for the
molecule with symmetric substituents and four conformers for asymmetric substituents
are possible.
The full geometry optimizations of these conformers were performed by B3LYP
[19], CAM-B3LYP[20], and M06-2X [21] methods with the 6-311G(d,p) basis set in the
vacuum and solvents (ethanol and DMSO). The results provided from these calculations
9

ACCEPTED MANUSCRIPT
are used for different analyses in this work. The geometries of the most stable conformers
for compound 4 according to B3LYP method are shown in Figure 3.
Figure 3. The most stable conformers of compound 4 calculated by B3LYP level of
theory.
The conformers have similar energies which differ from the energy of the most
stable isomer by less than 0.20 kcal/mol, depending on substituent and the method of
calculation. According to a Boltzmann analysis, they are simultaneously present in the
mixture. The properties of interest are simulated based on the static results (the most
stable conformer is used) and some of them as the dynamic results (Boltzmann averaging
procedure is applied), too.
Analyzing the results of calculations from default integration grid (FineGrid), the
disagreement between the prediction of the most stable conformers, as well as, the ratio
in the mixture were observed for different methods (see Table S1, Supplementary
material). The B3LYP and CAM-B3LYP methods in vacuum predict conformer A (or
A2) as the most stable conformer and in the same time B3LYP method predict conformer
B (or B2) as the most stable conformer in ethanol solution for all investigated
compounds. The CAM-B3LYP method for ethanol as a solvent, as well as the M06-2X
method, do not show agreement in the prediction of the ratio and the most stable
conformer of all investigated compounds.
As the influence of the integration grid can be potentially significant for the
accurate prediction of geometry and energy of conformers, especially for the M06 suite
of functionals [22], the additional reoptimization of geometries are done using the pruned
99,590 integration grid (UltraFineGrid).
Comparison of the results from FineGrid and UltraFineGrid calculations shows
negligible effects on the ratio of conformers from B3LYP and CAM-B3LYP methods in
vacuum. The impact of integration grid is notable for both methods in ethanol as solvent.
The M06-2X method shows a significant influence of the used integration grid which
change the ratio and detection of the most stable conformer to a large extent. After
implementing UltraFineGrid, all three methods show agreements in the detection of
10

ACCEPTED MANUSCRIPT
conformer A (or A2) as the most stable in vacuum. All three methods predict conformer
B (or B2) as the most stable in ethanol solution.
The torsion angle θ determines the deviation from planarity (Scheme 1). It can
serve as a rough measure for the efficient transfer of substituent's resonance effect from
phenyl to succinimide ring and its accurate determination is directly connected with the
accurate prediction of all properties altered by substituent. The analyses of the optimized
geometries show that the biggest difference between the applied methods occurs exactly
in predicting of this torsion angle (see Table S2).
The applied methods predict a torsion angle θ from 37 to 45 degrees in vacuum,
depending on the substituent. Methods with the CPCM model for simulation of solvent
(ethanol and DMSO) predict a much larger value for this angle, i.e., from 45 to 72
degrees. The B3LYP method predicts slightly higher values than the CAM-B3LYP and
M06-2X methods in vacuum, while the difference becomes much higher when the
simulation of the solvent is introduced (B3LYP - 47–72°, CAM-B3LYP - 46–64°, and
M06-2X - 45–53°). Simultaneously, implementation of UltraFineGrid leads to additional
changes in the torsion angle θ which becomes more reliable, especially for the M06-2X
method.
2.2.1.2. Theoretical calculation of UV-Vis spectral data
It is well known that an accurately determined geometry is essential for a good
prediction of physicochemical properties, especially if the investigated compounds can
exist in different tautomeric or conformational states. The solvent can change the
geometry and energy of conformers and on that way affect the ratio of conformers in a
mixture. The electronic property and UV/Vis absorption spectrum of a mixture depend on
all present particles as well as of their interactions. For the accurate determination of
conformers, their ratio in the mixture and their interaction with the solvent, the selection
of an appropriate computational system is of crucial importance [23–26].
In order to obtain the UV properties of the investigated succinimides, TD-DFT
methods are performed using different types of functionals: B3LYP [19], CAM-B3LYP
[20], and M06-2X [21] (hybrid GGA, hybrid with improved long-range properties and
hybrid meta-GGA functional) with the 6-311G(d,p) basis set have been tested. Jacquemin
11

ACCEPTED MANUSCRIPT
et al. have demonstrated that these functionals provide satisfying UV spectra compared to
experiments for twenty conjugated molecules [24]. Also, they have evaluated other
effects such as basis set, integration grid, temperature, and cavities in the Polarizable
Continuum Model scheme.
The static method ensures good results for rigid organic compounds for
application in analytical chemistry. On the other hand, when an organic compound exists
in more than one conformer, then for the accurate prediction of UV–Vis spectra
molecular dynamic (MD) methods [27,28] or simulations based on Boltzmann averaging
should be used since static calculations are unsatisfactory [29]. In molecules in which
flexibility is of importance, these calculations are often required. Therefore, the
theoretical explanation of the experimental UV spectra should not include only electronic
effects.
The impact of the integration grid can be potentially significant and has also been
studied [22]. For a benchmark purpose we used the Euler-Maclaurin integration and
pruned 75,302 grid, constituted of 75 radial and 302 angular points (FineGrid) and a
tighter pruned 99,590 grid (UltraFineGrid).
The electronic absorption spectrum of 1-phenyl-3-methylpyrrolidine-2,5-dione 4
in ethanol as solvent was recorded in the range of 200–400 nm. The experimental
spectrum is shown on Figure 4 together with the calculated spectrum and the electronic
transitions obtained with the B3LYP method.
Figure 4. Experimental UV spectrum of 1-phenyl-3-methylpyrrolidine-2,5-dione 4,
measured in ethanol, and the calculated TD-B3LYP electronic absorption spectrum in
vacuum.
As can be seen from Figs. 4 and S2, the electronic absorption spectrum of
compound 4 shows only one low-intensity band at 216 nm. Also, all TD-DFT
calculations predict only one low-intense electronic absorption with an oscillator strength
of f < 0.25. The shape of the calculated spectrum is in good agreement with the
experimental one. The electronic absorption corresponds to a transition from the ground
to the higher excited states (fourth or fifth, depending on the model used for calculation).
It can be described as a π→π* transition which corresponds to a one-electron excitation
12

ACCEPTED MANUSCRIPT
from the highest occupied molecular orbital (HOMO) to the lowest unoccupied molecular
orbital (LUMO) or to the second lowest unoccupied molecular orbital (LUMO+1) orbital
(see Fig. 4).
In accordance with the aforesaid, the correlations of experimental and calculated
absorption UV maxima for compounds 1–10 are done. Values for the experimentally
recorded and computationally determined position of the UV maxima for all (located)
conformers of compounds 1–10 by TD methods are listed in Table S3.
From the results presented in Tables S4 and S5, it can be seen that the best
correlation is provided between experimental and calculated UV maxima from a TD
calculation with the B3LYP method in vacuum. A graphical representation of the
relationship between them is shown in Figure 5. The CAM-B3LYP and M06-2X methods
in vacuum highly underestimate the values of the UV maxima but still produce good
correlation results R2 > 0.9.
Figure 5. Correlation between experimental and calculated UV maxima provided by the
B3LYP method in a vacuum.
Introduction of the solvent effect in the calculation produces a different impact on
the calculated UV maxima. The B3LYP calculations produce values that mostly
overestimate the experimental one with the lower correlation quality. The UV values
provided from CAM-B3LYP method shows a very low impact on the correlation quality.
However, the improvement of the slope and interception is evident. The M06-2X method
shows the largest positive effect from introduction of the solvent to the correlation
quality.
The influence of the integration grid on the quality of the obtained UV results is
tested and shows small but positive impacts on the correlation results. Only the M06-2X
method with the simulation of solvent shows the biggest benefit from an implementation
of a tighter integration grid. These results are in agreement with results of other authors
[22,23].
The used DFT methods show that conformers are simultaneously present in the
mixture but also TD DFT methods predict difference of less then 2 nm between
conformers A and B. In accordance with that, simulated UV values based on the static or
13

ACCEPTED MANUSCRIPT
dynamic results produce a relatively small impact but it is directly connected with the
introduction of the solvent effects. The dynamic UV values (produced with Boltzmann
averaging procedure) simulated in vacuum show a notable improvement in comparison to
static results. Contrary, the UV values simulated in ethanol as solvent, show better
correlations when the static values (simulated from most stable conformer) are used.
2.2.2. Electronic analysis
The frontier molecular orbitals (FMO), the highest occupied molecular orbital
(HOMO) and the lowest unoccupied molecular orbital (LUMO), define the UV–Vis
spectra and the electric and optical properties. The FMO is useful in the estimation of the
chemical reactivity of a molecule as well as that part of the molecule which interacts with
other molecules, and in characterization of the kinetic stability of the molecule. A low-
frontier orbital gap indicates a more polarizable molecule and a higher reactivity, but also
a low kinetic stability [30]. Energies of HOMO and LUMO and the energy gap between
them for 2, 4, 5, and 10 molecule calculated at the B3LYP method with 6-311G(d,p)
basis set are shown in Figure S3.
The energies of the frontier MOs (HOMO and LUMO) and the energy gap
between them indicate a significant influence of the nature of the substituent introduced
in 4-position of the phenyl ring. The introduction of electron-donating or electron-
withdrawing substituent leads to a decrease in the energy gap (a smaller decrease for
electron-donating and a larger decrease for electron-withdrawing substituents).
2.2.2.1. The torsion angle (θ)
For the quantitative assessment of the substituent effects on the torsion angle θ of
the most stable conformers (Table S6) a conventional LFER model (Eq. 2) have been
+
used. A better correlation of cos2θ is obtained with the σp substituent constants (Eq. 3)
compared to the σp constants. The results obtained indicate an intramolecular charge
transfer (ICT) from substituent to amid moiety of the succinimide ring. Moreover, the
positive sign of the proportional constant is in complete accordance with the mechanism
of transmission of the substituent electronic effects throughout the molecule given in Fig.
1, where electron-acceptor substituents induce a higher bathochromic band shift
14

ACCEPTED MANUSCRIPT
compared to the electron-donating substituents. It can be observed that the degree of
deviation from the planarity of molecule, between succinimid and phenyl ring, is in
relationship with the electronic character of the substituent.
+
cos2θ = 0.321 + 0.183σp
(3)
(R = 0.982 s = 0.020; F = 215; n = 10)
2.2.2.2. NBO and APT charge data (Correlations with partial atomic charges)
Calculated partial atomic charges are often utilized for an interpretation of the
changes of dipole moment, molecular polarizability, and molecular electronic structure.
Also, the chemical activity of the molecule can be interpreted on the basis of atomic
charges transfer [30].
The values of partial atomic charges from the atomic polar tenzors (APT) [31],
natural population analysis (NBO) procedures [32], and charges fit to the electrostatic
potential according to Merz-Singh-Kollman [33], CHelp [34], and CHelpG [35] schemes
using the B3LYP, CAM-B3LYP, and M06-2X methods are produced and correlated with
different experimental properties. An LFER analysis is applied to NBO, APT, and
Electrostatic Potentional (EP) data of investigated atoms (C2, C5, O1, O2, N1), using Eq.
(2) and the results obtained for APT data are given in Table 7. Results derived from Eq. 2
for NBO and EP data are statistically excellent and in the agreement with correlations of
APT charges (Table S7).
Table 7. Results of the APT data in DMSO as solvent, derived from B3LYP/6-311G(d,p)
correlated with Eq. (2).
+
The better correlation with σp than with σp for all investigated atoms (in DMSO)
is an excellent indicator that electron transmission through the molecule is in good
accordance with the mesomeric structures proposed in Figure 1. From Table 7, it can be
observed that electron-donating groups induce an increase in electron density on the
O1=C2–N1–C5=O2 part of the succinimide ring, while electron-accepting ones cause an
15

ACCEPTED MANUSCRIPT
opposite effect. Moreover, substituents in 4-position of the phenyl ring have no influence
on the changes of electron density on the rest part of the succinimide ring (C3, C4, C6).
2.2.2.3. Molecular electrostatic potential
The molecular electrostatic potential (MEP) is often used for prediction of the
reactivity of the chemical species in nucleophilic and electrophilic reactions, since MEP
is associated with the electron density. For this purpose, the MEP at the B3LYP/6-
311G(d,p) optimized geometry for succinimides 1–10 was calculated. Red and yellow
colored regions (negative) of the MEP represent sites of electrophilic reactivity, whereas
red regions (positive) refers to nucleophilic sites, as shown in Fig. 6 (MEP).
Figure 6. Electrostatic potentials map of 1-phenyl-3-methylpyrrolidine-2,5-dione 4 in
vacuum from B3LYP/6-311G(d,p) calculation plotted on isosurfaces of 0.002 a.u.
electron density); color coding: red (very negative), orange (negative), yellow (slightly
negative), green (neutral), turquoise (slightly positive), light blue (positive), dark blue
(very positive).
As it can be seen from Figure 6, the molecule of the 4 (in vacuum) has two
possible sites for electrophilic and one site for nucleophilic attack. Negative regions of 4
are located around the carbonyl oxygen atoms and a moderate positive region is localized
on the H atoms of the methylene group from the succinimide ring.
2.3. Computational vs spectroscopic data
The UV-Vis absorption spectra are highly dependent on the energy of frontier
molecular orbitals among which electronic transition occurs. The spreading and energy of
FMO's are dependent on the torsion angle θ, which reflect deviation from planarity of the
planes in which succinimide and phenyl rings lay. The relations of cos2θ derived from
B3LYP method for the most stable conformers of the investigated compounds vs. UV
absorption maxima data have been calculated (Figure 7). An excellent linear correlation
is obtained with ethanol as solvent. The result is in accordance with previously described
16

ACCEPTED MANUSCRIPT
results obtained from applied Hammett equation on the UV absorption wavenumbers
(Figure 2).
As NMR shifts depend on shielding effects, which are directly influenced by
substituent electronic effects and electron distribution, NBO, APT, and EP data obtained
with B3LYP method for investigated compounds in DMSO are correlated with
experimental SCS data (Table 8).
Excellent relations between experimental and theoretical data directly indicate
high validity and precision of proposed models.
Figure 7. νmax vs. cos2θ data for investigated 1-aryl-3-methylpyrrolidine-2,5-diones.
Table 8. APT, NBO, EP vs. SCS data for the 1-aryl-3-methylpyrrolidine-2,5-diones 1–
10.
3. Experimental
Commercially available materials were obtained from Aldrich and Fluka, and
were used without further purification. All investigated 1-aryl-3-methylpyrrolidine-2,5-
diones were synthesized from 2-methylsuccinic acid and the corresponding anilines under
solvent-free conditions (Scheme 2) as described previously [3]. The products were
purified by crystallization from an ethanol/water mixture. The FT-IR spectra were
determined using a Bomem Fourier Transform-Infrared (FT-IR) spectrophotometer, MB-
1
13
Series 100 in the form of the KBr discs. The H and C NMR spectral measurements
were performed on a Varian Gemini 2000 (200 MHz). The spectra were recorded at room
temperature in deuterated dimethyl sulfoxide (DMSO-d6). The chemical shifts are
expressed in ppm values referenced to TMS. The ultraviolet (UV) absorption spectra
were recorded on a Schimadzu 1700 spectrophotometer in the region 200–400 nm. The
UV absorption spectra were taken in spectroscopic quality solvents (Fluka) at a
concentration of 5·10–5 mol dm–3 and showed no dependence on the concentration. All
melting points were uncorrected and are in degree Celsius.
Scheme 2. Synthetic pathway for the 1-aryl-3-methylpyrrolidine-2,5-diones 1–10.
17

ACCEPTED MANUSCRIPT
1
The molecular structure of compounds 1–10 was confirmed by UV, FT-IR, H
NMR, and 13C NMR spectra. Characterization data are given below.
1-(4-Hydroxyphenyl)-3-methylpyrrolidine-2,5-dione (1, C11H11NO3). Black solid; yield:
64 %; m.p.: 228–229 °C (227–228 °C) [36]; FT-IR (KBr): ν/cm–1 = 1771 (C=O), 1695
1
(C=O); H NMR (200 MHz, DMSO): δ/ppm = 9.81 (s, 1H, –OH), 7.00 (d, J = 8.0 Hz,
2H, N–Ph), 6.84 (d, J = 8.0 Hz, 2H, N–Ph), 3.10-2.89 (m, 2H, –CH2–), 2.51-2.39 (m, 1H,
–CH) 1.27 (d, J = 8.0 Hz, 3H, –CH3) ; 13C NMR (50 MHz, DMSO): δ/ppm = 180.5 (C2),
176.4 (C5), 159.2, 128.3, 125.6, 114.3 (Ph), 36.5 (C3), 34.7 (C4), 16.0 (C6).
1-(4-Methoxyphenyl)-3-methylpyrrolidine-2,5-dione (2, C12H13NO3). Colorless solid;
yield: 38 %; m.p.: 92–93 °C (91–93 °C) [37]; FT-IR (KBr): ν/cm–1 = 1773 (C=O), 1704
(C=O); 1H NMR (200 MHz, DMSO): δ/ppm = 7.19 (d, J = 8.0 Hz, 2H, N–Ph), 7.02 (d, J
= 8.0 Hz, 2H, N-Ph), 3.78 (s, 3H, –OCH3), 3.02-2.90 (m, 2H, –CH– + –CH2–), 2.53-2.38
(m, 1H, –CH2–), 1.28 (d, J = 8.0 Hz, 3H, –CH3); 13C NMR (50 MHz, DMSO): δ/ppm =
180.4 (C2), 176.3 (C5), 159.1, 128.6, 125.5, 114.2 (Ph), 55.5 (C8, –OCH3), 36.4 (C3),
34.6 (C4), 16.0 (C6).
1-(4-Methylphenyl)-3-methylpyrrolidine-2,5-dione (3, C12H13NO2). Colorless solid; yield:
47 %; m.p.: 106–108 °C (107–108 °C) [38] (107 °C) [36]; FT-IR (KBr): ν/cm–1 = 1776
1
(C=O), 1708 (C=O); H NMR (200 MHz, DMSO): δ/ppm = 7.29 (d, J = 8.0 Hz, 2H, N–
Ph), 7.15 (d, J = 8.0 Hz, 2H, N-Ph), 3.07-2.91 (m, 2H, –CH– + –CH2–), 2.55-2.40 (m,
13
1H, –CH2–), 2.34 (s, 3H, Ph–CH3), 1.28 (d, J = 6.0 Hz, 3H, –CH3); C NMR (50 MHz,
DMSO): δ/ppm = 180.3 (C2), 176.2 (C5), 137.9, 130.4, 129.5, 127.1 (Ph), 36.4 (C3),
34.7 (C4), 20.9 (C8, –CH3), 16.0 (C6).
1-Phenyl-3-methylpyrrolidine-2,5-dione (4, C11H11NO2). Colorless solid; yield: 37 %;
m.p.: 105–107 °C (105–107 °C) [38]; FT-IR (KBr): ν/cm–1 = 1769 (C=O), 1702 (C=O);
1H NMR (200 MHz, DMSO): δ/ppm = 7.54-7.27 (m, 5H, N–Ph), 3.10-2.92 (m, 2H, –
CH– + –CH2–), 2.56-2.42 (m, 1H, –CH2–), 1.29 (d, J = 6.0 Hz, 3H, –CH3); 13C NMR (50
18

ACCEPTED MANUSCRIPT
MHz, DMSO): δ/ppm = 180.2 (C2), 176.1 (C5), 133.0, 129.0, 128.4, 127.4 (Ph), 36.4
(C3), 34.7 (C4), 16.0 (C6).
1-(4-Chlorophenyl)-3-methylpyrrolidine-2,5-dione (5, C11H10ClNO2). Colorless solid;
yield: 53 %; m.p.: 109–111 °C (110–111 °C) [39]; FT-IR (KBr): ν/cm–1 = 1774 (C=O),
1
1704 (C=O); H NMR (200 MHz, DMSO): δ/ppm = 7.56 (d, J = 8.0 Hz, 2H, N–Ph),
7.33 (d, J = 8.0 Hz, 2H, N-Ph), 3.10-2.92 (m, 2H, –CH– + –CH2–), 2.57-2.40 (m, 1H, –
13
CH2–), 1.29 (d, J = 8.0 Hz, 3H, –CH3); C NMR (50 MHz, DMSO): δ/ppm = 179.9
(C2), 175.9 (C5), 132.8, 131.8, 129.1 (Ph), 36.4 (C3), 34.7 (C4), 15.8 (C6).
1-(4-Bromophenyl)-3-methylpyrrolidine-2,5-dione (6, C11H10BrNO2). Colorless solid;
yield: 52 %; m.p.: 139–141 °C (143 °C) [36]; FT-IR (KBr): ν/cm–1 = 1774 (C=O), 1706
1
(C=O); H NMR (200 MHz, DMSO): δ/ppm = 7.69 (d, J = 8.0 Hz, 2H, N–Ph), 7.27 (d, J
= 8.0 Hz, 2H, N-Ph), 3.09-2.92 (m, 2H, –CH– + –CH2–), 2.57-2.41 (m, 1H, –CH2–),
1.29 (d, J = 8.0 Hz, 3H, –CH3); 13C NMR (50 MHz, DMSO): δ/ppm = 179.9 (C2), 175.9
(C5), 132.7, 131.9, 129.5 (Ph), 36.4 (C3), 34.7 (C4), 15.8 (C6).
1-(4-Iodophenyl)-3-methylpyrrolidine-2,5-dione (7, C11H10INO2). Colorless solid; yield:
34 %; m.p.: 149–151 °C; FT-IR (KBr): ν/cm–1 = 1764 (C=O), 1704 (C=O); 1H NMR
(200 MHz, DMSO): δ/ppm = 7.62 (d, J = 8.0 Hz, 2H, N–Ph), 7.41 (d, J = 8.0 Hz, 2H, N-
Ph), 3.11-2.95 (m, 2H, –CH– + –CH2–), 2.59-2.41 (m, 1H, –CH2–), 1.30 (d, J = 8.0 Hz,
13
3H, –CH3); C NMR (50 MHz, DMSO): δ/ppm = 179.9 (C2), 175.9 (C5), 134.4, 132.6,
130.8, 126.8 (Ph), 36.5 (C3), 34.8 (C4), 15.8 (C6).
1-(4-Carboxyphenyl)-3-methylpyrrolidine-2,5-dione (8, C12H11NO4). Colorless solid;
yield: 40 %; m.p.: 200–203 °C; FT-IR (KBr): ν/cm–1 = 1777 (C=O), 1704 (C=O); 1H
NMR (200 MHz, DMSO): δ/ppm = 13.07 (s, 1H, –COOH), 8.06 (d, J = 8.0 Hz, 2H, N–
Ph), 7.44 (d, J = 8.0 Hz, 2H, N–Ph), 3.12-2.94 (m, 2H, –CH– + –CH2–), 2.58-2.43 (m,
1H, –CH2–), 1,29 (d, J = 6.0 Hz, 3H, –CH3); 13C NMR (50 MHz, DMSO): δ/ppm = 179.8
(C2), 175.8 (C5), 167.0 (C8, –COOH), 136.8, 130.5, 130.1, 127.3 (C7), 36.5 (C3), 34.8
(C4), 15.9 (C6).
19

ACCEPTED MANUSCRIPT
1-(4-Acethylphenyl)-3-methylpyrrolidine-2,5-dione (9, C13H13NO3). Colorless solid;
yield: 47 %; m.p.: 165–167 °C; FT-IR (KBr): ν/cm–1 = 1779 (C=O), 1708 (C=O); 1H
NMR (200 MHz, DMSO): δ/ppm =8.07 (d, J = 8.0 Hz, 2H, N–Ph), 7.48 (d, J = 8.0 Hz,
2H, N–Ph), 3.13-2.92 (m, 2H, –CH– + –CH2–), 2.57-2.41 (m, 1H, –CH2–), 2.59 (s, 3H, –
13
COCH3), 1.33 (d, J = 8.0 Hz, 3H, –CH3); C NMR (50 MHz, DMSO): δ/ppm = 199.5
(C8, –COCH3) 179.8 (C2), 175.7 (C5), 141.3, 136.9, 131.1, 127.2 (Ph), 36.5 (C3), 34.8
(C4), 15.9 (C6).
1-(4-Nitrophenyl)-3-methylpyrrolidine-2,5-dione (10, C11H10N2O4). Yellow solid; yield:
25 %; m.p.: 158–160 °C (159–161.5 °C) [40]; FT-IR (KBr): ν/cm–1 = 1777 (C=O), 1709
1
(C=O); H NMR (200 MHz, DMSO): δ/ppm = 8.36 (d, J = 8.0 Hz, 2H, N–Ph), 7.63 (d, J
= 8.0 Hz, 2H, N–Ph), 3.18-2.96 (m, 2H, –CH– + –CH2–), 2.61-2.46 (m, 1H, –CH2–),1.31
(d, J = 8.0 Hz, 3H, –CH3); 13C NMR (50 MHz, DMSO): δ/ppm = 179.6 (C2), 175.6 (C5),
146.7, 138.6, 128.1, 124.3 (Ph), 36.5 (C3), 34.8 (C4), 15.7 (C6).
3.1. Method of calculation
All density functional theory (DFT) calculations were performed using the
Gaussian 09 program package [41] with B3LYP [19], CAM-B3LYP [20], and M06-2X
[21] methods with 6-311G(d,p) basis set. The default convergence criteria were used
without any constraint on the geometry. The stability of the optimized geometry was
confirmed by frequency calculations, which gave real values for all the experimentally
frequencies. In adition, the FineGrid (75,302)p and UltraFineGrid (99,590)p grids for
Euler-Maclaurin integration are used for benchmark of methods accuracy to determine
the conformer's ratio. The solvent effect was introduced by the Conductor Polarizable
Continuum Model (CPCM) [42].
In order to have a good description of the UV properties of the compounds, we
tested the performance of TD-DFT using different types of functionals: B3LYP [19],
CAM-B3LYP [20], and M06-2X [21] (i.e. hybrid GGA, hybrid with improved long-
range properties and hybrid meta-GGA functionals). The functionals were used with the
20

ACCEPTED MANUSCRIPT
6-311G(d,p) basis set. UV absorption energies of all compounds were calculated for the
fully optimized geometry in vacuum and ethanol as solvent.
The values of partial atomic charges from the atomic polar tenzors (APT) [31],
natural population analysis (NBO) procedures [32] and charges fit to the electrostatic
potential according to Merz-Singh-Kollman [33], CHelp [34], and CHelpG [35] schemes
are produced.
The frontier molecular orbital energies and energy gap of the investigated
compounds are also calculated at the same level of DFT theory. For obtaining the
chemical reactivity of the molecule, the Molecular Electrostatic Potential (MEP) surface
is plotted over the optimized geometry.
The GaussView 5.0 graphical interface was used for a visual presentation of the
UV spectra, molecular orbitals, and MEP [43].
4. Conclusion
In this work, ten 1-aryl-3-methylsuccinimides were synthesized in order to
characterize and study the influence of solvents and substituents on their electronic
absorptions and 13C NMR spectra. A Kamlet–Taft analysis shows that most of the
observed solvatochromism is due to solvent HBD acidity as compared to solvent
dipolarity/polarizability (non-specific interactions). In all cases the solvent hydrogen-
bond acceptor basicity has no influence on the solvatochromism. According to the
+
Hammett correlations, better results are obtained with the σp than with the σp substituent
constants, indicating extensive π-electron delocalization in the aryl amide group. Much
better insight into the transmission of substituent effects was obtained when the electron-
donating substituents on the phenyl ring were considered separately from the electron-
accepting substituents. The results obtained from UV spectroscopy measurements show
that the wavenumbers of the ultraviolet absorption maxima depend significantly on the
nature of the substituent on the phenyl ring, and less on the properties of the solvent.
From the 13C NMR chemical shifts, it can be observed that there is no significant
difference in the influence of electronic effects of substituents in 4-position on the phenyl
group on the C2 and C5 atoms.
21

ACCEPTED MANUSCRIPT
A conformational analysis was carried out for the of 1-aryl-3-methylsuccinimides
1–10. The performance of DFT methods using different types of functionals: B3LYP,
CAM-B3LYP, and M06-2X with the 6-311G(d,p) basis set, are tested for prediction of
the accurate energy and ratio of conformers, as well as the properties derived from them.
Special attention is directed towards an accurate prediction of the UV absorption.
The best correlation is provided between experimental and calculated maxima from TD
calculations with B3LYP method in vacuum. The CAM-B3LYP and M06-2X methods
highly underestimate the values of UV maxima but still produce good correlation results.
This result is in agreement with the fact that for small and rigid molecules the B3LYP
method produces more accurate results than other functionals. The solvent influence on
the correlation results is detectable but it is relatively small, compared to the theory-
experiment discrepancies. The influence of the integration grid on the quality of produced
results is tested and shows small but positive impacts. Only the results from M06-2X
method in solvent shows big benefit from applying a tighter integration grid. The impact
of values simulated based on the static or dynamic results is relatively small but it is
directly connected with the introduction of solvent effects.
Excellent relations between experimental and theoretical data directly indicate a
high validity and precision of the proposed models.
Acknowledgments
This work has been financially supported by the Ministry of Education, Science
and Technological Development, Republic of Serbia, under Grants no. 172013 and
172035. Numerical simulations were run on the PARADOX cluster at the Scientific
Computing Laboratory of the Institute of Physics Belgrade, supported in part by the
Ministry of Education, Science, and Technological Development of the Republic of
Serbia under project No. ON171017.
Appendix A. Supplementary material
Supplementary data associated with this article can be found in the online version
at http://
22

ACCEPTED MANUSCRIPT
Reference
[1] Obniska J, Zejc A, Karolak-Wojciechowska J (1999) Farmaco 54:423–429
[2] Perišić Janjić N, Kaliszan R, Wiczling P, Milošević N, Ušćumlić G, Banjac N (2011)
Mol. Pharm. 8:555–583
[3] Katritzky RA, Nair KS, Witeck MR, Hutchins MS (2003) Arkivoc. 9–18 (Issue in
Honor of Prof. Gábor Bernáth)
[4] Poupaert HJ, Vandervorst D, Guiot P, Moustafa MMM, Dumont PP (1984) J. Med.
Chem. 27:76–78
[5] Trišović N, Valentić N, Ušćumlić G (2011) Chem. Cent. J. 5:62–63
[6] Banjac NR, Trišović NP, Vitnik ŽJ, Vitnik VD, Valentić NV, Ušćumlić GS, Juranić
IO (2013) Monatsh. Chem. 144:1525–1535
[7] Kamiński K, Obniska J (2008) Bioorg. Med. Chem. 16:4921–4931
[8] Obniska J, Byrtus H, Kamiński K, Pawłowski M, Szczesio M, Karolak-
Wojciechowska J (2010) Bioorg. Med. Chem. 18:6134–6142
[9] Kamiński K, Rzepka S, Obniska J (2011) Bioorg. Med. Chem. Lett. 21:5800–5803
[10] Perišić Janjić N, Kaliszan R, Milošević N, Ušćumlić G, Banjac N (2013) J.
Pharmaceut. Biomed. 72:65–73
[11] Murray JS, Abu-Awwad F, Polityer P, Wilson LC, Troupin AS, Wall RE (1998) Int.
J. Quant. Chem. 70:1137–1143
[12] Hansch C, Björkrotahn JP, Leo A (1987) J. Pharm. Sci. 76:663–687
[13] Hadjipavlou-Litina D (1998) Med. Res. Rev. 18:91–119
[14] Kamlet MJ, Abboud JLM, Abraham MH, Taft RW (1983) J. Org. Chem. 48:2877–
2887
[15] Hammett L (1937) J. Am. Chem. Soc. 59:96–103
[16] Kamlet M, Abboud J, Taft R (1981) Prog. Phys. Org. Chem. 13:485–630
[17] Brown H, Okamoto Y (1958) J. Am. Chem. Soc. 80:4979–4987
[18] Charton M (1981) Prog. Phys. Org. Chem. 13:119–251
[19] Becke AD (1993) J. Chem. Phys. 98:5648–5652
[20] Yanai T, Tew D, Handy N (2004) Chem. Phys. Lett. 393:51–57
[21] Zhao Y, Truhlar DG (2008) Theor. Chem. Acc. 120:215–241
[22] Wheeler SE, Houk KN (2010) J. Chem. Theory. Comput. 6:395−404
23

ACCEPTED MANUSCRIPT
[23] Mori T, Inoue Y, Grimme S (2007) J. Phys. Chem. A 111:4222–4234
[24] Charaf-Eddin A, Planchat A, Mennucci B, Adamo C, Jacquemin D (2013) J. Chem.
Theory. Comput. 9:2749–2760
[25] Brasca R, Romero MA, Goicoechea HC, Kelterer AM, Fabian WMF (2013)
Spectrochim. Acta A 115:250–258
[26] Arsovski VM, Božićn BĐ, Mirković JM, Vitnik VD, Vitnik ŽJ, Fabian WMF,
Petrović SD, Mijin DŽ (2014) J. Mol. Model. 20:2384
[27] De Meyer T, Hemelsoet K, Van der Schueren L, Pauwels E, De Clerck K, Van
Speybroeck V (2012) Chem. Eur. J. 18:8120–8129
[28] Etienne T, Very T, Perpète EA, Monari A, Assfeld X (2013) J. Phys. Chem. B
117:4973–4980
[29] Uray G, Wagner UG, Fabian WMF (2011) Comp. Theor. Chem. 977:103–110
[30] Vitnik VD, Vitnik ŽJ (2015) Spectrochim. Acta A 138:1–12
[31] Cioslowski J (1989) J. Am. Chem. Soc. 111:8333–8336
[32] Reed AE, Weinhold F (1983) J. Chem. Phys. 78:4066–4073
[33] Singh UC, Kollman PA (1984) J. Comp. Chem. 5:129–145
[34] Chirlian LE, Francl MM (1987) J. Comp. Chem. 8:894–905
[35] Breneman CM, Wiberg KB (1990) J. Comp. Chem. 11:361–373
[36] Dane E, Blacke O, Hemmel H, Müller F (1957) Liebigs Ann. Chem. 607:92–108
[37] Li GQ, Li Y, Dai LX, You SL (2007) Org. Lett. 9:3519–3521
[38] Bellack E, De Witt JB (1954) J. Agric. Food. Chem. 2:1176–1179
[39] Hebeubrock KF (1978) Liebigs Ann. Chem. 1978:320–336
[40] Sumimoto Chemical Co. Ltd. (1971) GE Pat No 2,100,800
[41] Frisch MJ, Trucks GW, Schlegel HB, Scuseria GE, Robb MA, Cheeseman JR,
Scalmani G, Barone V, Mennucci B, Petersson GA, Nakatsuji H, Caricato M, Li X,
Hratchian HP, Izmaylov AF, Bloino J, Zheng G, Sonnenberg JL, Hada M, Ehara M,
Toyota K, Fukuda R, Hasegawa J, Ishida M, Nakajima T, Honda Y, Kitao O, Nakai H,
Vreven T, Montgomery JA, Peralta JE, Ogliaro F, Bearpark M, Heyd JJ, Brothers E,
Kudin KN, Staroverov VN, Keith T, Kobayashi R, Normand J, Raghavachari K, Rendell
A, Burant JC, Iyengar SS, Tomasi J, Cossi M, Rega N, Millam JM, Klene M, Knox JE,
Cross JB, Bakken V, Adamo C, Jaramillo J, Gomperts R, Stratmann RE, Yazyev O,
24

ACCEPTED MANUSCRIPT
Austin AJ, Cammi R, Pomelli C, Ochterski JW, Martin RL, Morokuma K, Zakrzewski
VG, Voth GA, Salvador P, Dannenberg JJ, Dapprich S, Daniels AD, Farkas O, Foresman
JB, Ortiz JV, Cioslowski J, Fox DJ (2010) Gaussian 09, revision C.01.Gaussian, Inc.,
Wallingford.
[42] Cossi M, Rega N, Scalmani G, Barone V (2003) J. Comput. Chem. 24:669–681
[43] Dennington R, Keith T, Millam J (2009) GaussView, Version 5.0.9. Shawnee
Mission, KS: Semichem Inc
25

ACCEPTED MANUSCRIPT
Scheme captions
Scheme 1. Molecular structure of the 1-aryl-3-methylpyrrolidine-2,5-diones 1–10 with
the atoms and torsion angle labeled.
Scheme 2. Synthetic pathway for the 1-aryl-3-methylpyrrolidine-2,5-diones 1–10.
Figure captions
Figure 1. Mesomeric structures of 1-aryl-3-methylsuccinimides.
+
Figure 2. Relationship between νmax and σp for the 1-aryl-3-methylsuccinimides 1–6 and
8–10.
Figure 3. The most stable conformers of compound 4 calculated by B3LYP level of
theory.
Figure 4. Experimental UV spectrum of 1-phenyl-3-methylpyrrolidine-2,5-dione 4,
measured in ethanol, and the calculated TD-B3LYP electronic absorption spectrum in
vacuum.
Figure 5. Correlation between experimental and calculated UV maxima provided by the
B3LYP method in a vacuum.
Figure 6. Electrostatic potentials map of 1-phenyl-3-methylpyrrolidine-2,5-dione 4 in
vacuum from B3LYP/6-311G(d,p) calculation plotted on isosurfaces of 0.002 a.u.
electron density); color coding: red (very negative), orange (negative), yellow (slightly
negative), green (neutral), turquoise (slightly positive), light blue (positive), dark blue
(very positive).
Figure 7. νmax vs. cos2θ data for investigated 1-aryl-3-methylpyrrolidine-2,5-diones.
Table captions
Table 1. Long-wavelength UV absorption maxima of 1-aryl-3-methylpyrrolidine-2,5-
diones 1–10.
Table 2. Regression fits to the solvatochromic parameters of Eq. 1.
26

ACCEPTED MANUSCRIPT
Table 3. Percentage contribution of the two solvatochromic parameters of Eq. 1.
+
Table 4. The results of the correlation between νmax 10–3/cm–1 and σp for the 1-aryl-3-
methylsuccinimides.
Table 5. 13C NMR substituent chemical shifts (δ/ppm).
Table 6. Results of the SCS data correlations according to Eq. (2).
Table 7. Results of the APT data in DMSO as solvent, derived from B3LYP/6-311G(d,p)
correlated with Eq. (2).
Table 8. APT, NBO, EP vs. SCS data for the 1-aryl-3-methylpyrrolidine-2,5-diones 1–
10.
27

ACCEPTED MANUSCRIPT
Table 1. Long-wavelength UV absorption maxima of 1-aryl-3-methylpyrrolidine-2,5-diones 1–10.
νmax × 10–3/cm–1
Solvent / Comp. No.
1
2
3
4
5
6
7
8
9
10
Methanol
43.71 43.63 45.41 46.99 44.48 43.71 42.05 41.81 39.49 36.36
42.59 42.64 44.52 46.79 44.40 43.40 40.72 40.36 38.67 35.60
42.23 42.27 44.27 46.71 44.27 43.90 40.75 41.25 38.55 36.21
42.83 42.63 44.19 46.75 44.17 43.46 40.77 40.82 39.06 36.15
43.92 43.46 45.52 46.25 44.66 43.22 41.79 42.27 39.90 36.18
43.74 43.38 45.48 46.21 44.48 43.80 42.09 42.48 39.59 36.55
44.09 43.98 45.54 45.91 45.23 44.27 41.81 43.07 40.97 36.23
43.80 43.82 45.54 46.08 45.60 43.96 42.25 43.37 41.41 36.09
42.92 43.48 43.94 45.87 44.54 43.50 41.89 40.95 42.03 36.47
44.01 44.29 45.37 46.06 44.70 43.69 42.00 41.82 39.68 36.67
43.78 43.74 45.50 45.91 44.76 44.35 42.23 41.65 40.13 36.59
44.01 43.92 45.31 46.23 44.80 44.21 42.18 42.11 39.98 36.51
43.18 42.84 45.25 46.23 44.68 43.46 41.82 41.27 40.40 36.42
44.01 43.67 45.37 46.02 44.84 43.86 41.96 41.41 39.65 37.04
43.73 43.57 45.29 45.87 44.44 43.73 41.91 41.24 39.25 35.69
Methanol/H2O (80:20)
Methanol/H2O (60:40)
Methanol/H2O (40:60)
Ethanol
Ethanol/H2O (80:20)
Ethanol/H2O (60:40)
Ethanol/H2O (40:60)
2-Propanol
2-Propanol/H2O (80:20)
2-Propanol/H2O (60:40)
2-Propanol/H2O (40:60)
2-Propanol/Ethanol(80:20)
2-Propanol/Ethanol(60:40)
2-Propanol/Ethanol(40:60)
2-Propanol/Methanol (80:20) 43.44 43.54 45.17 45.66 44.35 43.73 42.02 41.51 39.06 36.35
2-Propanol/Methanol (60:40) 43.59 43.63 45.33 45.96 44.58 43.84 42.05 41.46 39.26 36.64