The relationship between molecular structure and biological activity of alkali metal salts of vanillic acid: Spectroscopic, theoretical and microbiological studies

Article abstract of DOI:10.1016/j.saa.2012.01.044

In this paper we investigate the relationship between molecular structure of alkali metal vanillate molecules and their antimicrobial activity. To this end FT-IR, FT-Raman, UV absorption and ¹H, ¹³C NMR spectra for lithium, sodium, p

Full text of DOI:10.1016/j.saa.2012.01.044

Spectrochimica Acta Part A 100 (2013) 31–40
Contents lists available at SciVerse ScienceDirect
Spectrochimica Acta Part A: Molecular and
Biomolecular Spectroscopy
journal homepage: www.elsevier.com/locate/saa
The relationship between molecular structure and biological activity of alkali
metal salts of vanillic acid: Spectroscopic, theoretical and microbiological studies
∗
´
Renata Swisłocka, Jolanta Piekut, Włodzimierz Lewandowski
Division of Chemistry, Bialystok University of Technology, Zamenhofa 29, 15-435 Białystok, Poland
a r t i c l e i n f o
a b s t r a c t
Article history:
In this paper we investigate the relationship between molecular structure of alkali metal vanillate
molecules and their antimicrobial activity. To this end FT-IR, FT-Raman, UV absorption and 1H, 13C NMR
spectra for lithium, sodium, potassium, rubidium and caesium vanillates in solid state were registered,
assigned and analyzed. Microbial activity of studied compounds was tested against Escherichia coli, Pseu-
domonas aeruginosa, Staphylococcus aureus, Proteus vulgaris, Bacillus subtilis and Candida albicans. In order
to evaluate the dependence between chemical structure and biological activity of alkali metal vanillates
the statistical analysis was performed for selected wavenumbers from FT-IR spectra and parameters
describing microbial activity of vanillates. The geometrical structures of the compounds studied were
optimized and the structural characteristics were determined by density functional theory (DFT) using
at B3LYP method with 6-311++G** as basis set. The obtained statistical equations show the existence of
correlation between molecular structure of vanillates and their biological properties.
Received 30 October 2011
Received in revised form 4 January 2012
Accepted 18 January 2012
Keywords:
Vanillic acid
Alkali metal vanillates
Spectroscopic studies
Molecular structure
Microbial activity
© 2012 Elsevier B.V. All rights reserved.
1. Introduction
acid exhibited additive inhibitory effects, particularly at lower pH.
These natural antimicrobial compounds could prove useful either
Biological activity of benzoic acid derivatives depends on
their molecular and electronic structure. Such dependencies were
reported in our previous works [1–3]. Knowledge of such correla-
tions allows to replace trial and error method with a systematic
approach in the search for new antimicrobial substances.
Vanillic acid (4-hydroxy-3-methoxybenzoic acid) is a benzoic
acid derivative used as a flavoring agent. It is an oxidized form of
vanillin. It is also an intermediate in the production of vanillin from
ferulic acid. The highest amount of vanillic acid in plants known so
far is found in the root of Angelica sinensis, usually referred to as
“dong quai”, “dang gui” or “female ginseng”, a herb indigenous to
China, which is used in traditional Chinese medicine.
The antimicrobial effects of vanillin and vanillic acid were ver-
ified against several species and strains of Listeria monocytogenes,
Listeria innocua, Listeria grayi, and Listeria seeligeri in a laboratory
medium adjusted to pH values ranging from 5.0 to 8.0. Medium
pH had little influence on the MIC of vanillin as determined by a
broth dilution assay, and growth of all test strains was inhibited
by concentrations ranging from 23 to 33 mM. Bactericidal effects
increased with pH in media supplemented with vanillin. An inverse
relationship was found for vanillic acid, and the lethality of the com-
pound increased with declining pH. Mixtures of vanillin and vanillic
alone or in mixtures for the control of Listeria spp. in food products
[4].
The aim of this paper is to study: (1) the molecular struc-
ture of alkali metal salt of vanillic acid (one of naturally occurring
methoxy-derivatives of hydroxybenzoic acid); (2) the antimicro-
bial activities of the studied compounds; (3) dependencies between
the molecular structure of investigated molecules and their biolog-
ical activity. In this way it may become possible to find compounds
with specific activity not by a process of trial and error, but on the
basis of strictly described dependencies between the structure and
biological activity of compounds [2]. In the present work many dif-
ferent analytical methods, which complement one another, were
used: infrared (FT-IR), Raman (FT-Raman), electronic absorption
spectroscopy, nuclear magnetic resonance (1H and 13C NMR) and
quantum mechanical calculations carried out with the GAUSSIAN
09 [5]. Optimized geometrical structures of studied compounds
were calculated. The theoretical IR and NMR (1H and 13C) spectra
were obtained.
2. Experimental
Lithium, sodium, potassium, rubidium and caesium vanillates
were prepared by dissolving the powder of vanillic acid in the water
solution of the appropriate alkali metal hydroxide in a stoichio-
metric ratio. The mixed solution was slowly condensed at 70 ◦C.
Then, the remaining solvent was removed by drying at 120 ◦C. The
∗
Corresponding author.
E-mail address: w-lewando@wp.pl (W. Lewandowski).
1386-1425/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
doi:10.1016/j.saa.2012.01.044

´
R. Swisłocka et al. / Spectrochimica Acta Part A 100 (2013) 31–40
32
salts obtained were found to be anhydrous as indicated by the lack
of bonds characteristic of water of crystallization in the IR spec-
tra.
correlation between experimental and theoretical IR spectra was
noted. The correlation coefficients for IR spectra are: 0.9975 (acid),
0.9973 (Li), 0.9973 (Na) and 0.9979 (K). Raman and IR experimental
and calculated spectra of vanillic acid are presented in Fig. 1.
Characteristic vibration bands of the carboxylic group are
present in the spectra of vanillic acid, and are not observed
in the spectra of salt. There are very intense, broad stretching
The IR spectra were recorded with an Equinox 55, BRUKER FT-
IR spectrometer within the range of 400–4000 cm−1. Samples in
the solid state were measured in KBr matrix pellets which were
obtained with hydraulic press under 739 MPa pressure. Raman
spectra of solid samples in capillary tubes were recorded in the
range of 100–4000 cm−1 with a FT-Raman accessory of the Perkin-
−1
bands: ꢀ(C O): 1682 cm
(R); ꢀ(OH) in the range 2957–2571 cm
(IR), 1780 cm−1 (IRcalc.), 1642 cm−1
−1
(IR) and ꢀ C (OH):
Elmer system 2000. The resolution of spectrometer was 1 cm−1
.
1298 cm−1 (IR), 1357 cm−1 (IRcalc.), 1301 cm−1 (R). Deformation in
plane vibration bands ˇ(C O): 766 cm−1 (IR), 766 cm−1 (R) and
deformation out of plane vibration bands ꢁ(C O): 637 cm−1 (IR),
777 cm−1 (IRcalc.), 638 cm−1 (R). Moreover, in the spectra of acid
the bands of lower intensity are present, origi−n1ating from defor-
The NMR spectra of DMSO saturated solution were recorded with
the NMR AC 200 F, Bruker unit at room temperature. TMS was used
as an internal reference. The UV spectra in water solution were
recorded on a DR 5000 HACH-LANGE spectrophotometer between
190 and 400 nm. The compounds were studied in aqueous solutions
with concentrations 10−4 mol/dm3.
To calculate optimized geometrical structures the density func-
tional theory (DFT) hybrid method B3LYP with nonlocal correlation
provided by Lee–Yang–Parr expression was used. The 6-311++G**
basis set was used as the standard basis set available for all atoms
in studied molecules. The method has been successfully applied
to calculate geometrical parameters, IR spectra, atomic charges for
systems similar to these studied in this paper All theoretical calcu-
lations were performed using the GAUSSIAN 09 (rev. 11.2) package
of programs [5] running on a PC.
Five species of bacteria: Escherichia coli (PCM 2268), B. subtilis
(PCM 2021), P. vulgaris (PCM 2269), P. aeruginosa (PCM 2270), S.
aureus (PCM 2267) and one species of yeasts Candida albicans (PCM
2566) were used for antimicrobial tests. Bacterial cultures where
purchased from Polish Collection of Microorganisms (PCM). Micro-
biological analysis was carried out according to the literature [6,7].
Solutions of tested compounds were prepared by dissolving 0.2 g
of each of them in 9.8 ml of deionized water. The concentration of
compounds in the culture broth was 0.1%. 4.75 ml of broth inocu-
lated with bacteria was used and then 0.25 ml solution of tested
compound was added. The studied bacteria were inoculated on
enriched broth medium and yeast on Sabouraud broth medium and
stored in 35◦C (for yeast in 25 ◦C) for 24 h. The growth of tested cells
was standardized turbidimetrically by measuring optical density at
600 nm with Hach Lange UV/vis spectrophotometer DR 5000. The
microbiological tests for studied compounds were carried out in
deionized water medium. The samples were incubated in 35 ◦C for
bacteria and 25 ◦C for yeast. The number of colonies was directly
proportional to optical density, which was estimated similarly after
24 and 48 h incubation. Statistical calculations were performed
using Statistica 9.1 program [8].
mation vibrations ˇ(OH) located at 1205 cm
(IR), 1259 cm−1
(IRcalc.), 1206 cm−1 (R) and ꢁ(OH) 918 and 588 cm−1 (IR), 580 cm−1
(IRcalc.), 919 and 585 cm−1 (R) have also been observed. In the spec-
trum of the acid are also present bands from the aromatic ring
and functional groups of the ring (hydroxyl OHar and methoxy
OCH3 groups), which are also present in the spectra of salt.
Replacement of the carboxylic group hydrogen with a metal ion
brought about characteristic changes in the IR and Raman spectra
of the metal vanillates in comparison with the spectra of vanil-
lic acid. One can observe appearance of bands of the symmetric
and asymmetric vibrations of the carboxylate anion: ꢀas(COO−)
in the range: 1558–1549 cm−1 (IR), 1554–1535 cm−1 (IRcalc.),
1566–1546 cm−1 (R), ꢀs(COO−) in the range: 1395–1385 cm−1 (IR),
1426–1394 cm−1 (IRcalc.), 1396–1371 cm−1 (R); ˇs(COO−) in the
range: 961–957 cm−1 (IR), 835–810 cm−1 (IRcalc.), 968–953 cm−1
(R), ˇas(COO−) in the range: 488–461 cm−1 (IR), 490–461 cm−1
(R) and ꢁs(COO−) in the range: 781–777 cm−1 (IR), in the range
802–786 cm−1 (IRcalc.), in the range 785–778 cm−1 (R).
In the spectra of acid and salts there are present bands o
from the hydroxyl group attached to the aromatic ring: ꢀ(OH)ar
3485 cm−1 (IR), 3764 cm−1 (IRcalc.) in the spectrum of vanillic
acidand3455–3443 cm−1 (IR), 3768–3749 cm−1 (IRcalc.), ꢀC (OH)ar
1225–1219 cm−1 (IR), 1301–1270 cm−1 (IRcalc.), 1226–1217 cm−1
(R); ˇ(OH)ar 1253–1236 cm−1 (IRcalc.); ꢁ(OH)ar 928 cm−1 (IR),
932–929 and 645–640 cm−1 (R) in the spectra of salts.
Stretching vibrations ꢀas(CH3) give the bands in the ranges:
2948–2938 cm−1 (IR), 3147–3133 and 3088–3072 cm−1 (IRcalc.),
2944–2939 cm−1 (R) and ꢀs(CH3) in the ranges: 2841–2835 cm−1
(IR), 3023–3012 cm−1 (IRcalc.), 2841–2836 cm−1 (R).
The ıas(CH3) (in plane bending) bands that occur in the
ranges: 1474–1464, 1454–1447 and 1030–1026 cm−1 (IR),
1506–1494, 1492–1481 cm−1 (IRcalc.), 1474–1466, 1454–1448 and
1033–1026 cm−1 (R) and ıs(CH3) in the range: 1379–1354 cm−1
(IR), 1484–1478 cm−1 (IRcalc.), 1381–1351 cm−1 (R), as well as
rocking vibration bands ꢂ(CH3) in the range: 1222–1216 and
1171–1168 cm−1 (Rcalc.). The bands of the ꢀO (CH3) vibrations
are located in the range 1186–1182 cm−1 (IR), 1060–1043 cm−1
(IRcalc.), 1184–1178 and 1074–1064 cm−1 (R).
3. Results and discussion
3.1. Vibrational spectra
The vibrational spectra of vanillic acid and the synthesized
lithium, sodium, potassium, rubidium and caesium vanillates were
recorded and assigned. The observed Raman and IR bands together
with their relative intensities and band assignments of studied
metal salts are presented in Table 1 (vanillic acid) and 2 (vanil-
lates), respectively. The symbol “ꢀ” denotes stretching vibrations,
“ˇ”: in-plane bending modes, “ꢁ”: out-of-plane bending modes;
“ϕ(CCC)”: the aromatic ring out-of-plane bending modes, “˛(CCC)”:
the aromatic ring in-plane bending modes, “ı(CH3)”: the methyl
group bending vibrations, “ꢂ(CH3)”: the methyl group rocking
deformations. The bands are numbered along with the notation
used by Varsányi [9]. Theoretical calculations by B3LYP method
at 6-311++G** level were used to obtain values of wavenumbers
and intensities of IR and Raman spectra (Tables 1 and 2 ). Good
3.2. NMR spectra
Theoretical and experimental 1H and 13C NMR spectra of
lithium, sodium and potassium vanillates were obtained (Table 3).
Geometrical structures of studied compounds were optimized by
B3LYP/6-311++G** method. The atom positions were numbered as
in Fig. 2. The signals from protons no. 3 and 4 are shifted down-
field in comparison to the appropriate signals in the spectrum of
ligand. In the 1H NMR spectra of vanillates no regularity can be
observed in the chemical shifts of hydrogen along the series of
alkali metal vanillates. Chemical shift values of signals traced for
vanillic acid werer compared with those traced for m-anisic acid

´
R. Swisłocka et al. / Spectrochimica Acta Part A 100 (2013) 31–40
33
Table 1
Wavenumbers [cm−1], intensities and assignments of bands occurring in the experimental FT-IR, FT-Raman and theoretical FT-IR spectra of vanillic acid. The theoretical
wavenumbers were calculated in B3LYP/6-311++G** level.
Vanillic acid
Assignment
No.a
IR
Calc.
Raman
Exp.
Exp.
Theoret.b
Int.IR
98.71
137.80
2.21
1.50
2.06
Theoet.c
3776
3760
3230
3208
3192
Int.IR
3775
3764
3228
3209
3193
160.94
253.82
3.01
3.78
5.02
ꢀ(OH)
3485 vsd
3098 w
3016 sh
2988 w
ꢀ(OH)ar
ꢀ(CH)
ꢀ(CH)
3083 w
3027 vw
2991 vw
2
20b
20a
ꢀ(CH)
2957–2571
ꢀ(OH)
3142
3079
3016
1780
1643
1633
1544
1505
1492
1484
1457
16.98
28.78
43.58
443.02
30.91
140.73
176.54
48.84
10.73
1.44
3143
3080
3016
1758
1638
1628
1537
1497
1485
1486
1454
22.04
42.96
62.37
846.98
22.65
282.97
306.96
74.08
15.75
0.93
ꢀas(CH3)
ꢀas(CH3)
ꢀs(CH3)
ꢀ(C O)
ꢀ(CC)
1682 vs
1609 sh
1599 vs
1524 vs
1474 m
1454 vs
1642 m
1602 vs
8b
8a
19b
ꢀ(CC)
ꢀ(CC)
1518 w
1474 w
1454 w
ıas(CH3)
ıas(CH3)
ıs(CH3)
ꢀ(CC)
ıs(CH3)
ꢀC (OH)
ꢀ(CC)
ꢀ(CH)
ˇ(CH)
ꢀC (OH)ar
ˇ(OH)
ˇ(OH)ar
ꢂ(CH3)
1435 vs
1379 m
1298 vs
1283 vs
1238 s
13.94
18.99
1425 m
1381 m
1301 sh
1286 s
19a
1357
1415
208.25
37.02
1350
1411
296.77
63.80
14
7a
3
1238 w
1319
1301
1259
1227
1200
1181
1170
31.33
211.39
93.81
89.37
128.35
331.12
0.53
1319
1289
1246
1223
1196
1167
1175
41.36
429.10
133.94
117.92
179.92
519.85
0.90
1205 s
1206 m
ꢂ(CH3)
ꢀO (CH3)
ꢀ(CH)
ˇ(CH)
ˇ(CH)
ꢀO (CH3)
ıas−(CH3)
ꢁ(OH)
ꢀ(CH)
ꢁ(CH)
1186 m
1169 sh
1182 m
1170 m
13
18a
18b
1139
1087
1055
81.94
132.13
28.96
1133
1069
1051
192.75
419.81
18.02
1113 s
1123 vw
1030 s
918 m
883 m
820 m
806 m
766 s
758 sh
721 w
637 m
588 m
565 w
540 m
507 m
451 w
417 vw
1031w
919 m
884 vw
922
890
849
5.53
19.10
8.54
919
891
844
15.34
34.35
14.57
7b
11
808 m
766 w
742 s
ꢁ(CH)
ˇ(C O)
˛(CCC)
˛(CCC)
ꢁ(C O)
ꢁ(OH)
ϕ(CC)
˛(CCC)
ϕ(CC)
810
26.09
809
57.31
12
1
725 w
638 m
585 w
573 w
542 w
498 w
450 w
417 vw
777
588
730
719
63.37
23.34
7.46
771
578
735
717
93.88
10.59
15.70
29.79
16a
6a
16b
6b
17.44
570
481
34.21
82.23
568
426
67.49
˛(CCC)
ˇ(CH)
ꢁ(OH)ar
9b
180.98
a
Ref. [9].
b
calculations for isolated molecules.
calculations for the molecules in the solvent (water).
s, strong; m, medium; w, weak; v, very; sh, shoulder.
c
d
(4-metoxybenzoic acid) [10]. Lower values of ꢃ protons in vanillic
acid relative to the corresponding anisic acid protons were found.
The signals from C3, C4, C5, C6 and C8 carbon atoms in salts
are shifted upfield in comparison to the appropriate signals in
the spectrum of ligand, whereas the signals from carbons no. 2
and 7 are shifted downfield comparing to free ligand. This indi-
cates an increase in the electron density around C2 and C7 carbons
and a decrease in the electron density around the other car-
bon atoms. The differences between chemical shifts of carbon
atoms along metal series are small and irregular. Only the val-
ues of chemical shifts of C2 atoms regularly increase in the series:
acid → Li → Na → K → Rb → Cs. The same relationship describes the
calculated atomic charges mentioned before.
The linear correlation between calculated and experimental
data of 13C NMR is noted. Correlation coefficients of 13C NMR for
vanillic acid, lithium, sodium and potassium salts amount to 0.9816,
0.9767, 0.9779 and 0.9534, respectively. Corresponding values of
correlation coefficient of 1H NMR are lower.
3.3. UV spectra
The UV spectra of vanillic acid and vanillates were recorded. The
UV spectrum of vanillic acid was recorded in water solution (the
maxima of ␲ → ␲* bands are at 204, 256 and 289 nm) and ethanol
(the maxima are at 205, 216, 259 and 289 nm). Replacement of the
carboxylic group hydrogen with an alkali metal ion did not cause

´
R. Swisłocka et al. / Spectrochimica Acta Part A 100 (2013) 31–40
34
Table 2
Wavenumbers [cm−1], intensities and assignments of bands occurring in the IR and Raman spectra of lithium, sodium, potassium, rubidium and caesium vanillates. The
theoretical wavenumbers were calculated in B3LYP/6-311++G** level.
Li vanillate
Na vanillate
IR
Raman
Exp.
IR
Raman
Exp.
Exp.
Theoret.a
Int.IR
Theoret.b
3745
Int.IR
Exp.
Theoret.
3773
Int.IR
Theoret.b
3749
Int.IR
212.10
3445 s
3768
3217
3209
3187
3136
3075
3014
1645
1636
1535
1542
1505
1490
1484
1449
1426
123.23
4.19
2.00
218.92
4.69
5.91
3455 m
3072 w
3026 w
2976 w
2938 m
117.07
5.09
2.96
3088 w
3017 vw
2974 w
2948 w
3218
3205
3182
3149
3090
3024
1640
1628
1538
1531
1494
1482
1480
1440
1388
3080 w
2944 w
3216
3208
3184
3133
3072
3012
1646
1637
1554
1538
1506
1488
1482
1449
1411
3217
3205
3180
3148
3090
3024
1641
1628
1547
1529
1495
1481
1479
1441
1375
5.39
6.45
3094 w
3028 w
2977 w
2939 w
4.54
10.81
23.63
44.57
65.15
2.38
125.52
971.83
129.59
47.37
14.55
7.02
5.99
12.11
24.07
45.38
66.59
2.95
112.52
1007.03
158.54
61.65
14.65
1.75
20.87
31.12
47.75
0.21
101.23
481.91
49.54
32.15
9.88
23.40
33.05
50.73
0.38
84.78
492.26
84.20
41.60
9.40
2841 w
1626 sh
1601 m
1557 s
1518 s
1464 m
1454 m
2841 w
1648 w
1602 vs
1566 w
1519 m
2837 w
1616 sh
1599 m
1558 vs
1520 m
1468 w
1447 m
2836 w
1638 sh
1602 vs
1563 w
1525 w
1467 sh
1448 m
1453 s
17.61
130.91
536.10
4.38
59.38
309.68
1421 sh
1393 vs
1360 sh
1281 s
104.97
619.18
1417 w
1396 s
1369 m
1277 m
1250 w
1433 sh
1395 vs
1389 sh
1283 s
75.76
852.86
1413 w
1376 vs
1360 sh
1278 s
1402
207.50
1401
495.67
1394
395.87
1396
175.65
1250 m
1258 m
1254 w
1310
1288
1253
1222
1171
60.34
161.54
159.41
83.91
0.50
1300
1272
1238
1216
1169
80.56
339.50
306.25
107.99
0.91
1308
1285
1249
1222
1171
51.83
166.58
169.41
82.20
0.51
1300
1273
1236
1217
1169
67.44
342.77
325.81
111.27
0.97
1219 m
1217 w
1225 s
1226 w
1186 m
1153 sh
1178 m
1150 m
1186 m
1166 sh
1138 w
1113 s
1187 w
1158 vw
1134 w
1115 w
1064 vw
1027 m
958 vw
932 m
1135
1128
1057
25.37
44.45
46.33
1128
1118
1041
6.26
136.68
93.63
1133
1128
1060
35.53
34.23
46.21
1127
1119
1042
4.62
135.83
98.54
1113 s
1117 w
1030 s
961 vw
928 w
883 m
827 sh
793 sh
779 s
1033 w
967 vw
930 w
881 vw
832 vw
1026 s
957 w
928 w
878 m
826 vw
814 vw
777 s
835
12.10
819
1.10
825
1.25
814
0.23
938
911
854
798
798
4.34
16.61
4.12
54.78
88.98
929
904
848
790
786
6.11
31.41
5.59
113.20
143.50
933
914
855
802
793
5.18
16.27
3.24
52.65
80.39
925
907
848
792
784
6.48
30.46
3.76
96.75
163.77
881 w
822 vw
807 vw
778 m
743 w
725 w
645 m
569 w
542 w
498 vw
489 w
782 m
742 vw
722 vw
644 m
566 w
546 vw
501 w
490 w
461 w
414 vw
746 vw
727 w
746 w
723 w
565 vw
548 w
498 w
486 vw
457 vw
422 vw
734
672
583
4.00
122.33
0.78
734
651
577
6.64
44.83
9.38
567 w
540 w
494 w
488 w
455 w
419 vw
733
658
583
4.12
22.48
0.85
733
649
575
8.33
36.37
1.18
545
472
16.27
68.18
544
462
2.49
0.03
576
469
19.25
28.55
574
463
30.52
0.01
454 w
417 vw
K vanillate
IR
Rb vanillate
Cs vanillate
Assignment
No.
Raman
IR
Raman
Exp.
IR
Raman
Exp.
Exp.
Theoret.
Int.IR
Theoret.b
Int.IR
Exp.
Exp.
Exp.
3447 s
3749
3216
3203
3179
3147
3088
3023
1641
1628
1554
1527
1494
1481
1478
1440
1367
210.58
5.62
7.29
3749
3216
3203
3179
3147
3089
3023
1641
1628
1554
1527
1494
1481
1478
1440
1367
210.56
5.61
7.29
3447 s
3103 w
3022 vw
2972 vw
2940 w
3443 s
3078 w
3001 w
2963 w
2940 vw
ꢀ(OH)ar
ꢀ(CH)
ꢀ(CH)
3098 sh
3001 w
2970 w
2940 w
3079 w
3032 w
2974 w
2939 w
3074 w
3075 w
3004 w
2971 w
2943 w
2
20b
20a
3019 vw
2974 w
2940 w
12.51
24.71
46.01
67.32
4.24
107.18
983.39
175.95
63.44
14.50
1.00
12.51
24.73
46.02
67.32
4.25
107.17
983.44
175.93
63.43
14.50
1.00
ꢀ(CH)
ꢀas(CH3)
ꢀas(CH3)
ꢀs(CH3)
ꢀ(CC)
ꢀ(CC)
ꢀas(COO)
ꢀ(CC)
ıas(CH3)
ıas(CH3)
ıs(CH3)
ꢀ(CC)
ꢀs(COO)
ıs(CH3)
ꢀ(CC)
2837 w
1645 m
1597 m
1551 vs
1520 m
1464 w
1449 m
2836 w
1646 w
1602 vs
1558 w
1525 w
1466 w
1448 m
2837 w
1651 m
1599 m
1549 s
1516 s
1466 w
1450 m
2835 w
2837 w
1645 m
1599 m
1551 vs
1518 m
1465 m
1449 w
2836 w
1642 w
1601 vw
1546 w
1522 w
1469 w
1448 m
8b
8a
1603 vw
1549 w
1526 w
1470 w
1450 m
19b
1421 sh
1385 vs
1364 sh
1283 s
65.54
894.42
65.54
894.35
1418 w
1378 sh
1366 sh
1276 s
1418 sh
1385 vs
1354 sh
1284 s
1422 vw
1375 sh
1355 m
1278 m
1250 vw
1418 sh
1387 vs
1356 sh
1283 s
19a
1371 sh
1351 m
1275 m
1255 vw
1394
133.68
1394
133.67
14
7a
1256 m
1255 vw
1254 m
1254 m
ꢀ(CH)

´
R. Swisłocka et al. / Spectrochimica Acta Part A 100 (2013) 31–40
35
Table 2 (continued).
K vanillate
IR
Rb vanillate
Cs vanillate
Assignment
No.
Raman
Exp.
IR
Raman
Exp.
IR
Raman
Exp.
Exp.
Theoret.
Int.IR
72.93
353.10
332.37
102.84
42.44
Theoret.b
Int.IR
72.91
353.03
332.40
102.82
42.39
Exp.
Exp.
1299
1270
1236
1216
1190
1299
1270
1236
1216
1190
ˇ(CH)
3
1221 vs
1218 w
1219 s
1218 w
1221 s
1217 w
ꢀC (OH)ar
ˇ(OH)ar
ꢂ(CH3)
ꢂ(CH3)
ꢀO (CH3)
ꢀ(CH)
1182 m
1183 w
1151 w
1131 w
1115 w
1074 vw
1026 m
959 vw
931 m
1184 m
1183 w
1150 vw
1129 w
1115 w
1074 vw
1030 w
953 w
1182 m
1184 w
1150 vw
1133 vw
1113 w
1074 vw
1026 w
968 w
13
18a
18b
1130 m
1113 s
1127
1117
1043
23.32
123.41
95.47
1127
1117
1043
23.31
123.43
95.40
1130 sh
1115 s
1129 sh
1113 vs
ˇ(CH)
ˇ(CH)
ꢀO (CH3)
ıas−(CH3)
ˇs(COO)
ꢁ(OH)ar
ꢀ(CH)
1026 s
1026 s
1026 vs
810
1.90
810
1.90
928 w
881 m
836 vw
810 sh
781 vs
746 sh
727 w
928 w
880 m
827 w
931 m
882 w
826 w
928 w
881 m
826 vw
808 sh
781 s
929 m
922
906
847
786
779
7.19
31.78
5.63
92.69
179.10
922
906
847
786
779
7.20
31.78
5.63
92.67
176.11
886 w
882 vw
825 vw
796 m
782 s
745 w
729 w
640 m
566 w
538 w
490 w
462 w
456 vw
428 vw
7b
11
ꢁ(CH)
ꢁ(CH)
797 m
783 s
777 s
746 m
727 w
785 s
739 w
726 w
643 m
574 w
542 w
498 w
461 w
458 w
417 vw
ꢁs(COO)
˛(CCC)
˛(CCC)
ꢁ(OH)ar
ϕ(CC)
˛(CCC)
ϕ(CC)
ˇas(COO)
˛(CCC)
ˇ(CH)
750 w
727 w
641 m
574 w
538 w
490 w
465 w
442 vw
417 w
745 w
727 w
12
1
571 vw
536 w
494 vw
461 w
438 vw
420 w
731
647
575
574
543
7.44
40.22
1.41
731
647
575
574
543
7.44
40.22
11.41
30.37
4.67
567 w
538 w
492 w
462 w
457 w
422 w
569 w
538 w
494 w
461 w
16a
6a
16b
30,36
4.68
6b
9b
430 vw
417
174.09
417
174.10
ꢁ(OH)ar
a
Calculations for isolated molecules.
Calculations for the molecules in the solvent (water).
b
significant changes in the UV spectrum. For vanillates the maxima
of ␲ → ␲* bands were found to be shifted hipsochromically and
located within the range of 202–203, 251–254 and 284–285 nm.
are gathered in Table 4. In acid molecule almost all bonds
are of similar length, being longer than the conjugated C
C
˚
bond of the benzene ring (approx, 1.390 A for benzene). The
C2 C3 distance is the shortest one and the C3 C4 bond is the
longest one in the ring. Alkali metal ions influence mostly the
bond lengths and angles in the carboxylic group (which was
expected) and they influence the methoxy and hydroxy group
geometry.
3.4. Structural parameters
The geometrical parameters calculated at the B3LYP/6-
311++G** level for vanillic acid and Li, Na and K vanillates
Fig. 1. Calculated (a) and experimental FT-IR (b), Raman (c) spectra of vanillic acid.

´
R. Swisłocka et al. / Spectrochimica Acta Part A 100 (2013) 31–40
36
Table 3
The experimental and theoretical chemical shifts of protons in the 1H NMR and carbons in the 13C NMR spectra of vanillic acid and alkali metal vanillates, ı [ppm].
Atom positiona
Vanillic acid
Vanillates
Li
Na
K
Rb
Cs
Exp.
12.47
7.45
3.80
3.80
9.82
Calc.
Exp.
Calc.
–
7.77
3.78
4.02
5.55
6.96
8.01
113.93
130.31
150.05
155.87
116.62
130.15
188.75
55.37
Exp.
–
7.44
3.74
3.74
9.01
6.64
7.31
122.17
113.24
146.12
147.66
113.92
120.86
169.67
55.32
Calc.
–
7.79
3.73
3.97
5.37
6.88
8.01
114.69
132.50
149.73
154.58
116.09
130.30
182.29
55.05
Exp.
–
7.45
3.74
3.74
–
6.75
7.32
131.20
113.31
146.30
148.03
114.14
122.28
169.20
55.32
Calc.
–
7.65
3.71
3.92
5.27
6.86
7.94
114.20
135.91
149.31
153.69
116.03
130.07
180.02
55.39
Exp.
–
7.46
3.75
3.75
–
6.84
7.36
129.09
113.31
146.67
149.33
114.53
122.60
169.61
55.40
Exp.
–
7.46
3.73
3.73
–
6.87
7.36
130.05
113.40
146.69
149.28
114.58
122.53
169.99
55.37
H1
H2
H3a,b
H3c
H4
H5
H6
C1
C2
C3
C4
C5
5.39
7.50
3.75
4.09
5.76
6.05
7.72
3.88
4.27
6.21
–
7.44
3.75
3.75
9.15
6.67
7.33
122.43
113.16
146.34
148.14
114.12
120.86
168.61
55.35
–
7.77
3.88
4.21
5.87
7.09
7.81
133.16
114.30
151.18
154.50
116.10
127.84
183.11
56.57
–
7.77
3.84
4.20
5.77
7.01
7.80
134.78
114.57
151.11
153.86
115.81
127.78
180.05
56.35
–
7.73
3.83
4.20
5.75
6.97
7.79
135.95
114.35
151.05
153.71
115.74
127.31
179.57
56.17
6.68
7.33
7.12
8.04
7.20
7.95
123.56
112.80
147.29
151.16
115.10
121.69
167.28
55.60
113.79
124.23
150.16
157.57
117.99
132.70
169.41
55.60
123.99
115.41
151.62
157.96
117.47
131.02
171.73
56.26
C6
C7
C8
a
Atom numbering in Fig. 2.
In the series of vanillic acid → Li → Na → K vanillates, the C1 C7,
C4 O4, O1 M/H1, O2 M, and O3—C8 distances increase, whereas
the C7 O2 and C7 O1 (Li → K) bonds shorten. Moreover, the
C2 C1 C6 and C1 C7 O1 angles decrease, and the C1 C2 C3,
as well as C5 C6 C1 angles increase along the series: vanillic
acid → Li → Na → K vanillates. The existence of some regular
changes in the distances and angles of bonds suggests that other
parameters, such as atomic charges of molecules or wavenumbers
in the IR or Raman spectra, should also change regularly in the same
fashion for vanillates. Nevertheless one should bear in mind that
the calculations were performed for isolated molecules, whereas
in real systems the intermolecular interactions exist, which affect
the electronic systems of compounds.
(1 hartree = 2625.5 kJ/mol). Obtained results show that the degree
of ionic bonding increases from H to Na atom, because of an increase
symmetrization of carboxylate ion and in aromaticity of molecule.
Atomic charges. The alkali metal cations affect the electronic den-
sity of ligand (Table 6), The substitution of metal ions in the place
of the carboxylic group hydrogen causes a decrease in the elec-
tronic charge density on the C1 atom and an increase on C2, C3,
C4, C5, C6, C8, O3, O4, H2, H4, H5 and H6 atoms compared with
the appropriate atomic charges calculated for acid. Electron density
around these atoms increases in the series: acid → Li → Na → K. The
atomic charges on other atoms change insignificantly, depending
on the applied method. Among the three applied methods in atomic
charge calculations the best statistical correlations were obtained
for NBO and ATP analysis and the changes of carbon and proton sig-
nals found in experimental NMR spectra (exemplary correlation for
atomic charges and values of chemical shifts from 1H NMR spectra
are presented in Fig. 3). This suggests that NBO and ATP methods
describe the electronic charge distribution in vanillate molecules
and similar systems in a better way than Mulliken method. Metal
substitution in the carboxyl group causes an increase in a total
charge of the carboxylic ion and a decrease on the remaining func-
tional group.
Correlation between theoretical and experimental [11] obtained
bond lengths and angles of vanillic acid molecule was studied and
good agreement was found, R = 0.9398 for bonds, R = 0.8782 for
angles.
Geometric aromaticity indices (HOMA, Aj, BAC and I6—Bird‘s
index) were calculated and presented in Table 5. Substitution of
the carboxylic group hydrogen with the metal atom causes an
increase in the aromaticity indices. Aromatics regularly increase
in the series: acid → Li → K. Comparing the aromaticity indices for
vanillic acids with those for m-anisic and benzoic acids [10] it can be
seen that the aromatics of these compounds increase in the series:
vanillic acid → m-anisic acid → benzoic acid.
Figs. 4–5.
3.5. Microbial activity
Comparing the values of dipole moment for molecule of
vanillic acid and its salts one can observe an increasing ten-
dency: 4.198 D for free acid; 4.799 D for lithium salt; 7.441 D for
sodium; for potassium salt 11,458 D. The energy (hartree) shows
a decreasing tendency: −610.76; −617.76; −772.52; −1210.20
Vanillic acid and its salts preparations tested at a concentra-
tion of 0.1% showed a varied impact on the growth of tested
1
O2
OH1
C
7
6H
5H
1
H2
6
2
5
3
3
3
4
O-8CH3
4OH4
Fig. 3. The relation between the obtained Mulliken, ATP and NBO atomic charges
Fig. 2. Vanillic acid, numbering scheme used in the present work.
and change of proton signals from experimental NMR spectra of vanillic acid.

´
R. Swisłocka et al. / Spectrochimica Acta Part A 100 (2013) 31–40
37
Table 4
The bond lengths and angles of the structures of vanillic acid and its alkali metal salts calculated in B3LYP/6-311++G** level (atom numbering in Fig. 2).
Vanillic acid
Vanillates
Li
Na
K
Calc.
Exp. [11]
Calc.
Calc.
Calc.
Bond lengths [Å]
C1 C2a
C2 C3
C3 C4
C4 C5
C5 C6
C1 C6
C1 C7
C7 O1
1.407
1.384
1.410
1.390
1.390
1.397
1.478
1.210
1.364
0.968
1.389
1.381
1.397
1.373
1.383
1.391
1.476
1.274
1.266
1.137
1.404
1.385
1.409
1.389
1.392
1.395
1.490
1.275
1.277
1.850
1.853
1.081
1.375
1.423
1.095
1.095
1.089
1.360
0.967
1.083
1.082
1.403
1.405
1.385
1.408
1.389
1.394
1.395
1.501
1.270
1.272
2.203
2.207
1.081
1.378
1.422
1.095
1.095
1.089
1.363
0.967
1.084
1.082
1.388
1.408
1.388
1.396
1.396
1.519
1.264
1.265
2.715
2.717
1.081
1.371
1.430
1.093
1.093
1.088
1.366
0.969
1.084
1.082
C7 O2
O1 H1/M
O2
M
C2 H2
C3 O3
O3 C8
C8 H3a
C8 H3b
C8 H3c
C4 O4
O4 H4
C5 H5
1.080
1.372
1.425
1.094
1.094
1.088
1.355
0.968
1.083
1.082
1.003
1.366
1.426
0.975
1.012
0.954
1.364
0.838
0.995
1.027
C6 H6
Angles [◦]
C1 C2 C3
C2 C3 C4
C3 C4 C5
C4 C5 C6
C5 C6 C1
C2 C1 C6
C1 C7 O1
C1 C7 O2
O1 C7 O2
C7 O1 H1/M
119.58
120.25
119.95
119.91
120.36
119.96
125.50
113.08
121.43
106.36
–
119.23
120.07
120.61
119.78
120.14
120.16
119.14
118.68
122.18
114.93
–
119.75
120.20
119.90
119.91
120.43
119.89
119.85
119.64
120.51
83.10
119.92
120.27
119.76
119.93
120.53
119.59
118.52
118.42
123.06
88.15
120.34
120.03
119.84
119.88
120.82
119.10
117.75
117.73
124.52
93.47
C7 O2
M
82.88
87.90
93.34
C1 C2 H2
C2 C3 O3
C3 O3 C8
O3 C8 H3a
O3 C8 H3b
O3 C8 H3c
C3 C4 O4
C4 O4 H4
C4 C5 H5
C5 C6 H6
119.31
126.10
118.52
110.98
110.98
105.97
120.05
108.03
118.66
120.71
119.84
125.92
117.35
113.07
110.07
105.10
119.79
108.23
118.24
119.95
118.20
126.08
118.24
111.04
111.04
106.03
120.02
107.73
118.61
120.79
117.96
125.96
118.15
111.09
111.09
106.06
120.05
107.53
118.60
120.80
118.42
126.14
118.55
110.87
110.87
105.89
119.98
107.88
118.92
120.20
a
1 A = 10−10 m.
˚
1,4
1,2
1,0
0,8
0,6
0,4
0,2
0,0
-0,2
Average
Standard error
Min-Max
Fig. 4. The degree of growth inhibition exhibited by studied compounds in relation to E. coli (EC), P. vulgaris (PV), P. aeruginosa (PA), B. subtilis (BS), S. aureus (SA) and C.
albicans (CA) evaluated by means of turbidimetry method by measuring optical density of water solutions at 600 nm after 24 h incubation.

´
R. Swisłocka et al. / Spectrochimica Acta Part A 100 (2013) 31–40
38
2,2
2,0
1,8
1,6
1,4
1,2
1,0
0,8
0,6
0,4
0,2
0,0
Average
Standard error
Min-Max
Fig. 5. The degree of growth inhibition exhibited by studied compounds in relation to E. coli (EC), P. vulgaris (PV), P. aeruginosa (PA), B. subtilis (BS), S. aureus (SA) and C.
albicans (CA) evaluated by means of turbidimetry method by measuring optical density of water solutions at 600 nm after 48 h incubation.
Table 5
bacteria (E. coli, P. aeruginosa, B. subtilis, S. aureus and P. vulgaris)
The calculated aromaticity indices for vanillic acid and vanillates lithium, sodium
and potassium.
and the fungus C. albicans, depending on the incubation time 24
and 48 h. In E. coli the highest growth inhibition after 24 h of
Aromaticity indices
Vanillic acid
Vanillates
Li
incubation was induced by lithium vanillate 24.8%, and the least
by rubidium vanillate: 8.6%. However, after 48 h incubation, the
observed growth inhibition was low, within 0.9–7.0%. In the case
of P. vulgaris after 24 h incubation growth inhibition was found
only in the case of vanillic acid, sodium and potassium vanillate.
Vanillic acid and its salts, in the case of bacteria, P. aeruginosa
showed a similar degree of inhibition, both after 24 h and 48 h and
it ranged within the 12.1–23%. In case of the bacteria B. subtilis
growth inhibition after 24 h of incubation was low—from 0.7% to
8.1%. After 48 h incubation, the strongest effect was obtained for
lithium vanillate, up to 34.4%. In the case of S. aureus it has been
demonstrated only after 48 h incubation, the strongest in the case
of lithium vanillate 33.7% and the weakest in the case of rubidium
vanillate—15.7%. After 24 h incubation, growth inhibition was not
Na
0.9691
93.4541
0.9929
0.8751
0.023
K
HOMA
I6
Aj
BAC
ꢄ(CC)a
0.9592
92.1141
0.9897
0.8551
0.026
0.9668
93.0157
0.9919
0.8689
0.024
0.9649
93.6131
0.9933
0.8785
0.020
a
Differences between the longest and the shortest bonds in the ring.
observed. For the fungus C. albicans growth inhibition after incu-
bating for 24 h was the strongest in the case of sodium vanillate.
After 48 h incubation, the stimulation of growth was found at the
level of 31–37%.
Table 6
The atomic charge [ea] calculated by B3LYP/6-311++G** method for vanillic acid and lithium, sodium and potassium vanilliates.
Vanillic acid
Vanillates
Li
Atom
Na
K
Mulliken
ATP
NBO
Mulliken
ATP
NBO
Mulliken
ATP
NBO
Mulliken
ATP
NBO
C1
C2
C3
C4
C5
C6
C7
C8
O1
O2
O3
O4
H1/M
H2
0.629
−0.047
−0.541
−0.189
−0.174
−0.296
0.215
−0.304
−0.301
−0.207
−0.263
−0.242
0.299
−0.360
−0.007
0.374
−0.182
−0.236
0.256
1.958
0.061
−0.335
−0.033
0.386
−0.166
−0.236
0.251
2.133
0.073
−0.282
−0.054
0.395
−0.159
−0.242
0.249
0.831
0.372
−0.493
−0.541
0.007
−0.585
−0.084
−0.313
−0.539
−0.549
−0.277
−0.300
1.015
−0.326
−0.091
0.562
−0.154
−0.247
0.248
−0.675
−0.188
−0.244
−0.602
−0.351
−0.327
−0.346
−0.340
−0.264
−0.251
0.188
−0.694
−0.279
−0.171
−0.653
−0.447
−0.324
−0.478
−0.468
−0.265
−0.257
0.445
0.611
0.313
0.581
0.302
0.560
0.295
0.726
0.288
−0.147
−0.246
−0.151
0.787
−0.202
−0.607
−0.700
−0.571
−0.661
0.485
−0.135
−0.250
−0.161
0.765
−0.200
−0.821
−0.830
−0.575
−0.670
0.887
−0.122
−0.003
1.369
−0.252
−0.169
0.765
−0.200
−0.820
−0.828
−0.577
−0.674
0.917
−0.141
−0.036
1.910
−0.254
−0.176
0.777
−0.199
−0.821
−0.829
−0.578
−0.680
0.942
0.045
0.022
1.461
0.508
1.422
0.509
0.511
0.636
−0.873
−0.769
−0.878
−0.799
0.302
−1.043
−1.029
−0.883
−0.798
0.828
−1.026
−1.012
−0.887
−0.796
0.854
−1.358
−1.353
−1.157
−1.032
1.029
0.211
0.092
0.232
0.206
0.096
0.236
0.213
0.098
0.235
0.195
0.127
0.231
H3a
H3b
H3c
H4
0.166
0.167
0.190
0.297
−0.024
−0.024
0.008
0.175
0.175
0.193
0.488
0.165
0.165
0.170
0.296
−0.026
−0.026
0.002
0.174
0.174
0.189
0.486
0.163
0.163
0.168
0.294
−0.028
−0.028
−0.002
0.329
0.173
0.173
0.187
0.485
0.177
0.177
0.195
0.317
−0.034
−0.034
0.004
0.173
0.173
0.186
0.483
0.338
0.333
0.418
H5
0.193
0.060
0.222
0.183
0.050
0.217
0.179
0.044
0.215
0.207
0.057
0.212
H6
0.193
0.083
0.231
0.200
0.079
0.228
0.204
0.079
0.227
0.187
0.092
0.225
COOH/M
OCH3
OH
0.006
0.121
−0.035
−0.230
−0.173
−0.849
−0.091
0.045
0.178
0.001
−0.948
−0.095
0.037
0.185
0.034
−0.157
−0.041
0.017
0.228
0.069
−0.044
−0.410
−0.424
−0.238
−0.434
−0.244
−0.585
−0.245
0.055
−0.461
−0.465
−0.184
−0.467
−0.189
−0.614
−0.197
a
e = 1.6021892 × 10—19 C.

´
R. Swisłocka et al. / Spectrochimica Acta Part A 100 (2013) 31–40
39
Table 7
Results of statistical analysis performed for spectral (i.e. wavenumbers of selected bands) and biological data (percent growth inhibition) for vanillic acid and vanillinates
alkali metals.
Tested bacteria
Time
incubat.
Assignment
(correlation
coefficients)
MODEL (regression function)
P-Value
R-SQ ADJ.
R-SQ
Sign. F
E. coli
E. coli
24
48
8b (0.72)
18b (0.72)
18b (0.71)
ˇas(COO) (−0.8)
0.260·8b − 0.306·18b
0.122
0.191
0.015
0.131
0.998
0.747
0.999
0.749
7.65 × 10−5
8.51 × 10−6
0.158·18b − 0.174·ˇas(COO)
P. vulgaris
P. vulgaris
24
48
19b (−0.77)
−11.832·19b + 16.254·18b
0.699·8b − 0.667·ꢀas(COO)
0.034
0.033
0.046
0.061
0.964
0.705
0.991
0.739
4.41 × 10−3
5.60 × 10−4
18b (0.78)
8b (0.81)
ꢀas(COO) (−0.81)
P. aeruginosa
P. aeruginosa
24
48
8b (0.65)
18b (0.92)
ıas(CH3) (0.76)
ꢀC (OH)ar (−0.60)
0.111·8b − 0.083·18b
0.198
0.474
0.079
0.164
0.999
0.749
0.999
0.749
1.02 × 10−5
1.27 × 10−5
0.209·ıas(CH3) − 0.181·ꢀC (OH)ar
B. subtilis
B. subtilis
24
48
8a (−0.37)
18b (0.38)
19a (0.63)
14 (0.49)
−0.968·8a + 1.480·18b
0.933·19a − 0.975·14
0.289
0.264
0.251
0.275
0.999
0.749
0.985
0.731
1.03 × 10−5
1.24 × 10−3
S. aureus
S. aureus
24
48
14 (0.87)
18b (0.73)
14 (0.80)
18b (0.87)
5.597·14 − 6.300·18b
−0.194·14 + 0.290·18b
0.723
0.729
0.958
0.945
0.980
0.725
0.994
0.743
1.83 × 10−3
2.72 × 10−4
C. albicans
C. albicans
24
48
8a (−0.83)
0.113·8a − 0.180·ˇas(COO)
0.126·19b − 0.073·ꢀas(COO)
0.010
0.094
0.001
0.030
0.999
0.749
0.999
0.750
5.21 × 10−6
2.30 × 10−7
ˇas(COO) (0.76)
19b (0.95)
ꢁs(COO) (−0.84)
In the multiple regression analysis the most important problem
is the choice of a proper set of two or more predictor variables
(highest intensity of the bands). The general rule is based on
the need to consider variables which have statistically significant
correlation with the predicted variable (percent of growth inhi-
bition) and at the same time they are not statistically connected
between themselves. This rule can be obeyed using analysis of cor-
relation coefficients matrix and the elimination of those predictors
which coefficients are:
Concluding there are models that are positively verified with
the above criteria (grey rows in Table 7).
It is important to mention that here only the simplest linear
statistical model was presented. Better solutions could be obtained
using more samples, additional analyses will be performed in the
nearest future.
4. Conclusion
In this work we have studied the relationship between molecu-
lar structure and biological activity of vanillic acid and alkali metal
vanillates.
•
nonsignificant for connections with predicted variable;
significant for connections between predictors.
•
The intensities and wavenumbers of the majority of aromatic
system bands in IR and Raman spectra increased in the case of
salts comparing to the free ligand, although for some of them the
intensity and wavenumbers decreased. Calculations showed a good
correlation between experimental and theoretical IR spectra. The
correlation coefficients for IR spectra are: 0.9975 (acid), 0.9973 (Li),
0.9973 (Na) and 0.9979 (K).
Alkali metal ions influence on the bond lengths and angles in
the carboxylic group and they influence the methoxy and hydroxy
group geometry.
In fact there exist a lot of econometric methods, but quite often
the method of trials and computers simulations should be used. If to
much predictors are included for a small number of samples it can
be wrongly concluded that all predictors are significant In our case
analysis of correlation coefficients between predictors for highest
intensity of the bands and percent of growth inhibition was per-
formed. In each case we choose two predictors that were the best
in the sense of the maximum correlation but not strongly connected
between themselves.
Good agreement was obtained between the experimental and
calculated data of 13C NMR is noted, but corresponding values of
correlation coefficient of 1H NMR are lower.
Table 7 displays models with two predictors for each for the
tested bacteria at 24 and 48 h of incubation separately. Models were
verified taking into account:
The calculated aromaticity indices (HOMA, Aj, BAC and I6—Bird’s
index) for vanillic acid and vanillates lithium, sodium and potas-
sium increase in series acid → Li → Na → K. Obtained results of total
energy of studied molecules decreases along the series: vanillic
acid → potassium vanillate. The values of dipole moment show
increasing tendency in this series. The aromaticity of salts is higher
for the acid molecule compared to salts: HOMA, Aj, BAC and I6
aromaticity indices.
1) Values of coefficients of linear determination (R2) for models: all
are about 99% which means that variability of growth inhibition
percent could be explained by the regression models in 99%.
2) Analysis of variance with test F which says if jointed effect of
the two variables on predicted values is statistically significant.
Here all the significance levels F are small enough that the action
of two predictors is strong.
Alkali metal vanillinate showed strong activity against S. aureus
and B. subtilis after 48 h of incubation and E. coli after 24 h of
incubation. Caesium and rubidium vanillinates show relatively
3) Verification of significance of regression coefficients with p-
values. In our models the significance level was 0.1.

´
40
R. Swisłocka et al. / Spectrochimica Acta Part A 100 (2013) 31–40
´
higher antimicrobial properties than the other tested alkali metal
salts.
[3] M. Kalinowska, R. Swisłocka, M.H. Borawska, J. Piekut, W. Lewandowski, Spec-
trochim. Acta A 70 (2008) 126–135.
[4] P. Delaquis, K. Stanich, P. Toivonen, J. Food Prot. 68 (7) (2005) 1472–1476.
[5] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman,
G. Scalmani, V. Barone, B. Mennucci, G.A. Petersson, H. Nakatsuji, M. Caricato,
X. Li, H.P. Hratchian, A.F. Izmaylov, J. Bloino, G. Zheng, J.L. Sonnenberg, M. Hada,
M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda,
O. Kitao, H. Nakai, T. Vreven, J.A. Montgomery Jr., J.E. Peralta, F. Ogliaro, M.
Bearpark, J.J. Heyd, E. Brothers, K.N. Kudin, V.N. Staroverov, R. Kobayashi, J.
Normand, K. Raghavachari, A. Rendell, J.C. Burant, S.S. Iyengar, J. Tomasi, M.
Cossi, N. Rega, J.M. Millam, M. Klene, J.E. Knox, J.B. Cross, V. Bakken, C. Adamo,
J. Jaramillo, R. Gomperts, R.E. Stratmann, O. Yazyev, A.J. Austin, R. Cammi, C.
Pomelli, J.W. Ochterski, R.L. Martin, K. Morokuma, V.G. Zakrzewski, G.A. Voth,
P. Salvador, J.J. Dannenberg, S. Dapprich, A.D. Daniels, Ö. Farkas, J.B. Foresman,
J.V. Ortiz, J. Cioslowski, D.J. Fox, Gaussian 09, Revision A.11.2, Gaussian, Inc.,
Wallingford, CT, 2009.
Satisfactory correlations between molecular structure of vanil-
lates and their biological activity have been revealed. The use
of spectral data for structure describing is sufficient to create
statistically important regression models that characterize and
allow to predict antimicrobial properties of structure similar com-
pounds. High determination coefficients obtained for regression
functions show good dependency between parameters used in
these equations.
Acknowledgements
[6] J. Howard, D.M. Whitcombe, Methods in Molecular Biology, vol. 46,
Diagnostic Bacteriology Protocols, Humana Press, Inc., Totowa, NJ,
1995.
[7] A. Wilcks, N. Jayaswal, D. Lereclus, L. Andrup, Microbiology 1 (1998) 1263.
[8] STATISTICA 9.1 program, StatSoft, Poland.
This work was supported by grant no. N N312 111838.
The authors are grateful to Izabela Swie˛cicka for allowing the
implementation of research in the Department of Microbiology,
University of Bialystok.
´
[9] G. Varsányi, Assignments for Vibrational Spectra of 700 Benzene Derivatives,
Akademiai Kiado, Budapest, Hungary, 1973.
´
[10] M. Kalinowska, R. Swisłocka, Z. Rza˛czyn´ ska, J. Sienkiewicz, W. Lewandowski, J.
References
Phys. Org. Chem. 23 (2010) 37–47.
[11] B. Kozlevcar, D. Odlazek, A. Golobic, A. Pevec, P. Strauch, P. Segedin, Polyhedron
25 (2006) 1161–1166.
´
[1] P. Koczon´ , J. Piekut, M. Borawska, R. Swisłocka, W. Lewandowski, Spectrochim.
Acta A 61 (2005) 1917–1922.
[2] P. Koczon´ , J. Piekut, M. Borawska, R. Swisłocka, W. Lewandowski, Anal. Bioanal.
´
Chem. 384 (2006) 302–308.