Characterization of some amino acid derivatives of benzoyl isothiocyanate: Crystal structures and theoretical prediction of their reactivity

Article abstract of DOI:10.1016/j.molstruc.2015.05.053

The reaction of benzoyl isothiocyanate with l-serine, l-proline, d-methionine and l-alanine gave 2-[(benzoylcarbamothioyl)amino]-3-hydroxypropanoic acid (I), 1-(benzoylcarbamothioyl)pyrrolidine-2-carboxylic acid (II), 2-[(benzoylcarbamothioyl)amino]-4-(methylsulfanyl)butanoic acid (III) and 2-[(benzoylcarbamothioyl)amino]propanoic acid (IV), respectively. The compounds have been characterized by IR, NMR, microanalyses and mass spectrometry. The crystal structures of all the compounds have also been discussed. Compound II showed rotamers in solution. DFT calculations of the frontier orbitals of the compounds have been carried out to ascertain the groups that contribute to the HOMO and LUMO, and to study their contribution to the reactivity of these compounds. The calculations indicated that the carboxylic acid group in these compounds is unreactive hence making the conversion to benzimidazoles via cyclization on the carboxylic acids impractical. This has been further confirmed by the reaction of compounds I-IV, respectively, with o-phenylene diamine which was unsuccessful but gave compound V.

Full text of DOI:10.1016/j.molstruc.2015.05.053

Journal of Molecular Structure 1099 (2015) 38e48
Contents lists available at ScienceDirect
Journal of Molecular Structure
journal homepage: http://www.elsevier.com/locate/molstruc
Characterization of some amino acid derivatives of benzoyl
isothiocyanate: Crystal structures and theoretical prediction
of their reactivity
,
*
Felix Odame a, Eric C. Hosten a, Richard Betz a, Kevin Lobb b, Zenixole R. Tshentu a
a Department of Chemistry, Nelson Mandela Metropolitan University, P.O. Box 77000, Port Elizabeth 6031, South Africa
b Department of Chemistry, Rhodes University, P.O. Box 94, Grahamstown 6140, South Africa
a r t i c l e i n f o
a b s t r a c t
Article history:
The reaction of benzoyl isothiocyanate with L-serine, L-proline, D-methionine and L-alanine gave 2-
Received 17 September 2014
Received in revised form
21 May 2015
Accepted 22 May 2015
Available online 17 June 2015
[(benzoylcarbamothioyl)amino]-3-hydroxypropanoic acid (I), 1-(benzoylcarbamothioyl)pyrrolidine-2-
carboxylic acid (II), 2-[(benzoylcarbamothioyl)amino]-4-(methylsulfanyl)butanoic acid (III) and 2-
[(benzoylcarbamothioyl)amino]propanoic acid (IV), respectively. The compounds have been character-
ized by IR, NMR, microanalyses and mass spectrometry. The crystal structures of all the compounds have
also been discussed. Compound II showed rotamers in solution. DFT calculations of the frontier orbitals
of the compounds have been carried out to ascertain the groups that contribute to the HOMO and LUMO,
and to study their contribution to the reactivity of these compounds. The calculations indicated that the
carboxylic acid group in these compounds is unreactive hence making the conversion to benzimidazoles
via cyclization on the carboxylic acids impractical. This has been further confirmed by the reaction of
compounds IeIV, respectively, with o-phenylene diamine which was unsuccessful but gave compound V.
© 2015 Elsevier B.V. All rights reserved.
Keywords:
Benzoyl isothiocyanate
Amino acids
Crystal structures
Frontier orbitals
1. Introduction
et al. [4]. Pyrazole acyl thiourea derivatives have also been syn-
thesized from monomethylhydrazine (or phenylhydrazine) and
Amino acid derivatives of benzoyl isothiocyanate can be used as
intermediates to construct other biologically active molecules with
a thiourea backbone, and in this case the title compounds were
envisaged to be converted to benzimidazoles by reacting the car-
boxylic acid with 1,2-diaminobenzene. The isothiocyanates can be
generated by a reaction of the acylated intermediates with
ammonium thiocyanate in the presence of polyethylene gylcol 400
(PEG 400) [1], and then reacted with amines to generate the thio-
urea derivatives. Thiourea derivatives have been synthesized
through several ways. For example, ethyl isothiocyanates and aro-
matic amines were mixed and stirred at room temperature in
acetone for 15 h to give the corresponding thioureas in high yields
[2]. A simple and efficient method for the synthesis of thiourea
derivatives in high purity and high yield has been reported using
tetrabutyl ammonium bromide (TBAB) as a phase transfer catalyst
ethyl acetoacetate [1]. N-(5-Aryl-2-furonyl) thiourea derivatives
containing substituted pyrimidine rings have been synthesized in
good yield using PEGe400 as a phase transfer catalyst under ul-
trasonic irradiation [5]. Fluorinated pyrazoles, benzene sulfonyl-
urea and thiourea derivatives as well as their cyclic
sulfonylthioureas have been prepared by the reaction of bromi-
nated trifluoromethyl diketones with isocyanates and iso-
thiocyanates [6]. Benzimidazoles conjugated to thioureas have
been synthesized by the refluxing of isothiocyanates and benz-
imidazoles in dimethyl formamide [7].
Versatile and expeditious syntheses of taurine-derived thio-
ureas, ureas, and guanidines using taurine isothiocyanate as the key
intermediate have also been reported [8]. The thioureas were ob-
tained by a one-pot two-step procedure starting from taurine by
the isothiocyanation reaction with thiophosgene in aqueous
tetrahydrofuran, followed by coupling with aliphatic and aromatic
amines. Desulfurization of the thiourea derivatives with mercu-
ry(II) oxide gave either taurine-containing ureas or guanidine [8].
Urea and thiourea derivatives of diphenylphoshoramidate may be
accessed by the reaction of diphenylphosphoramidate with
[3].
A
series of thiourea derivatives containing the
quinazolinee4(3H) framework have been synthesized by Saeed
* Corresponding author.
E-mail address: zenixole.tshentu@nmmu.ac.za (Z.R. Tshentu).
http://dx.doi.org/10.1016/j.molstruc.2015.05.053
0022-2860/© 2015 Elsevier B.V. All rights reserved.

F. Odame et al. / Journal of Molecular Structure 1099 (2015) 38e48
39
aromatic isocyanates and isothiocyanates in tetrahydrofuran in the
2.2. General method for the synthesis of the compounds
presence of triethylamine [9]. Multifunctional thioureas bearing a
variety of functional groups at a position remote from the thiourea
moiety have been synthesized via ruthenium catalyzed Huisgen
cycloaddition [10]. The aspartic acid dimethyl ester and alanine
methyl ester derivatives of benzoyl isothiocyanate have also been
synthesized in acetone from the corresponding thione [11,12]. A
methoxy substituted derivative of benzoyl isothiocyanate has been
accessed by boiling the thione directly in methanol [13]. N-[Ben-
zoylamino)thioxomethyl]amino acid derivatives have been pre-
pared by the reaction of benzoyl isothiocyanate with various amino
acids in acetone, namely, histidine, alanine, phenylalanine, serine
and cysteine, however the alanine and serine derivatives were
characterised by NMR only [14]. The cadmium(II) and zinc(II)
complexes of the phenylalanine derivatives have also been re-
ported and characterized by IR, NMR, and microanalysis [14]. The
cobalt(II), copper(II) and nickel(II) complexes of the aspartic acid,
glutamic acid, methionine, leucine and tryptophan derivatives of
benzoyl isothiocyanates have been synthesized and tested for their
antibacterial activity [15]. Some thiourea derivatives of various
amines have also been synthesized in acetone and tested for their
anti-amoebic properties [16]. Cyclohexanecarbonyl isothiocyanate
has been reacted with various amines in acetone, with a slightly
different work-up procedure which involved the addition of 0.1 N
hydrochloric acid to the mother liquor before filtration, and this
was reported to give higher yields of between 86 and 93% [17].
Benzothiazole derivatives of various carbonyl isothiocyanates have
been synthesized in acetone in the presence of 3% tetrabutyl
ammonium bromide (TBAB) [18].
Ammonium thiocyanate (0.10 mol, 7.6 g) was dissolved in
100 mL of acetone. Benzoyl chloride (0.10 mol, 11.3 mL) was added
followed by heating under reflux at 100e120 ꢀC for 2 h. The product
was filtered and the respective amino acid was added to the filtrate
and refluxed at 100e120 ꢀC for 6 h 20 mL of water was then added
and the mixture was refluxed for a further 2 h. The reaction mixture
was extracted with diethyl ether, and the solvent removed via ro-
tary evaporation followed by drying the compound in a high vac-
uum. GCeMass spectra were recorded for the synthesized
compounds and they showed molecular ion (Mþ) peaks, in addition
to others, which confirmed the formation the products.
2.2.1. 2-[(Benzoylcarbamothioyl)amino]-3-hydroxypropanoic acid
(I)
The product was recrystallized and obtained as a colourless solid
from acetone:water (4:1). Yield ¼ 71%, Mp ¼ 163e165 ꢀC. 1H NMR
(ppm): 11.43 (s, N(2)eH), 11.49 (s, 1H, O(2)eH), 7.96 (d, 2H, C(12)e
H, C(16)eH)), 7.65 (m, 1H, C(14)eH), 7.53 (m, 2H, C(13)eH, and
C(15)eH), 5.34 (br, 1H, N(1)eH), 4.94 (t, 1H, C(3)eH), 3.88 (s, 2H,
C(5)eH). 13C NMR (ppm): 180.3 (C(2)eS(1), 170.7 (C(4)eO(3)), 168.3
(C(1)e(O(1)), 128.4 (C(13) and C(15)), 128.5 (C(12) and C(16)), 132.1
(C(14), 133.0C(11)), 60.5 (C(6)), 60.3 (C(3). IR (nmax, cmꢂ1): 3229
(NeH), 2980 (aliphatic CeH), 1725 (C]S), 1654 (C]O), 1509 (aro-
matic C]C), 1164 (CeN). Anal. calcd. for C11H12N2O4S.H2O: C, 46.15;
H, 4.93; S, 11.20; N, 9.78. Found: C, 46.92; H, 4.87; S, 10.67; N, 9.76.
LRMS (m/z, Mþ): Found for C11H12N2O3S ¼ 268.00, Expected
mass ¼ 268.29.
Compounds IeIV can be useful intermediates for further syn-
thesis of biologically relevant compounds, and in this case benz-
imidazole formation via cyclization on the amino acid carboxylic
acid using polyphosphoric acid in toluene at 165 ꢀC. It is therefore
important to study their chemical properties in order to understand
their reactivity since the attempted cyclization reaction was not
successful but gave compound V. The computational studies of the
frontier orbitals have been carried out and the HOMO and LUMO
gaps discussed. Compounds IeIV have also been characterized by
single crystal XRD. The NMR analysis (1H, 1D NOESY, 2D NOESY, 13C,
HMBC, HSQC) of compound II showed the formation of rotomers in
solution which was not observed in the solid state.
2.2.2. 1-(Benzoylcarbamothioyl)pyrrolidine-2-carboxylic acid (II)
The product recrystallized as a white solid from acetone:water
(4:1). Yield ¼ 75%, Mp ¼ 114e117 ꢀC. 1H NMR (ppm): 10.90 (s, 1H,
O(3)eH), 7.94 (d, 1H, C(12)eH), 7.85 (d, 1H, C(16)eH), 7.58 (m, 1H,
C(14)eH), 7.50 (m, 2H, C(13)eH and C(15)eH), 4.68 (t, 1H, C(21)e
H)), 3.68 (m, 2H, C(24)eH), 2.02 (m,2H, C(23)eH), 1.99 (m, 2H,
13
C(22)eH). C NMR (ppm): 179.1 (C(2)eS(1)), 171.9 (C(30)eO(20)),
171.2 (C(3)eO(2)), 164.0 (C(1)eO(1)), 132.9 (C(11), 132.2 (C(14),
128.4 (C(12) and C(16)), 128.2 (C(13) and C(15)), 65.4 (C(21), 62.7
(C(24), 61.7 (C(240), 31.0 (C(23), 29.4 (C(230), 24.19 (C(22), 22.94
(C(22). IR (nmax, cmꢂ1): 3270 (NeH), 2987 (aliphatic CeH), 1730
(C]S), 1659 (C]O), 1491 (aromatic C]C), 1182 (CeN). Anal. calcd.
for C13H14N2O3S.H2O: C, 52.69; H, 5.44; S, 10.83; N, 9.45. Found: C,
52.55; H, 5.64; S, 9.80; N, 9.11. LRMS (m/z, Mþ): Found for
C13H14N2O3S ¼ 278.40, Expected mass ¼ 278.33.
2. Experimental
2.1. Reagent and instrumentation
Analytical grade reagents and solvents for synthesis such as
ammonium thiocyanate, L-serine, D-methionine, L-proline and L-
2.2.3. 2-[(Benzoylcarbamothioyl)amino]-4-(methylsulfanyl)
butanoic acid (III)
alanine were obtained from Sigma Aldrich (USA) whilst acetone
and benzoyl chloride were obtained from Merck Chemicals (SA).
The chemicals were used as received (i.e. without further purifi-
cation). 1H NMR and 13C NMR spectra were recorded on a Bruker
Avance AV 400 MHz spectrometer operating at 400 MHz for 1H and
100 MHz for 13C using DMSO-d6 as solvent and tetramethylsilane as
internal standard. Chemical shifts are expressed in ppm. FTeIR
spectra were recorded on a Bruker Platinum ATR Spectrophotom-
eter Tensor 27. Elemental analyses were performed using a Vario
Elementar Microcube ELIII. Melting points were obtained using a
Stuart Lasec SMP30 whilst the masses were determined using an
Agilent 7890A GC System connected to a 5975C VLeMSC with
electron impact as the ionization mode and detection by a triple-
Axis detector.The GC was fitted with a 30 m ꢁ 0.25 mm x
The product recrystallized as a white solid from acetone:water
(4:1). Yield ¼ 77%, Mp ¼ 123e125 ꢀC. 1H NMR (ppm): 11.54 (s, 1H,
O(2)eH), 11.28 (s,1H, N(1)eH), 7.96 (m, 2H, C(12)eH and C(16)eH),
7.65 (m, 1H, C(14)eH), 7.52 (d, 2H, C(13)eH and C(15)eH), 5.03 (1H,
C(3)eH), 2.26 (m, 2H, C(7)eH), 2.17 (2H, C(5)eH),1,99 (3H, C(6)eH).
13C NMR (ppm): 180.5 (C(2)eS(1)), 171.8 (C(4)eO(3)), 168.4 (C(1)e
O(1),133.1 (C(11)), 132.1 (C(14)), 128.6 (C(12) and C(16)), 128.2
(C(13) and C(15)), 66.6 (C(3)), 30.5 (C(7)), 29.2 (C(5)), 14.6 (C(6)). IR
(
nmax, cmꢂ1): 3282 (NeH), 3202 (NeH), 2914 (aliphatic CeH), 1715
(C]S), 1664 (C]O), 1519 (aromatic C]C), 1193 (CeN) Anal. calcd.
for C13H16N2O3S2: C, 49.98; H, 5.16; S, 20.53; N, 8.97. Found: C,
50.40; H, 5.52; S, 20.03; N, 8.97. LRMS (m/z, Mþ): Found for
C13H16N2O3S2 ¼ 312.90, Expected mass ¼ 312.41.
0.25
m
m DB-5 capillary column. Helium was used as carrier gas at a
2.2.4. 2-[(Benzoylcarbamothioyl)amino]propanoic acid (IV)
The product recrystallized as a yellow solid from acetone:water
(4:1). Yield ¼ 65%, Mp ¼ 161e163 ꢀC 1H NMR (ppm): 11.51 (s, 1H,
flow rate of 1.6 mL minꢂ1 with an average velocity of 30.2 cm sꢂ1
and a pressure of 63.7 KPa.

40
F. Odame et al. / Journal of Molecular Structure 1099 (2015) 38e48
O(2)eH), 11.30 (s, 1H, N(1)eH.)), 7.93 (d, 2H, C(12)eH, and C(16)e
H), 7.64 (m, 1H, C(14)eH), 7.52 (m, 2H, C(13)eH and C(15)eH), 4.85
((q, 1H, C(3)eH), 1.49 (d, 3H, C(5)eH). 13C NMR (ppm): 13C NMR
(ppm): 179.9 (C(2)eS(1)), 172.9 (C(4)eO(3)), 168.5 (C(1)e(O1)),
133.0C(11), 132.1C(14), 128.7 (C(12) and C(16), 128.4 (C(13) and
C(15)), 63.1 (C(3)eH), 17.2C(5). IR (nmax, cmꢂ1): 3384 (NeH), 2992
(aliphatic CeH), 1726 (C]S), 1678 (C]O), 1489 (aromatic C]C),
1196 (CeN). Anal. calcd. for C11H12N2O3S: C, 52.37; H 4.79; S 12.71;
N 11.10. Found: C, 52.59; H 5.11; S 12.56 N; 11.07. LRMS (m/z, Mþ):
Found for C11H12N2O3S ¼ 252.00, Expected mass ¼ 252.29.
3. Results and discussion
3.1. Synthesis and characterization
The syntheses of compounds I, III and IV were readily achieved
by the reaction of benzoyl isothiocyanate with the different amino
acids (Scheme 1). Scheme 2 gives the reaction path for compound
II. The lone pair of electrons on the nitrogen atom of the amino acid
attacks the carbon of the thione. This shifts the electron density
onto the nitrogen of the isothiocyanate allowing it to abstract a
proton from the amine of the amino acid. The FTIR, 1H NMR and 13
C
NMR confirmed the formation of the amino acid derivatives.
Compound V (Schemes 1 and 2) was accessed by the reaction of
compounds IeIV, respectively, with o-phenylene diamine with the
diamine attacking the carbonyl close to the benzene ring (benzoyl
group) instead of the carboxylic acid to form VI.
2.2.5. 2-Phenyl-1H-benzimidazole (V)
15 g of polyphosphoric acid, in 5 mL of toluene, was heated at
120 ꢀC for 30 min. Compounds I, II, III and IV (0.20 mol) and o-
phenylenediamine (0.2 mol) were added and the mixture was
heated at 165 ꢀC for 6 h. After cooling, 20 mL of water was added
with stirring and then sodium hydrogen carbonate was added until
effervescence ceased. The dark brown precipitate obtained was re-
dissolved in little methanol and placed on the silica gel column and
eluted with methanol:ethyl acetate (1:1). The product was recrys-
tallized from ethanol:toluene (1:1) and obtained as a brown solid.
Yield ¼ 55e68%, Mp ¼ 240e242 ꢀC. 1H NMR (ppm): 8.20 (m, 2H),
7.61 (m, 2H), 7.56 (m, 2H), 7.48 (m, 1H), 7.21 (m, 2H), 13C NMR
(ppm): 151.2 (C]N), 130.0 (PheH), 129.9 (PheH), 128.9 (PheH),
126.4 (PheH), 122.1 (PheH), IR (nmax, cmꢂ1): 3048 (NeH), 2961
(CeH), 2850 (CeH), 1539 (C]N), 1461 (C]C), 1443 (CeN). Anal.
Calcd for C13H10N2: C, 80.39; H, 5.19; N, 14.42. Found: C, 80.29; H,
The FTIR spectra for compounds IeV are provided in Figs. S1eS5
in the Supplementary Information. The 1H NMR spectra for com-
pounds 1eV and 13C NMR spectra for compounds 1, III,1V and V are
also provided in Figs. S6eS15 in the Supplementary Information.
The nitrogen protons which do not appear in the 1H NMR spectra of
compounds II, III and IV exchange with the residual water mole-
cules in DMSO hence are undetectable. The 13C NMR spectrum of
compound II (Fig. 1) showed six methylene groups instead of three
which occurred at 54.33, 51.72, 30.96, 29.44, 24.19, and 22.94. This
was also confirmed in the DEPT spectrum (Fig. 2). This showed that
compound II possibly exists as two rotamers in solution. Fig. S16
gives the 1He1H COSY for compound I. Figs. S17eS19 present the
1He1H COSY, 1D NOESY and HMBC spectra of compound II, and
they showed further evidence of formation of rotamers. The two
rotamers in solution are formed by restricted rotation around the
nitrogen of the proline, specifically around the N(2)-C(2) bond,
making the carbon signals for all the four carbon atoms in the
proline ring double. This also gives rise to two carbonyl signals for
the proline part of compound II as evident in the CNMR spectrum.
The 2D NOESY spectrum (Fig. 3) showed an exchange between the
signals at 4.42 and 4.62 ppm, and both signals are attached to the
same carbon that changes position as indicated in the HSQC spec-
trum (Fig. 4).
5.32; N, 14.64. Found for
C
13H10N2
¼
194.10, Expected
mass ¼ 194.23.
2.3. X-ray crystallography
X-ray diffraction analyses of IeIV were performed at 200 K using
a Bruker Kappa Apex II diffractometer with monochromated MoK
radiation (
¼ 0.71073 Å). APEXII [19] was used for data collection
a
l
and SAINT [19] for cell refinement and data reduction. The struc-
tures were solved by direct methods using SHELXS-2013 [20] and
refined by least-squares procedures using SHELXL-2013 [20] with
SHELXLE [21] as the graphical interface. All non-hydrogen atoms
were refined anisotropically. Carbon-bound H atoms were placed in
calculated positions (CeH 0.95 Å for aromatic carbon atoms and
CeH 0.99 Å for methylene groups) and were included in the
refinement in the riding model approximation, with Uiso(H) set to
1.2Ueq(C). The H atoms of the methyl groups were allowed to rotate
with a fixed angle around the CeC bond to best fit the experimental
electron density (HFIX 137 in the SHELX program suite [19]), with
Uiso(H) set to 1.5Ueq(C). The H atom of the hydroxyl group was
allowed to rotate with a fixed angle around the CeO bond to best fit
the experimental electron density (HFIX 147 in the SHELX program
suite [20]), with Uiso(H) set to 1.5Ueq(O). Nitrogen-bound H atoms
were located on a difference Fourier map and refined freely. Data
were corrected for absorption effects using the numerical method
implemented in SADABS [19].
3.2. Crystal structures
Compounds I, II and III were recrystallized from acetone:water
(4:1) as colourless crystals, whilst compound IV was recrystallized
from acetone:water (4:1) as yellow crystals. The crystallographic
data, selected bond lengths, bond angles and torsion angles for the
structures of IeIV are provided in Tables 1 and 2. Compounds I and
II crystallized in orthorhomic space group P212121, whilst com-
pound III crystallized in monoclinic space group P21 and compound
IV in orthorhomic space group C2221. The absence of rotamers in
the solid state was confimed during the refinement procedure by
means of the Flack parameter in each case.
The ORTEP diagram of compound I is presented in Fig. 5. The
experimental bond distance of S(1)eC(2) ¼ 1.670(2) Å in compound
I is consistent with the C]S of the thione [22], whilst the C(1)eO(1)
and C(4)eO(3) bond distances of 1.233(2) Å and 1.202(2) Å show
the presence of the carbonyl bond of the amide and the carbonyl of
the carboxylic acid, respectively. The bond distance of C(1)e
N(1) ¼ 1.372(2) Å is indicative of the CeN bond of the amide. The
O(2)eC(4)eO(3) bond angle of 124.8(2)ꢀ is consistent with the
bond angle around the sp2 carbon of the carboxylic acid whilst the
S(1)eC(2)eN(2) bond angle of 125.2(1)ꢀ indicated the presence of
the sp2 carbon of the thione. The O(1)eC(1)eN(1) bond angle of
121.6(1)ꢀ also showed the presence of the sp2 carbon of the amide.
The C(2)eN(1)eC(1)eO(1) torsion angle of 6.1(2)ꢀ confirmed the
2.4. Computational studies
The calculations were carried out using Gaussian 09 program.
Molecular geometries of the singlet ground state of all the com-
pounds were fully optimised in the gas phase at the density func-
tional theory (DFT) level of theory using B3LYP/6-31þþG. For each
compound a frequency calculation was carried out to ensure that
the optimized molecular structure corresponded to a minimum,
thus only positive frequencies were expected. The structures were
generated using Chemcraft.

F. Odame et al. / Journal of Molecular Structure 1099 (2015) 38e48
41
O
S
O
S
R
H
R
C
O
C
O
H
N
H
N
H
(I-IV)
N
acetone
N
H
OH
H
OH
PPA or HCl
1,2-diaminobenzene
(VI)
O
S
(V)
H
N
R
H
N
N
H
N
C
H
N
H
N
Scheme 1. Synthesis scheme for compounds I and IIIeV.
HO
O
O
S
HO
O
S
O
N
N
H
N
HN
acetone
PPA
o
r HCl
1,2-diaminobenzene
O
N
H
N
S
HN
N
N
H
N
(V)
(VI)
Scheme 2. Synthesis scheme for compounds II and V.
rigidity around the carbonyl of the amide whilst the C(3)eN(2)e
C(2)eS(1) torsion angle of ꢂ7.3(2)ꢀ confirmed the rigidity around
the thione.
of the amide O(1) and the nitrogen from the amino acid N(2)
(Fig. S18). The carbonyl (O(1)) also forms an intermolecular
hydrogen bond with the hydroxyl group O(4) of the adjoining
molecule which also forms a bond with the water molecule. The
There is an intramolecular hydrogen bond between the carbonyl
Fig. 1. 13C NMR spectrum of compound II.

42
F. Odame et al. / Journal of Molecular Structure 1099 (2015) 38e48
Fig. 2. DEPT spectrum of compound II.
Fig. 3. 2D NOESY spectrum of compound II.
thione S(1) also forms a bond with a water molecule. The packing of
compound I showed the presence of four molecules in the unit cell
(Fig. S19).
The structure of compound II is presented in Fig. 6. The bond
distances of O(2)eC(3) and C(1)eO(1) which were 1.201(2) Å and
1.229(2) Å are consistent with the carbonyl of the carboxylic acid
Fig. 4. HSQC spectrum of compound II.

F. Odame et al. / Journal of Molecular Structure 1099 (2015) 38e48
43
Table 1
Crystallographic data and structure refinement summary for compounds IeIV.
Property
Compound I
Compound II
Compound III
Compound IV
CCDC number
Empirical formula
Formula weight
Crystal system
Space group
a (Å)
1014079
1014083
C13H14N2O3S,H2O
296.35
Orthorhombic
P212121
8.8347(5)
11.8182(7)
13.3896(8)
90
1014284
C13H16N2O3S2
312.42
Monoclinic
P21
5.1663(2)
1.4170(5)
12.7108(5)
90
96.843(1)
90
1014322
C11H12N2O3S
252.30
Orthorhombic
C2221
8.2584(3)
21.7331(8)
13.4905(6)
90
C11H12N2O4S,H2O
286.31
Orthorhombic
P212121
7.0613(3)
11.0752(4)
16.5713(7)
90
b (Å)
c (Å)
a
b
g
(ꢀ)
(ꢀ)
(ꢀ)
90
90
90
90
90
90
V [Å^3]
Z
1295.96(9)
4
1398.01(14)
4
744.39(5)
2
2421.28(17)
8
D(calc) (g/cm^3)
1.467
1.408
1.394
1.384
m
(MoKa) (/mm)
0.268
0.246
0.365
0.265
F(000)
600
624
328
1056
Crystal size (mm)
Temperature (K)
Radiation (Å)
0.26 ꢁ 0.39 ꢁ 0 0.52
0.47 ꢁ 0.48 ꢁ 0.56
0.16 ꢁ 0.37 ꢁ 0.49
0.23 ꢁ 0.53 ꢁ 0.72
200
200
200
200
0.71073
2.2e28.3
ꢂ7:9; ꢂ14:9; ꢂ22:20
6893, 3088
0.010
0.71073
2.8e28.4
ꢂ1:8; ꢂ15:13; ꢂ17:17
10872, 3334
0.014
0.71073
2.4e28.3
ꢂ6:4; ꢂ14:15; ꢂ16:16
6845, 3344
0.017
0.71073
1.9e28.3
ꢂ11:9; ꢂ25:28; ꢂ15:17
6792, 3009
0.014
q
MineMax (ꢀ)
Dataset
Tot., Uniq. Data
Rint
[I > 2.0 sigma(I)]
2991
3224
3176
2879
Nref, Npar
R
3088,180
0.0227
3334, 195
0.0240
3344,185
0.0252
3009, 164
0.0238
wR2
0.0643
0.0650
0.0654
0.0659
S
1.09
1.03
1.03
1.04
Max. and Av. shift/error
Min. residual. density. [e/Å^3]
Max. residual. density. [e/Å^3
Flack parameter
0.00, 0.00
ꢂ0.17
0.00, 0.00
ꢂ0.17
0.000,0.00
ꢂ0.20
0.000,0.00
ꢂ0.14
025
0.24
0.26
0.25
0.020(13)
0.002(14)
0.00(2)
0.012(19)
Table 2
Selected bond length (Å), bond angles and torsion angles of compounds IeIV.
Bond lengths
I
II
III
IV
Expt
Expt
Expt
Expt
S(1)eC(2)
O(1)eC(1)
N(1)eC(2)
N(1)eC(1)
N(2)eC(2)
C(1)eC(11)
S(2)eC(7)
S(2)eC(6)
1.670(2)
1.232(2)
1.399(2)
1.372(2)
1.331(2)
1.489(2)
-
1.661(2)
1.229(2)
1.420(2)
1.367(2)
1.326(2)
1.488(2)
-
1.676(2)
1.226(2)
1.385(3)
1.383(3)
1.331(3)
1.486(3)
1.793(3)
1.801(2)
1.677(1)
1.222(2)
1.323(2)
1.386(2)
1.487(2)
1.487(2)
-
-
-
-
Bond angles
I
II
III
IV
O(1)eC(1)eN(1)
O(1)eC(1)eC(11)
S(1)eC(2)eN(1)
S(1)eC(2)eN(2)
N(1)eC(1)eC(11)
N(1)eC(2)eN(2)
C(1)eN(1)eC(2)
121.8(1)
121.6(1)
118.3(1)
125.2(1)
116.6(1)
116.6(1)
127.8(1)
121.8(1)
120.9(1)
119.2(1)
124.4(1)
117.3(1)
116.3(1)
123.2(1)
121.5(2)
121.3(2)
120.9(2)
122.3(2)
117.2(2)
116.8(2)
126.9(2)
122.0(1)
121.8(1)
119.6(1)
123.5(1)
116.2(1)
116.8(1)
127.4(1)
Fig. 5. An ORTEP view of compound I showing 50% probability displacement ellipsoids
and the atom labelling.
compound I which was 1.372(2) Å. The bond distance of S(1)e
C(2) ¼ 1.661(2) Å showed the presence of the C]S of the thione and
was comparable to that of compound I (1.670(2) Å). The O(3)ꢂ
C(3)ꢂO(2) bond angle of 125.0(1)ꢀ is indicative of the sp2 carbon of
the carbonyl of the carboxylic acid and the N(1)eC(1)eO(1) bond
angle of 121.8(1)ꢀ also confirmed the presence of the sp2 carbon of
the carbonyl and these were comparable to the bond angles
observed in compound I. The S(1)eC(2)eN(1) bond angle of
119.2(1)ꢀ also indicated the presence of the sp2 carbon of the thi-
one. The C(2)eN(1)eC(1)eO(1) torsion angle of 7.1(2)ꢀ was
consistent with the rigidity in the amide whilst the C(21)eN(2)e
C(2)eS(1) torsion angle of 1.3(2)ꢀ confirmed the rigidity around the
thione. There exists an intermolecular hydrogen bond in compound
II between the carbonyl of the amide O(1) and a water molecule
which is also bonded to the carbonyl O(2) of the carboxylic acid of
Torsion angles
I
II
III
IV
N(1)eC(1)eC(11)eC(16)
C(2)eN(1)eC(1)eO(1)
C(1)eN(1)eC(2)eS(1)
C(3)eN(2)eC(2)eS(1)
O(1)eC(1)eC(11)eC(16)
C(1)eN(1)eC(2)eN(2)
O(1)eC(1)eC(11)eC(12)
ꢂ158.7(2)
6.1(2)
176.0(1)
ꢂ7.3(2)
1.0(2)
ꢂ21.1(2)
7.1(2)
121.5(1)
-
159.2(2)
ꢂ60.3(2)
ꢂ20.2(2)
ꢂ174.2(2)
3.5(3)
179.2(2)
6.1(3)
6.7(3)
0.4(3)
ꢂ15.8(3)
ꢂ6.0(3)
179.3(1)
2.8(2)
162.7(2)
ꢂ0.4(2)
ꢂ14.4(3)
ꢂ3.2(2)
ꢂ156.7(2)
ꢂ173.4(2)
and the carbonyl of the amide, respectively. This also compared
favourably with the carbonyl bond distances in compound I. The
bond distance of C(1)eN(1) ¼ 1.367(2) Å indicated the CeN bond of
the amide and it was comparable to the CeN bond distance in

44
F. Odame et al. / Journal of Molecular Structure 1099 (2015) 38e48
with those of compounds I and II (Table 2). There is an intra-
molecular bond between the carbonyl O(1) of the amide and the
nitrogen N(2) from the amino acid. There exists an intermolecular
hydrogen bond in compound III between the thione S(1) and the
hydroxyl O(2) of the adjoining molecule and also between the ni-
trogen N(1) of the amide and the carbonyl O(3) of the carboxylic
acid of the adjoining molecule (Fig. S22). The packing of compound
III shows two molecules in the unit cell (Fig. S23).
The ORTEP diagram of compound IV is presented in Fig. 8. The
bond distances and angles as well as the torsion angles were
comparable with those of compounds I-III (Table 2). There is an
intramolecular hydrogen bond in compound IV between the
carbonyl O(1) of the amide and the nitrogen N(2) from the amino
acid (Fig. S24). There also exists an intermolecular bond between
the thione S(1) and the hydroxyl O(2) of the adjoining molecule.
The packing of the molecules of compound IV in a unit cell shows
eight molecules in each unit cell (Fig. S25).
Fig. 6. An ORTEP view of compound II showing 50% probability displacement ellip-
soids and the atom labelling.
3.3. HOMO-LUMO analysis
In view of the fact that compounds IeIV are useful in-
termediates in accessing other complex molecules, it was impor-
tant to carry out some theoretical studies to characterise the nature
of these compounds in terms of chemical reactivity. In fact, a
chemical reaction that was attempted on compounds IeIV, i.e. a
cyclization reaction with 1,2-diaminobenzene at the carboxylic acid
site to form benzimidazoles, has been unsuccessful yielding rather
a benzimidazole through the diamine attacking the carbonyl of the
benzoyl group (compound V) instead of the carboxylic acid site.
This prompted us to carry out some computational studies in order
to predict the reactivity of these compounds. The highest occupied
molecular orbital and lowest unoccupied molecular orbital energy
separations, HOMOeLUMO energy gaps, have been evaluated by
DFT methods. The frontier orbitals are useful in predicting the most
Fig. 7. An ORTEP view of compound III showing 50% probability displacement ellip-
soids and the atom labelling.
reactive position in p-electron systems [23] and to explain several
types of reactions in
isothiocyanate.
a conjugated system like benzoyl
Benzoyl isothiocyanate is characterised by a small highest
occupied molecular orbital-lowest unoccupied molecular orbital
(HOMOeLUMO) separation, which is the result of a significant
degree of intramolecular charge transfer from the end capping
electron-donor groups through a
p-conjugated path [24]. The
HOMO and LUMO are the main orbitals that determine chemical
stability of the species [25]. The HOMO, which represents the
ability to donate an electron, is delocalised over the entire molecule
except the carbonyl and carbon of the thione whilst the LUMO
(which represent the ability to accept an electron) shows delocal-
ization of it orbitals over both the carbon atoms of the carbonyl and
the thione indicating that both are susceptible to attack by an
incoming group but because the carbonyl is sterically hindered the
carbon of the thione is the preferred site of attack and this attack is
Fig. 8. An ORTEP view of compound IV showing 50% probability displacement ellip-
soids and the atom labelling.
the same molecule and the hydroxyl O(3) of the adjoining molecule
(Fig. S20). The packing of the molecules of compound II in a unit cell
shows four molecules in each unit cell (Fig. S21).
The ORTEP diagram of compound III is presented in Fig. 7. The
bond lengths and angles as well as tosion angles compared well
Table 3
Summary of the HOMO-LUMO energies for compounds IeIV and the starting materials.
HOMO (kJ/mol)
LUMO (kJ/mol)
HOMOeLUMO gap (kJ/mol)
Benzoyl isothiocyanate
Compound I
Serine
Compound II
Proline
Compound III
Methionine
Compound IV
Alanine
ꢂ7I5.16
ꢂ675.93
ꢂ692.24
ꢂ616.00
ꢂ638.97
ꢂ564.17
ꢂ599.43
ꢂ651.52
ꢂ678.72
ꢂ498.40
ꢂ243.75
ꢂ85.64
ꢂ205.55
ꢂ74.83
ꢂ214.50
ꢂ91.34
ꢂ233.12
ꢂ86.35
216.76
432.18
606.60
410.45
564.14
349.67
508.09
418.40
592.37

F. Odame et al. / Journal of Molecular Structure 1099 (2015) 38e48
45
Fig. 9. The atomic orbitals compositions of the frontier molecular orbitals for benzoyl isothiocyanate.
Fig. 10. The atomic orbitals compositions of the frontier molecular orbital for compound I.
further stabilized by charge transfer through the benzene ring. The
energy of the HOMO is directly related to the ionization potential
whilst the energy of the LUMO is related to the electron affinity. The
energy difference between HOMO and LUMO orbitals, known as
the energy gap, determines the stability or reactivity of molecules
[26]. A narrow energy gap for benzoyl isothiocyanate is consistent
with its high reactivity. The energy gap is a critical parameter in
determining molecular electrical transport properties because it is
a measure of electron conductivity [27]. The hardness of a molecule
also corresponds to the gap between the HOMOeLUMO orbitals
[28]. Large HOMOꢂLUMO gap indicates high stability and resis-
tance to charge transfer, therefore, hard molecules have a large
HOMOꢂLUMO gap.
Fig. 11. The atomic orbitals compositions of the frontier molecular orbital for serine.
Table 3 gives the computed energy gap between the LUMO and
HOMO of compounds IeIV and benzoyl isothiocyanate as well as
Fig. 12. The atomic orbitals compositions of the frontier molecular orbitals for compound II.

46
F. Odame et al. / Journal of Molecular Structure 1099 (2015) 38e48
Fig. 13. The atomic orbitals compositions of the frontier molecular orbitals for proline.
Fig. 14. The atomic orbitals compositions of the frontier molecular orbitals for compound III.
the amino acids. The computed energy gap between the LUMO and
HOMO of benzoyl isothiocyanate is narrow which implies that it is a
soft molecule [29]. Compound I with a HOMOeLUMO gap of
432.18 kJ/mol is lower compared to that of serine which is
receive than compound IV which is 418.40 kJ/mol. From this in-
formation, it is not clear why the carboxylic acid site is less reactive
in the benzoyl isothiocyanates derivatives compared with the
amino acids since these molecules are must softer than the parent
amino acids. This prompted the study of the frontier orbitals.
Fig. 9 shows the HOMO and LUMO orbitals of benzoyl isothio-
cyanate. The HOMO is delocalised over the thione, nitrogen and the
ring with the oxygen atom making no contribution. whilst the
LUMO is delocalised over nearly the entire molecule. This suggests
that the entire molecule is involved in the acceptance of electrons
which makes this species very reactive.
606.60 kJ/mol, confirming that
L-serine is harder and less reactive
than compound I, even though their sites of reactivity differ.
L-
Proline with a HOMOeLUMO gap of 564.14 kJ/mol is harder but less
reactive than compared II which has a HOMOeLUMO gap of
410.45 kJ/mol, but the reactivity of these compounds are concen-
trated on different sites of the molecules. Compound III with a
HOMOeLUMO gap of 349.67 kJ/mol is harder but less reactive than
D
-methionine which has a HOMOeLUMO gap of 508.09 kJ/mol, and
The frontier orbitals of compound I (Fig. 10) shows that the
HOMO is mostly concentrated on the sulfur of the thione, the
alanine with a HOMOeLUMO gap of 592.37 kJ/mol is hard and less
Fig. 15. The atomic orbitals compositions of the frontier molecular orbitals for methionine.

F. Odame et al. / Journal of Molecular Structure 1099 (2015) 38e48
47
Fig. 16. The atomic orbitals compositions of the frontier molecular orbitals for compound IV.
nitrogen atoms and the phenyl ring, whilst the LUMO is largely
delocalised over the thione, the nitrogen atoms, the carbonyl and
over the benzene ring which aids in stabilising the compound
during charge transfer via delocalization over the benzene ring. The
carboxylic acid has no contribution to the frontier orbitals (LUMO)
hence is unreactive in the compound. This further confirmed the
inability to convert it to a benzimidazole by reacting it with 1,2-
diaminobenzene. The frontier orbitals of serine (Fig. 11) are
distributed over the entire molecule with most of the atoms mak-
ing a contribution. The carboxylic acid makes a little contribution to
the LUMO hence the cyclization should be possible on a free serine.
The frontier orbitals of compound II are shown in Fig. 12. The
HOMO is mostly concentrated on the sulfur of the thione and is
slightly delocalised on the nitrogen atoms whilst the LUMO is
largely delocalised over the entire molecule except the carboxylic
the acid and the methylene groups. Once again, the carboxylic acid
was unreactive because it makes no contribution to the frontier
orbitals which confirmed why it could not be converted to a
benzimidazole. The frontier orbitals of proline are delocalized over
nearly the entire molecule in LUMO confirming the possibility for
benzimidazole formation whilst in the HOMO the orbital is delo-
calized over the nitrogen and the surrounding carbon atoms
(Fig. 13).
the sulfur and methyl group attached to it, whilst its HOMO is
delocalized on the entire molecule except the carboxylic acid.
The frontier orbitals of compound IV are shown in Fig. 16. The
HOMO is mostly concentrated on the sulfur of the thione and is
slightly delocalised on the nitrogen and the carboxylic acid whilst
the LUMO is largely delocalised over the entire molecule except the
carboxylic acid and the methyl group. The carboxylic acid could not
be functionalized to a benzimidazole by the attachment of 1,2-
diamonobenzene due to the fact that the carboxylic acid has no
contribution to the LUMO, which would receive the incoming
electrons. The frontier orbitals of alanine in Fig. 17 shows delocal-
ization over the entire molecule in the LUMO whilst the HOMO is
also delocalized over the entire molecule with little contribution
from the carboxylic acid.
The cyclization to form a benzimidazole using the alanine has
been reported to be accessed by solvent free melting [30] and also
in the presence of a mixture of orthophosphic acid and poly-
phosporic acid via microwave irradiation [31]. Hence, the com-
pounds were not reproduced here. However, the theoretical
calculations indicate that the amino acid derivatives of benzoyl
isothiocyanate are not reactive to the cyclization reaction with 1,2-
diaminobenzene at the carboxylic acid site (Scheme 1).
Fig. 14 shows the frontier orbitals of compound III. The HOMO is
mostly concentrated on the sulfur atoms and the methylene groups
from the methionine whilst the LUMO is mostly delocalised over
the benzene ring, the carbonyl, the nitrogen atoms and the thione.
The delocalisation of the LUMO over the benzene ring aids in sta-
bilising the compound during charge transfer. The carboxylic acid
makes no contribution to the frontier orbitals (LUMO) which again
confirmed its unreactivity. In frontier orbitals of methionine
(Fig. 15) the LUMO is delocalised over the entire molecule except
4. Conclusions
2-[(Benzoylcarbamothioyl)amino]-3-hydroxypropanoic acid (I),
1-(benzoylcarbamothioyl)pyrrolidine-2-carboxylic acid (II), 2-
[(benzoylcarbamothioyl)amino]-4-(methylsulfanyl)butanoic acid
(III), and 2-[(benzoylcarbamothioyl)amino]propanoic acid (IV)
have been synthesized from the reaction of
methionine and -alanine with benzoyl isothiocyanate, respec-
tively. The compounds have been characterized by IR, NMR, mi-
croanalyses and mass spectrometry. The crystal structures of all the
compounds have also been discussed. Compound II exhibited
rotamers in solution. DFT calculations of the frontier orbitals
showed that the carboxylic acid group in these compounds was
unreactive which confirmed the unsuccessful attempts to convert
these compounds to benzimidazoles via cyclization with 1,2-
diaminobenzene.
L-serine, L-proline, D-
L
Acknowledgements
The authors are grateful to the Medical Research Council (MRC)
of South Africa for financial assistance and Felix Odame would also
like to thank the Faculty of Science (NMMU) for awarding him a
bursary for his studies (2013 and 2014). The NMMU RCD is also
acknowledged by Felix Odame for the bursary funding (PGRS) for
2015. We would also like to express our gratitude to the Centre for
Fig. 17. The atomic orbitals compositions of the frontier molecular orbitals for alanine.

48
F. Odame et al. / Journal of Molecular Structure 1099 (2015) 38e48
329e330.
High Performance Computing (CHPC) in Cape Town for the use of
their modelling resources.
[14] A.T. Kabbani, H. Ramadam, H.H. Hammud, A.M. Ghannoum, JUCTM 40 (2005)
339e344.
[15] D.M.H. AL-Mudhaffar, D.S. Al-Edani, S.M. Dawood, J. Korean Chem. Soc. 54
(2010) 506e514.
[16] M.A. Ibrahim, M.S.M. Yusof, N.M. Amin, Molecules 19 (2014) 5191e5204.
[17] C.K. Ӧzer, H. Arslan, D. VanDerveer, N. Külcü, Molecules 14 (2009) 655e666.
[18] A.G. Patel, K.N. Raval, S.P. Patel, K.S. Patel, S.V. Patel, J. Pharma Res. 4 (2012)
1e4.
[19] APEX2, SADABS and SAINT. Bruker AXS Inc., Madison, WI, USA, 2010.
[20] G.M. Sheldrick, Acta Cryst. A 64 (2008) 112e122.
[21] C.B. Hübschle, G.M. Sheldrick, B. Dittrich, J. Appl. Cryst. 44 (2011) 1281e1284.
[22] S. Saeed, N. Rashid, P.G. Jones, M. Ali, R. Hussain, Eur. J. Med. Chem. 45 (2010)
1323e1331.
[23] C.H. Choi, M. Kertez, J. Phys. Chem. A 101 (1997) 3823e3831.
[24] M. Kurt, P. Chinna Babu, N. Sundaraganesan, M. Cinar, M. Karabacak, Spec-
trochim. Acta A 79 (2011) 1162e1170.
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.molstruc.2015.05.053.
References
[1] J. Wu, Q. Shi, Z. Chen, M. He, L. Jin, D. Hu, Molecules 17 (2012) 5139e5150.
[2] A. Yahyazadeh, Z. Ghasemi, Eur. Chem. Bull. 2 (2013) 573e575.
[3] S. Saeed, N. Rashid, M. Ali, R. Hussain, P. Jones, Eur. J. Chem. 1 (2010) 221e227.
[4] S. Saeed, R. Hussain, Eur. Chem. Bull. 2 (2013) 465e467.
[5] S.J. Xue, S.Y. Ke, T.B. Wei, J. Chin. Chem. Soc. 51 (2004) 1013e1018.
[6] H.M. Fadallah, K.A. Khan, A.M. Asir, J. Fluor. Chem. 132 (2011) 131e137.
[7] G.W. Lee, N. Singh, D.O. Jang, Tetrahedron Lett. 490 (2008) 1952e1956.
[8] J.M. Marquez, O. Lopez, I. Maya, J. Fuentes, J.G. Fernandez-Bolanos, Tetrahe-
dron Lett. 49 (2008) 3912e3915.
[25] D.F.V. Lewis, C. Ioannides, C.D.V. Xenobiotica 24 (1994) 401e408.
[26] L. Padmaja, R.C. Kunar, D. Sajan, I.H. Joe, V.S. Jayakunmar, G.R. Pettit,
O.F. Nielsen, J. Raman Spectrosc. 40 (2009) 419e428.
[27] A. Poiyamozhi, N. Sundaraganesan, M. Karabacak, O. Tanriverdi, M. Kurt,
J. Mol. Struc 1024 (2012) 1e12.
[9] G. Madhava, V.K. Subbaiah, R. Srenivasulu, C.N. Raju, Der Pharm. Lett. 4 (2012)
1194e1201.
[28] P. Udhayakala, A. Jayanthi, T.V. Rajendiran, S. Gunasekaran, Arch. Appl. Sci.
Res. 3 (2011) 424e439.
[10] K. Takasu, T. Azuma, I. Enkhtaivan, Y. Takemoto, Molecules 15 (2010)
8327e8348.
[11] F. Odame, E. Hosten, Z.R. Tshentu, R. Betz, Z. Kristallogr. NCS 229 (2014)
337e338.
[12] F. Odame, E. Hosten, Z.R. Tshentu, R. Betz, Z. Kristallogr. NCS 230 (2015) 9e10.
[13] F. Odame, E. Hosten, Z.R. Tshentu, R. Betz, Z. Kristallogr. NCS 229 (2014)
[29] R.G. Pearson, Proc. Natl. Acad. Sci. U. S. A. 83 (1986) 8440e8441.
[30] R.H. Chen, J.F. Xiang, P. Peng, G.Z. Mo, X.S. Tang, Z.Y. Wang, X.F. Wang, Asian J.
Chem. 26 (2014) 926e932.
[31] P. Peng, J.F. Xiong, G.Z. Mo, J.L. Zheng, R.H. Chen, X.Y. Chen, Z.Y. Wang, Amino
Acids (2014) 2427e2433.