The mixed diol-dithiol 2,2-bis-(sulfanylmethyl)propane-1,3-diol: Characterization of key intermediates on a new synthetic pathway

Article abstract of DOI:10.1107/S0108270110049371

A new synthetic route to 2,2-bis(sulfanylmethyl)propane-1,3-diol, (II), is described starting from the commercially available 2,2-bis(hydroxymethyl) propane-1,3-diol. The structures of two intermediates on this route are described. 5,5-Dimethenyl-2,2-dimethyl-1,3-dioxane bis(thiocyanate) (systematic name: {[5-(cyanosulfanyl)-2,2-dimethyl-1,3-dioxan-5-yl]sulfan-yl}formo-nitrile), C₁₀H₁₄N₂O₂S₂, (X), crystallizes in the space group P2₁/c with no symmetry relationship between the two thio-cyanate groups. There is a short intra-molecular N...S contact for one thio-cyanate group, while the second group is positioned such that this type of interaction is not possible. 1,3-(Hydroxymethyl)propane-1,3- diyl bis(thiocyanate), C₇H₁₀N₂O ₂S₂, (XI), also features a single short N...S contact in the solid state. Hydrogen bonding between two mol-ecules of compound (XI) results in the formation of dimers in the crystal, which are then linked together by a second hydrogen-bond interaction between the dimers. In addition, the structures of two inter-mediates from an unsuccessful alternative synthesis of (II) are reported. 2,2-Bis(chloromethyl)propane-1,3-diol, C₅H ₁₀Cl₂O₂, (VI), crystallized as an inversion twin with a minor twin fraction of 0.43 (6). It forms a zigzag structure as a result of intermolecular hydrogen bonding. The structure of 9,9-di-methyl-2,4,8, 10-tetraoxa-3λ⁴-thiaspiro-[5.5]undecan-3-one, C ₈H₁₄O₅S, (VII), shows evidence for a weak S...O contact with a distance of 3.2529 (11) A.

Full text of DOI:10.1107/S0108270110049371

organic compounds
Acta Crystallographica Section C
Crystal Structure
Communications
large degree of hydrogen bonding in the solid state. In
contrast, the mixed diol–dithiol 2,2-bis(sulfanylmethyl)pro-
pane-1,3-diol, (II) (see Scheme 2), has received much less
attention in the literature. The synthesis of (II) was first
reported by Backer & Tamsma (1938), with further routes
reported by Peppell & Signaigo (1946), Bladon & Owen
ISSN 0108-2701
The mixed diol–dithiol 2,2-bis-
(sulfanylmethyl)propane-1,3-diol:
characterization of key intermediates
on a new synthetic pathway
˚
(1950) and Nygard (1967). Compound (II) has clear potential
as a starting material for functionalized dithiol ligands, and it is
therefore highly surprising that there have been no reports of
˚
(II) in the literature since Nygard’s paper. None of the
reported syntheses of (II) involves a direct route from (I) and,
given the availability of the latter, this seemed to be an
obvious avenue to explore in producing large amounts of (II)
with minimal synthetic effort.
Trevor R. Simmons, Christopher J. Pickett and Joseph A.
Wright*
Energy Materials Laboratory, School of Chemistry, University of East Anglia, Norwich
NR4 7TJ, England
Correspondence e-mail: joseph.wright@uea.ac.uk
Received 27 September 2010
Accepted 25 November 2010
Online 8 December 2010
A new synthetic route to 2,2-bis(sulfanylmethyl)propane-1,3-
diol, (II), is described starting from the commercially available
2,2-bis(hydroxymethyl)propane-1,3-diol. The structures of two
intermediates on this route are described. 5,5-Dimethenyl-
2,2-dimethyl-1,3-dioxane bis(thiocyanate) (systematic name:
{[5-(cyanosulfanyl)-2,2-dimethyl-1,3-dioxan-5-yl]sulfanyl}-
formonitrile), C10H14N2O2S2, (X), crystallizes in the space
group P21/c with no symmetry relationship between the two
thiocyanate groups. There is a short intramolecular Nꢀ ꢀ ꢀS
contact for one thiocyanate group, while the second group is
positioned such that this type of interaction is not possible. 1,3-
(Hydroxymethyl)propane-1,3-diyl bis(thiocyanate), C7H10-
N2O2S2, (XI), also features a single short Nꢀ ꢀ ꢀS contact in
the solid state. Hydrogen bonding between two molecules of
compound (XI) results in the formation of dimers in the
crystal, which are then linked together by a second hydrogen-
bond interaction between the dimers. In addition, the
structures of two intermediates from an unsuccessful alter-
native synthesis of (II) are reported. 2,2-Bis(chlorometh-
yl)propane-1,3-diol, C5H10Cl2O2, (VI), crystallized as an
inversion twin with a minor twin fraction of 0.43 (6). It forms
a zigzag structure as a result of intermolecular hydrogen
bonding. The structure of 9,9-dimethyl-2,4,8,10-tetraoxa-3ꢀ4-
thiaspiro[5.5]undecan-3-one, C8H14O5S, (VII), shows
evidence for a weak Sꢀ ꢀ ꢀO contact with a distance of
˚
3.2529 (11) A.
Comment
The structure of 2,2-bis(hydroxymethyl)propane-1,3-diol
[‘pentaerythritol’, (I), see Scheme 1] was first reported by
Llewellyn et al. (1937), and a search of the Cambridge Struc-
tural Database (CSD, Version 5.31; Allen, 2002) reveals an
additional 12 structural reports to date. This is unsurprising as
(I) is a cheap commercially available material and exhibits a
The first strategy examined for the synthesis of (II) is
outlined in Scheme 1. Conversion of (I) first to the acetal,
Acta Cryst. (2011). C67, o1–o5
doi:10.1107/S0108270110049371
# 2011 International Union of Crystallography o1

organic compounds
Figure 1
Figure 3
The structure of (VI), showing the atom-numbering scheme. Displace-
ment ellipsoids are drawn at the 50% probability level and H atoms are
shown as small spheres of arbitrary radii.
The structure of (VII), showing the atom-numbering scheme. Displace-
ment ellipsoids are drawn at the 50% probability level and H atoms are
shown as small spheres of arbitrary radii.
Figure 4
The structure of (X), showing the atom-numbering scheme. Displacement
ellipsoids are drawn at the 50% probability level and H atoms are shown
as small spheres of arbitrary radii.
component of 0.43 (6). The hydroxy groups in (VI) lead to an
extended two-dimensional hydrogen-bonding network (Fig. 2
and Table 1), which links adjacent molecules to form a layer
structure perpendicular to the b axis. This comprises a zigzag
arrangement of hydrogen bonds, which hold the alternating
molecules of (VI) together. The location of the hydrogen-
bonding framework at the centre of the layer leaves the Cl
atoms on the ‘outside’ of the layer.
The reaction to form a cyclic sulfur-containing ester, (VII)
(Fig. 3), is similar to that reported by Rao et al. (2009), who
showed that the reaction of (II) directly with SOCl2 yields the
bis(ester), (VIII) (see below). The geometry of compound
Figure 2
The two-dimensional hydrogen-bond network in (VI). H atoms, except
for H2 and H3, have been omitted for clarity. The view is approximately
along (100).
(III), and then to the dichloride, (IV), was expected to be
readily achieved. From (IV), the next target would be the
bis(isothiouronium) salt, (V), with the intention of one-pot
conversion of (V) to (II) in analogy to the formation of
ethane-1,2-dithiol (Speziale, 1963).
Formation of acetal (III) from (I) presented no difficulties.
However, the reaction of (III) with thionyl chloride (SOCl2)
was more challenging than initially anticipated. Reaction of
(III) with SOCl2 in pyridine at room temperature led to the
isolation of two distinct crystalline solids after column chro-
matography. Both materials appeared to give spectroscopic
data which would be consistent with the desired material.
However, examination by X-ray methods revealed that the
two solids were in fact (VI) and (VII) (Scheme 1). Presumably
the formation of (VI) was due to hydrolysis of the desired
material, (IV), on standing for crystallization (which took
place in the air). Compound (VI) (Fig. 1) crystallizes in the
space group Cc and the chosen crystal was found to exhibit
inversion twinning, with a final occupancy for the minor
(VII) is similar to that reported for (VIII) and is unremark-
able. Both (VII) and (VIII) show evidence of weak inter-
molecular interactions. In (VII), there is a close contact
between symmetry-related S O groups, with an S41ꢀ ꢀ ꢀO410
0
˚
distance of 3.2529 (11) A [symmetry code: ( ) 2 ꢁ x, 1 ꢁ y,
ꢂ
o2 Simmons et al.
C5H10Cl2O2, C8H14O5S, C10H14N2O2S2 and C7H10N2O2S2
Acta Cryst. (2011). C67, o1–o5

organic compounds
sented no significant difficulty. Compound (IX) could then be
reacted with potassium thiocyanate in dimethyl sulfoxide
(DMSO) to yield (X). This reaction proceeded much more
readily in dimethyl sulfoxide than in the commonly employed
dimethylformamide (DMF), presumably because of the
differing basicity of the two solvents. Compound (X) could be
crystallized as white needles from hexane–ethyl acetate
(Fig. 4). Notably, compound (X) crystallizes in the space group
P21/c with no symmetry relationship between the two thio-
cyanate groups. In contrast, the only other reported bis(thio-
cyanate) structure, for methylene bis(thiocyanate), crystallizes
in the space group I2/c and does exhibit a symmetry rela-
tionship between the two groups (Konnert & Britton, 1971).
The latter structure exhibits a short intermolecular Nꢀ ꢀ ꢀS
Figure 5
The structure of (XI), showing the atom-numbering scheme. Displace-
ment ellipsoids are drawn at the 50% probability level and H atoms are
shown as small spheres of arbitrary radii.
˚
contact [3.17 (1) A]. In (X), there is one intramolecular
˚
contact [S4ꢀ ꢀ ꢀN51 = 3.2377 (16) A], while the locations of S5
and N41 preclude contacts for these atoms. The molecular
geometry of (X) is unremarkable.
2 ꢁ z], while in (VIII) there are a number of Sꢀ ꢀ ꢀO distances
˚
in the range 3.308 (3)–3.315 (3) A.
Treatment with acid removed the acetal-protecting group
from (X) to yield (XI) (Fig. 5). Compound (XI) crystallizes in
the space group Pbca and as in (X) there is no symmetry
relationship between the two thiocyanate groups. The intra-
An alternative route to the desired chloride was explored
by reacting (III) with 4-toluenesulfonyl chloride in pyridine.
This led to a crystalline material with spectroscopic data fully
in agreement with the structure of (IV). While a diffraction
study on this new material showed the correct connectivity for
(IV), high residual values prevented a satisfactory completion
of the refinement; application of a twin law failed to improve
the residual values obtained.
With (IV) available, reaction to form the desired salt, (V),
was attempted. Reaction under a range of conditions failed to
yield any material with spectroscopic characteristics matching
those expected for the desired material. Given this failure, an
alternative route to (II) was devised via the isothiocyanate
compound, (XI) (see Scheme 2).
˚
molecular contact distance S5ꢀ ꢀ ꢀN41 of 3.264 (2) A is similar
to that in (X) and again larger than in the previously reported
methylene bis(thiocyanate) (Konnert & Britton, 1971).
Hydrogen bonding between molecules of (XI) occurs in two
ways. Firstly, a pair of symmetry-related O—Hꢀ ꢀ ꢀO hydrogen
bonds create ‘dimers’ of molecules. These dimers are then
linked in an extended ribbon parallel to the a axis by a second
set of symmetry-related hydrogen bonds between a hydroxy
group and atom N51 of the thiocyanide group (Fig. 6 and
Table 2). The ribbons are essentially flat without twists, and
the ribbon plane lies parallel to (012).
Starting from (III), the formation of (IX) by reaction with
4-toluenesulfonyl chloride in the presence of pyridine pre-
Final reduction of (XI) to (II) with lithium aluminium
hydride proceeded cleanly, and (II) was crystallized from
Figure 6
The hydrogen-bond network in (XI). H atoms, except for H2 and H3, have been omitted for clarity. The view is approximately along (010).
ꢂ
Acta Cryst. (2011). C67, o1–o5
Simmons et al.
C5H10Cl2O2, C8H14O5S, C10H14N2O2S2 and C7H10N2O2S2 o3

organic compounds
Table 1
Hydrogen-bond geometry (A, ) for (VI).
dichloromethane–diethyl ether. The resulting material gave
spectroscopic data consistent with the formulation as (II).
In summary, a new synthesis of (II) has been disclosed
starting from the commercial tetraol, (I). The structures of a
number of key intermediates on this pathway have been
presented.
ꢃ
˚
D—Hꢀ ꢀ ꢀA
D—H
Hꢀ ꢀ ꢀA
Dꢀ ꢀ ꢀA
D—Hꢀ ꢀ ꢀA
O2—H2ꢀ ꢀ ꢀO3i
0.77 (4)
0.73 (3)
1.91 (4)
1.93 (3)
2.653 (2)
2.657 (2)
165 (3)
172 (3)
O3—H3ꢀ ꢀ ꢀO2ii
1
2
1
2
1
2
Symmetry codes: (i) x þ 1; y; z; (ii) x ꢁ ; ꢁy þ ; z þ .
Experimental
Data collection
Compounds (VI) and (VII) were obtained by reaction of (III) with
thionyl chloride in pyridine at 273 K. Column chromatography on
silica (hexane–ethyl acetate 4:1 v/v) yielded two fractions, one of
which was initially an oil and the other of which was a solid. The oil
crystallized after standing in the air for several days to yield yellow
crystals of (VI), while sublimation of the solid fraction at 1.3 Pa gave
colourless crystals of (VII). Compound (X) was formed by the
reaction of (IX) with potassium thiocyanate in dimethyl sulfoxide at
398 K. After aqueous work-up, column chromatography on silica
(hexane–ethyl acetate 1:2 v/v) gave a clear oil which crystallized after
standing overnight. Reaction of compound (X) with lithium alu-
minium hydride in tetrahydrofuran at 313 K followed by aqueous
work-up gave (XI) as an off-white solid. This was recrystallized from
hot chloroform. Full details of the syntheses of all the reported
compounds and spectroscopic data are available in the archived CIF.
Oxford Diffraction Xcalibur 3/CCD
diffractometer
Absorption correction: multi-scan
(CrysAlis Pro; Oxford Diffrac-
tion, 2009)
12831 measured reflections
2222 independent reflections
1805 reflections with I > 2ꢄ(I)
Rint = 0.028
Tmin = 0.771, Tmax = 1.000
Refinement
R[F2 > 2ꢄ(F2)] = 0.029
wR(F2) = 0.079
S = 1.06
129 parameters
H-atom parameteꢁrs3 constrained
˚
Áꢅmax = 0.31 e A
Áꢅmin = ꢁ0.39 e A
ꢁ3
˚
2222 reflections
Compound (X)
Crystal data
3
˚
V = 1231.38 (5) A
Z = 4
C10H14N2O2S2
Mr = 258.35
Compound (VI)
Monoclinic, P21=c
Mo Kꢂ radiation
ꢃ = 0.42 mmꢁ1
T = 140 K
0.70 ꢄ 0.08 ꢄ 0.08 mm
˚
a = 9.3934 (2) A
Crystal data
˚
b = 8.5696 (2) A
3
˚
C5H10Cl2O2
Mr = 173.03
Monoclinic, Cc
a = 6.1635 (3) A
b = 19.6495 (10) A
˚
c = 6.3889 (4) A
ꢁ = 96.617 (5)ꢃ
V = 768.60 (7) A
Z = 4
˚
c = 15.7918 (4) A
ꢁ = 104.380 (2)ꢃ
Mo Kꢂ radiation
ꢃ = 0.77 mmꢁ1
T = 140 K
0.40 ꢄ 0.20 ꢄ 0.10 mm
˚
Data collection
˚
Oxford Diffraction Xcalibur 3/CCD
diffractometer
Absorption correction: multi-scan
(CrysAlis Pro; Oxford Diffrac-
tion, 2009)
17624 measured reflections
2791 independent reflections
2222 reflections with I > 2ꢄ(I)
Rint = 0.030
Data collection
Oxford Diffraction Xcalibur 3/CCD
diffractometer
Absorption correction: multi-scan
(CrysAlis Pro; Oxford Diffrac-
tion, 2009)
5628 measured reflections
1707 independent reflections
1609 reflections with I > 2ꢄ(I)
Rint = 0.028
Tmin = 0.840, Tmax = 1.000
Refinement
R[F2 > 2ꢄ(F2)] = 0.029
wR(F2) = 0.077
S = 1.01
147 parameters
H-atom parameteꢁrs3 constrained
Tmin = 0.776, Tmax = 1.000
˚
Áꢅmax = 0.37 e A
ꢁ3
˚
2791 reflections
Áꢅmin = ꢁ0.25 e A
Refinement
R[F2 > 2ꢄ(F2)] = 0.028
wR(F2) = 0.075
S = 1.08
1707 reflections
91 parameters
2 restraints
H atoms treated by a mixture of
independent and constrained
refinement
Compound (XI)
ꢁ3
˚
Áꢅmax = 0.46 e A
ꢁ3
˚
Áꢅmin = ꢁ0.23 e A
Crystal data
Absolute structure: Flack (1983),
825 Friedel pairs
Flack parameter: 0.43 (6)
3
˚
V = 1964.8 (2) A
Z = 8
C7H10N2O2S2
Mr = 218.29
Orthorhombic, Pbca
Mo Kꢂ radiation
ꢃ = 0.51 mmꢁ1
T = 140 K
0.39 ꢄ 0.13 ꢄ 0.01 mm
˚
a = 8.6971 (5) A
˚
b = 10.3972 (5) A
Compound (VII)
˚
c = 21.7280 (15) A
Crystal data
Data collection
3
˚
V = 984.49 (15) A
Z = 4
Mo Kꢂ radiation
ꢃ = 0.32 mmꢁ1
T = 140 K
C8H14O5S
Mr = 222.25
Oxford Diffraction Xcalibur 3/CCD
diffractometer
Absorption correction: multi-scan
(CrysAlis Pro; Oxford Diffrac-
tion, 2009)
25270 measured reflections
2245 independent reflections
1498 reflections with I > 2ꢄ(I)
Rint = 0.082
Monoclinic, P21=c
˚
a = 13.1287 (10) A
˚
b = 6.0588 (5) A
˚
c = 12.5024 (12) A
0.80 ꢄ 0.20 ꢄ 0.06 mm
ꢁ = 98.135 (8)ꢃ
Tmin = 0.926, Tmax = 1.000
ꢂ
o4 Simmons et al.
C5H10Cl2O2, C8H14O5S, C10H14N2O2S2 and C7H10N2O2S2
Acta Cryst. (2011). C67, o1–o5

organic compounds
Table 2
Hydrogen-bond geometry (A, ) for (XI).
ware used to prepare material for publication: SHELXL97, enCIFer
(Allen et al., 2004) and publCIF (Westrip, 2010) for (VI) and (VII);
SHELXL97, enCIFer and PARST (Nardelli, 1995) for (X) and (XI).
ꢃ
˚
D—Hꢀ ꢀ ꢀA
D—H
Hꢀ ꢀ ꢀA
Dꢀ ꢀ ꢀA
D—Hꢀ ꢀ ꢀA
O2—H2ꢀ ꢀ ꢀO3i
0.78 (2)
0.75 (2)
1.97 (3)
2.08 (2)
2.744 (2)
2.832 (2)
170 (3)
177 (3)
The authors thank the BBSRC and the EPSRC for funding.
O3—H3ꢀ ꢀ ꢀN51ii
Symmetry codes: (i) ꢁx þ 1; ꢁy þ 1; ꢁz þ 2; (ii) x þ 1; y; z.
Supplementary data for this paper are available from the IUCr electronic
archives (Reference: SK3390). Services for accessing these data are
described at the back of the journal.
Refinement
R[F2 > 2ꢄ(F2)] = 0.040
wR(F2) = 0.069
S = 0.94
2245 reflections
124 parameters
H atoms treated by a mixture of
independent and constrained
refinement
Áꢅmax = 0.31 e A
References
Allen, F. H. (2002). Acta Cryst. B58, 380–388.
Allen, F. H., Johnson, O., Shields, G. P., Smith, B. R. & Towler, M. (2004).
J. Appl. Cryst. 37, 335–338.
Altomare, A., Cascarano, G., Giacovazzo, C. & Guagliardi, A. (1993). J. Appl.
Cryst. 26, 343–350.
ꢁ3
˚
ꢁ3
˚
Áꢅmin = ꢁ0.26 e A
All structure solutions and refinements were carried out from
within the WinGX suite of programs (Farrugia, 1999). All C-bound H
atoms were refined using a riding model (SHELXL97; Sheldrick,
2008), and with Uiso(H) = 1.2Ueq(C) for CH and CH2 groups or
1.5Ueq(C) for methyl groups. Methyl groups were allowed additional
rotational freedom. In structures (VII) and (X), the oxygen-bound H
atoms were initially positioned using the CALC-OH program
(Nardelli, 1999), and both coordinates and Uiso values were freely
refined. The chosen crystal of (VI) was found to exhibit inversion
twinning and the refinement of the absolute structure parameter
yielded a value of 0.43 (6) for the minor twin component.
Backer, H. J. & Tamsma, A. F. (1938). Recl Trav. Chim. Pays-Bas, 57, 1183–
1210.
Bladon, P. & Owen, L. N. (1950). J. Chem. Soc. pp. 591–597.
Farrugia, L. J. (1997). J. Appl. Cryst. 30, 565.
Farrugia, L. J. (1999). J. Appl. Cryst. 32, 837–838.
Flack, H. D. (1983). Acta Cryst. A39, 876–881.
Konnert, J. H. & Britton, D. (1971). Acta Cryst. B27, 781–786.
Llewellyn, F. J., Cox, E. G. & Goodwin, T. H. (1937). J. Chem. Soc. pp. 883–894.
Nardelli, M. (1995). J. Appl. Cryst. 28, 659.
Nardelli, M. (1999). J. Appl. Cryst. 32, 563–571.
˚
Nygard, B. (1967). Ark. Kemi. 28, 89–97.
Oxford Diffraction (2009). CrysAlis Pro. Version 1.171. Oxford Diffraction
Ltd, Yarnton, Oxfordshire, England.
Peppell, W. J. & Signaigo, F. K. (1946). US Patent 2 402 665.
Rao, Z.-Y., Xiao, X., Zhang, Y.-Q., Xue, S.-F. & Tao, Z. (2009). Acta Cryst. E65,
o36.
Sheldrick, G. M. (2008). Acta Cryst. A64, 112–122.
Speziale, A. J. (1963). Org. Synth. Coll. Vol. 4, 401.
Westrip, S. P. (2010). J. Appl. Cryst. 43, 920–925.
For all compounds, data collection: CrysAlis Pro (Oxford
Diffraction, 2009); cell refinement: CrysAlis Pro; data reduction:
CrysAlis Pro; program(s) used to solve structure: SIR92 (Altomare et
al., 1993); program(s) used to refine structure: SHELXL97 (Shel-
drick, 2008); molecular graphics: ORTEP-3 (Farrugia, 1997). Soft-
ꢂ
Acta Cryst. (2011). C67, o1–o5
Simmons et al.
C5H10Cl2O2, C8H14O5S, C10H14N2O2S2 and C7H10N2O2S2 o5