Compounds N,N′-bis(9-cyclohexyl-9-xanthenyl)ethylenediamine and its thio derivative, N,N′-bis(9-cyclohexyl-9-thioxanthenyl)ethylenediamine, as potential hosts in the presence of xylenes and ethylbenzene: Conformational analyses and molecular modelling considerations

Article abstract of DOI:10.1016/j.tet.2019.04.065

Two novel crystalline compounds, N,N′-bis(9-cyclohexyl-9-xanthenyl)ethylenediamine (OED) and its thio derivative, N,N′-bis(9-cyclohexyl-9-thioxanthenyl)ethylenediamine (SED), were designed and synthesized in our laboratories, and assessed for their potential as host compounds for the four C8 aromatic compounds, namely o-, m- and p-xylene (o-Xy, m-Xy, p-Xy), and ethylbenzene (EB). Despite the only difference between the two compounds being the heteroatoms in their B rings, immense behaviour differences were noted: only OED displayed host behaviour in these conditions, clathrating all but m-Xy, while SED failed to form complexes with any of the four organic solvents. These observations prompted an investigation into the conformations of OED and SED through single crystal diffraction (SCXRD) analyses as well as computational studies with surprising results. SCXRD was also employed to analyse the three complexes that successfully formed with OED, and thermal analyses (TA) assisted in understanding the selectivity behaviour of OED when presented with mixed guests.

Full text of DOI:10.1016/j.tet.2019.04.065

Accepted Manuscript
Compounds N,N′-bis(9-cyclohexyl-9-xanthenyl)ethylenediamine and its thio
derivative, N,N′-bis(9-cyclohexyl-9-thioxanthenyl)ethylenediamine, as potential hosts
in the presence of xylenes and ethylbenzene: Conformational analyses and molecular
modelling considerations
Benita Barton, Ulrich Senekal, Eric C. Hosten
PII:
S0040-4020(19)30486-7
https://doi.org/10.1016/j.tet.2019.04.065
TET 30310
DOI:
Reference:
To appear in: Tetrahedron
Received Date: 4 April 2019
Revised Date: 24 April 2019
Accepted Date: 25 April 2019
Please cite this article as: Barton B, Senekal U, Hosten EC, Compounds N,N′-bis(9-
cyclohexyl-9-xanthenyl)ethylenediamine and its thio derivative, N,N′-bis(9-cyclohexyl-9-
thioxanthenyl)ethylenediamine, as potential hosts in the presence of xylenes and ethylbenzene:
Conformational analyses and molecular modelling considerations, Tetrahedron (2019), doi: https://
doi.org/10.1016/j.tet.2019.04.065.
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.

ACCEPTED MANUSCRIPT
GRAPHICAL ABSTRACT
Compounds N,N’-bis(9-cyclohexyl-9-xanthenyl)ethylenediamine and its thio derivative, N,N’-
bis(9-cyclohexyl-9-thioxanthenyl)ethylenediamine, as potential hosts in the presence of
xylenes and ethylbenzene: conformational analyses and molecular modelling considerations
Benita Barton,* Ulrich Senekal and Eric C. Hosten
Department of Chemistry, PO Box 77000, Nelson Mandela University, Port Elizabeth, 6031,
South Africa. E-mail: benita.barton@mandela.ac.za

ACCEPTED MANUSCRIPT
1
Compounds N,N’-bis(9-cyclohexyl-9-xanthenyl)ethylenediamine and its thio derivative, N,N’-
bis(9-cyclohexyl-9-thioxanthenyl)ethylenediamine, as potential hosts in the presence of
xylenes and ethylbenzene: conformational analyses and molecular modelling considerations
Benita Barton,* Ulrich Senekal and Eric C. Hosten
Department of Chemistry, PO Box 77000, Nelson Mandela University, Port Elizabeth, 6031,
South Africa. E-mail: benita.barton@mandela.ac.za
Abstract
Two novel crystalline compounds, N,N’-bis(9-cyclohexyl-9-xanthenyl)ethylenediamine (OED)
and its thio derivative, N,N’-bis(9-cyclohexyl-9-thioxanthenyl)ethylenediamine (SED), were
designed and synthesized in our laboratories, and assessed for their potential as host
compounds for the four C8 aromatic compounds, namely o-, m- and p- xylene (o-Xy, m-Xy, p-
Xy), and ethylbenzene (EB). Despite the only difference between the two compounds being the
heteroatoms in their B rings, immense behaviour differences were noted: only OED displayed
host behaviour in these conditions, clathrating all but m-Xy, while SED failed to form complexes
with any of the four organic solvents. These observations prompted an investigation into the
conformations of OED and SED through single crystal diffraction (SCXRD) analyses as well as
computational studies with surprising results. SCXRD was also employed to analyse the three
complexes that successfully formed with OED, and thermal analyses (TA) assisted in
understanding the selectivity behaviour of OED when presented with mixed guests.
Keywords
Host-guest Chemistry; Supramolecular Chemistry; Xanthenyl; Xylene; Ethylbenzene.

ACCEPTED MANUSCRIPT
2
1. Introduction
Host-guest chemistry is a field of science that is characterized by non-covalent interactions
between at least two different molecular compounds, namely the host and the guest,1,2 and
may be subdivided into two major classes known as the clathrands and the cavitands.
Cavitands are those host compounds that contain permanent intramolecular cavities, and
guests are able to interact with the host both in solution and in the crystalline state.
Clathrands, on the other hand, only possess extramolecular cavities which are present only in
the crystalline or solid states.3 The non-covalent forces that are responsible for entrapping the
guest within the host crystal are, more usually, hydrogen bonding, ion pairing (ion∙∙∙ion,
dipole∙∙∙dipole, ion∙∙∙dipole), π∙∙∙π, C−H∙∙∙π, and van der Waals interactions.1−4 A few examples
of cavitands include the calixarenes, carcerands, cyclodextrins, cyclophanes, cryptophanes,
crown ethers and cryptands, while typical clathrands are the hexa-, roof, scissor, urea and
wheel-and-axle hosts.4 Chemistry of this type has a number of applications in fields such as the
chemical, biological and pharmaceutical industries, and include the separation of enantiomers
and other isomers,5 chromatography,6,7 employment in stationary phases for chromatographic
separations,4,6 storage of toxic substances and gases,3 and various environmental applications,7
amongst numerous others.
Owing to the challenges associated with the separation of the components of the C8 aromatic
fraction through distillation [o-, m-, p- xylene and ethylbenzene (o-Xy, m-Xy, p-Xy, EB), Scheme
1] due to their near-identical physical properties (boiling points: o-Xy 144.5ꢀ°C, m-Xy 139.1ꢀ°C, p-
Xy 138.2ꢀ°C, EB 136.2ꢀ°C), alternative methods for their separation are attractive. To this end,
many researchers have spent much time and capacity on investigating this challenge. As early
as the 1950s, Clark et al8 examined the separation of these compound types by means of the
formation of a 1:1 solid p-Xy:CCl4 complex from a ternary mixture containing this xylene and the
meta isomer. Schaeffer and his co-workers9 also considered complex formation but using
inorganic compounds as complexing material. Other strategies have employed zeolites,10 gas
chromatography,11,12 and metal-organic frameworks comprising vanadium(IV) in the

ACCEPTED MANUSCRIPT
3
structure,13 while Lusi and Barbour made use of a nickel-based Werner complex in their work.14
Organic host-guest chemistry has also been considered here, and calixarenes have been shown
to behave selectively in the presence of these compounds during crystallization experiments.15
Furthermore, Toda et al16 demonstrated that the host compound 1,1,2,2-tetraphenylethane-
1,2-diol possesses the ability to selectively include p-Xy from a p-Xy/m-Xy mixture, and
Nassimbeni and co-workers17 employed host compounds 9,9′-bianthryl, 9,9′-spirobifluorene and
trans-2,3-dibenzoylspiro[cyclopropane-1,9-fluorene] to achieve this separation. A recent review
commendably summarises much of the work that has taken place in the area of the xylene
separations.18
During ongoing investigations in our laboratories that focusses on this challenge, we reported that a
tartaric acid-derived host compound, (R,R)-(−)-2,3-dimethoxy-1,1,4,4-tetraphenylbutane-1,4-diol,
was able to extract only moderate amounts (54%) of p-Xy from a ternary mixture of the
xylenes,19 while N,N′-bis(9-phenyl-9-thioxanthenyl)ethylenediamine fared significantly better,
and 95% of the same isomer was observed in the crystals that emanated during the
recrystallization process.20 The success of this latter host prompted further examination and,
consequently, novel host derivatives were prepared based on a substitution of the
ethylenediamine linker for a trans-cyclohexane-1,4-diamine one, and this afforded two
compounds, namely trans-N,N'-bis(9-phenyl-9-xanthenyl)cyclohexane-1,4-diamine and trans-
N,N'-bis(9-phenyl-9-thioxanthenyl)cyclohexane-1,4-diamine.21 The recrystallization of these
from a quaternary o-Xy/m-Xy/p-Xy/EB mixture showed that both hosts were efficient and
displayed selectivity for the para isomer once more (74 and 71%, respectively).
Owing to the encouraging results obtained for the xanthenyl-type systems, in the present work
we have further modified these structures to afford two novel species, N,N’-bis(9-cyclohexyl-9-
xanthenyl)ethylenediamine (OED) and N,N’-bis(9-cyclohexyl-9-thioxanthenyl)ethylenediamine
(SED), and examined their behaviour in this context (Scheme 1). We were surprised to observe
that, despite the only difference between the two compounds being the heteroatom in their B
rings, OED and SED displayed very different host ability when presented with these guest types.

ACCEPTED MANUSCRIPT
4
In fact, SED did not function as a host and was not able to form complexes with any of the four
organic solvents, while OED was very capable, with the exception of m-Xy, which OED did not
clathrate. This surprising observation prompted an in-depth analysis into the conformations of
the two compounds, both in their crystal structures and through molecular modelling, in order
to elucidate the reasons for the immense host behaviour differences. All successfully-formed
single solvent complexes were also subjected to SCXRD and thermal analyses and these assisted
in explaining the selective behaviour that OED displayed when recrystallized from various
mixtures of these organic solvents. We now report on these findings.
X
NH
m-Xy
o-Xy
NH
X
OED (X = O)
SED (X = S)
EB
p-Xy
Scheme 1. Structures of host N,N’-bis(9-cyclohexyl-9-xanthenyl)ethylenediamine (OED) and N,N’-bis(9-cyclohexyl-
9-thioxanthenyl)ethylenediamine (SED), and guests o-xylene (o-Xy), m-xylene (m-Xy), p-xylene (p-Xy) and
ethylbenzene (EB).
2. Experimental
2.1 General
All starting materials and solvents were purchased from Merck and used without further
purification. The melting points of all synthesized solids were measured on a Stuart SMP10
melting point apparatus and are uncorrected. The 1H- and 13C- NMR spectra were recorded by
means of a 400 MHz Bruker 400 Ultrashield Plus spectrometer, and CDCl3 was used as the
deuterated solvent. These spectra were analyzed using Bruker TopSpin 3.2 data software. The

ACCEPTED MANUSCRIPT
5
infrared spectra were obtained by means of a Bruker Tensor 27 FT-IR spectrometer. All spectra
are provided in the Supplementary Information section (Figures S1–S6).
GC-MS was an appropriate method with which to analyse complexes that were crystallized
from mixed guests. This was achieved by means of an Agilent 7890A gas chromatograph fitted
with an Agilent 5975C VL mass spectrometer, equipped with an Agilent J&W Cyclosil-B column.
An initial temperature of 50 °C was continued for 1 min, which was followed by a heating ramp
of 0.5 °C∙min-1 until 60 °C was reached, and the temperature was maintained there for 1 min.
SCXRD experiments were conducted on suitable crystals of apohosts OED and SED, and the
three single solvent complexes with OED (2OED•o-Xy, 2OED•p-Xy and 2OED•EB). These data
were obtained at 200 K using a Bruker Kappa Apex II diffractometer utilizing a graphite-
monochromated Mo Kα radiation (λ = 0.71073 Å), and analysed by means of APEXII data
software, while the SAINT program was used for data reduction and cell refinement.22 The
structures were solved using SHELXT-2018.23 SHELXL-2018,24 with SHELXLE25 as a graphical
interface, was used to refine the structures by employing least-squares methods. All atoms,
excluding hydrogen, were refined anisotropically. Absorption effects present in the data were
corrected for using SADABS.22 These crystal structures were deposited at the Cambridge
Crystallographic Data Centre, and CCDC numbers are as follows: 1895818 (OED), 1895819
(SED), 18958202 (OED•o-Xy), 1895821 (2OED•p-Xy) and 1895822 (2OED•EB). The Mercury
3.10.2 software package26 was used to construct figures displaying the host‒guest packing as
well as for calculating voids present in the crystal after removal of the guest from the packing
calculation.
DSC and TG experiments were conducted on all of the complexes synthesized in this work.
These thermal data were obtained using a Perkin Elmer Simultaneous Thermal Analyzer (STA)
6000 and analyzed using Pyris Series data software. Open ceramic pans were employed here
with an empty pan serving as reference. The DSC and TG experiments were carried out under

ACCEPTED MANUSCRIPT
6
high purity inert nitrogen gas atmosphere. Samples were heated at 10 °C∙min-1 from
approximately 45 °C to 260 °C.
2.2 General methods for the synthesis of OED and SED
A standard Grignard addition reaction of cyclohexylmagnesium bromide was performed on the
xanthone 1 (Scheme 2), in anhydrous THF, to afford the respective alcohol 2 in moderate to
good yields. To the alcohol, dissolved in a mixture of acetic anhydride and dichloromethane
and cooled in an ice/water bath, was added perchloric acid dropwise. Diethyl ether was then
added to the solution in order to ensure crystallization of the perchlorate salt 3, which was
filtered and washed with diethyl ether, and dried overnight under high vacuum. Afterwards,
the perchlorate salt was dissolved in dichloromethane, and ethylenediamine was added
dropwise. The resulting solution was stirred overnight and then washed with water (3 x 300
mL). The organic layer was separated and dried over sodium sulfate, and the dichloromethane
removed using a rotary evaporator. The resulting gum was crystallized using dichloromethane
and petroleum ether (bp 40–60 °C) to afford the host material, OED or SED, as applicable, which
was dried under high vacuum before it was used in subsequent host-guest experiments.
MgBr
1.
X
X
2. Acidic workup
OH
O
1
2
HClO4
X
a: X = O
b: X = S
X
6
NH
3
2
1
H2NCH2CH2NH2
0
ClO4
1
HN
X
3
OED (X = O)
SED (X = S)

ACCEPTED MANUSCRIPT
7
Scheme 2. Synthetic strategy towards compounds OED and SED
2.2.1 Synthesis of OED
2.2.1.1
9-Cyclohexylxanthen-9-ol (2a).
Magnesium turnings (3.52 g, 144.8 mmol),
bromocyclohexane (23.30 g, 142.9 mmol) and xanthone (13.90 g, 70.8 mmol) yielded 9-
cyclohexylxanthen-9-ol (2a) (8.66 g, 30.9 mmol, 43.6%) as a white solid, mp 160‒162 °C;
ν
max(solid)/cm-1 3394 (OH), 2930 (CH), 2852 (CH), 1600 (Ar) and 1572 (Ar); δH(CDCl3)/ppm 0.69‒
1.16 (5H, m, 2CH2CH and CHCOH), 1.46‒1.76 (6H, m, 3CH2CH2), 2.28 (1H, s, CHCOH), 7.12‒7.41
(6H, m, ArH) and 7.68 (2H, d, ArH); δC(CDCl3)/ppm 26.21 (CH2CH), 26.25 (CH2CH2), 26.82
(CH2CH2), 52.68 (CHCOH), 71.87 (COH), 115.74 (ArC), 122.96 (ArC), 126.91 (ArC), 127.12
(quaternary ArC), 128.52 (ArC) and 151.03 (quaternary ArC) (see Figure S1 in the
Supplementary Information).
2.2.1.2 9-Cyclohexyl-9-xanthenylium perchlorate (3a). 9-Cyclohexylxanthen-9-ol (5.02 g, 17.9
mmol), perchloric acid (1.40 mL), acetic anhydride (5 mL) and diethyl ether (30 mL) afforded 9-
cyclohexyl-9-xanthenylium perchlorate (3a) (6.05 g, 16.7 mmol, 92.7%) as a bright orange solid,
mp 179‒182 °C; νmax(solid)/cm-1 2937 (CH), 2865 (CH) and 1620 (Ar); δH(CDCl3)/ppm 1.84‒2.64
(10H, m, 5CH2), 4.36 (1H, t, CH-C+), 8.06‒8.11 (2H, m, Ar), 8.34‒8.36 (2H, m, Ar), 8.47‒8.49 (2H,
m, Ar), 8.98‒9.00 (2H, m, Ar); δC(CDCl3)/ppm 18.38 (CH2CH), 26.53 (CH2CH2), 28.46 (CH2CH2),
31.22 (CHC+), 58.54 (CHC+), 116.35 (ArC), 122.46 (ArC), 126.82 (quaternary ArC), 127.03 (ArC),
128.07 (ArC) and 154.85 (quaternary ArC) (see Figure S2 in the Supplementary Information).
2.2.1.3
N,N’-Bis(9-cyclohexyl-9-xanthenyl)ethylenediamine (OED).
9-Cyclohexyl-9-
xanthenylium perchlorate (5.09 g, 14.03 mmol) and ethylenediamine (1.16 g, 19.3 mmol)
afforded N,N’-bis(9-cyclohexyl-9-xanthenyl)ethylenediamine (OED) (3.82 g, 6.53 mmol, 46.5%)

ACCEPTED MANUSCRIPT
8
as a white solid, mp 209‒211 °C; νmax(solid)/cm-1 2926 (CH), 2852 (CH), 1598 (Ar) and 1573 (Ar);
δH(CDCl3)/ppm 0.76‒1.14 (10H, m, 4CH2CH and 2CHCNH), 1.56‒1.70 (12H, m, 6CH2CH2), 2.15‒
2.25 (6H, m, 2NHCH2) and 7.12‒7.50 (16H, m, ArH); δC(CDCl3)/ppm 26.40 (CH2CH), 26.83
(CH2CH2), 27.31 (CH2CH2), 44.05 (NHCH2), 54.31 (CHCNH), 60.52 (CHCNH), 115.99 (ArC), 122.52
(ArC), 124.41 (ArC), 127.56 (ArC), 127.80 (quaternary ArC) and 152.72 (quaternary ArC). (Found
for C40H44O2N2: 82.04% C, 7.75% H, 4.71% N. Calculated 82.15% C, 7.58% H, 4.79% N) (see
Figure S3 in the Supplementary Information).
2.2.2 Synthesis of SED
2.2.2.1 9-Cyclohexylthioxanthen-9-ol (2b). Magnesium turnings (3.44 g, 141.5 mmol),
bromocyclohexane (23.04 g, 141.3 mmol) and thioxanthone (15.04 g, 70.9 mmol) yielded 9-
cyclohexylthioxanthen-9-ol (2b) (7.75 g, 26.1 mmol, 37.0%) as a light cream solid, mp 144‒146
°C; νmax(solid)/cm-1 3445 (OH), 2929 (CH), 2850 (CH), 1583 (Ar) and 1561 (Ar); δH(CDCl3)/ppm
1.03‒1.64 (10H, m, 3CH2CH2 and 2CH2CH), 2.02 (1H, t, CHCOH), 2.25 (1H, s, CHCOH), 7.26‒7.44
(6H, m, ArH) and 7.78‒7.79 (2H, m, ArH); δC(CDCl3)/ppm 26.29 (CH2CH), 26.47 (CH2CH2), 26.70
(CH2CH2), 40.90 (CHCOH), 77.75 (CHCOH), 125.84 (ArC), 126.48 (ArC), 126.59 (ArC), 126.85
(ArC), 130.39 (quaternary ArC) and 139.75 (quaternary ArC) (see Figure S4 in the
Supplementary Information).
2.2.2.2 9-Cyclohexyl-9-thioxanthenylium perchlorate (3b). 9-Cyclohexylthioxanthen-9-ol (5.01
g, 16.9 mmol), perchloric acid (1.50 ml), acetic anhydride (5 mL) and diethyl ether (35 mL)
afforded 9-cyclohexyl-9-thioxanthenylium perchlorate (3b) (5.50 g, 14.5 mmol, 86.1%) as a
crimson red solid, mp 124‒126 °C; νmax(solid)/cm-1 2915 (CH), 2857 (CH) and 1628 (Ar);
δH(CDCl3)/ppm 2.04‒2.63 (10H, m, 5CH2), 4.49 (1H, t, CH-C+), 8.17‒8.33 (4H, m, ArH), 8.64‒8.72
(2H, m, ArH) and 9.22–9.24 (2H, d, ArH). [This perchlorate was unstable in the atmosphere and
converted back to the alcohol form (9-cyclohexylthioxanthen-9-ol). Therefore, resonance peaks
in the 13C-NMR spectrum could not be confidently assigned. (See Figure S5 in the
Supplementary Information).]

ACCEPTED MANUSCRIPT
9
2.2.2.3
N,N’-Bis(9-cyclohexyl-9-thioxanthenyl)ethylenediamine (SED).
9-Cyclohexyl-9-
thioxanthenylium perchlorate (5.01 g, 13.2 mmol) and ethylenediamine (1.08 g, 18.0 mmol)
afforded N,N’-bis(9-cyclohexyl-9-thioxanthenyl)ethylenediamine (SED) (2.70 g, 4.38 mmol,
33.2%) as a white solid, mp 238‒240 °C; νmax(solid)/cm-1 2919 (CH), 2848 (CH), 1584 (Ar) and
1558 (Ar); δH(CDCl3)/ppm 1.02 (10H, m, 4CH2CH and 2CHCNH), 1.57‒1.74 (12H, m, 6CH2CH2),
2.20 (2H, s, 2NHCH2), 2.44 (4H, s, 2NHCH2), 7.25‒7.33 (12H, m, ArH) and 7.68‒7.69 (4H, m ArH);
δC(CDCl3)/ppm 26.51 (CH2CH2), 26.81 (CH2CH2), 27.21 (CH2CH), 43.75 (CHCNH), 50.15 (NHCH2),
65.77 (CHCNH), 125.20 (ArC), 125.82 (ArC), 126.58 (ArC), 129.41 (ArC), 131.96 (quaternary ArC)
and 135.64 (quaternary ArC). (Found for C40H44S2N2: 77.60% C, 7.08% H, 4.48% N, 10.11% S.
Calculated 77.88% C, 7.19% H, 4.54% N, 10.39% S) (See Figure S6 in the Supplementary
Information).
2.3 Single solvent experiments
The single solvent experiments were conducted in glass vials. Hosts OED (0.050 g, 0.085 mmol)
or SED (0.050 g, 0.081 mmol) were independently dissolved in an excess of each of the
potential guest compounds (5 mmol). These mixtures were gently heated using a hot water
bath, if necessary, to ensure dissolution of all of the host material. The vials were then left
open at room temperature and at ambient pressure until crystallization occurred. The crystals
were collected using vacuum filtration, washed with petroleum ether (bp 40–60 °C) and
analysed by means of 1H-NMR spectroscopy, with CDCl3 as the solvent, in order to determine if
inclusion had occurred. When complexation was successful, the host:guest (H:G) ratios were
determined through integration of relevant host and guest resonance signals.
2.4 Competition experiments

ACCEPTED MANUSCRIPT
10
2.4.1 Equimolar guest mixtures. Competition experiments were carried out in order to
determine whether hosts OED and SED possessed the ability to discriminate between the
various guests when presented with a guest mixture. The hosts were thus individually
recrystallized, in glass vials, from equimolar binary, ternary and quaternary mixtures of the
xylenes and EB. The vials were closed and stored in a refrigerator at 0 °C. Crystals that formed
were treated in the same way as previously mentioned. The guest:guest (G:G) ratios of the
1
inclusion complexes were determined using GC-MS, and the overall H:G ratio by means of H-
NMR spectroscopy.
2.4.2 Non-equimolar guest mixtures. The host was independently recrystallized from various
binary combinations of the four guest solvents, but the concentrations of these guests were
now varied. The G1:G2 mixtures that were employed varied from molar ratios of 20:80 to
80:20. The vials were treated in the same manner as for the equimolar experiments before,
and the mother liquors as well as the crystals that formed from these mixtures were analysed
using GC-MS. The molar fraction of guest in the crystals (Y) was plotted against the molar
fraction of the same guest in the mother liquor (X) to obtain the selectivity profiles provided in
Figure 1.
2.5 Computational methods
Using Spartan ’10 V1.1.0,27 supplied by Wavefunction Inc., computational methods were carried
out on both OED and SED. A conformational search was thus conducted on each host
compound on the molecular mechanics level using the Monte-Carlo algorithm and MMFF force
field. Both hosts yielded 100 conformers of which only the three lowest energy conformers
(OEDa−c and SEDa−c, in ascending energy order) were used in subsequent calculations with
ground state energetics at the DFT (density functional theory) BLYP/6-31G* through to the
higher DFT B3LYP/6-311G* levels.
3. Results and discussion

ACCEPTED MANUSCRIPT
11
Hosts compounds OED and SED were readily prepared by reacting the xanthone 1 with the
Grignard reagent cyclohexylmagnesium bromide (Scheme 2). After work-up, the resulting
alcohols 2 were treated with perchloric acid to yield salts 3, which were then linked by means
of a reaction with ethylenediamine to afford the title host compounds, OED and SED, in low to
moderate yields.
3.1 Single solvent experiments
Table 1 contains data obtained after recrystallization of OED and SED from each of o-Xy, m-Xy,
p-Xy and EB, and analysis of the resulting crystals by means of 1H-NMR spectroscopy.
Table 1
H:G ratiosa for the single solvent
Host
OED
OED
OED
OED
SED
SED
SED
SED
Guest
o-Xy
m-Xy
p-Xy
EB
o-Xy
m-Xy
p-Xy
EB
H:G ratioa
experiments using OED and SED
2:1
b
2:1
2:1
b
b
b
b
aDetermined using 1H-NMR spectroscopy, with CDCl3 as the deuterated solvent
bNo inclusion occurred
Surprisingly, OED and SED displayed significantly different host ability in the presence of the C8
aromatics. While OED was efficient, clathrating all (with 2:1 H:G ratios) but the m-Xy isomer,
SED failed to form complexes with any of these solvents (Table 1). We also investigated
whether it would be possible to encourage SED to enclathrate p-Xy by preparing a nearly
saturated solution of SED in p-Xy, carefully filtering this solution, and seeding it with crystals of
1
the complex OED∙p-Xy. The crystals that so-formed were analysed by means of H-NMR

ACCEPTED MANUSCRIPT
12
spectroscopy (see Supplementary Information, Figure S7) which showed that only apohost SED
crystallized out and, hence, p-Xy remained uncomplexed despite the presence of the OED∙p-Xy
complex.
We consequently investigated the selectivity behaviour of OED and SED when presented with
various mixtures of these guests.
3.2 Competition experiments
3.2.1 Equimolar guest mixtures. Unsurprisingly, owing to the lack of host ability observed in
the single solvent experiments, SED failed to form any complexes, mixed or otherwise, when
presented with the guest mixtures. As expected, however, OED remained efficient in these
conditions, and formed mixed complexes in most of these experiments when recrystallized
from the equimolar binary, ternary and quaternary guest mixtures. Table 2 summarises the
data obtained (for OED), where preferred guests are highlighted in red.
Table 2
Competition experiment results for OED when recrystallized from equimolar mixtures of o-Xy, m-Xy, p-Xy and EBa
o-Xy
m-Xy
p-Xy
EB
Guest ratios
(% e.s.d.s)b
81.2:18.8
(0.4)
Overall H:G ratio
X
X
X
X
2:1
2:1
X
64.9:35.1
(0.2)
c
c
c
c
X
c
c
X
X
X
X
X
X
X
58.5:41.5
(0.2)
2:1
2:1
2:1
2:1
X
X
X
X
X
59.3:10.6:30.1
(1.6)(0.6)(1.0)
66.2:10.9:22.9
(1.6)(0.9)(0.7)
53.4:27.6:19.0
(1.4)(0.6)(0.8)
X
X
X

ACCEPTED MANUSCRIPT
13
X
X
X
X
X
16.5:48.3:35.2
(0.5)(0.0)(0.5)
49.2:8.1:24.8:17.9
(1.5)(0.2)(0.4)(0.9)
2:1
2:1
X
X
aGC-MS and 1H-NMR spectroscopy were used to obtain the guest and overall H:G ratios, respectively
bThe competition experiments were conducted in triplicate; % estimated standard deviations (% e.s.d.s) are
provided in parentheses
cNo inclusion occurred
For each binary competition experiment in which o-Xy was present (o-XyXy/m-Xy and o-Xy/p-
Xy), with the exception of o-Xy/EB, this was the preferred guest (81.2% and 64.9%, respectively)
(Table 2). The o-Xy/EB experiment afforded crystals, but no guest was included. OED,
furthermore, displayed a slight affinity for p-Xy (58.5%) when it was recrystallized from a p-
Xy/EB solution. Mixtures comprising m-Xy/p-Xy and m-Xy/EB also yielded no complexes in
these experiments.
Competition experiments involving ternary guest mixtures (o-Xy/m-Xy/p-Xy, o-Xy/m-Xy/EB and
o-Xy/p-Xy/EB) all showed that o-Xy remained preferred (59.3%, 66.2% and 53.4%, respectively),
as was the case in the binary experiments. The ternary equimolar m-Xy/p-Xy/EB competition
experiment, in the absence of o-Xy, permitted p-Xy (48.3%) to now be preferred followed by
35.2% EB and 16.5% m-Xy. Therefore, results from these ternary equimolar experiments
suggested a host selectivity order of o-Xy > p-Xy > EB > m-Xy, which was in agreement with the
quaternary competition experiment that was conducted (49.2% o-Xy >24.8% p-Xy > 17.9% EB >
8.1% m-Xy).
The preference of OED for the ortho isomer is unprecedented in our laboratories, and (R,R)-(−)-
2,3-dimethoxy-1,1,4,4-tetraphenylbutane-1,4-diol,19
N,N′-bis(9-phenyl-9-
thioxanthenyl)ethylenediamine,20 trans-N,N'-bis(9-phenyl-9-xanthenyl)cyclohexane-1,4-diamine
and trans-N,N'-bis(9-phenyl-9-thioxanthenyl)cyclohexane-1,4-diamine21 always favoured p-Xy
when these guests competed.

ACCEPTED MANUSCRIPT
14
3.2.2 Non-equimolar guest mixtures. OED was recrystallized from binary guest mixtures in
which the concentration of the constituent guests were varied in order to observe whether the
selectivity of the host, obtained from the equimolar experiments, is affected in such changing
conditions. Thus Figures 1a−c represent the selecꢁvity profiles obtained for the o-Xy/m-Xy, o-
Xy/p-Xy and p-Xy/EB experiments; o-Xy/EB, m-Xy/EB and m-Xy/p-Xy mixtures were not
considered here since these furnished no complexes in the equimolar experiments. Note that
the line that represents an unselective host is denoted by the black dotted line in each figure.

ACCEPTED MANUSCRIPT
15
(a)
(b)
100
80
60
40
20
0
100
80
60
40
20
0
0
20
40
60
80
100
0
20
40
60
80
100
Xo-Xy
Xo-Xy
(c)
100
80
60
40
20
0
0
20
40
60
80
100
Xp-Xy
Figure
1.
Selectivity
profiles
for
the
(a)
o-Xy/m-Xy,
(b)
o-Xy/p-Xy
and
(c)
p-Xy/EB
competition
experiments

ACCEPTED MANUSCRIPT
16
Figure 1a displays the selectivity profile for OED when recrystallized from various o-Xy/m-Xy
mixtures. o-Xy was significantly favoured over m-Xy over the entire concentration range. At
22.3% o-Xy in the mother liquor, OED already contained 59.6% of this guest. The selectivity for
o-Xy persisted as the concentration of this guest in the mother liquor increased further, and the
complex that resulted when 79.3% o-Xy was present in the mother liquor contained 94.0% o-
Xy. The binary o-Xy/p-Xy competition experiment resulted in the selectivity profile that is
provided in Figure 1b. At a low concentration of o-Xy (21.9%) in this solution, this guest was
preferred by OED, and 31.0% o-Xy was present in the resultant complex. However, as the
concentration of o-Xy was increased, only subtle changes in the selectivity for o-Xy was
observed: a complex containing 88.3% of this guest was obtained when the mother liquor
contained 79.6% o-Xy. In the pXy/EB experiment (Figure 1c), the host displayed only a slight
preference for the p-Xy guest relative to EB. The selectivity for p-Xy was 24.5% when the
solution contained 19.2% p-Xy, while a 78.5% p-Xy mixture resulted in a complex containing
83.8% p-Xy.
While the selectivity of OED for o-Xy in all applicable competition experiments may only be
described as adequate, what is striking is the vast differences in behaviour between OED and
SED despite the only variance being in their B rings. In order to better understand why these
compounds behave so differently, we conducted an in-depth analysis of their conformations by
means of computational calculations, and compared these results with the crystal structures of
apohosts OED and SED.
3.3 Computational methods
Both OED and SED displayed similar calculated geometries for the three lowest energy
conformers OEDa−c and SEDa−c. (The atomic coordinates of the calculated structures have
been deposited in the Supplementary Information.) The two lowest energy conformers a and b
had their the nitrogen atoms in a gauche conformation when observing along the C1‒C2 bond of
the ethylenediamine linker (Figures 2.1a and 2.2a, and 2.1b and 2.2b, respectively) (see Scheme

ACCEPTED MANUSCRIPT
17
2 for atomic labelling) while, in the instance of the third lowest energy conformer
c, the
nitrogen atoms were arranged in an anti-orientation relative to this C1‒C2 bond (Figures 2.1c
and 2.2c). This energy order of the conformers at the higher computational levels was
maintained, with the exception that at the BLYP/6-31+G* level for the geometry of host SED,
conformers and were reversed energetically. These energies together with N0∙∙∙N3 and
H1∙∙∙N3 distances, and N0‒C1‒C2‒N3 (N0∙∙∙N3) dihedral angles, are provided in Table 3.
a‒c
b
c

ACCEPTED MANUSCRIPT
18
Table 3
Calculated
energies, distances
and relevant angles
for the OEDa‒c and
SEDa‒c conformers
Computational level
DFT (BLYP/6-31G*)
Conformer Energy
(kJ∙mol-1)
‒4751758.33
Relative Energy Dihedral angle
Distance
N0∙∙∙N3 (Å)
Distance
H1∙∙∙N3 (Å)
(kJ∙mol-1)
0
N0‒C1‒C2‒N3 (°)
‒62.29
‒59.38
180
OEDa
OEDb
OEDc
SEDa
SEDb
SEDc
OEDa
OEDb
OEDc
SEDa
SEDb
SEDc
OEDa
OEDb
OEDc
SEDa
SEDb
SEDc
OEDa
OEDb
OEDc
SEDa
SEDb
SEDc
‒4751754.25 4.09
‒4751752.78 5.56
‒6447579.49
‒6447576.01 3.48
‒6447574.06 5.42
‒4751920.10
‒4751919.00 1.1
‒4751918.30 1.8
‒6447730.79
‒6447728.45 2.34
‒6447728.63 2.16
‒4754236.04
‒4754231.53 4.51
‒4754229.92 6.12
‒6450112.37
‒6450108.88 3.49
‒6450107.20 5.17
‒4755040.93
‒4755039.21 1.71
‒4755038.05 2.88
‒6450940.77
‒6450937.97 2.8
‒6450936.64 4.14
0
62.01
‒67.75
180
DFT (BLYP/6-31+G*)
DFT (B3LYP/6-31G**)
DFT (B3LYP/6-311G*)
0
‒66.75
‒62.48
180
3.008
2.956
3.774
2.973
2.973
3.771
2.905
2.863
3.733
2.888
2.903
3.732
2.943
2.897
3.733
2.903
2.922
3.732
3.425
2.544
4.064
0
64.11
‒63.96
180
0
‒62.4
‒58.95
180
3.263
2.428
4.026
0
60.32
‒61.42
180
0
‒64.3
‒60.84
180
3.356
2.484
4.029
0
60.9
‒62.3
180

ACCEPTED MANUSCRIPT
19
The N0∙∙∙N3 dihedral angles of the computed conformers of OED ranged between –62.29 and
–66.75° for
from 60.32 to 64.11° (
nitrogen atoms of the
these are on opposite sides of the C1–C2 bond. For conformers
other hand, these angles are comparable. In the structures of conformers OED
after the final DFT level calculation (Figures 2.1a‒c and 2.2a‒c), the shape of the
ethylenediamine linker (N0‒C1‒C2‒N3) in conformers and is bucket-like, whereas this linker
of the conformers of OED and SED exhibits a zig-zagged orientation.
a, –58.95 and –62.48° for
b, and 180° for
c
(Table 3). In SED, these angles ranged
a
), ‒61.42 to ‒67.75° (
b) and 180° (c). It is therefore evident that the two
a
conformers in each host are situated in the gauche orientation but
and for both hosts, on the
and SED
b
c
a‒c
a‒c,
a
b
c
The N0∙∙∙N3 and H1∙∙∙N3 distances were calculated at the DFT (BLYP/6-31+G*) and higher levels
for conformers OED , while only the former were calculated for conformers SED since
a‒c
a‒c
this host did not experience an intramolecular N0‒H1∙∙∙N3 hydrogen bond in the crystal
(discussed later) (Table 3). The N0∙∙∙N3 distances for corresponding conformers for both hosts
were similar. For example, this distance for OED
correspondingly, SED ranged from 2.888 and 2.973 Å. The H1∙∙∙N3 distances of conformers
of host SED differed extensively, as expected, with the anti conformer having the longest of
these (4.026–4.064 Å) and which, as a result of the orientations of the nitrogen groups, are not
able to hydrogen bond. Interestingly, conformer OED experienced the shortest hydrogen
(3.263–3.425 Å) and (4.026–4.064
a ranged between 2.905 and 3.008 Å while,
a
a‒c
c
b
bond (H1∙∙∙N3 2.428–2.544 Å) compared with conformers
Å).
a
c
The electrostatic potential maps for conformers OEDa‒c and SEDa‒c, after the DFT B3LYP/6-
311G* level calculation, are provided in Figures 2.1a‒c and 2.2a‒c, respectively. Electron-rich
areas are red while the dark blue regions indicate electron-poor regions. The oxygen and sulfur
atoms are shown in red and yellow, respectively.

ACCEPTED MANUSCRIPT
20
a)
0°
270°
0°
90°
180°
b)
90°
180°
270°
c)
0°
90°
180°
270°
Figure 2.1 The structures of the computed OEDa‒c conformers, displayed in tube form and spacefill
representations, and their electrostatic potential maps at 90° rotation intervals

ACCEPTED MANUSCRIPT
21
a)
0°
90°
180°
270°
b)
0°
90°
180°
270°
c)
0°
90°
180°
270°
Figure 2.2 The structures of conformers SEDa‒c, displayed in a tube and spacefill form, accompanied by their
electrostatic potential maps which are displayed at 90° rotation intervals

ACCEPTED MANUSCRIPT
22
We subsequently investigated the geometries of OED and SED in their pure crystal forms by
means of SCXRD experiments for comparison purposes with these calculated geometries.
3.4 SCXRD analyses of apohosts OED and SED
SCXRD analyses were conducted on suitable crystals of the pure hosts OED and SED when they
were recrystallized from
enclathrated in this way). The crystallographic data and refinement parameters for both
compounds are provided in Table 4. Both crystallized in the triclinic crystal system and -1
m-xylene and 2-methylpyridine, respectively (these solvents were not
P
space group but the host packing differed in each. Furthermore, both hosts displayed disorder
of the nitrogen hydrogens over two positions, but only one of the two molecules in the unit cell
of SED displayed this type of disorder. Unit cells and host packing diagrams are provided in
Figures 3 and 4.
A summary of the non-covalent interactions that hosts OED and SED experience in their crystal
structures are provided in Table 5. Interestingly, SED experienced shorter π∙∙∙π interactions
than OED [3.988(1)‒5.307(1) Å vs 4.858(1)‒5.878(1) Å]. Intermolecular C‒H∙∙∙π interactions
were also observed in both structures, and OED was involved in three of these and SED in only
one (2.71‒2.87 Å, 148‒161° and 2.82 Å, 174°, respectively). Only host OED experienced an
intramolecular N‒H∙∙∙N bond [2.909(2) Å, 104.8(2)°].
Both hosts displayed intramolecular C‒H∙∙∙N interactions, and these were shorter in SED than
OED (2.40‒2.53 Å, 102‒103° and 2.57‒2.61 Å, 102°, respectively). Interestingly, OED was
involved in two intermolecular C‒H∙∙∙O‒C interactions whereas SED experienced three
intermolecular C‒H∙∙∙H‒C contacts (2.66‒2.69 Å, 133‒156° and 2.32‒2.36 Å, 121‒156°,
respectively). All of these interactions assisted in the stabilization of the host packing in the
crystals. Furthermore, the distances of the interactions that host SED experienced were
considerably shorter than those of OED.

ACCEPTED MANUSCRIPT
23
Table 4
Crystallographic data for apohosts
OED
SED
OED
and
SED
Chemical formula
Formula weight
Crystal system
Space group
µ (Mo-Kα)/mm‒1
a/Å
b/Å
c/Å
alpha/°
beta/°
C40H44N2O2
584.77
triclinic
P-1
C40H44N2S2
616.89
triclinic
P-1
0.073
0.196
9.0748(5)
12.8877(8)
14.6778(9)
94.717(3)
96.443(2)
106.537(2)
1623.43(17)
2
9.5375(3)
13.0385(4)
14.7133(5)
64.928(1)
80.988(1)
78.669(1)
1619.41(9)
2
gamma/°
V/Å 3
Z
D(calc)/g.cm‒3
F(000)
1.196
628
1.265
660
Temp./K
200
200
Restraints
Nref
0
8075
0
8045
Npar
415
403
R
wR2
S
0.0405
0.1095
1.02
0.0385
0.0970
1.03
θ min‒max/°
Tot. data
Unique data
Observed data
[I > 2.0 sigma(I)]
Rint
Completeness
Min. resd. dens. (e/Å3)
Max. resd. dens. (e/Å3)
2.0, 28.4
43830
8075
1.5, 28.4
42664
8045
6405
6304
0.021
1.000
‒0.22
0.31
0.026
0.998
‒0.25
0.31

ACCEPTED MANUSCRIPT
24
(a)
(b)
Figure 3. Unit cells of (a) OED and (b) SED
(a)
(b)
Figure 4. Host packing for hosts (a) OED and (b) SED

ACCEPTED MANUSCRIPT
25
Table 5
Interaction
OED
SED
Symmetry
(host)π∙∙∙π(host)
4.858(1)–5.878(1) Å
(7 contacts)
3.988(1)–5.307(1) Å
(6 contacts)
C−H∙∙∙π
(H∙∙∙Cg, X‒H∙∙∙Cg)
2.83 Å, 160°
2.71 Å, 148°
(H∙∙∙Cg, X‒H∙∙∙Cg)
(host)C−H∙∙∙ π(host)c
(host)C−H∙∙∙ π(host)c
(host)C−H∙∙∙ π(host)c
(host)C−H∙∙∙ π(host)c
Hydrogen bonding
(host)N‒H∙∙∙N(host)d
Other short contacts
(host)m-ArH∙∙∙O‒C(host)e
(host)m-ArH∙∙∙O‒C(host)e
(host)C‒H∙∙∙N(host)f
(host)C‒H∙∙∙N(host)f
1‒x, 1‒y, 1‒z
2‒x, 1‒y, 1‒z
‒1+x, y, z
2.87 Å, 161°
2.82 Å, 174°
None
x, y, z
2.909(2) Å, 104.8(2)°, <<
2.69 Å, 156°, <
2.66 Å, 133°, <
2.61 Å, 102°, <
2.57 Å, 102°, <
2‒x, 1‒y, 1‒z
X, 1+y, z
(host 1)C‒H∙∙∙H‒C(host 1)e
(host 2)m-ArH∙∙∙H‒C(host 1)e
(host 2)C‒H∙∙∙m-ArH(host 1)e
(host 1)o-ArH∙∙∙N(host 1) f
(host 1)o-ArH∙∙∙N(host 1) f
(host 2)o-ArH∙∙∙N(host 2) f
(host 2)o-ArH∙∙∙N(host 2) f
2.32 Å, 126°, <
2.36 Å, 121°, <
2.35 Å, 156°, <
2.40 Å, 103°, <<
2.53 Å, 102°, <<
2.47 Å, 102°, <<
2.46 Å, 103°, <<
‒x, 1‒y, 2‒z
x, y, z
x, 1+y, ‒1+z
Summary of the non-covalent host∙∙∙host interactions in the crystals of OED and SEDa,b
a< denotes contacts less than the sum of the van der Waals radii and << contacts less than this sum minus 0.2 Å
bOne SED host in the crystal contained nitrogen hydrogens that were disordered over two positions; thus the host
with no disorder was labelled ‘host 1’ while the one with disorder was labelled ‘host 2’
cIntermolecular C−H∙∙∙π interaction between two host molecules
dIntramolecular H‒bond within each host molecule
eIntermolecular interaction
fIntramolecular interaction

ACCEPTED MANUSCRIPT
26
Side-to-side, overlaid and stereoviews of the geometries of the calculated conformers that
bared a resemblance to that of the geometries of hosts OED and SED in the crystals are
provided in Figures 5 and 6, respectively. The calculated conformers have a blue hue in the
overlaid view of these structures. Conformer OEDb closely resembled host OED from the
crystal, while conformer SEDc was similar to that of host SED in the crystal (both conformers
were obtained at the DFT B3LYP/6-311G* level).
(a)
(b)
(c)
(d)
Figure 5. Geometry of (a) the calculated OEDb conformer and (b) OED from the crystal; (c) is an overlay of these
two (the calculated structure has the blue hue), and (d) a stereoview showing the geometry of each host molecule
from the crystal; molecules are shown in ball-and-stick form

ACCEPTED MANUSCRIPT
27
(a)
(b)
(c)
(d)
Figure 6. Geometry of (a) the calculated SEDc conformer and (b) SED from the crystal; (c) is an overlay of these
two (the calculated structure, once more, has the blue hue), and (d) a stereoview displaying the molecular
geometry of SED from the crystal; molecules are shown in ball-and-stick form
Conformer OEDb has only a slightly shorter H1∙∙∙N3 bond distance (measured using the Spartan
’10 software27) compared with OED from the crystal (2.484 and 2.548 Å, respectively).
However, this distance in conformers OEDa and OEDc is significantly larger than that observed
in the crystal (3.356 and 4.029 Å vs 2.548 Å). The SEDc conformer, on the other hand, had the
same N∙∙∙N dihedral angle as SED from the crystal (180°) as well as a comparable distance
between the two nitrogen atoms (3.732 and 3.691 Å, respectively). Both host molecules from
the crystals therefore occupied higher energy states than conformers a (Table 3), but energy
differences were small.
This comparative investigation has demonstrated that OED and SED from the crystals possess
very different geometries relative to one another, in the orientation of their cyclohexyl and
xanthenyl moieties and, most notably, in the geometry of their ethylenediamine linkers. This

ACCEPTED MANUSCRIPT
28
observation explains the vastly different host behaviours of the two compounds in the presence
of the C8 aromatic fraction. However, an additional experiment was conducted in which a
nearly saturated and carefully filtered solution of apohost OED in m-Xy (which OED does not
clathrate) was seeded with crystals of apohost SED in order to encourage OED to crystallize in
the same conformation as SED. However, analysis (SCXRD) of the crystals of OED thus formed
revealed, surprisingly, that the conformation of OED was identical to that initially obtained (see
Supplementary Information where the .cif file for the obtained structure is provided). Hence
this experiment showed that OED could not be encouraged to crystallize in the same
conformation as SED despite there being SED crystals present.
3.5 SCXRD analyses of successfully-formed complexes with OED
Table 6 contains the crystallographic data and refinement parameters for the 2OED•o-Xy,
2OED•p-Xy and 2OED•EB complexes. In all three of these, the guest is disordered around an
inversion point, while the nitrogen hydrogens of the host are disordered over two positions, but
only in 2OED•EB. All three complexes are isostructural and crystallized in the triclinic crystal
system and P-1 space group.