Host proficiency of N,N′-bis(9-phenyl-9-thioxanthenyl)ethylenediamine for pyridine and the methylpyridine guests–a competition study

Article abstract of DOI:10.1080/10610278.2017.1361037

N,N′-Bis(9-phenyl-9-thioxanthenyl)ethylenediamine proved to be an extremely efficient host compound for pyridine and the isomeric methylpyridines. Furthermore, this host displayed selective behaviour during equimolar guest competitions, consistently favouring 3-methylpyridine in binary, ternary and quaternary experiments. Selectivity orders were 3-methylpyridine >> 4-methylpyridine?>?2-methylpyridine, and 3-methylpyridine?>?pyridine?>?4-methylpyridine?>?2-methylpyridine, for equimolar ternary and quaternary solutions, respectively. When guest concentrations in binary competitions were varied, 3-methylpyridine remained the favoured guest, even at low 3-methylpyridine concentrations. Single crystal X-ray diffraction showed that all four complexes were isostructural (monoclinic, P2₁/n) while guests occupy discrete cavities in the crystal. Only 3-methylpyridine experiences (guest methyl)C–H???π(host) and (guest methyl)C–H???H–Ar(host) interactions, explaining the observed affinity of the host for this guest. DSC analyses provided further affirmation for the host preference: endotherm peak temperatures for the guest release processes correlated directly with the selectivity order for the three methylpyridine isomers.

Full text of DOI:10.1080/10610278.2017.1361037

Supramolecular Chemistry
ISSN: 1061-0278 (Print) 1029-0478 (Online) Journal homepage: http://www.tandfonline.com/loi/gsch20
Host proficiency of N,N′-bis(9-phenyl-9-
thioxanthenyl)ethylenediamine for pyridine and
the methylpyridine guests – a competition study
Benita Barton, Lize de Jager & Eric C. Hosten
To cite this article: Benita Barton, Lize de Jager & Eric C. Hosten (2018) Host proficiency of
N,N′-bis(9-phenyl-9-thioxanthenyl)ethylenediamine for pyridine and the methylpyridine guests – a
competition study, Supramolecular Chemistry, 30:1, 61-71, DOI: 10.1080/10610278.2017.1361037
To link to this article: https://doi.org/10.1080/10610278.2017.1361037
View supplementary material
Published online: 02 Aug 2017.
Submit your article to this journal
Article views: 38
View related articles
View Crossmark data
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=gsch20
Download by: [University of Florida]
Date: 19 December 2017, At: 20:12

Supramolecular chemiStry, 2018
Vol. 30, No. 1, 61–71
https://doi.org/10.1080/10610278.2017.1361037
Host proficiency of N,N′-bis(9-phenyl-9-thioxanthenyl)ethylenediamine for
pyridine and the methylpyridine guests – a competition study
Benita Barton, Lize de Jager and Eric C. Hosten
Department of chemistry, Nelson mandela metropolitan university, port elizabeth, South africa
ABSTRACT
ARTICLE HISTORY
received 23 may 2017
accepted 23 July 2017
N,N′-Bis(9-phenyl-9-thioxanthenyl)ethylenediamine proved to be an extremely efficient host
compound for pyridine and the isomeric methylpyridines. Furthermore, this host displayed
selective behaviour during equimolar guest competitions, consistently favouring 3-methylpyridine
in binary, ternary and quaternary experiments. Selectivity orders were 3-methylpyridine
>> 4-methylpyridine > 2-methylpyridine, and 3-methylpyridine > pyridine >4-methylpyridine >
2-methylpyridine, for equimolar ternary and quaternary solutions, respectively. When guest
concentrations in binary competitions were varied, 3-methylpyridine remained the favoured guest,
even at low 3-methylpyridine concentrations. Single crystal X-ray diffraction showed that all four
complexes were isostructural (monoclinic, P21/n) while guests occupy discrete cavities in the crystal.
Only 3-methylpyridine experiences (guest methyl)C–H∙∙∙π(host) and (guest methyl)C–H∙∙∙H–Ar(host)
interactions, explaining the observed affinity of the host for this guest. DSC analyses provided further
affirmation for the host preference: endotherm peak temperatures for the guest release processes
correlated directly with the selectivity order for the three methylpyridine isomers.
KEYWORDS
host-guest chemistry;
inclusion; selectivity;
pyridine; supramolecular
chemistry
Stereoview showing the (guest methyl)C–H∙∙∙π(host) interaction experienced only by the preferred
guest, 3-methylpyridine
specifically, host-guest chemistry, as a means of achiev-
ing structural isomer separations which may have appli-
cation on an industrial scale. As an example, we have
demonstrated that host (R,R)-(–)-2,3-dimethoxy-1,1,4,4-
tetraphenylbutane-1,4-diol 2 may be utilized in the sep-
aration of isomers of the xylenes by means of selective
inclusion (2).
In this present work, we report on the use of N,N′-
bis(9-phenyl-9-thioxanthenyl)ethylenediamine 3, a host
compound initially designed and synthesised in our
1. Introduction
The applications of host-guest chemistry are manifold,
and the extensive review article featuring the host com-
pound TADDOL 1, and derivatives thereof, by Seebach
et al. (1) commendably conveys this fact. These include
racemate resolutions, enantiomer enrichment, and uses
in chromatography and asymmetric syntheses, amongst
many others.
Our research team has, in recent times, focussed our
energies on using supramolecular chemistry and, more
CONTACT Benita Barton
benita.barton@nmmu.ac.za
the supplemental data for this article can be accessed at https://doi.org/10.1080/10610278.2017.1361037.
© 2017 informa uK limited, trading as taylor & Francis Group

62
B. BARTON ET AL.
laboratories (3), for the inclusion of pyridine and 2-, 3- and
4- methylpyridine (also known as the picolines). We also
report on the plausibility of applying this host compound
in separations of these isomeric picoline guests. We are
not alone in our quest for highly selective host species: for
example, the wheel-and-axle host compound 1,1,6,6-tet-
raphenylhexa-2,4-diyne-1,6-diol was observed to display
selectivity for 4-methylpyridine when recrystallised from
an equimolar binary mixture of this guest and pyridine
(4). Furthermore, the team of Nassimbeni et al. carried out
competition experiments where their host compound,
comprising both rigid fluorenyl and binaphthyl units,
exhibited a selectivity preference for these guests in the
order 4- > 3- > 2- methylpyridine (5).
2. Results and discussion
Synthesis of the title host compound N,N′-bis(9-phenyl-9-
thioxanthenyl)ethylenediamine 3 was achieved by consid-
ering a previously published procedure (3), and involved
first the preparation of 9-hydroxy-9-phenylthioxanthene
5 from thioxanthone 4 using a Grignard addition reaction
with phenylmagnesium bromide followed by an aqueous
acidic work-up. The so-formed alcohol is then treated with
70% perchloric acid and, as a result, a striking ruby red salt
6 precipitates out of the solution. This salt is finally reacted
with ethylenediamine to afford the desired compound 3
in appreciable yield (90%) (Scheme 1).
The focus on methylpyridine separations is not as a result
of a trivial interest entertained by bored chemists, but is
of substantial industrial importance. The boiling points of
the three isomers are very similar and, consequently, their
separation and purification by means of fractional distil-
lation is a cumbersome, expensive and time-consuming
undertaking (6). These compounds have widespread
industrial application as solvents, colourants and as pre-
cursors to a myriad pharmaceutical and agrochemical
compounds, as well as in the preparation of various pol-
ymers and textiles (6). Therefore, a facile and inexpensive
method of separating them from one another after their
recovery from coal tar, based on host-guest chemistry, may
provide an attractive alternative to current methodologies.
In addition, host compounds may readily be recycled and
reused, a characteristic that holds much appeal when the
cost of the separation procedure becomes an important
consideration (Molecules 1–3).
2.1. Assessment of the host proficiency of N,N′-
bis(9-phenyl-9-thioxanthenyl)ethylenediamine 3 for
guests pyridine and 2-, 3- and 4- methylpyridine
As is usual practice, the ability of compound 3 to behave
as a host for pyridine and the three isomers of methylpyri-
dine was assessed by dissolving 3, a solid material, in each
of these potential guests. These experiments were con-
ducted in open vials, thus facilitating slow evaporation of
the guests at ambient temperature and pressure with the
passage of time. In this way, crystallisation ensued, and
the crystals so-formed were collected by suction filtration,
washed thoroughly with petroleum ether (40–60 °C) to
1
remove superficial guest solvent, and analysed by H-
NMR spectroscopy. Table 1 summarises the results of the
host:guest (H:G) ratios that were thus obtained through
integration of relevant host and guest resonances.
These experiments have therefore shown that com-
pound 3 is a proficient host species for the individual
enclathration of pyridine and 2-, 3- and 4- methylpyridine,
including each of these consistently with a 1:1 H:G ratio as
determined by means of 1H-NMR spectroscopy (Table 1).
O
O
MeO
MeO
OH
OH
OH
OH
2.2. Competition studies
2.2.1. Using equimolar binary, ternary and
quaternary mixtures of pyridine and the three
methylpyridine isomers
1
2
Since host 3 is able to individually include pyridine and
the isomeric methylpyridines, we subsequently conducted
competition experiments to ascertain whether the host
shows selectivity in the presence of any of these guests
in mixtures. In order to achieve this, we mixed together,
in equimolar quantities, the various combinations of
any two of these guest species and dissolved the host
compound in each of these binary mixtures. In order to
preserve the equimolar condition, the vials were closed
with their lids, parafilmed and kept at approximately 0 °C
S
HN
NH
S
3

SUPRAMOLECULAR CHEMISTRY
63
S
S
HClO4
ClO4
OH
5
6
1) PhMgBr
2) H (aq)
H2NCH2CH2NH2
S
host compound 3
O
4
Scheme 1. Synthesis of N,N'-bis(9-phenyl-9-thioxanthenyl)ethylenediamine 3.
Table 1. h:G ratiosa of complexes formed by host 3 with pyridine
and the three methylpyridine isomers.
Table 2. results of competition experiments using 3 and various
equimolar binary, ternary and quaternary mixtures of pyridine
and the methylpyridine isomersa,b,c
.
Guest (G)
H:G ratio
Overall H:G
ratio
pyridine
1:1
1:1
1:1
1:1
Py
2MP
3MP
4MP
Guest ratios
2-methylpyridine
3-methylpyridine
4-methylpyridine
x
x
x
0.17:0.83
0.24:0.76
0.70:0.30
0.09:0.64:0.27
0.77:0.23
0.39:0.61
0.62:0.38
0.35:0.11:0.54
0.57:0.12:0.31
0.21:0.51:0.28
0.25:0.07:0.47:0.21
≈1:1
≈1:1
≈1:1
≈1:1
≈1:1
≈1:1
≈1:1
≈1:1
≈1:1
≈1:1
≈1:1
x
x
x
x
x
aDetermined using 1h-Nmr spectroscopy using cDcl3 as the solvent.
x
x
x
x
x
x
x
x
x
x
x
to await crystallisation. Once this occurred, the crystals
were treated in an identical manner as in the single sol-
vent recrystallisation experiments. Due to the overlap of
certain host and guest resonances, GC-MS was selected
over 1H-NMR spectroscopy as the more prudent method
of analysis of the resultant complexes. We also considered
equimolar ternary and quaternary variations in this way,
and hence Table 2 was populated with the obtained data,
where the preferred guest is shown in bold italic font face
for ease of examination.
x
x
x
x
x
x
x
x
x
ah:G ratios were determined using Gc.
bexperiments were conducted in triplicate for confirmation purposes.
cpy = pyridine and mp = methylpyridine.
A ternary experiment involving the three MP isomers,
once again, highlighted the host’s preferential enclathra-
tion of 3-MP, and the preference order noted was 3-MP
(64%) >> 4-MP (27%) > 2-MP (9%), in accordance with
that which might have been expected when considering
results from the binary experiments alone.
Addition of pyridine to these competitions does not
alter the selectivity order of the host for the methylpyri-
dines other than to insert, as its preference, this guest
over 4- and 2- MP. 3-MP remains the favoured guest in
all instances. Equimolar binary experiments with Py/2-MP,
Py/3-MP and Py/4-MP resulted in the preferential extrac-
tion of Py (77%), 3-MP (61%) and Py (62%), respectively.
The overall H:G ratio remains 1:1 in these binary exper-
iments, which was also the favoured ratio in the single
solvent experiments (Table 2).
From this table, it is evident that host 3 consistently
favours 3-MP whenever it is present. In the equimolar
binary experiments involving the methylpyridines, 83 and
70% of 3-MP was extracted by the host when recrystallised
from mixtures of 2- and 3-, and 3- and 4- MP, respectively.
In the absence of 3-MP, 3 remained particular in its behav-
iour, and selected significantly more of 4-MP (76%) than
2-MP (24%).

64
B. BARTON ET AL.
Figure 1. General selectivity curves.
The three ternary experiments where pyridine was
present, Py/2-MP/3-MP, Py/2-MP/4-MP and Py/3-MP/4-MP,
showed selective extractions of 3-MP (54%), Py (57%) and
3-MP (51%), correspondingly.
to 1. Plot (b) indicates a situation where the host com-
pound consistently prefers guest A irrespective of its
concentration in the solution, while Plot (c) is one that
represents a guest concentration-dependent selectivity
profile.
After a careful consideration of the above information,
we obtained the following selectivity plots for each binary
combination of guest species (Figure 2). (Note that the
vials in which these experiments were carried out were
treated in an identical fashion to those in the equimo-
lar experiments mentioned previously.) The straight line
plots in each figure are theoretical and have been inserted
for comparative purposes only, and represent the host’s
behaviour if it were completely unselective over the entire
concentration range assessed.
An experiment where the host was recrystallised from
an equimolar mixture of all four guests, and a consid-
eration of the results of all binary and ternary experi-
ments, therefore allowed us to conclude that the overall
selectivity of this host compound may be written as in
the order 3-MP > Py > 4-Mp > 2-MP. The only experi-
ment which is not in agreement with this summary is
when the crystallising mixture contained Py/3-MP/4-MP,
where 4-MP was (only) slightly favoured over Py (28 ver-
sus 21%); however, here 3-MP remained the preferred
guest solvent.
It is clear from these selectivity plots that host 3,
unsurprisingly, has a considerable preference for 3-MP
whenever it is present: analyses of the binary mixture
experiments where differing molar ratios of 3-MP/2-MP,
3-MP/4-MP and Py/3-MP were utilised shows that the
amount of 3-MP in each crystalline complex was always
greater than the amount present in the solution from
which the crystals had formed (Figures 2(a), (b) and (e),
respectively). Only the last of these experiments, where Py
and 3-MP were mixed together in varying ratios, did the
host display some ambivalence in its selectivity but only
at low concentrations of 3-MP: here, K = 1 (i.e. there was no
selectivity) up until the solution contained approximately
33% 3-MP, after which point the host displayed clear prej-
udice towards Py in favour of 3-MP once more.
In the absence of 3-MP and presence of Py, the host’s
selectivity was consistently for Py, regardless of this guest’s
concentration in the solution [Figures 2(d) (Py/2-MP) and
2f (Py/4-MP), respectively], while experiments involving
2-MP and 4-MP showed the host to have clear partiality
towards 4-MP.
These selectivity experiments further confirm the
preference of the host for these guests to be in the order
3-MP > Py > 4-Mp > 2-MP.
2.2.2. Using non-equimolar amounts of pyridine and
the methylpyridine guests in binary mixtures
Previously we considered only equimolar binary, ternary
and quaternary mixtures of the four guest species. We
subsequently conducted experiments in which the host
compound was recrystallised from binary mixtures of any
two guest species, but where the ratio of each guest in the
mixture was varied systematically. In this way, we could
determine whether the host compound shows a consist-
ent preference for a particular guest irrespective of this
guest’s concentration in the crystallising mixture or not.
In order to achieve this, we were required to analyse both
the crystalline H:G complex and also the solution from
which crystallisation had ensued. Therefore a plot of the
mole fraction of guest A present in the crystal complex (ZA)
against the mole fraction of A in the solution (XA) provided
selectivity curves, and these may be of three general types
(Figure 1) (5).
In Figure 1, Plot (a) represents the scenario where
no selectivity is displayed by the host compound, and
the amount of guest A in the crystal equals that pres-
ent in the solution. Here, the selectivity constant (5),
given by KA:B = ZA/ZB × XB/XA, where XA + XB = 1, is equal

SUPRAMOLECULAR CHEMISTRY
65
Figure 2. (colour online) Selectivity curves for (a) 3-mp/2-mp, (b) 3-mp/4-mp, (c) 2-mp/4-mp, (d) py/2-mp, (e) py/3-mp and (f) py/4-mp as
guest concentrations are varied systematically.
The four complexes were subjected to single crystal
X-ray diffraction (SCXRD) experiments in order to establish
the reasons for the observed selectivity order of host 3.
located on the difference Fourier map. Data were cor-
rected for absorption effects using the numerical method
implemented in SADABS.(8) [CCDC reference numbers
1,549,682 (3•Py), 1,549,683 (3•2-MP) 1,549,684 (3•3-MP)
and 1,549,685 (3•4-MP).]
2.3. Single crystal X-ray crystallographic analyses
of the pyridine and three methylpyridine complexes
with host 3
Table 3 summarises the relevant crystallographic data.
Each of the complexes crystallises in the monoclinic
crystal system and P21/n space group (Table 3). In fact,
upon closer examination of these data, it is clear that all
four of the crystalline compounds are isostructural with
respect to the host packing, but pseudo-isostructural when
considering the guests, since the host molecules may be
overlayed exactly but, due to the differing structures of the
guests, these cannot be overlayed with the same precision.
Unit cells are provided in Figures 3(a)–(d). Note that the
angle β for the 2-MP complex was necessarily selected as
an acute angle (which is not the normal convention) to
ensure that its isostructurality with the other three com-
plexes is correctly represented.
These experiments were conducted at 200 K using a
Bruker Kappa Apex II diffractometer with graphite-mon-
ochromated Mo Kα radiation (λ = 0.71073 Å). APEXII and
SAINT were used for data collection, and cell refinement
and data reduction, respectively (7). SHELXT-2014 (8) was
used to solve the structures, and refined by least-squares
procedures using SHELXL-2016 (8) together with SHELXLE
(9) as a graphical interface. All non-hydrogen atoms were
refined anisotropically. Carbon bound hydrogen atoms
were added in idealised geometrical positions in a rid-
ing model while nitrogen bound hydrogen atoms were

66
B. BARTON ET AL.
Figure 2. (Continued).
The 4-MP guests are seriously disordered over a num-
ber of positions, to the extent that this disorder could not
be modelled satisfactorily (Figure 3(d)). Each of the other
guest molecules, however, assume only one orientation in
their respective crystals (Figures 3(a)–(c)).
contacts for each, ranging overall from 4.782(1) to 5.987(1)
Å] than 2-MP [8 contacts, 4.960(1)–5.950(1)], but this differ-
ence is not at all striking. However, and more significantly,
the most preferred guest (3-MP) is the only guest to expe-
rience a (guest methyl)C–H∙∙∙π(host) (H∙∙∙Cg 2.76 Å, C–H∙∙∙π
138°) and a (guest methyl)C–H∙∙∙H–Ar(host) (2.32 Å, 158°)
interaction; H–G contacts of this type are largely insignif-
icant in the other two complexes. This observation thus
explains the observed affinity for 3-MP.
2.3.1. Significant host-guest interactions
We consequently considered the host–guest interactions
present in each of these complexes in order to attempt
to ascertain the reasons for the host’s obvious partiality
towards 3-MP. Note that, due to the difficulty in modelling,
adequately, the significant disorder in the 4-MP guests, the
resultant host-guest bond distances and angles were dis-
counted in this present investigation. Consequently, only
the interactions present in the other three complexes have
been summarised in Table 4.
A close analysis of the intermolecular host-guest inter-
actions revealed some interesting truths (Table 4). Firstly,
there exist more host–guest π–π interactions in the case of
the 3-MP and Py complexes, the more favoured guests [9
Surprisingly, any intermolecular H-bonds present are
non-classical, despite the H-bond acceptor and donor
capabilities of the guest and host species, respectively.
2.3.2. Guest accommodation type
Of some interest to us was the nature of the guest
accommodation, since the relative thermal stabilities
of complexes have been related to whether the guests
find themselves in channels or cavities (10). We therefore
omitted the guest molecules from the packing calculation
using the Mercury CSD 3.5.1 software package. Figure 4 is

SUPRAMOLECULAR CHEMISTRY
67
Figure 2. (Continued).
a representative example of the resultant void diagram for
the host 3•Py inclusion complex.
In each complex, the guests occupy discrete voids with
two guest molecules contained in each void. We therefore
expect some thermal stability for all four complexes.
The thermal events observed for each complex after
being heated at 10 °C.min−1 are not unexpected (Figures
5(a)–(d)): each DSC trace (medial curve) has mainly two
endotherms, the first representing the guest release step,
and the second the melting of the host compound (≈
177 °C). (An exception here is that the first endotherm
of the pyridine complex appears to be overlapping with
another.) Furthermore, the mass loss expected for each
complex upon complete guest removal is in reasonable
agreement with that expected (Table 5).
2.4. Thermogravimetric (TG) and differential
scanning calorimetry (DSC) analyses of complexes
of host 3 with the four pyridyl guests
We carried out thermal analyses by heating the com-
plexes in open ceramic pans at 10 °C/min. In this way, we
obtained DSC (medial), TG (upper) and the derivative of the
TG (lower) traces. These diagrams are provided in Figures
5(a)–(d).
The relative thermal stabilities of complexes may be
attained by considering the term Ton–Tb (11), where Ton is
the temperature at which the first of the guest is released
from the host crystal and Tb the boiling point of pure guest
solvent. Complexes with higher relative thermal stabilities
are associated with more positive Ton–Tb values. This is even
more applicable when the complexes being compared are
Table 5 contains a summary of the thermal data
obtained from the traces provided in Figures 5(a)–(d).

68
B. BARTON ET AL.
Table 3. relevant single crystal X-ray crystallographic data for complexes of host 3 with py, 2-mp, 3-mp and 4-mp.
Host 3•Py
Host 3•2-MP
Host 3•3-MP
Host 3•4-MP
chemical formula
Formula weight
crystal system
c40h32N2S2•c5h5 N
683.90
monoclinic
P21/n
c40h32N2S2•c6h7 N
697.92
monoclinic
P21/n
c40h32N2S2•c6h7 N
697.92
monoclinic
P21/n
c40h32N2S2•c6h7 N
697.92
monoclinic
P21/n
Space group
μ (mo-Kα)/mm−1
0.195
0.189
0.191
0.190
a/Å
10.1347(3)
13.3006(3)
25.3821(7)
90
91.964(2)
90
10.5651 (6)
13.0737(7)
25.7522(15)
90
87.947(3)
90
10.3996 (4)
13.3612 (6)
25.3108 (11)
90
91.310 (2)
90
10.4008 (8)
13.2791 (10)
25.5934 (19)
90
90.039 (3)
90
b/Å
c/Å
α/°
β/°
γ/°
V/Å 3
3419.44(2)
4
3554.7 (3)
4
3516.1 (3)
4
3534.8 (5)
4
Z
F(000)
temp./K
1440
200
1472
200
1472
200
1472
200
restraints
Nref
0
8508
0
8847
0
8747
0
8822
Npar
R
wr2
459
469
469
436
0.0343
0.0970
1.06
0.0381
0.1041
1.06
0.0368
0.0999
1.03
0.0569
0.1648
1.03
S
θ min-max/°
tot. data
unique data
1.6, 28.3
74,142
8508
1.6, 28.3
160,065
8847
1.6, 28.3
78,757
8747
1.6, 28.4
84,737
8822
observed data [i > 2.0 sigma(i)]
rint
Dffrn measured fraction θ full
min. resd. dens. (e/Å3)
max. resd. dens. (e/Å3)
6832
0.035
1.000
−0.25
7654
0.019
0.998
−0.28
7099
0.023
1.000
−0.26
7448
0.018
0.999
−1.02
0.31
0.35
0.36
0.93
Figure 3. (colour online) unit cells for (a) 3•py, (b) 3•2-mp, c) 3•3-mp, and (d) 3•4-mp; host molecules are shown in ball-and-stick
representation and guests in space-fill form; only the 4-mp guests are disordered; the host packing in each is isostructural.

SUPRAMOLECULAR CHEMISTRY
69
Table 4. host–guest interactions for host 3•py, host 3•2-mp and host 3•3-mpa,b
.
Non-covalent interaction
Host 3•Py
Host 3•2-MP
Host 3•3-MP
Symmetry
−1+x, y, z
π-π
ch∙∙∙π
h-bonding
Short contacts
4.782(1)–5.975(1) Å (9 contacts) 4.960(1)–.950(1) Å (8 contacts)
5.008(1)–5.987(1) Å (9 contacts)
2.76 Å, 138° (guest methyl)c–h∙∙∙π(host)
None
2.32 Å, 158° (<) (guest methyl)c−h∙∙∙h– 1−x, 1−y, 1−z
ar(host)
None
None
None
None
None
None
aDistances denoted by < are contacts that measure less than the sum of the van der Waals radii of the atoms involved.
b4-mp experienced significant disorder which could not be modelled satisfactorily, and hence these h−G interactions are not included here.
Figure 4. (colour online) Discrete voids are present in the host crystal after guest removal; these were calculated using the host 3•py
inclusion complex; two guests are accommodated in each void.
isostructural, as is the case in all four of the complexes in
the present investigation. The value of Ton is estimated from
the derivative of the TG (lower curve).
Considering this term, the relative thermal stabilities
of the complexes may be written as in the order Py >>
3-Mp > 2-Mp > 4-MP. In this present case, the selectivity
and relative thermal stability orders according to Ton–Tb
data therefore do not correlate.
The other temperature value of interest, in our opinion,
is that at which guest release is most rapid, and this may
be obtained by considering the peak temperature of the
endotherm (from the DSC, medial curve) that represents
the guest release process (Tp, Table 5). This temperature is
related to how strongly a guest is held in the crystal, and
more energy in the form of heat is required to mobilise
the host structure and facilitate guest release when guest
molecules are tightly constrained. The pyridine complex
Table 5. thermal properties of complexes formed by host 3 with
pyridine and the methylpyridines.
Actual
mass
loss
Guest
(G)
Mass loss
meas-
Ton/°C Tb/°C (Ton−Tb)/°C Tp/°C expected/% ured/%
py
91
116
−25
137,
142
95
133
119
11.7
12.4
2-mp
3-mp
4-mp
64
83
78
128
144
145
−64
−61
−67
13.4
13.4
13.4
11.2
14.0
10.8
produces a DSC that has two overlapping endotherms
representing the escape of this guest, the first at 137 and
the second at 142 °C, which is not that different to that
for the release of 3-MP (133 °C). We noted from the earlier
selectivity curve that pyridine alone competed at least
to some extent with 3-MP (Figure 2(e)), and the similar Tp

70
B. BARTON ET AL.

SUPRAMOLECULAR CHEMISTRY
71
values may be a reflection of this fact. The other Tp values
follow an identical trend to the selectivity order for this
host: Tp values decrease in the order 3-MP (133 °C) > 4-MP
(119 °C) > 2-MP (95 °C).
400 MHz Bruker Avance Ultrashield Plus 400 spectrometer
or Varian Gemini 200 spectrometer. Thermal experiments
were conducted using a TA SDT Q600 Module system and
analysed using TA Universal Analysis 2000 data analysis
software. Samples were placed in open ceramic pans with
an empty ceramic pan functioning as a reference. High
purity nitrogen gas was used as purge gas. The GC-MS
experiments were conducted using an Agilent 7890A gas
chromatograph coupled with an Agilent 5975C VL mass
spectrometer. The column was an Agilent J&W DB-wax col-
umn [30 m × 250 μm × 0.25 μm (calibrated)], and helium
was used as carrier gas.
3. Conclusions
N,N′-Bis(9-phenyl-9-thioxanthenyl)ethylenediamine, read-
ily prepared from thioxanthone, was recrystallised from
pyridine and the methylpyridines individually, and enclath-
ration of each one occurred consistently with a 1:1 H:G
ratio. Recrystallisation of this host from various equimolar
binary mixtures showed the host to display discrimination
in favour of 3-MP. From an equimolar ternary mixture of
the three methylpyridines, the host extracted 3-MP (64%)
>> 4-MP (27%) > 2-MP (9%), while from a similar mix-
ture with added pyridine, the extraction was in the order
3-MP (47%) > Py (25%) > 4-MP (21%) > 2-MP (7%). Binary
guest:guest ratios were also varied, but the host remained
consistently selective for 3-MP, even at low concentrations
of this guest, except in the 3-MP/Py experiment where Py
was favoured up until a 33 mol % concentration of 3-MP.
SCXRD experiments revealed that all four of the com-
plexes are isostructural (monoclinic, P21/n). Furthermore,
the 4-MP guest had significant disorder and much of this
disorder could not be modelled.
4.2. Synthesis of N,N′-bis(9-phenyl-9-thioxanthenyl)
ethylenediamine 3
Thetitlehostcompound,N,N′-bis(9-phenyl-9-thioxanthenyl)
ethylenediamine, was synthesized according to a pub-
lished procedure (3).
Acknowledgements
Financial support is acknowledged from the Nelson Mandela
Metropolitan University, and L. Bolo and K. Ngobo are thanked
for thermal analyses. Prof. Mino Caira (University of Cape Town)
is gratefully acknowledged for fruitful SCXRD discussions.
In each complex, guest occupation takes place in dis-
crete cavities, not channels, and each void accommodates
two guest molecules.
Disclosure statement
No potential conflict of interest was reported by the authors.
A study of the host–guest interactions obtained from
SCXRD analyses showed that 3-MP experiences both a
(guest methyl)C–H∙∙∙π(host) (H∙∙∙Cg 2.76 Å, C–H∙∙∙π 138°)
and a (guest methyl)C–H∙∙∙H–Ar(host) (2.32 Å, 158°) inter-
action which are absent in the other two complexes. Also,
2-methylpyridine, the least favoured guest species, is
involved in fewer π–π stacking contacts than pyridine and
3-MP, accounting for the discrimination against this guest.
Thermal analyses showed that the selectivity order
of the host for the methylpyridine isomers was in direct
accordance with the DSC endotherm peak temperatures
for the guest release processes (Tp).
References
(1) Seebach, D.; Beck, A.K.; Heckel, A. Angew. Chem., Int. Ed.
2001, 40, 92–138.
(2) Barton, B.; Hosten, E.C.; Pohl, P.L. Tetrahedron 2016, 72,
8099–8105.
(3) Barton, B.; McCleland, C.W.; Taljaard, B. S. Afr. J. Chem. 2002,
55, 144–148.
(4) Bacsa, J.; Caira, M.R.; Jacobs, A.; Nassimbeni, L.R.; Toda, F.
Cryst. Eng. 2000, 3, 251–261.
(5) Nassimbeni, L.R.; Marivel, S.; Su, H.; Weber, E. RSC. Adv.
2013, 3, 25758–25764.
(6) https://chemicalland21.com/industrialchem/organic/
PICOLINES.htm
(7) APEX2, SADABS and SAINT; Bruker AXS: Madison, WI, 2010.
(8) Sheldrick, G.M. Acta Crystallogr. C 2015, 71, 3–8.
(9) Hübschle, C.B.; Sheldrick, G.M.; Dittrich, B. J. Appl.
Crystallogr. 2011, 44, 1281–1284.
4. Experimental
4.1. General methods
(10) Tanaka, K.; Hori, K.; Tsuyuhara, S.; Motoki, S.; Shide, S.;
Arakawa, R.; Caira, M.R. Tetrahedron 2013, 69, 1120–
1127.
(11) Caira, M.R.; Nassimbeni, L.R.; Niven, M.L.; Schubert, W.;
Weber, E.; Dörpinghaus, N. J. Chem. Soc., Perkin Trans. 2,
1990, 2129–2133.
Melting points were recorded on a Stuart SMP10 melting
point apparatus and are uncorrected. Infrared spectra
were recorded on a Bruker Tensor 27 Fourier Transform
Infrared spectrometer or a Perkin Elmer 1600 Series
Fourier Transform Infrared spectrometer, and 1H-NMR on a