Synthesis, structural characterization and quantum chemical studies of silicon-containing benzoic acid derivatives

Article abstract of DOI:10.1016/j.molstruc.2016.05.038

The present paper is concerned with the synthesis and molecular structure investigation of two new benzoic acid derivatives having trimethylsilyl tails, 4-((trimethylsilyl)methoxy) and 4-(3-(trimethylsilyl)propoxy)benzoic acids. The structures of the nove

Full text of DOI:10.1016/j.molstruc.2016.05.038

Accepted Manuscript
Synthesis, structural characterization and quantum chemical studies of silicon-
containing benzoic acid derivatives
Mirela-Fernanda Zaltariov, Corneliu Cojocaru, Sergiu Shova, Liviu Sacarescu, Maria
Cazacu
PII:
S0022-2860(16)30484-7
10.1016/j.molstruc.2016.05.038
MOLSTR 22552
DOI:
Reference:
To appear in: Journal of Molecular Structure
Received Date: 1 March 2016
Revised Date: 11 May 2016
Accepted Date: 12 May 2016
Please cite this article as: M.-F. Zaltariov, C. Cojocaru, S. Shova, L. Sacarescu, M. Cazacu, Synthesis,
structural characterization and quantum chemical studies of silicon-containing benzoic acid derivatives,
Journal of Molecular Structure (2016), doi: 10.1016/j.molstruc.2016.05.038.
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
Synthesis, structural characterization and quantum chemical studies of
silicon‒containing benzoic acid derivatives
*
Mirela-Fernanda Zaltariov , Corneliu Cojocaru, Sergiu Shova, Liviu Sacarescu, Maria Cazacu
Department of Inorganic Polymers, “Petru Poni” Institute of Macromolecular Chemistry,
Romanian Academy, Aleea Grigore Ghica Voda 41 A, 700487 Iasi, Romania.
∗
Corresponding author: Mirela-Fernanda Zaltariov
Email: zaltariov.mirela@icmpp.ro
Tel: +40-232-217454
Fax: +40-232-211299
Electronic Supplementary Information (ESI) available
1

ACCEPTED MANUSCRIPT
Abstract
The present paper is concerned with the synthesis and molecular structure investigation of
two new benzoic acid derivatives having trimethylsilyl tails, 4‒((trimethylsilyl)methoxy) and
4
‒(3‒(trimethylsilyl)propoxy)benzoic acids. The structures of the novel compounds have been
confirmed by X‒ray crystallography, Fourier‒transform infrared spectroscopy (FTIR) and
1
13
nuclear magnetic resonance ( H and C NMR). The theoretical studies of molecules were
conducted by using the quantum chemical methods, such as Density Functional Theory (DFT
B3LYP/6‒31+G**), Hartree‒Fock (HF/6‒31+G**) and semiempirical computations (PM3, PM6
and PM7). The optimized molecular geometries have been found to be in good agreement with
experimental structures resulted from the X‒ray diffraction. The maximum electronic absorption
bands observed at 272‒287 nm (UV‒vis spectra) have been assigned to πꢀπ* transitions, which
were in reasonable agreement with the time dependent density functional theory (TD‒DFT)
calculations. The computed vibrational frequencies by DFT method were assigned and compared
with the experimental FTIR spectra. The mapped electrostatic potentials revealed the reactive
sites, which corroborated the observation of the dimer supramolecular structures formed in the
crystals by hydrogen‒bonding. The energies of frontier molecular orbitals (HOMO and LUMO),
energy gap, dipole moment and molecular descriptors for the new compounds were calculated
and discussed.
Keywords: Benzoic acid derivatives; Organosilicon compounds; Computational chemistry;
Vibrational spectra; Electronic absorption spectra; HOMO‒LUMO.
2

ACCEPTED MANUSCRIPT
1. Introduction
Benzoic acid derivatives are versatile compounds used as chemicals, pharmaceuticals,
agrochemicals and consumer products [1]. Benzoic acid inhibits bacterial development and its
substituted compounds are very important for the development of new materials in food and
pharmaceutical industries [2]. In the last years, considerable efforts have been devoted to
theoretically and experimentally study of the benzoic acid derivatives [1‒14]. Thus, a number of
these compounds were investigated: p‒hydroxybenzoic, m‒anisic, vanillic, and syringic acids [1];
4
‒butyl benzoic acid [2]; methyl‒ and methoxybenzoic acids [3]; m‒trifluoromethylbenzoic acid
[
4]; 4‒(2,5‒di‒2‒thienyl‒1H‒pyrrol‒1‒yl)benzoic acid [5]; p‒(p‒hydroxyphenoxy)benzoic acid
6]; toluic acid [7]; 2‒(4‒hydroxyphenylazo)benzoic acid [8]; 2‒amino‒5‒bromo‒benzoic acid
[
methyl ester [9]; 4‒(trimethylammonium)benzoic acid chloride [10]; 2,3,4‒tri‒fluoro‒benzoic
acid dimer [11]; 2‒[(2‒hydroxyphenyl)carbonyloxy]benzoic acid [12]; Cu(II) complex with
4
‒{(Z)‒[(2‒hydroxybenzoyl)hydrazono]methyl}benzoic acid [13] and 4‒[[(substituted
phenyl)imino]methyl]benzoic acids [14]. All of the above mentioned molecules were mainly
modeled by density functional theory (DFT), and the simulation outcomes were compared with
the available experimental results (e.g. X‒ray crystallography; FTIR, Raman, UV‒vis and NMR
spectra).
The current work aims to report on the synthesis pathway, molecular structure characterization
and theoretical quantum chemical investigations of two novel benzoic acid derivatives having
trimethylsilyl
‒(3‒(trimethylsilyl)propoxy)benzoic acids as potential ligands for the metal ions. The presence
of TMS groups creates premises for obtaining metal complexes with increased solubility in
non‒polar media, such as supercritical carbon dioxide (scCO ), a non‒toxic, non‒flammable,
(TMS)
tails,
i.e.
4‒((trimethylsilyl)methoxy)
and
4
2
eco‒friendly and relatively cheap solvent, where they can act as homogeneous catalyst for
3

ACCEPTED MANUSCRIPT
different reactions. The number of metal complexes that are soluble in this environment is very
limited [15]. In addition, due to the higher hydrophobicity of the trimethylsilyl group, these
compounds have a high potential for application as bioactive compounds in enzymatic reactions
[16].
Previously published articles described the experimental investigations regarding the synthesis
and characterization of benzoic acid derivatives containing organosilicon groups [17‒19].
However, to the best of our knowledge no report has been focused, until now, on both theoretical
and experimental investigations of the molecular structures of silicon‒containing benzoic
derivatives [20]. Molecular modeling is an extremely useful tool to complement experimental
techniques providing a potential structural context for understanding the experimental results
[21]. Note that, some organosilicon compounds have been tested as pharmaceuticals [20].
2. Experimental
2.1. Materials
4
‒Hydroxybenzoic acid (Aldrich), (chloromethyl)trimethylsilane 98% (Aldrich),
3‒chlororpropyl)trimethylsilane 97% (Aldrich), anhydrous K CO (Aldrich) Na SO (Aldrich),
dimethylformamide (Aldrich), chloroform (Aldrich) were used as received.
.2. Methods
Fourier transform infrared (FT‒IR) measurements were carried out using a Bruker Vertex
0 FT‒IR spectrometer. Spectra were recorded in the transmission mode in the range 400‒4000
(
2
3
,
2
4
2
7
‒
1
‒1
cm at room temperature with a resolution of 2 cm and accumulation of 32 scans.
The NMR spectra were recorded on a Bruker Avance DRX 400 MHz Spectrometer
equipped with a 5 mm QNP direct detection probe and Z‒gradients. Spectra were recorded in
CDCl
3
, at room temperature. The chemical shifts are reported as δ values (ppm). The assignments
4

ACCEPTED MANUSCRIPT
of all the signals in the 1D NMR spectra were done using 2D NMR experiments like
H,H‒COSY, H,C‒HMQC and H,C‒HMBC.
3
UV‒vis absorption spectra measurements were carried out in CHCl and DMSO solutions
on a Specord 200 spectrophotometer.
X‒Ray crystallographic measurements for 1 and 2 were carried out with an
Oxford‒Diffraction XCALIBUR E CCD diffractometer equipped with graphite‒monochromated
Mo‒Kα radiation. Single crystals were positioned at 40 mm from the detector and 848, and 208
frames were measured each for 25, and 40 s over 1° scan width for 1, and 2, respectively. The
unit cell determination and data integration were carried out using the CrysAlis package of
Oxford Diffraction [22]. The structures were solved by direct methods using Olex2 [23] and
refined by full‒matrix least‒squares on F² with SHELXL‒97 [24]. Atomic displacements for
non‒hydrogen atoms were refined using an anisotropic model. All H atoms were introduced in
idealized positions (dCH = 0.96 Å) using the riding model with their isotropic displacement
parameters fixed at 120% of their riding atom. The molecular plots were obtained using the
Olex2 program. The crystallographic data and refinement details are quoted in Table 1, while
bond lengths, interatomic angles and dihedral angles are summarized in ESI (Tables B1-B6).
Table 1. Crystallographic data and structure refinement parameters for the compounds 1 and 2.
Compound
1
2
Empirical formula
C H O Si
C H O Si
1
1
16
3
13 20
3
Formula weight
Temperature
Crystal system
Space group
224.331 g/mol
200.00(14) °K
triclinic
252.385 g/mol
200.05(10) °K
monoclinic
P-1
1
P2 /c
5

ACCEPTED MANUSCRIPT
a = 6.6120(3) Å
b = 7.6894 (4) Å
c = 12.8140(10) Å
a = 16.508(13) Å
b = 10.026(2) Å
c = 8.8482(15) Å
90.00°
Unit cell dimensions
1
01.928(6)°
8.644(6)°
9
95.89(3)°
1
00.425(4) °
90.00°
3
3
Volume
614.86(6) Å
2
1456.8(12) Å
Z
4
3
3
Calculated density
Absorption coefficient, ꢀ
F(000)
1.212 g/cm
1.151 g/cm
-
1
-1
0.177 mm
240.0
0.156 mm
544.0
Crystal size
0.15 × 0.1 × 0.1 mm
MoKα (λ = 0.71073 Å)
5.54 to 46.5°
0.1 × 0.05 × 0.05 mm
MoKα (λ = 0.71073 Å)
4.76 to 46.5°
Radiation (wavelength)
2
Θ range for data collection
-
7 ≤ h ≤ 7, -8 ≤ k ≤ 8,
-12 ≤ h ≤ 18, -11 ≤ k ≤ 11,
-9 ≤ l ≤ 9
Index ranges
-14 ≤ l ≤ 14
Reflections collected
5839
4770
1
736 [R = 0.0334,
2080 [R = 0.0568,
int
int
Independent reflections
Rsigma = 0.0303]
1736/0/140
1.064
Rsigma = 0.1031]
2080/0/158
1.051
Data/restraints/parameters
Goodness-of-fit (GOF) on F
Final R indexes [I>=2σ (I)]
2
R
1
2 1 2
= 0.0897, wR = 0.2172 R = 0.0933, wR = 0.1862
6

ACCEPTED MANUSCRIPT
Final R indexes [all data]
R
1
= 0.0931, wR
2
= 0.2187 R
1
= 0.1745, wR
2
= 0.2275
-3
-3
Largest diff. peak and hole
0.88 and -0.38 e Å
0.34 and -0.25 e Å
R = Σ||F | – |F ||/Σ|F |;
1
o
c
o
2
o
2 2
2 2 1/2
wR = {Σ[w(F – F ) ]/Σ[w(F ) ]} ;
2
c
o
2
o
2 2
1/2
GOF = {Σ[w(F
c
– F ) ]/(n – p)} , where n is the number of reflections and p is the total
number of parameters refined.
CCDC‒1438222 (1), CCDC‒1438223 (2) contain the supplementary crystallographic
data for this contribution. These data can be obtained free of charge via
www.ccdc.cam.ac.uk/conts/retrieving.html (or from the Cambridge Crystallographic Data Centre,
1
2 Union Road, Cambridge CB2 1EZ, UK; fax: (+44) 1223-336-033; or deposit@ccdc.ca.ac.uk).
2
2
.3. Synthesis
.3.1. Synthesis of 1
In a 100 mL round bottom flask equipped with magnetic stirring and reflux condenser
protected
with
CaCl
2
tube,
4‒hydroxybenzoic
(0.98
g,
7
mmol)
and
(trimethylsilyl)chlorometylsilane (0. 86 g, 1 mL, 7 mmol) were disolved together in DMF (15
mL). Anhydrous K
2
CO
3
(1.45 g, 10.5 mmol) was added to the resulting solution and the mixture
o
was stirred under argon athmosphere at 110 C for 8 hours. The resulting precipitate was filtered
off and the filtrate was poured into water (60 mL) and extracted with chloroform (3 x 15 mL).
The combined organic phases were dried (Na SO ) and evaporated under reduced pressure. The
2
4
resulted beige solid product was dried in vacuo. X‒ray diffraction‒quality single crystals were
obtained by recrystallization from methanol, washed with methanol and diethylether and dried in
7

ACCEPTED MANUSCRIPT
air at room temperature. Yield: 1.2 g, 75%. Anal. Calcd for C H O Si (M 224.33): C, 58.89, H,
1
1
16
3
r
7
.19. Found: C, 58.67, H, 7.23.
‒
1
IRνmax (KBr), cm : 3551w, 3474w, 3416w, 2957w, 2901w, 2882w, 2822w, 1688s,
1
8
601vs, 1578s, 1506s, 1421m, 1292s, 1250s, 1215m, 1157s, 1130vw, 1109w, 1007m, 858vs,
47vs, 775m, 758w, 706w, 636m, 619m, 604w, 546w, 513w, 505w, 480vw, 451vw, 378vw.
1
H NMR (CDCl
3
, 400.13 MHz, δ, ppm): 9.87 (s, 1H, ‒COOH), 7.82 (d, J=8.7 Hz, 2H,
‒), 0.16 (s, 9H, ‒Si‒CH ).
, 100.16 MHz, δ, ppm): 190.88 (‒COOH), 166.72 (Ar‒C), 131.91
Ar‒C), 129.63 (Ar‒C), 114.53 (Ar‒C), 61.78 (‒CH ‒), ‒3.20 (‒Si‒CH ).
Ar‒H), 7.05 (d, J=8.7 Hz, 2H, Ar‒H), 3.66 (s, 2H,‒CH
2
3
1
3
C NMR (CDCl
3
(
2
3
‒
1
‒1
‒1
‒1
UV‒vis (CHCl ), λ (ε, L M cm ): 272 (21816); UV‒vis (DMSO), λ (ε, L M cm ):
3
max
max
2
75 (20823).
2
.3.2. Synthesis of 2
In a 100 mL round bottom flask equipped with magnetic stirrer and reflux condenser
protected with CaCl tube, 4‒hydroxybenzoic (0.83 g, mmol) and
trimethylsilyl)chloropropylsilane (0. 88 g, 1 mL, 6 mmol) were disolved together in DMF (15
mL). Anhydrous K CO (1.45 g, 10.5 mmol) was added to the resulting solution and the mixture
2
6
(
2
3
o
was stirred under argon athmosphere at 110 C for 8 hours. The resulting precipitate was filtered
off and the filtrate was poured into water (60 mL) and extracted with chloroform (3 x 15 mL).
The combined organic phases were dried (Na SO ) and evaporated under reduced pressure. The
2
4
solid product was dried in vacuo. X‒ray diffraction‒quality single crystals were obtained by
recrystallization from chloroform/methanol (1:2, v:v), washed with methanol and diethylether
8

ACCEPTED MANUSCRIPT
and dried in air at room temperature. Yield: 1.3 g, 86 %. Anal. Calcd for C H O Si (M 252.38):
1
1
16
3
r
C, 61.87, H, 7.99. Found: C, 61.67, H, 8.13.
‒
1
IRνmax (KBr), cm : 3366m, 3289m, 3155w, 3119w, 3074m, 3040w, 2951vs, 2895s,
878s, 2826s, 2904s, 2735s, 2573w, 2515vw, 2428vw, 2012w, 1936vw, 1906w, 1693vs, 1601vs,
578vs, 1510vs, 1470s, 1427s, 1393s, 1313s, 1250vs, 1215vs, 1184s, 1159vs, 1109s, 1063s,
047s, 1009s, 895s, 854vs, 835vs, 787s, 758s, 694s, 650s, 617s, 606s, 515s, 453w, 436w, 422w,
2
1
1
3
82w.
1
3
H NMR (CDCl , 400.13 MHz, δ, ppm): 9.86 (s, 1H, ‒COOH), 7.82 (d, J=8.7 Hz, 2H,
Ar‒H), 6.97 (d, J=8.7 Hz, 2H, Ar‒H), 3.98 (t, J=6.8 Hz, ‒CH ‒CH ‒CH ‒), 1.84‒1.77 (m, 2H,
2
2
2
‒
CH ‒CH ‒CH ‒), 0.63‒0.58 (m, 2H, ‒CH ‒CH ‒CH ‒), 0.02 (s, 9H, ‒Si‒CH ).
2 2 2 2 2 2 3
1
3
C NMR (CDCl , 100.16 MHz, δ, ppm): 190.85 (‒COOH), 164.24 (Ar‒C), 132.00
3
(
Ar‒C), 129.68 (Ar‒C), 114.72 (Ar‒C), 70.99 (‒CH ‒CH ‒CH ‒), 23.66 (‒CH ‒CH ‒CH ‒),
2 2 2 2 2 2
1
2.50 (‒CH ‒CH ‒CH ‒), ‒1.79 (‒Si‒CH ).
2
2
2
3
‒
1
‒1
‒1
‒1
UV‒vis (CHCl
87 (26176).
3
), λmax (ε, L M cm ): 285 (28235); UV‒vis (DMSO), λmax (ε, L M cm ):
2
3. Results and discussion
A reaction of Williamson type was used to prepare two silicon‒containing dicarboxylic
acids by treating 4‒hydroxybenzoic acid with (chloromethyl)trimethylsilane and
3‒chlororpropyl)trimethylsilane, respectively in 1:1 molar ratio, in DMF, in the presence of
CO (Figure 1). This is a S 2 reaction between the deprotonated alcohol and the alkyl halide
(
K
2
3
N
moiety attached to the silicon atom. The alcohol deprotonation occurs in situ in the presence of
K CO . The reaction products were isolated in pure, crystalline state, in reasonable yields (75 and
2
3
8
6 wt%, respectively).
9

ACCEPTED MANUSCRIPT
CH3
O
C
OH Cl
Si CH3
HO
n
CH3
o
1
10 C DMF
K CO
8h
2
3
CH3
O
C
O
Si CH3
CH3
n
HO
n=1, (1); n=3, (2)
Figure 1. Scheme of the synthesis route of the titled compounds,
n=1): 4‒((trimethylsilyl)methoxy)benzoic acid;
(
(n=3): 4‒(3‒(trimethylsilyl)propoxy)benzoic acid.
1
13
The H NMR and C NMR spectra of the silicon‒containing carboxylic acids are given
1
in ESI (Figures C1‒C4). The H NMR spectrum of 1 confirms the proposed structure by the
presence of the four aromatic protons at 7.82 and 7.05 ppm, a singlet for the carboxylic acid
proton at 9.87 ppm, a singlet assigned to –CH ‒O‒ at 3.66 ppm, while the peak corresponding to
2
1
3
–
Si‒CH protons appears at 0.16 ppm. The C NMR spectrum is also in concordance with the
3
proposed structure.
1
In the H NMR spectrum of 2, all protons resonated at appropriate positions: a singlet for
carboxylic acid proton at 9.86 ppm, doublets for aromatic protons at 7.82 ppm and 6.97 ppm,
while the aliphatic protons and those of the trimethylsilyl groups appear in the region 3.98-0.58
ppm and at 0.02 ppm respectively. Their intensity ratio corresponds with the presumed structure.
1
3
The C NMR spectra also confirm their structure (ESI, Figures C2 and C4).
1
0

ACCEPTED MANUSCRIPT
3
.1. X‒ray crystallography
The molecular structures determined by X‒ray crystallography are given as ORTEP plots
with the atom‒numbering scheme in Figure 2.
Figure 2. Molecular structures of the investigated compounds determined by X‒ray structure
analysis (Olex2 view): (a) 1 (C H O Si), 4‒((trimethylsilyl)methoxy)benzoic acid;
1
1
16
3
(
b) 2 (C H O Si ), 4‒(3‒(trimethylsilyl)propoxy)benzoic acid. Thermal ellipsoids are drawn at
13 20 3
50% probability level.
The single‒crystal X‒ray study has revealed that both compounds have molecular crystal
structures resulted from the packing of dimeric supramolecular architectures, by formation of the
stable cyclic hydrogen bonded system, as depicted in Figure 3. No co‒crystallized solvent has
been found in the crystals of compounds 1 and 2.
1
1

ACCEPTED MANUSCRIPT
Figure 3. Partial packing diagrams for the compounds 1 (a) and 2 (b) showing the dimmers
structures formed via hydrogen-bond bridges (O…H-O). Thermal ellipsoids are drawn at 50%
probability level. H-bonds parameters of 1.: O4−HꢁꢁꢁO3 [O4−H 0.82 Å, HꢁꢁꢁO3 1.83 Å, O4ꢁꢁꢁO3
(3 – x, 3 - y, 1 - z) 2.625(6) Å, O4−HꢁꢁꢁO3 161.5°]. H-bonds parameters of 2: O2−HꢁꢁꢁO6 [O2−H
0
.82 Å, HꢁꢁꢁO6 1.80 Å, O2ꢁꢁꢁO6 (2 – x, 1 - y, 2 - z) 2.613(6) Å, O2−HꢁꢁꢁO6 171.0°].
Apart from the H‒bonding, in the crystal of 1, the system of intermolecular interactions is
completed by the significant stacking between centrosymmetrically related C6‒C15 aromatic
π‒π
1
2

ACCEPTED MANUSCRIPT
rings, which is evidenced by the centroid‒to‒centroid distance of 3.723 Å and shift distance of
1
.364 Å. These interactions are responsible for the supramolecular aggregation of the dimeric
units into the ribbon‒like architecture running parallel to [110] direction (ESI, Figure A1).
4. Computational details
The silicon‒containing benzoic acid derivatives 1 and 2 were modeled by computational
chemistry approach using the density functional theory (DFT), ab‒initio (Hartree‒Fock, HF) and
semiempirical (PM3, PM6 and PM7) methods. To this end, the quantum chemical calculations
(
DFT, HF and semiempirical) were performed in order to study the molecular structure
characteristics and electronic properties of the titled compounds. The software package Gaussian
9 [25] was employed for DFT and HF calculations of the molecules, using the split‒valence
0
basis set 6‒31+G** with added polarization and diffuse functions. The density functional method
was implemented at the level of B3LYP functional (i.e. B3LYP/6-31+G**). The computations by
PM3, PM6 and PM7 methods were performed by means of MOPAC semiempirical quantum
chemistry program [26‒31]. The molecular modeling results were analyzed using
graphical‒interface computational chemistry software, such as GaussView 5 [32], Gabedit [33]
and HyperChem [34].
Geometry optimization is an important part of computational chemistry, and it deals with
searching by energy minimization of the optimized molecular structures [35]. In this work, the
geometry optimization of molecules 1 and 2 was performed by minimizing the energies with
respect to all geometrical parameters without imposing any symmetry constrains. The input
structures for optimization were taken from the CIF‒files obtained from X‒ray single‒crystal
measurements. Further, the geometry optimization was performed on single molecules, in vacuo.
1
3

ACCEPTED MANUSCRIPT
Frequency calculations at the optimized geometries were done to confirm the optimum
conformations (i.e. no imaginary frequencies) and to analyze the vibration modes. The electronic
absorption spectra of the investigated compounds were calculated by the time dependent density
functional theory (TD‒DFT).
4.1. Optimized molecular structures
Since the properties of a molecule depend on its structure, an important goal in
computational chemistry is to find the most probable geometry of the molecule by optimization.
In this work, the molecular geometries of the investigated acids 1 and 2 were optimized at
different theoretical levels using B3LYP/6‒31+G**, HF/6‒31+G** as well as semiempirical
(PM3, PM6 and PM7) methods. Figure 4 illustrates as an example, the optimized geometries of
compounds 1 and 2 computed by B3LYP/6‒31+G** method. The ball and stick rendering
models with full atomic numbering have been used for the spatial representations. The
investigated compounds possess C symmetry (point group).
1
1
4

ACCEPTED MANUSCRIPT
Figure 4. Optimized molecular geometries of the investigated compounds 1 (a) and 2 (b),
computed at ground state (S
0
) by DFT method (B3LYP/6‒31+G**); balls and sticks rendering
models with full atomic numbering.
The optimized structure parameters such as bond lengths, interatomic angles and
dihedrals of the studied compounds were calculated by computational chemistry methods and
compared with the experimental values observed from X‒ray crystallography. The full analysis
of the optimized structure parameters and the corresponding experimental counterparts is given in
1
5

ACCEPTED MANUSCRIPT
the electronic supplementary information ESI (Tables B1
‒
B6). The representative geometric
parameters are summarized in Table 2.
o
Table 2. Summary of selected structural parameters: bond lengths (Å), angles ( deg) and
o
dihedrals ( deg), for the investigated molecules 1 (C H O Si) and 2 (C H O Si ).
11
16
3
13 20
3
a
b
c
Geometry: 1
Exp.
PM3
1.936
1.888
1.376
1.412
PM6
1.924
1.865
1.362
1.465
PM7
HF
DFT
1
1
Si-17C
Si-28C
1.884(6)
1.849(8)
1.358(8)
1.436(7)
106.3(3)
118.8(4)
109.0(4)
166.0(5)
170.8(4)
Exp.
1.898
1.866
1.350
1.431
1.911
1.891
1.340
1.419
1.918
1.892
1.360
1.440
2
O-7C
2O-17C
17C-1Si-28C
105.649 103.611 106.251 106.789 106.730
116.844 118.130 117.644 121.103 119.777
109.061 112.740 105.145 108.136 108.179
178.079 179.632 179.761 -179.971 -179.954
162.090 179.488 179.255 -179.927 -179.934
7C-2O-17C
1
Si-17C-2O
2
8C-1Si-17C-2O
7
C-2O-17C-1Si
Geometry: 2
PM3
1.890
1.908
1.371
1.425
PM6
1.870
1.892
1.361
1.469
PM7
1.870
1.886
1.348
1.435
HF
DFT
1.894
1.905
1.358
1.434
1
1
Si-24C
Si-32C
1.821(10)
1.907(7)
1.362(6)
1.440(6)
106.8(4)
117.1(4)
114.5(6)
171.7(6)
1.892
1.900
1.338
1.411
7
O-11C
O-12C
7
2
8C-1Si-32C
1C-7O-12C
2C-35C-32C
8C-1Si-32C-35C
108.426 107.118 107.665 108.213 108.324
117.410 118.917 118.063 120.932 119.492
112.885 110.199 109.833 113.913 114.186
-179.584 179.051 -179.832 179.342 -179.319
1
1
2
1
6

ACCEPTED MANUSCRIPT
1
1C-7O-12C-35C
-179.2(5)
178.110 179.968 177.023 179.312 179.062
Histograms have been ploted on the basis of complete analysis of structure parameters
Tables B1‒B6) and following the approach proposed in Refs. [36,37] the histograms have been
(
plotted as shown in Figure 5. Such histograms give the general image concerning the
distributions of the bond lengths and interatomic angles in the molecule based on experimental
observations and theoretical results. As shown in Figure 5, some models overestimate whereas
others underestimate in reasonable limits the observed structural parameters of the studied
molecules.
Figure 5. Histograms of interatomic angles distributions (a) and bond lengths distributions (b) for
the investigated compounds 1 and 2.
1
7

ACCEPTED MANUSCRIPT
All theoretical models overestimated the shortest bond lengths values for O‒H and C‒H.
As presented in histograms (Figure 5), the highest frequencies for these bonds were within 0.82‒
1
.19 Å range. For Si‒C, C‒O, C=O and C‒C both the calculated and measured bond lengths were
in good agreement (Figure 5 and Tables B1, B4 / ESI).
All values of interatomic angles for the studied benzoic acid derivatives are reported in
Tables B2 and B5 (ESI). Figure 5 also shows the histograms of interatomic angles distributions
plotted on the basis of experimental data and theoretical models. As shown in Table 2 the bond
angles are reasonable predicted by the quantum chemistry models.
Some structural disparities have been observed in terms of torsion angles (dihedrals). The
information about dihedrals (torsion angles) for 1 and 2 is given extensively in Tables B3 and B6
(ESI), respectively. In Table 2 only the selected torsion angles are summarized.
Such differences could be explained by taking into account the conformational constraints
that occur when the molecules are assembled in crystalline structures. However, in order to
identify the overall conformational similitude between observed molecular structures and the
modeled ones, the overlap plots and the calculation of root‒mean‒square‒deviation (RMSD) are
commonly used as goodness‒of‒fit estimators [38‒40]. The RMSD [40] of the atomic positions is
a similarity measure for the quantitative comparison of one structure with another, and it is given
by:
N
2
1
i
( )
j
k
(
)
RMSD
(
r
)
=
r − r
(1)
∑
(
k
)
N
k =1
Where N denotes the number of atoms in the molecule; k is an index over these atoms;
rk( ) – coordinates of atom k in conformations i and j. The minimal value of RMSD is
rk( )
i
,
j
1
8

ACCEPTED MANUSCRIPT
desirable and it is obtained by the optimal superposition of the two structures. In this work, the
RMSD values have been calculated to compare the observed molecular structures (from X‒ray
crystallography) with their optimized geometries given by quantum chemistry models. The
calculated values of RSMD are summarized in Table 3.
Table 3. Root mean square deviation (RMSD) values to compare the goodness‒of‒fit between
observed and predicted structures of the investigated molecules 1 and 2.
PM3
PM6
PM7
HF
DFT
1
2
(C H O Si)
11 16 3
RMSD, Å
0.2553
0.4304
0.4173
0.4151
0.4212
(C H O Si )
13
20
3
RMSD, Å
0.2029
1.4959
0.3499
0.2328
0.2302
In addition, the overlap plots are given in Figure 6 and ESI (Figures A2‒A3), showing
the superposition of the X‒ray structures of the compounds and their optimized counterparts.
Generally, the predicted geometries are quite similar with the experimental molecular structures
(Figure 6 and Figures A2‒A3/ESI).
1
9

ACCEPTED MANUSCRIPT
2
0

ACCEPTED MANUSCRIPT
Figure 6. Superposition of the X‒ray structures (blue) of the investigated compounds and their
optimized counterparts (red) computed by different methods: (a) 1 (HF/6‒31+G**); (b) 1 (PM3);
(c) 2 (B3LYP/6‒31+G**); (d) 2 (PM6).
The semiempirical method PM3 has yield the lowest RSMD values of 0.2553 Å and
.2029 Å (Table 3), for the compounds 1 and 2, respectively, showing the best similitude with
0
the observed structures. This may be attributed to the fact that PM3 method has better predicted
the dihedral angles. Likewise, the methods HF, DFT and PM7 were appropriate for the
optimization of molecular structures of the benzoic‒acid derivatives, providing the low values of
RMSD, i.e. RMSD<0.43 Å (Table 3). The semiempirical method PM6 indicated the greatest
values of RMSD, suggesting some conformational discrepancy between modeled structures and
observed ones. For example, in case of compound 2, the RMSD value given by PM6 method is
equal to 1.4959 Å, which is significantly greater than RMSD values provided by the other
methods. This structural disparity in terms of atomic positions can be also seen in Figure 6 (d).
In summary, the optimized molecular geometries computed by DFT, HF, PM3 and PM7 methods
were in good agreement with the observed molecular structures obtained from X‒ray
crystallography. In turn, the optimized geometries calculated by PM6 method fairly matched with
the observed structures.
4.2. Molecular orbital analysis and electronic properties
The molecular orbital (MO) analysis is extensively employed to describe the electronic
structure and chemical behavior of the investigated molecule. The theoretical results concerning
the modeled molecules 1 and 2 are summarized in Tables 4 and 5, respectively. This information
2
1

ACCEPTED MANUSCRIPT
refers to the optimized molecular electronic structures of the investigating compounds. Thus,
Tables 4 and 5 provide theoretical outcomes regarding the number of double occupied molecular
orbitals treated by quantum chemistry methods; total energy; dipole moment; molecular size;
energies of frontier molecular orbitals; band gap energy and molecular descriptors.
Table 4. Summary of theoretical results obtained for optimized geometries of the investigated
molecule 1 (C H O Si) using different QM methods.
1
1
16
3
DFT
60
HF
PM3
41
PM6
41
PM7
41
No. of occ. MO
Et, (kcal / mol)
60
-592411
4.324
11.967
458.61
735.71
-0.623
-1.396
-6.528
-7.332
5.132
6.709
6.528
1.396
3.962
-589522
4.131
-57765
3.968
11.812
464.16
734.35
0.163
-0.324
-9.287
-10.534
8.963
10.697
9.287
0.324
4.806
-57958
6.093
11.598
462.98
734.84
0.124
-0.347
-9.117
-10.172
8.770
10.296
9.117
0.347
4.732
-58936
4.674
11.865
456.92
730.44
-0.039
-0.412
-9.310
-10.285
8.898
10.246
9.310
0.412
4.861
D
µ , (Debye)
dmax, (Å)
11.909
454.29
730.11
1.697
2
S , (Å )
m
3
V , (Å )
m
E LUMO+1 (eV)
E LUMO (eV)
1.212
E
HOMO (eV)
-8.851
-9.587
10.063
11.284
8.851
E
HOMO-1 (eV)
∆E, (eV)
∆
E
1
, (eV)
IP (eV)
EA (eV)
-1.212
3.820
χ
(eV)
µ (eV)
-3.962
-3.820
-4.806
-4.732
-4.861
2
2

ACCEPTED MANUSCRIPT
η
(eV)
(eV)
2.566
3.059
5.032
1.450
4.482
2.576
4.385
2.553
4.449
2.656
ω
Table 5. Summary of theoretical results obtained for optimized geometries of the investigated
molecule 2 (C H O Si ) using different QM methods.
13
20
3
DFT
68
HF
68
PM3
47
PM6
47
PM7
47
No. of occ. MO
Et, (kcal / mol)
µD, (Debye)
-641757
4.098
-638518
3.937
-64663
3.645
-64873
5.674
-65853
4.329
d
max, (Å)
13.946
522.38
837.37
-0.647
-1.418
-6.555
-7.358
5.137
13.892
515.47
831.29
1.643
13.869
520.50
833.35
0.088
14.410
520.46
838.23
-0.009
-0.472
-9.307
-9.984
8.835
13.814
511.10
832.64
-0.088
-0.459
-9.384
-10.332
8.925
2
S , (Å )
m
3
V , (Å )
m
E LUMO+1 (eV)
E LUMO (eV)
E HOMO (eV)
1.311
-0.388
-9.406
-10.156
9.018
-8.878
-9.616
10.189
E
HOMO-1 (eV)
∆
E, (eV)
∆
E1 , (eV)
6.711
11.259
10.244
9.975
10.244
IP (eV)
EA (eV)
6.555
1.418
3.986
8.878
-1.311
3.784
9.406
0.388
4.897
9.307
0.472
4.890
9.384
0.459
4.922
χ
(eV)
µ (eV)
-3.986
-3.784
-4.897
-4.890
-4.922
2
3

ACCEPTED MANUSCRIPT
η
(eV)
(eV)
2.568
3.094
5.095
1.405
4.509
2.659
4.418
2.706
4.463
2.714
ω
According to the theoretical results listed in Tables 4 and 5, the total energies of
molecules computed by HF and DFT methods are lower with about one order of magnitude than
the total energies calculated by semiempirical methods (PM3, PM6 and PM7). This can be
explained by the fact that semiempirical methods use a simpler Hamiltonian along with empirical
data to approximate the Schrödinger equation, and thus treating only the valence electrons. In
turn, the core electrons are considered frozen. In contrast, HF and DFT methods treat more
precisely all type of electrons, i.e. core and valence electrons. Therefore, the number of double
occupied molecular orbitals taken into account by HF and DFT methods is always greater than
that considered by semiempirical models (Tables 4, 5).
The heats of formation computed by semiempirical methods for the titled compounds 1
and 2 are summarized in Table B7 (ESI). According to PM3 method, the heats of formation for 1
and 2 were equal to ‒144.37 kcal/mol and ‒156.14 kcal/mol, respectively. Other semiempirical
techniques (i.e. PM6 and PM7) converged to similar results. The theoretical calculations revealed
significant values of the dipole moment for the investigated compounds (Tables 4 and 5). For the
molecule 1, the dipole moment ranged from 4.131 to 6.093 Debyes, whereas for the molecule 2,
the calculated dipole moment varied in the extent 3.645‒5.674 Debyes, depending on the
quantum chemistry method applied. Thus, the dipole moment values for compound 2 are lower
than for the compound 1, due to additional content of hydrophobic (‒CH ‒) groups. The
2
3
calculated polarizability volumes of titled compounds 1 and 2 were found to be equal to 23.28 Å
3
and 26.95 Å , respectively. The molecular size (dmax) is given in Tables 4 and 5 as the measure of
2
4

ACCEPTED MANUSCRIPT
the maximum atomic distance in the molecules. For instance, according to DFT method, the
molecular dimensions for the compounds 1 and 2 are equal to 11.967 Å and 13.946 Å,
respectively. The molecular surface area (S ) and molecular volume (V ) have been calculated
m
m
by numerical integration grid techniques implemented in HyperChem program. The calculated
values of these parameters are also reported in Tables 4 and 5.
The energy spectra of molecular orbitals (MOs) in ground state (S ) were computed for
0
both molecules by means of quantum chemistry methods. Figure 7 shows as an example the
energy spectra of MOs calculated for the compounds 1 and 2 at the level of PM3 model. As
shown in Figure 7, each molecular orbital has an orbital index (Ψ ) and a corresponding energy
j
value (E – eigenvalue).
j
Figure 7. Energy spectra of molecular orbitals (eigenvalues profiles) computed by PM3 method
0
for the investigated molecules 1 (a) and 2 (b) in their ground state (S ).
2
5

ACCEPTED MANUSCRIPT
Here, the blue lines indicate the eigenvalues of occupied molecular orbitals (OMO), and
the red ones denote the energy levels of unoccupied (virtual) orbitals (UMO). Such spectra give
useful details about the energy values of the main molecular orbitals. The information about
highest occupied (HOMO) and lowest unoccupied (LUMO) molecular orbitals are important,
since they are mostly responsible for the chemical reactivity and spectroscopic properties of the
molecule [36, 41].
Likewise,
the
energy
difference
between
frontier
molecular
orbitals
(HOMO-LUMO), known as energy gap (∆E), is a measure of the intramolecular charge transfer
and is frequently used in chemical and biochemical activity studies [42].
The energy values of frontier molecular orbitals (HOMO, LUMO, HOMO‒1, LUMO+1)
and the corresponding energy gaps (∆E, ∆E1) computed by different quantum chemistry methods
are summarized in Tables 4 and 5. Thus, the semiempirical and HF methods suggested high
energy gap (∆E) values for the titled compounds, i.e. from 8.8 to 10.2 eV. The DFT method
provided smaller energy gap values of about 5.1 eV. Such disparity can be attributed to the fact
that DFT method (B3LYP/6‒31+G**) uses the exchange‒correlation functional [43] and
therefore it treats in explicit manner the electronic correlation effects comparing with HF and
semiempirical methods.
In our case, the large HOMO‒LUMO gaps (> 5 eV) suggest the high excitation energies
for many excited states, good chemical stability and moderate reactivity for the investigated
compounds. The molecular descriptors such as ionization potential (IP), electron affinity (EA),
electronegativity (
can be computed from the eigenvalues of the frontier molecular orbitals. These estimators are
important parameters for quantum chemistry. Note that, electronegativity ( ) is a chemical
χ), chemical potential (µ), chemical hardness (η) and electrophilicity index (ω)
χ
2
6

ACCEPTED MANUSCRIPT
property describing the tendency of a molecule to attract electrons (or electron density) towards
itself. By contrast, the electronic chemical potential (
electron density from the molecule. The chemical hardness (
molecule to lose electrons. In turn, the electrophilicity index (
of a molecule, its propensity to be saturated with electrons [44].
µ
) determines the escaping tendency of the
) measures the resistance of the
) indicates the electrophilic power
η
ω
Thus, by using the energy values of HOMO and LUMO the molecular descriptors have
been calculated and reported in Tables 4 and 5. The relationships employed for the calculation of
the molecular descriptors are given in Ref. [37] and in Table B8 (ESI). According to the
theoretical results, the chemical hardness values greater than 2 eV and the low chemical potential
values suggested a good stability and resistance of the molecules to lose electrons. The
electronegativity values higher than 3.7 eV indicated the tendency of the molecules to attract the
electron density. However, the electrophilicity index values ranging from 1.4 eV to 3.1 eV,
revealed a moderate electrophilic power.
The plots of the frontier molecular orbitals for the studied molecules (1 and 2) are shown
in Figure 8. In this figure, the spatial distributions of the molecular orbitals were computed at the
level of DFT theory (B3LYP/6‒31+G**).
2
7

ACCEPTED MANUSCRIPT
Figure 8. Plots of frontier molecular orbitals for optimized structures of investigated compounds
1
and 2 computed at the level of DFT method (B3LYP/6‒31+G**): (a), (b), (c), (d) molecular
orbitals for 1; (e), (f), (g), (h) molecular orbitals for 2.
As one can see from Figure 8 (a, e), the HOMO is a delocalized π‒orbital distributed on
the aromatic ring and over the ether and carboxylic groups. By contrast, the HOMO‒1 levels
2
8