Structural insights into the polymorphism of bismuth(III) di-n-butyldithiocarbamate by X-ray diffraction, solid-state (13C/15N) CP-MAS NMR and DFT calculations

Article abstract of DOI:10.1016/j.poly.2017.03.018

Two crystalline polymorphs of a binuclear tris(di-n-butyldithiocarbamato)bismuth(III) complex, I and II, with an empirical formula of [Bi{S₂CN(n-C₄H₉)₂}₃] were synthesised and characterised by X-ray diffraction (XRD), solid-state NMR and density functional theory (DFT) calculations. At the supramolecular level, these mononuclear molecular units interact in pairs via secondary Bi?S bonds, yielding binuclear formations of [Bi₂{S₂CN(n-C₄H₉)₂}₆]. The polymorph I (P1ˉ) contains two isomeric non-centrosymmetric binuclear molecules of [Bi₂{S₂CN(n-C₄H₉)₂}₆], which are related to each other as conformers, therefore having four structurally inequivalent bismuth atoms and twelve inequivalent dithiocarbamate ligands. In contrast, the structurally simpler polymorph II (P2₁/n) exists as a single molecular form of the corresponding centrosymmetric binuclear formation, comprising two structurally equivalent bismuth atoms and three structurally different dithiocarbamate groups. The polymorphs I and II were found to be interconvertible by altering the solvent system during the recrystallisation process. Sun et al. (2012) has reported a crystalline form of the title compound which resembles, but is not identical with, polymorph II. Experimental solid-state ¹³C and ¹⁵N cross-polarisation (CP) magic-angle-spinning (MAS) NMR spectra of both polymorphs I and II were in accord with the direct structural data on these complexes. Assignments of the resonance lines in the solid-state ¹³C and ¹⁵N NMR spectra were assisted by chemical shift calculations of the crystals using periodic DFT.

Full text of DOI:10.1016/j.poly.2017.03.018

Accepted Manuscript
Structural Insights into the Polymorphism of Bismuth(III) Di-n-Butyldithiocar-
bamate by X-Ray Diffraction, Solid-State (13C/15N) CP-MAS NMR and DFT
Calculations
Vasantha Gowda, Bipul Sarma, Risto S. Laitinen, Anna-Carin Larsson,
Alexander V. Ivanov, Dinu Iuga, Perttu Lantto, Oleg N. Antzutkin
PII:
S0277-5387(17)30198-5
DOI:
http://dx.doi.org/10.1016/j.poly.2017.03.018
POLY 12533
Reference:
Polyhedron
To appear in:
Received Date:
Revised Date:
Accepted Date:
10 December 2016
1 March 2017
4 March 2017
Please cite this article as: V. Gowda, B. Sarma, R.S. Laitinen, A-C. Larsson, A.V. Ivanov, D. Iuga, P. Lantto, O.N.
Antzutkin, Structural Insights into the Polymorphism of Bismuth(III) Di-n-Butyldithiocarbamate by X-Ray
Diffraction, Solid-State (13C/15N) CP-MAS NMR and DFT Calculations, Polyhedron (2017), doi: http://dx.doi.org/
10.1016/j.poly.2017.03.018
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.

Structural Insights into the Polymorphism of Bismuth(III) Di-n-
Butyldithiocarbamate by X-Ray Diffraction, Solid-State (13C/15N) CP-MAS
NMR and DFT Calculations
Vasantha Gowda,a,b,* Bipul Sarma,c Risto S. Laitinen,d Anna-Carin Larsson,a Alexander V.
Ivanov,e Dinu Iuga,f Perttu Lantto,b Oleg N. Antzutkina,f,*
a Chemistry of Interfaces, Luleå University of Technology, SE-97187 Luleå, Sweden
b NMR Research Unit, University of Oulu, P.O. Box 3000, FI-90014 Oulu, Finland
c Department of Chemical Sciences, Tezpur University, 784028, Tezpur, Assam, India
d Laboratory of Inorganic Chemistry, University of Oulu, P.O. Box 3000, FI-90014, Finland
e Institute of Geology and Nature Management, Far Eastern Branch of the Russian Academy
of Sciences, Blagoveshchensk, 675000, Russia
f Department of Physics, Warwick University, Coventry CV4 7AL, UK
*Email address: gowda.vasantha.k@gmail.com (Vasantha Gowda), Oleg.Antzutkin@ltu.se
and O.N.Antzutkin@warwick.ac.uk (Oleg N. Antzutkin)
1

ABSTRACT
Two crystalline polymorphs of a binuclear tris(di-n-butyldithiocarbamato)bismuth(III)
complex, I and II, with an empirical formula of [Bi{S2CN(n-C4H9)2}3] were synthesised and
characterised by X-ray diffraction (XRD), solid-state NMR and density functional theory
(DFT) calculations. At the supramolecular level, these mononuclear molecular units interact
in pairs via secondary Bi⋅⋅⋅S bonds, yielding binuclear formations of [Bi2{S2CN(n-C4H9)2}6].
ꢁ
The polymorph I (ꢀ1) contains two isomeric non-centrosymmetric binuclear molecules of
[Bi2{S2CN(n-C4H9)2}6], which are related to each other as conformers, therefore having four
structurally inequivalent bismuth atoms and twelve inequivalent dithiocarbamate ligands. In
contrast, the structurally simpler polymorph II (ꢀ2ꢂ/ꢃ) exists as a single molecular form of
the corresponding centrosymmetric binuclear formation, comprising two structurally
equivalent bismuth atoms and three structurally different dithiocarbamate groups. The
polymorphs I and II were found to be interconvertible by altering the solvent system during
the recrystallisation process. Sun et al. (2012) has reported a crystalline form of the title
compound which resembles, but is not identical with, polymorph II. Experimental solid-state
13C and 15N cross-polarisation (CP) magic-angle-spinning (MAS) NMR spectra of both
polymorphs I and II were in accord with the direct structural data on these complexes.
Assignments of the resonance lines in the solid-state 13C and 15N NMR spectra was assisted by
chemical shift calculations of the crystals using periodic DFT.
Keywords: Bismuth(III) di-n-butyldithiocarbamate; Polymorphism; Single-crystal X-ray
structure; 13C and 15N CP-MAS NMR; Density functional theory calculations.
1. Introduction
2

Bismuth (Bi) is the heaviest among the pnictogen (Group 15) elements and it is called a
‘green’ metal, because known bismuth compounds are benign or are considerably less toxic
compared to other pnictogens, such as antimony and arsenic [1,2]. Moreover, bismuth
compounds may even provide therapeutic responses and antibacterial activities [3]. Metal
dithiocarbamate compounds find wide-ranging applications including the synthesis of
nanomaterials [4,5], solvent extraction of rare earths [6], as additives to lubricating oils [7], as
anticarcinogenic drugs [8,9] and as pesticides or fungicides [10-12]. Dithiocarbamates are
highly versatile mono-anionic ligands, which can form stable complexes with almost all
metals in the periodic table [13]. The metal-dithiocarbamate complexes show a wide
structural diversity, ranging from mononuclear and binuclear to polymeric supramolecular
assemblies [14-16], depending on several factors, such as the metal centre, alkyl substituents
at the nitrogen atom and also on solvent properties.
Intermolecular interactions such as van der Waals, secondary and hydrogen bonding,
steric repulsions or weak cooperative interactions like C–H⋯O hydrogen bonding, ꢄ ⋯ ꢄ, C–
H⋯ ꢄ, mono- and polyhapta, M⋅⋅⋅H–C agostic (attractive) and anagostic (repulsive)
interactions can play a vital role in stabilising one particular molecular arrangement in a
crystal. Because of such weak interactions, various molecular arrangements can have nearly
the same energy and the molecules can crystallise in different forms under different
experimental conditions, such as temperature, solvent properties, solute concentration, crystal
growth rate and the presence of impurities. Different polymorphic forms of a given
compound can exhibit differences in solubility, melting point, density, optical and electrical
properties, etc. Being able to distinguish different polymorphs and their individual properties
is of utmost importance because of their various pharmaceutical applications [17]. Solid-state
13C/15N cross-polarisation (CP) magic-angle-spinning (MAS) NMR spectroscopy has proven
itself as a powerful structural tool for studies of polycrystalline polymorphic systems [18,19].
15
In solid NMR (13C, N, etc.), depending on the number of inequivalent molecules in the
3

asymmetric unit of the crystal structure, one can often observe multiple resonance lines for
each symmetrically inequivalent atom. Since different polymorphs of a compound are packed
differently in 3-dimensional space, the NMR resonance lines for the symmetrically
inequivalent nuclei will differ in different solid polymorphs [20].
In this work, we apply the so called “SMARTER Crystallography” approach [21-26],
which enables a combination of experimental solid-state NMR and XRD techniques with
theoretical calculations of structure and NMR chemical shifts of periodic systems, aiming to
gain much greater insight into the structure and properties of materials than by any individual
approach considered alone. In our previous reports, we have successfully used a similar
multidisciplinary approach for structural investigations of yttrium(III) [27] and
lanthanum(III) [28] complexes with diethyldithiocarbamate and phenanthroline ligands.
It has been previously reported by Sun et al. [29] that recrystallisation of the tris(di-n-
butyldithiocarbamato)bismuth(III) compound from acetonitrile resulted in monoclinic single
crystals (ꢀ2ꢂ/ꢅ), having a centrosymmetric binuclear molecule of [Bi2{S2CN(n-C4H9)2}6] and,
therefore, three inequivalent dithiocarbamate ligands in the unit cell. In our case, a
13
comparatively different PXRD pattern and a missing correlation with the solid-state C/15N
NMR spectra for the polycrystalline material of the title compound, also recrystallised from
acetonitrile, indicated a different crystal structure type. On the contrary, recrystallisation from
dichloromethane of the title compound resulted in a crystalline form of the compound which
showed a good correlation to, in terms of the PXRD pattern and solid-state NMR spectral
data, the crystal structure by Sun et al. [29]. Herein, we report the synthesis and
characterisation
of
two
crystalline
polymorphs
of
the
tris(di-n-
butyldithiocarbamato)bismuth(III) complex, which were obtained by recrystallisation of the
synthesised title compound from (i) acetonitrile (polymorph I) and (ii) dichloromethane
13
15
(polymorph II), by solid-state C and N NMR spectroscopy, powder and single crystal
XRD, and further explored by employing first principle DFT calculations using gauge
4

including the projector augmented wave (GIPAW) method for NMR shieldings. Polymorph I
ꢁ
crystallises in the triclinic system (ꢀ1 space group), having four structurally inequivalent
mononuclear [Bi{S2CN(n-C4H9)2}3] units, which form two non-centrosymmetric isomeric
molecules of [Bi2{S2CN(n-C4H9)2}6] stabilised in pairs via non-symmetric secondary Bi⋅⋅⋅S
bonds. These binuclear molecules are related to each other as conformational isomers
(conformers). The crystal structure of polymorph II was solved and refined in the setting of a
monoclinic system with the space group ꢀ2ꢂ/ꢃ, which presents significant differences in the
bond parameters, especially for the alkyl chains when compared to the structure reported by
Sun et al. (hereafter IIs) [29]. The polymorph II (as for IIs) is also presented by a single
centrosymmetric binuclear molecule of [Bi2{S2CN(n-C4H9)2}6]: there are two such
crystallographically independent molecules per unit cell. The single crystal structures and
13
15
experimental solid-state C and N CP-MAS NMR spectra of I and II were compared with
DFT/GIPAW [30-32] calculated structural parameters and chemical shifts.
2. Materials and methods
Dibutylamine, carbon disulfide, sodium hydroxide, ethanol, dichloromethane, diethyl ether,
acetonitrile and bismuth trinitrate pentahydrate were purchased from Sigma-Aldrich and were
used without further purification.
2.1. Synthesis of the sodium salt of di-n-butyldithiocarbamate
Sodium di-n-butyldithiocarbamate was obtained by the reaction of dibutylamine with carbon
disulfide in an alkaline medium at a temperature below 5 °C. In short, di-n-butylamine (100
mmol) and sodium hydroxide (100 mmol) were dissolved in ethanol (100 mL) and stirred at 0
°C. Carbon disulfide (101 mmol) was added dropwise to the above stirred mixture. Stirring
was continued for another 2 hours for completion of the reaction. The solvent was evaporated
5

in a rotary evaporator at low pressure. The crude product was washed with diethyl ether and
recrystallised from ethanol. The yield was 90%.
2.2. Synthesis of the tris(di-n-butyldithiocarbamato)bismuth(III) complex
Bismuth trinitrate pentahydrate [Bi(NO3)3·5H2O] (4.85 g, 10 mmol) was stirred for about 1
hour in 50 mL ethanol, whereupon a curdy white solution was obtained. A solution of the
sodium salt of di-n-dibutyldithiocarbamate (31 mmol) in ethanol was then added and the
reaction mixture was stirred for about 4 hours at room temperature. The resulting solution
was suction washed with ethanol, filtered and dried. The yield was 80 %.
In contrast to the previously reported synthetic method by Sun et al. [29], the title compound
ꢁ
crystallised in the triclinic system (ꢀ1) upon recrystallisation from acetonitrile at room
temperature over a few days (polymorph I). Polymorph II was prepared by the
recrystallisation of the dried crude product from dichloromethane. It was also found by NMR
and PXRD that polymorph I can be readily converted into polymorph II by switching the
recrystallisation solvent from acetonitrile to dichloromethane and vice versa.
Melting point: Polymorph I 101-102 °C and polymorph II 96-97 °C.
2.3. Powder X-ray diffraction (PXRD)
The PXRD patterns [see Fig. 1] were collected using a PANalytical Empyrean X-ray
diffraction spectrometer (45 kV and 40 mA) and Cu Kα radiation with λ = 1.54 Å. The data
collection was performed at ambient room temperature over the range 2θ = 4-40° with a scan
interval of θ = 0.013°. The theoretical PXRD patterns of both polymorphs I and II [see Fig.
1] were calculated using Mercury crystal structure visualisation software (version 3.5) [33].
2.4. Solution-state NMR of [Bi{S2CN(-n-C4H9)2}3]
6

1
One-dimensional H and 13C (1H decoupled) NMR spectra were measured on a Bruker
1
13
Avance III spectrometer operating at 400.13 and 100.61 MHz for H and C, respectively.
1
The spectral parameters for the H/13C experiments were as follows: ꢄ/2 pulse lengths
1
13
8.8/12.0 ꢆꢇ, recycle delay 3/5 s, number of transients 32/1024. Both H and C spectra were
referenced to TMS ( 0.0 ppm).
1H NMR (solution in CDCl3, see Fig. S1a in ESI) , ppm: 3.72 (-N-CH2-), 1.74 (-CH2-), 1.35
(-CH2-), 0.94 (-CH3).
13C NMR (solution in CDCl3, see Fig. S1b in ESI) , ppm: 200.95 (-S2CN), 54.32 (-N-CH2-),
29.33 (-CH2-), 20.49 (-CH2-), 14.06 (-CH3).
2.5. Solid-state NMR measurements
Solid-state 13C/15N CP-MAS NMR spectra of polymorph I were obtained on a Bruker Avance
13
III 400 MHz (B0 = 9.4 T) NMR spectrometer operating at 100.61/40.56 MHz for C/15N,
respectively, while high resolution solid-state 13C/15N CP-MAS NMR spectra of polymorph II
were recorded on a Bruker Avance III 850 MHz (B0 = 19.975 T) NMR spectrometer
operating at 213.8/86.1 MHz for 13C/15N, respectively. The NMR spectra were recorded using
cross-polarisation (CP) from protons together with broadband decoupling using the spinal64
sequence [34] at the proton resonance to average out dipole-dipole 13C-1H/15N-1H interactions.
13
The C/15N spectra were recorded under MAS conditions at spinning frequencies of 6-10
kHz; proton /2 pulse duration 2.7/2.5 µs (at 400 MHz) and 3.9/3.5 µs (at 850 MHz) µs; 1H-
13C/1H-15N contact time 3.0/2.5 ms (at 400 MHz) 1.0/2.0 ms (at 850 MHz); the repetition time
was 3.0 s. Isotropic chemical shifts were referenced to adamantane ( 38.48 ppm w.r.t TMS at
13
0.0 ppm) [35] in C NMR and crystalline NH4Cl ( 39.2 ppm w.r.t. NH3 at 0.0 ppm) [36]
in 15N NMR.
7

Both non-centrosymmetric (polymorph I) and centrosymmetric (polymorph II) binuclear
13
tris(di-n-butyldithiocarbamato)bismuth(III) complexes were characterised by C and 15N CP-
MAS NMR.
Polymorph I, [Bi2{S2CN(n-C4H9)2}6]:
13C CP-MAS NMR , ppm(see Fig. 2a): 206.1, 205.4, 202.3, 201.4, 201.1, 200.2 (1:1:1:1:1:1,
-S2CN=); 55.7, 55.5, 55.1, 54.7, 54.0, 53.7, 52.7 (2:1:1:3:2:2:1, =NCH2-); 31.6, 30.9, 30.6,
29.6, 29.3, 29.2, 28.9, 28.7 (1:1:3:2:1:2:1:1, -CH2-); 21.9, 21.5, 21.1, 20.4 (1:1:9:1, -CH2-
CH3); 16.8, 15.9, 15.4, 15.0, 14.7, 14.1, 13.8 (1:1:1:3:3:1:2, -CH3);
15N CP-MAS NMR , ppm (see Fig. 3a): 180.8, 180.2, 178.0, 174.4, 169.1 (2:1:1:1:1, =N-).
Polymorph II, [Bi2{S2CN(n-C4H9)2}6]:
13C CP-MAS NMR , ppm (see Fig. 2b): 204.7, 203.6, 199.9 (1:1:1, -S2CN=); 56.6, 55.8,
54.9, 54.2, 53.6 (2:1:1:1:1, =NCH2-); 30.8, 30.5, 29.6, 29.1, 28.8 (1:2:1:1:1, -CH2-); 21.6,
21.3, 20.9, 20.8 (1:2:1:2, -CH2-CH3); 15.5, 15.4, 15.1, 14.2 (2:1:1:2, -CH3).
15N CP-MAS NMR , ppm (see Fig. 3b): 177.3, 171.2, 167.7 (1:1:1)
2.6. X-ray diffraction (XRD) structure determination
The X-ray reflections of the single crystals of polymorphs I and II were collected on Bruker
APEX-II and Bruker Nonius Kappa CCD diffractometers, respectively, using Mo K ( =
0.71073 Å) radiation. Data reduction was performed using the Bruker SAINT software, while
the intensities for the absorptions were corrected using SADABS. The structures were solved
and refined using SHELS-2013 and SHELXL-2013, respectively, for non-H atoms [37,38].
The non-H atoms were refined with anisotropic displacement parameters. The H-atoms were
positioned geometrically (methyl groups 0.96 Å and methylene groups 0.97 Å) and refined
using riding models in which the isotropic thermal parameters of the H-atoms are
proportional to those of the carbon atoms to which they are attached. In the case of II, the site
8

occupation factors of the butyl ligands were refined along with positional and displacement
parameters. The final site occupancy factors for the six butyl groups were:
Bu(11): C11A-C12A-C13A-C14A [0.444(16)] and C11B-C12B-C13B-C14B [0.556(16)],
Bu(12): C15A-C16A-C17A-C18A [0.543(13)] and C15B-C16B-C17B-C18B [0.457(13)],
Bu(21): C21A-C22A-C23A-C24A [0.608(13)] and C21B-C22B-C23B-C24B [0.392(13)],
Bu(22): C25A-C26A-C27A-C28A [0.805(15)] and C25B-C26B-C27B-C28B [0.195(15)],
Bu(31): C31A-C32A-C33A-C34A [0.723(19)] and C31B-C32B-C33B-C34B [0.277(19)] and
Bu(32): C35A-C36A-C37A-C38A [0.713(17)] and C35B-C36B-C37B-C38B [0.287(17)].
Further information in the form of a CIF file has been deposited as depository CCDC
1458699 and 1494946 for forms I and II, respectively. A check of the final CIF file using
PLATON did not show any missed symmetry.
2.7. Computational details of the periodic calculation
Geometry optimisations were carried out at the DFT level of theory using the Perdew Burke
Ernzerhof (PBE) functional [39] with a Tkatchenko-Scheffler (TS) dispersion correction [40],
as implemented in the plane wave code CASTEP [31,41]. The structures were described as an
infinite periodic solid using periodic boundary conditions. For NMR shielding constant
calculations, the all-electron information was reconstructed using the GIPAW [30] method,
as implemented in NMR CASTEP 7.0 code [30-32]. The geometry optimisations were
performed using ultra-soft pseudopotentials [32]. The total energy of the system was
converged within 1E-7 eV. For the geometry optimization, Monkhorst-Pack k-point meshes
with a grid of 0.03Å-1 for the Brillouin zone sample with a cut-off energy of 500 eV were
chosen to achieve the total energy convergence of < 0.05 eV/atom. For NMR calculations, we
used a slightly higher energy cut-off (550 eV), which was tested to give sufficiently
converged NMR parameters (< 0.1 ppm). The integrals over the Brillouin zone were
performed using a 2 x 2 x 1 (polymorph I) and 4 x 4 x 2 (polymorphs II and IIs) Monkhorst-
9

Pack grid of k-points corresponding to less than 0.03 Å-1 separation between them in each
reciprocal lattice direction.
Within the periodic boundary conditions, we have performed geometry optimisation of the
crystal structures of I, II, and IIs, either by relaxing the positions of only the hydrogen atoms
(H-opt) or all the atoms (full-opt) in the simulation cell. In both cases, the lattice parameters
were kept fixed. The forces acting on the atoms were converged below 0.01 eV/Å in the
geometry optimisation.
As was explained in section 2.6 of the Material and methods, because of the partial atom
occupancy of some of the butyl carbons in polymorph II, the disordered dithiocarbamate
ligands were considered as having two conformations (conformer-A and conformer-B, see
Fig. 4b for the labelling scheme). The geometry optimisation of the crystal structure was
performed by considering these two conformations separately. The calculated chemical shift
is, therefore, a weighted average of the chemical shifts for the six disordered butyl ligands.
The site occupancy factors (s.o.f.) of these ligands are also given in section 2.6.
13
For a direct comparison of the GIPAW calculated C shieldings with the experimentally
measured isotropic chemical shift values the following expression was used:
ꢈꢉꢊꢋ,ꢌꢍꢎꢌ ꢏ=ꢏꢐꢑꢒꢓ ꢏ−ꢏꢐꢉꢊꢋ
(1)
where ꢐꢑꢒꢓ = 170.9 ppm and 234.7 ppm for 13C (TMS) and 15N (NH ), respectively [28].
3
3. Results and discussion
3.1 Powder X-ray diffraction (PXRD)
The experimental PXRD patterns for polymorphs I and II were compared with the calculated
PXRD patterns for the corresponding experimental X-ray structures (see Figs. 1a and 1b).
Although there are small discrepancies in the intensities and positions of certain peaks in the
experimental vs. calculated PXRD patterns, here we only focus on the main features of
resemblance between them. A qualitative comparison based on a visual assessment reveals a
10

good match between the experimental and calculated PXRD patterns for both polymorphs,
which ensures that the PXRD patterns of the synthesised materials and hence the solid-state
NMR spectra correspond to their single crystal structures. It is known that the experimental
PXRD pattern can be influenced, apart from the features of the instrumentation and method
of data collection, by several factors related to the microstructural characteristics of the
powder sample such as the size, shape and orientational distribution of the crystallites in the
powder [42]. It is important that for a quantitative fitting of the experimental and the
calculated peak profiles, a rigorous consideration of the Rietveld refinement is needed
[42,43].
3.2 Single crystal X-ray diffraction
High quality single crystals of polymorphs I and II were obtained and their structures
determined as explained earlier (see section 2.6). The crystallographic data are given in Table
1. The ORTEP representation of the molecular structure of polymorph I is depicted in Fig. 4.
Some selected interatomic distances and bond angles for polymorph I are given in Tables S1
and S2 in the ESI, respectively. The molecular structures of polymorphs II and IIs [29], along
with the crystallographic labelling scheme, are depicted in Figs. 5a and S2 (see the ESI),
respectively. In Tables S3 and S4 in the ESI, we have compared some selected bond
distances and bond angles, respectively, for the polymorphs II with the corresponding bonds
of IIs. A molecule of the composition [Bi{S2CN(n-C4H9)2}3] is the main structural unit of these
binuclear compounds. There are four structurally inequivalent (isomeric) mononuclear
molecules of bismuth di-n-butyldithiocarbamate in polymorph I and a single kind of
[Bi{S2CN(n-C4H9)2}3] in polymorph II. Therefore, in the former case, a combination of these
mononuclear species yields two isomeric non-centrosymmetric binuclear molecules of
[Bi2{S2CN(n-C4H9)2}6] (polymorph I), whereas there is a single kind of centrosymmetric
binuclear molecule of the same composition, [Bi2{S2CN(n-C4H9)2}6], in the latter case
11

(polymorph II). The bismuth atoms form [BiS6] chromophores of distorted pentagonal
bipyramidal geometries, in which the lone pair of stereochemically active electrons
originating from the central metal atom occupy one of the axial positions. Each bismuth atom
in the considered molecules is coordinated in the S,S′-aniso-bidentate manner to three non-
equivalent dithiocarbamate ligands: for all the di-n-butyldithiocarbamate ligands, one of the
Bi-S bonds is markedly stronger (polymorph I: 2.642(3)-2.798(3) Å and polymorph II:
2.582(2)-2.806(3) Å) than the other (polymorph I: 2.840(3)-3.110(3) Å and polymorph II:
2.790(3)-2.906(3)). The C-N bond lengths, which range from 1.312(14) to 1.370(13) Å, are
shorter than the ideal C-N single bonds (1.47 Å) because of the double bond contribution in
the dithiocarbamate ligands.
At the supramolecular level of polymorph I, isomeric non-centrosymmetric binuclear
molecules of [Bi2{S2CN(n-C4H9)2}6] are formed by the pairs of relatively weak unsymmetrical
secondary bonds between the mononuclear moieties: Bi(1)⋅⋅⋅S(7) (3.088 Å), Bi(2)⋅⋅⋅S(1)
(3.039 Å) and Bi(3)⋅⋅⋅S(19) (3.141 Å), Bi(4)⋅⋅⋅S(18) (3.110 Å). For each dimer, the equatorial
planes, where the mononuclear molecules come close to each other, are mutually parallel.
The combination of mononuclear [Bi{S2CN(n-C4H9)2}3] molecules is purely based on
structural selectivity. In polymorph I, two inequivalent molecules come together because of
the pairs of weak Bi⋅⋅⋅S interactions to form a non-centrosymmetric dimer, unlike in
polymorph II where a centrosymmetric dimer is formed because of the interaction of two
equivalent molecular units with a Bi⋅⋅⋅S distance of 3.3898(10) Å (see Fig. 5b). The reported
secondary Bi⋅⋅⋅S distance in the polymorph IIs was 3.3629(34) Å [29].
The crystal structure of II has been solved and refined in the setting ꢀ2ꢂ/ꢃ of space group 14,
rather than in ꢀ2ꢂ/ꢅ as reported by Sun et al. [29]. This gives a beta angle closer to 90°,
which provides more accurate computations, since the rounding errors are smaller. The unit
cell parameters are as follows:
12

Polymorph IIs [29]: a = 16.802(9) Å, b = 10.256(6) Å, c = 25.083(10) Å, ꢔ = 121.77(3)°;
Polymorph II: a = 16.932(4) Å, b = 10.366(2) Å, c = 21.817(4) Å, ꢔ = 99.44(3)°. Apart from
that, polymorph IIs does not display any conformational disorders [29]. However, in our case,
it was found that all six butyl groups of II were disordered, because the butyl chains of all the
three dithiocarbamate ligands could adopt several different conformations. In some cases,
there were even more possibilities of minor site occupancy factors, but we have only taken
the two most common conformations (referred to as conformer-A and conformer-B in the
following sections) (see Fig. 5a). The site occupation factors of the butyl groups of our
structure are given in section 2.6 of Materials and methods. The heavy atom root-mean-
square deviation (RMSD) between the two conformers is 1.3 Å.
3.3 Solid-state 13C and 15N CP-MAS NMR spectra
13
In the C NMR spectrum (solution in CDCl3) of tris(di-n-butyldithiocarbamato)bismuth(III)
one observes resonances for each inequivalent carbon atom in the molecule, i.e. single
resonance lines for the –S2CN, -NCH2-, -CH2-, -CH2-, and -CH3 groups (see Fig. S1b in ESI).
On the other hand, Fig. 2a shows the solid-state 13C CP-MAS NMR spectrum of a
polycrystalline sample of polymorph I, depicting multiple resonance lines representing
inequivalent carbon atoms. The signal enhancement of a CP-MAS experiment depends
H
largely on the proton spin-lattice relaxation time (T1 ) and the competition between TCP and
T1ꢕH during CP transfer, where TCP and T1ꢕH represents CP time (contact pulse width) and
proton spin-lattice relaxation time during CP. Since these relaxation parameters are usually
different for atoms in chemically inequivalent sites, the observed signal enhancements may
vary significantly. In our case, to quantify the number of sites of each kind, the peaks
corresponding to atoms in the same position in the molecule are compared. The most
informative region of the dithiocarbamate ligands, i.e. the -S2CN- carbon atoms, resulted in
six well resolved 13C resonance signals with chemical shifts over the range 200.2-206.1 ppm
13

with relative integral intensities of 1:1:1:1:1:1, indicating the presence of at least six
structurally inequivalent n-butyldithiocarbamate ligands in the unit cell. However, we
expected twelve resonance lines for the dithiocarbamate groups, as there are twelve
inequivalent sites revealed from the X-ray diffraction structure, but there is insufficient
13
resolution in the NMR spectrum for this observation. Additionally, in the C CP-MAS
spectrum of polymorph I, one observes 13C resonances corresponding to twelve carbon atoms
each for the four alkyl carbon atoms of the butyl chains. On the other hand, five resonance
signals with chemical shifts over the range 169.1-180.8 ppm, with an integrated intensity
15
ratio of 2:1:1:1:1, were observed in the N CP-MAS NMR spectrum (see Fig. 3a), again
suggesting the presence of at least six inequivalent nitrogen sites in the dithiocarbamate
ligands of the complex.
On the contrary, both 13C and 15N CP-MAS spectra for the centrosymmetric binuclear
[Bi2{S2CN(-n-C4H9)2}6] complex (polymorph II) are more simple: both spectra show only
three resonance signals for the dithiocarbamate groups, with relative integral intensities of
1:1:1 for both 13C and 15N sites, in each case indicating the presence of only three inequivalent
di-n-butyldithiocarbamate ligands (see Figs. 2b and 3b). The solid-state NMR spectral
assignments for selected carbon and nitrogen sites are suggested in Tables 6 and 7,
respectively.
Overall, the 13C and 15N CP-MAS NMR data of polymorph I agree well with the single crystal
XRD data, indicating that the system under consideration contains a minimum of six
structurally inequivalent dithiocarbamate groups. Similarly, the single crystal XRD data for
polymorph II show the presence of a centrosymmetric binuclear molecule of [Bi2{(S2CN(n-
C4H9)2}6], which is in accord with the 13C and 15N chemical shift data, suggesting at least three
structurally inequivalent dithiocarbamate groups in this conformer.
3.4 Geometry optimisation and NMR calculations
14

The experimental chemical shifts for both polymorphs I and II are compared with those
13
calculated from the DFT/GIPAW approach (see Tables 6 and 7). The calculated C/15N
nuclear shielding values for some selected atoms of polymorphs I, II and IIs are listed in
13
15
Tables S1 and S2 (see the ESI), respectively. In CASTEP, the C and N NMR chemical
shift calculations were performed at both the “H-opt” and “full-opt” geometries of
polymorphs I, II and IIs. The strong forces on the atoms indicate that the structural model
derived from diffraction requires optimisation of the atomic coordinates. The forces on non-
proton atoms of “H-opt” structures were fairly large (> 2 eV/Å), while in the “full-opt”
structures the forces on all atoms were reduced to below 0.01 eV/Å. The CASTEP calculated
bond lengths and bond angles are compared with the actual (experimental) values in Tables
S1-S4 in the ESI. The heavy atom RMSD between the “H-opt” and “full-opt” structures of
polymorph I was 0.07 Å.
In the case of polymorph I, the calculated bond parameters are very close to the experimental
values. However, the NMR chemicals shifts are known to be very sensitive even to small
variations in the bond parameters. Therefore, the geometry optimisation of the X-ray
structures will have a large and direct effect on the calculated chemical shifts. Based on the
geometrical close resemblance and the calculated NMR chemical shift results of the “full-
opt” structure, we have grouped the dithiocarbamate ligands as either ‘terminal’ or ‘bridged’
(see Tables 6 and 7). In polymorph I, the experimental NMR spectral assignments of different
dithiocarbamate sites were made based on the observations that both the bonding parameters
and the calculated NMR chemical shifts show a close resemblance between Bi1 and Bi4 as
well as Bi2 and Bi3 molecular units. Although both “H-opt” and “full-opt” structures produce
13
15
almost similar C chemical shifts, the N chemicals shifts of “H-opt” structure were rather
random (see Tables 2) indicating that the latter are more sensitive to the subtle structural
changes. Therefore, it is not possible to show any meaningful correlation between the
15
calculated and the experimental N NMR chemical shifts for different dithiocarbamate sites
15

based on the “H-opt” structure. Moreover, it does not allow us to group them as ‘bridged’ or
‘terminal’ dithiocarbamate ligands. As can be seen from Tables 2, for the “full-opt” structure,
the 13C chemical shifts for the bridged dithiocarbamate carbon atoms appear more deshielded
13
than for the terminal dithiocarbamate carbon atoms. In the C CP-MAS NMR spectra of
polymorph I (Figure 3a), the higher (more deshielded) chemical shifts at 206.1 and 205.4
ppm can be attributed to the bridging dithiocarbamate carbon atoms, while the terminal
ligands have lower (more shielded) chemical shifts between 202.3 and 200.2 ppm (see Table
15
2). On the contrary, the N CP-MAS NMR resonance lines for the nitrogen sites in the
bridging dithiocarbamate ligands are more shielded (169.1 and 174.3 ppm) as compared to
the terminal ligands (178.0-180.8 ppm). The experimental chemical shift ranges for the
13C/15N dithiocarbamate sites in polymorph I are 6.1/11.7 ppm while the calculated chemical
shift ranges are 6.9/18.5 ppm.
There is clearly a good agreement between the experimental and theoretical 13C and 15N data,
as confirmed by the linear regression plots (with slopes and R2 very close to unity) for the
experimental vs. calculated (GIPAW) chemical shifts for the “full-opt” structure of
13
polymorph I (Figs. 6a and 6b, respectively). A similar C correlation plot corresponding to
the “H-opt” structure is shown in Fig. S3a in the ESI. The root-mean-square deviation
13
(RMSD) between the experimental and the calculated C/15N chemical shifts for the “full-
opt” and the “H-opt” structures are also compared in Table 2. The lower RMSDs for the
“full-opt” structure, again, indicate a better correspondence with the experimental values.
The sources of the discrepancies in the calculated 13C/15N chemical shifts observed for the “H-
opt” structure can be any or all of the following: (i) The reported uncertainties in atomic
positions from diffraction measurements can lead to significant variations in the computed
chemical shifts; (ii) The DFT calculations were done at zero K (without zero-point
vibrations) whereas the experimental determinations are performed under very different
circumstances (100.0 K for structure determination and at room temperature (298 K) for all
16

NMR measurements) [45]. The improved NMR results in case of the “full-opt” structure are
most probably because of the converged forces on the atoms, which were still rather large in
the X-ray structure, indicating a non-equilibrium state of the system [44,45]. The close-to
equilibrium structure is a prerequisite to get reasonable assignments of NMR spectra, as
without vibrational averaging of modelled parameters it is the best approximation of the real
(temperature) averaged structure corresponding to the experimentally observed NMR
chemical shifts.
In Table 3 we have compared the experimental NMR chemical shifts with those calculated
for the “H-opt” and “full-opt” structures of polymorph II. As explained before, the geometry
optimisation of polymorph II was performed by considering the disordered pairs for each of
the dithiocarbamate ligands separately. Therefore, in all cases, the chemical shifts for
polymorph II were calculated as the average over the conformationally disordered ligands.
13
15
Although the calculated C and N chemical shifts for the “H-opt” structure qualitatively
represent the experimental spectral pattern, the dithiocarbamate ligands show deviations in
terms of the chemical shift range as well as the relative chemical shift difference between the
13
inequivalent sites. For example, the experimental C/15N chemical shift ranges for the –S2CN
sites are relatively smaller (4.7/9.6 ppm) compared to the calculated ones (10.4/11.4 ppm).
Furthermore, the relative difference between the calculated chemical shifts for two of the
dithiocarbamate carbon atoms, viz. C3 and C2, is much larger (8.7 ppm) than the
experimentally observed difference between them (1.0 ppm). Similarly, the calculated and
experimental 15N chemical shift difference between atoms N2 and N3 are 10.6 and 3.5 ppm,
respectively. On the other hand, as indicated from the lower RMSD values (see Tables 3), the
calculated 13C/15N chemical shifts for the “full-opt” structure of polymorph II showed a much
better agreement with the experimental values than for the “H-opt” structure. Also, in terms
of both the NMR chemical shift range and the relative difference between the resonance lines
for inequivalent ligand sites, the calculated chemical shifts for the “full-opt” structure better
17

represents the observed chemical shifts. The plots of experimental vs. calculated 13C chemical
shifts for the “full-opt” structure of II displays an excellent linear correlation (R2 close to
13
unity) (see Fig. 6c). The C experimental vs. calculated correlation plot for the “H-opt”
structure of II is depicted in Figs. S3b (see the ESI).
As can be seen from Tables S3 and S4 in the ESI, there is a significant difference in some of
the experimental bond parameters between the crystal structures of II and IIs. Many of the
bond distances are not equal within the experimental error limits, especially for the C-N,
alkyl chain N-C and C-C bonds. For example, the reported C-N bond lengths for IIs (1.31(1)-
1.33(2) Å) are shorter compared to our structure II (1.341(12)-1.354(12) Å), while the DFT
calculated bond distances vary between 1.349 and 1.356 Å, which is much closer to our
experimental results. Much shorter C-C bond lengths have been reported for IIs, e.g. C21-C22
is 1.38(3) Å, while the corresponding C-C bond distances in our structure are 1.556(10)-
1.562(10) Å, which are again very close to the DFT calculated bond distances. Many such
differences were also observed in the alkyl N-C/C-C bond lengths and N-C-C/C-C-C bond
angles of the other two dithiocarbamate ligands (not shown in Table 5). The heavy atom
RMSD between the dimers of II and IIs is 2.1 Å (see Fig. S4 in the ESI). Also, a slightly
longer Bi⋅⋅⋅S bond length of 3.3898(10) Å for II (see Fig. 5b) vs. 3.3629(34) Å for IIs [29] and
Bi-Bi distance of 4.850 Å for II (see Fig. 5b) vs. 4.807 Å for IIs [29] indicate a slight variation
in the intermolecular interaction between the mononuclear molecules at the supramolecular
level. In addition, as mentioned in section 2.6, the three dithiocarbamate ligands for
polymorph II have conformational disorders, while the reported crystal structure for IIs does
not. The DFT calculated bond parameters for the “full-opt” structure of IIs are very close, but
not identical, to those calculated for polymorph II. In Fig. S5 (see the ESI), we have overlaid
the “full-opt” mononuclear molecular units for polymorph IIs and the two conformers of
polymorph II. The “full-opt” structure of IIs follows conformer-A of polymorph II with an
RMSD of 0.06 Å (see Tables S3 and S4 in the ESI).
18

In order to understand how the experimentally observed differences in the bond parameters
between the crystal structures for the polymorphs II and IIs (see Tables S3 and S4 in the ESI)
affect their calculated NMR chemical shifts, we have compared the GIPAW calculated
chemical shifts for the “H-opt” structures of both II and IIs (see Tables 3). Although the
calculated 13C chemical shifts of the butyl carbon atoms (averaged for different
dithiocarbamate ligands) for the “H-opt” structure of IIs are somewhat closer to those
15
calculated for II, both the 13C and N chemical shifts of the dithiocarbamate sites show much
larger deviations. On the other hand, the calculated chemical shifts for the “full-opt”
structures of IIs and II are much closer to each other (see Tables 3). In addition, we have also
13
calculated the RMSD between the experimental C/15N chemical shifts that corresponds to
polymorph II and those calculated for IIs (see Tables 3). It is observed that the RMSD
decreases from 6.1/12.7 (H-opt) to 0.8/2.5 (full-opt), indicating that the “full-opt” structure
for IIs, having bond parameters similar to that of the “full-opt” structure for II, provides a
better comparison than the “H-opt” structure for IIs.
4. Conclusions
Two crystalline polymorphs of the tris(di-n-butyldithiocarbamato)bismuth(III) complex,
referred to as I and II, were synthesised and characterised by using a combination of
13
experimental solid-state C/15N CP-MAS NMR, powder and single crystal X-ray diffraction
methods, and periodic DFT/GIPAW computations. The single crystal structure of polymorph
I, having two isomeric non-centrosymmetric binuclear molecules of [Bi2{S2CN(n-C4H9)2}6] at
ꢁ
the supramolecular level, was described in the triclinic system, ꢀ1 space group. On the other
hand, the crystal structure of polymorph II comprising of one kind of centrosymmetric
molecule of [Bi2{S2CN(n-C4H9)2}6] was solved in the monoclinic system, ꢀ2ꢂ/ꢃ space group,
which is similar to, but not identical with, a previously reported crystal structure of the title
complex by Sun et al. [29]. It was also found that polymorph I was readily convertible to
polymorph II by changing the solvent system from acetonitrile to dichloromethane and vice
19

versa during the recrystallisation. We have suggested the crystallographic assignments for the
13C and 15N resonance lines in NMR spectra of both polymorphs by comparing them with the
GIPAW calculated chemical shifts. Because of the large forces on the atoms of the
experimental XRD structures, the best correspondence with the experimental NMR chemical
shifts were obtained when the atomic positions of all atoms in the single crystal structures of
both polymorphs were optimised.
Acknowledgements
Computations were carried out at CSC - the Finnish IT center for science Ltd (Espoo,
Finland). The current project is financed by the Center for Advanced Mining and Metallurgy
(CAMM) at Luleå University of Technology (LTU), Sweden, the Nordic Mining School and
Svenska Kulturfonden at the University of Oulu, Finland and the Academy of Finland
(#218191, #255641, and #285666). We thank the foundation in memory of J. C. and Seth M.
Kempe and LTU labfonden for a grant from which the 400 MHz equipment has been
purchased. The UK 850 MHz solid-state NMR Facility was provided by the University of
Warwick.
Appendix A. Supplementary data
CCDC 1458699 and 1494946 contains the supplementary crystallographic data for
polymorphs
I
(non-centrosymmetric molecules of [Bi2{(S2CN(n-C4H9)2}6] and II
(centrosymmetric molecules of [Bi2{(S2CN(n-C4H9)2}6]. These data can be obtained free of
charge via http://www.ccdc.cam.ac.uk/conts/retrieving.html. or from the Cambridge
Crystallographic Data Centre, 12 Union Road, Cambridge CB2 1EZ, UK; fax: (+44) 1223-
336-033; or e-mail: deposit@ccdc.cam.ac.uk.
References
20

[1] S. Patai, The chemistry of organic arsenic, antimony and bismuth compounds, John Wiley
and Sons, Chichester, 1994.
[2] H. Li, H. Sun, Curr. Opin. Chem. Biol., 16 (2012) 74-83.
[3] G.G. Briand, N. Burford, Chem. Rev., 99 (1999) 2601-2658.
[4] M.D. Regulacio, N. Tomson, S.L. Stoll, Chem. Mater., 17 (2005) 3114-3121.
[5] O.C. Monteiro, H.I.S. Nogueira, T. Trindade, M. Motevalli, Chem. Mater., 13 (2001)
2103-2111.
[6] S.A. Afrillya, Y. Mulyasih, I. Hastiawan, In Proceeding of the International Seminar on
Chemistry, Jatinangor, (2008) 186-190.
[7] F.U. Shah, S. Glavatskih, O.N. Antzutkin, Tribol. Lett., 45 (2012) 67-78.
[8] J. Sedlacek, L.M. Martins, P. Danek, A.J. Pombeiro, B. Cvek, J. Appl. Biomed., 12
(2014) 301-308.
[9] H. Li, C.S. Lai, J. Wu, P.C. Ho, D. de Vos, E.R.T. Tiekink, J. Inorg. Biochem., 101
(2007) 809-816.
[10] G. Hogarth, Mini-Rev. Med. Chem., 12 (2012) 1202-1215.
[11] R.A. Ludwig, G.D. Thorn, Adv. Pest Control Res., 3 (1960) 219-252.
[12] G.D. Thorn, R.A. Ludwig, The dithiocarbamates and related compounds, Elsevier,
Amsterdam, 1962, pp. 224-260.
[13] G. Hogarth, Transition metal dithiocarbamates: 1978–2003, Prog. Inorg. Chem., 53
(2005) 71-561.
[14] A.V. Ivanov, I.V. Egorova, M.A. Ivanov, O.N. Antzutkin, R.S. Tsvykh, Dokl. Phys.
Chem., 454 (2014) 16-20.
[15] S.S. Garje, V.K. Jain, Coord. Chem. Rev., 236 (2003) 35-56.
[16] C.S. Lai, E.R.T. Tiekink, Z. Kristallogr., 222 (2007) 532-538.
[17] H.G. Brittain, Polymorphism in pharmaceutical solids, second ed., New York, CRC
Press 2016.
[18] R.K. Harris, J. Pharm. Pharmacol., 59 (2007) 225-239.
[19] A.V. Ivanov, O.N. Antzutkin, Top. Curr. Chem., 246 (2005) 271-337.
[20] R. Purohit, P. Venugopalan, Reson., 14 (2009) 882-893.
[21] S.E. Ashbrook, D. McKay, Chem. Comm., 52 (2016) 7186-7204.
21

[22] R.K. Harris, R.E. Wasylishen, M.J. Duer, NMR Crystallography, Wiley, New York,
2009.
[23] S. Sneddon, D.M. Dawson, C.J. Pickard, S.E. Ashbrook, Phys. Chem. Chem. Phys., 16
(2014) 2660-2673.
[24] C. Martineau, Solid State Nucl. Magn. Reson., 63 (2014) 1-12.
[25] C. Martineau, A. Cadiau, B. Bouchevreau, J. Senker, F. Taulelle, K. Adil, Dalton Trans.,
41 (2012) 6232-6241.
[26] R.K. Harris, Solid State Sci., 6 (2004) 1025-1037.
[27] V. Gowda, B. Sarma, S. Öberg, V.-V. Telkki, A.-C. Larsson, P. Lantto, O.N. Antzutkin,
Eur. J. Inorg. Chem., 20 (2016) 3278-3291.
[28] V. Gowda, R.S. Laitinen, V.-V. Telkki, A.-C. Larsson, O.N. Antzutkin, P. Lantto,
Dalton Trans. 45 (2016) 19473-19484.
[29] R.Z. Sun, Y.S. Guo, W.M. Liu, S.Y. Chen, Y.Q. Feng, Chinese J. Struct. Chem., 31
(2012) 655-660.
[30] C.J. Pickard, F. Mauri, Phys. Rev. B., 63 (2001) 245101.
[31] S.J. Clark, M.D. Segall, C.J. Pickard, P.J. Hasnip, M.I. Probert, K. Refson, M.C. Payne,
Z. Kristallogr., 220 (2005) 567-570.
[32] J.R. Yates, C.J. Pickard, F. Mauri, Phys. Rev. B, 76 (2007) 024401.
[33] C.F. Macrae, P.R. Edgington, P. McCabe, E. Pidcock, G.P. Shields, R. Taylor, M.
Towler, J. van de Streek, J. Appl.Crystallogr., 39 (2006) 453-457.
[34] B. Fung, A. Khitrin, K. Ermolaev, J. Magn. Reson., 142 (2000) 97-101.
[35] C.R. Morcombe, K.W. Zilm, J. Magn. Reson., 162 (2003) 479-486.
[36] P. Bertani, J. Raya, B. Bechinger, Solid State Nucl. Magn. Reson., 61 (2014) 15-18.
[37] G.M. Sheldrick, Acta Crystallogr. Sect. A, 46 (1990) 467-473.
[38] G.M. Sheldrick, Acta Crystallogr. Sect. A, 64 (2007) 112-122.
[39] Y. Zhang, W. Yang, Phys. Rev. Lett., 80 (1998) 890.
[40] A. Tkatchenko, M. Scheffler, Phys. Rev. Lett., 102 (2009) 073005.
[41] M.D. Segall, P.J. Lindan, M.A. Probert, C.J. Pickard, P.J. Hasnip, S.J. Clark, M.C.
Payne, J. Phys. Condens. Matter., 14 (2002) 2717-2744.
[42] I.Y. Bagryanskaya, Y.V. Gatilov, J. Struct. Chem., 24 (1983) 150-151.
22

[43] D.W Bruce, D. O'Hare, R.I. Walton, Structure from Diffraction Methods: Inorganic
Materials Series, John Wiley and Sons, New York, 2014.
[44] K. Rissanen, Advanced X-ray crystallography, Springer, Heidelberg, 2012.
[45] R.K. Harris, P. Hodgkinson, C.J. Pickard, J.R. Yates, V. Zorin, Magn. Reson. Chem., 45
(2007) S174-S186.
Table 1. X-ray diffraction crystallographic data for polymorphs I and II.
Crystal Data
Empirical formula
Formula wt.
Crystal system
T [K]
I
II
C H54N3S Bi
27822.076
Monoclinic
298 (2)
16.932(3)
10.366(2)
21.817(4)
90.0
99.44(3)
90.0
3777.3(14)
ꢀ2ꢂ/ꢃ
C H54N3S Bi
27822.076
Triclinic
150
a [Å]
b [Å]
c [Å]
α [°]
β [°]
13.7919(3)
16.7642(5)
31.3282(7)
95.316(2)
92.6420(10)
94.0720(10)
7183.9(3)
γ [°]
Volume [Å3]
Space group
Z
ꢁ
ꢀ1
8
1.520
4
Dcalc [g cm–3]
1.446
5.019
1
µ/mm−
5.278
Reflns. collected
Unique reflns.
Observed reflns.
R1 [I>2σ(I)], wR2
Instrument
104145
28205
17957
23705
6644
4396
0.0520, 0.0683
0.0568, 0.1369
Bruker APEX-II CCD Bruker Nonius Kappa CCD
23

Table 2. Experimental and theoretical (GIPAW/PBE) 13C and 15N chemical shifts of
polymorph I.
Sites
Labels a
H-opt b Full-opt b Expt.
13C Chemical shifts
C1, C82
C28, C73
205.4
203.6
207.5
205.3
206.1
205.4
-S2C-N
(bridged)
C19, C100 204.1
203.7
202.8
202.2
200.6
202.3
201.4
201.1
200.2
-S2C-N
(terminal)
C46, C55
C37, C64
C10, C91
201.8
201.7
198.2
Avg.
Avg.
Avg.
Avg.
56.0
28.6
19.6
12.5
1.6
54.8
27.9
20.1
12.8
1.3
54.5
29.9
21.2
15.1
-N-CH2-
-CH2-CH2-
-CH -CH3
-CH23
RMSD c
15N Chemical shifts
N2, N11
180.8
183.4
175.0
169.9
183.0
184.8
8.4
186.6
183.8
182.1
180.5
173.9
168.1
3.0
180.8
180.8
180.2
178.0
174.4
169.1
-S2C-N-
(terminal)
N3, N12
N5, N8
N6, N7
N1, N10
N4, N9
-S2C-N-
(bridged)
RMSD c
a For atom labelling scheme, see Fig. 4
b The stated chemical shifts are averages of the calculated values for the two sites. For
isotropic shielding values of each atom, see Table S5 in the ESI.
c RMSD: Root-mean-square deviation between the experimental and calculated chemical
shifts.
Tables 3. Experimental and theoretical (CASTEP/PBE) 13C and 15N chemical shifts for
polymorphs II and IIs.
Sites
13C Chemical shifts
C3
(Terminal) C2
Labels H-opta Full-opta H-optb Full-optb Expt.b
196.1 204.3
193.1 203.2
207.1 205.6
198.4 205.0
204.7
203.7
-S2C-N
C1
191.4 199.9
196.7 201.3
200.0
-S2C-N
(Bridged)
-N-CH2-
Avg.
55.7
28.1
20.4
14.4
6.1
55.7
28.1
20.4
14.4
0.8
57.7
33.2
21.8
15.3
3.1
56.7
28.6
20.0
14.1
1.2
55.0
29.5
21.2
15.1
-
-CH -CH - Avg.
-CH2-CH23 Avg.
-CH23
Avg.
RMSDc
15N Chemical shifts
24

N1
195.4 176.9
177.3 177.7
177.3
-S2C-N
(Bridged)
N3
N2
172.6 175.0
157.3 169.8
170.1 175.8
163.5 169.9
171.2
167.7
-
-S2C-N
(Terminal)
RMSDc
12.7
2.5
3.3
2.9
a Polymorph IIs (atom labelling scheme: C3=C1, C2=C10, C1=C19, N1=N3, N2=N2, and
N3=N1, see Fig. 5a and S3 in the ESI)
b Polymorph II (for atom labelling scheme, see Fig. 5a)
c
RMSD: Root-mean-square deviation between the experimental and calculated chemical
shifts.
(a)
Calc.
Expt.
10
20
30
40
Position (2θ)
(b)
25

Calc.
Expt.
10
20
30
40
Position (2θ)
Fig. 1 Experimental and theoretically calculated powder X-ray diffraction (PXRD) patterns
of polymorph I (a) and polymorph II (b).
(a)
(b)

13
Fig. 2 C CP-MAS spectra for polymorphs (a) I (100.6 MHz) and (b) II (213.8 MHz). The
intensity of the dithiocarbamate carbon atoms displayed along the y-axis has been doubled
for clarity.
(a)

(b)
15
Fig. 3 N CP-MAS spectra for polymorph I (40.6 MHz) (a) and polymorph II (86.1 MHz)
(b).

Fig. 4 Molecular structures of two non-centrosymmetric dimers of polymorph I, showing the
atom labelling scheme and thermal ellipsoids at the 50% probability level. For clarity, the
alkyl chain carbon atoms have been drawn as ball-and-stick models while the H-atoms are
omitted.
(a)