Sorption properties toward environmentally important VOCs of half-sandwich Ru(II) complexes containing perylene bisimide ligands

Article abstract of DOI:10.1080/15533174.2016.1186068

Two py-functionalized perylene bisimide ligands were synthesized and used to make bimetallic halfsandwich Ru(II) complexes. These were characterized by IR, ¹H NMR, ¹³C CPMAS SSNMR spectroscopy, and elemental analysis. The complexes are wheel-and-axle (waa) compounds, where the axle is the divergent ligand and the wheels are the [(p-cymene)RuCl₂] units. The complexes, but not the free ligands, showed absorption of volatile organic compounds such as toluene and xylenes through heterogeneous solid/gas uptakes. The reactivity is ascribable to the waa geometry, likely by an upset of the high stacking characterizing the crystalline frameworks of the free ligands. The kinetic profiles of the uptake reactions were determined

Full text of DOI:10.1080/15533174.2016.1186068

Synthesis and Reactivity in Inorganic, Metal-Organic, and
Nano-Metal Chemistry
ISSN: 1553-3174 (Print) 1553-3182 (Online) Journal homepage: http://www.tandfonline.com/loi/lsrt20
Sorption Properties towards Environmentally
Important VOCs of Half-Sandwich Ru(II) Complexes
Containing Perylene-Bis-Imide Ligands
Paolo Pagano, Paolo Pelagatti, Alessia Bacchi, Michele R. Chierotti, Susan A.
Bourne & Gift Mehlana
To cite this article: Paolo Pagano, Paolo Pelagatti, Alessia Bacchi, Michele R. Chierotti,
Susan A. Bourne & Gift Mehlana (2016): Sorption Properties towards Environmentally
Important VOCs of Half-Sandwich Ru(II) Complexes Containing Perylene-Bis-Imide Ligands,
Synthesis and Reactivity in Inorganic, Metal-Organic, and Nano-Metal Chemistry, DOI:
10.1080/15533174.2016.1186068
To link to this article: http://dx.doi.org/10.1080/15533174.2016.1186068
View supplementary material
Accepted author version posted online: 03
Aug 2016.
Published online: 03 Aug 2016.
Submit your article to this journal
View related articles
View Crossmark data
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=lsrt20
Download by: [Stephen F Austin State University]
Date: 03 August 2016, At: 23:23

ACCEPTED MANUSCRIPT
SORPTION PROPERTIES TOWARDS ENVIRONMENTALLY IMPORTANT VOCs of HALF-
SANDWICH Ru(II) COMPLEXES CONTAINING PERYLENE-BIS-IMIDE LIGANDS
a
a,*
a
b
c
Paolo Pagano, Paolo Pelagatti, Alessia Bacchi, Michele R. Chierotti, Susan A. Bourne, Gift
c
Mehlana
a
Dipartimento di Chimica, Università degli Studi di Parma, Parco Area Scienze 17/A, 43124
Parma, Italy.;
b
Department of Chemistry and NIS Centre of Excellence, University of Torino, Via P. Giuria 7,
10125, Torino, Italy.
c
Chemistry Department, University of Cape Town, 7700 Rondebosch, Cape Town, South Africa.
*
Corresponding author. E-mail: paolo.pelagatti@unipr.it
Abstract
Two py-functionalized perylene-bis-imide ligands were synthesized and used to make bimetallic
1
13
half-sandwich Ru(II) complexes. These were characterized by IR, H NMR, C CPMAS SSNMR
spectroscopy and elemental analysis. The complexes are wheel-and-axle compounds where the
axle is the divergent ligand and the wheels are the [(p-cymene)RuCl ] units. The complexes, but
2
not the free ligands, showed absorption of VOCs such as toluene and xylenes through
heterogeneous solid-gas uptakes. The reactivity is ascribable to the waa geometry, likely by an
upset of the high stacking characterizing the crystalline frameworks of the free ligands. The
kinetic profiles of the uptake reactions were determined.
1
ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
Introduction
Perylene tetracarboxylic acid bisimides, hereafter referred as PBIs, are highly insoluble
compounds, characterized by high thermal stability, chemical inertness and light- weather-
fastness, characteristics which have made them useful pigments especially for surface-coating
in the automotive industry. PBI pigments receive a lot of attention as n-type semiconductor
too, due to their high electron-affinity which makes them highly promising materials for organic
[1-3]
field effect transistors.
Functionalization of the imide nitrogens or of the polycyclic scaffold,
the so-called perylene-bay, with coordinating functions, allows the preparation of PBI-
containing organometallic compounds with different supramolecular architectures, which are
[4-7]
of interest in the development of fluorescent based sensors
or supramolecular
[8,9]
aggregates. A review concerning the metal-complexes containing rylene (and then also
[10]
perylene) ligands has recently appeared. N,N’-bis(4-aminopyridine)-3,4,9,10-tetracarboxy-
bis-imide and N,N′-bis(4-aminomethylene)-3,4,9,10-tetracarboxy-bis imide (L1 and L2 in Figure
1) are divergent bidentate ligands which can give rise to dinuclear complexes. With L1 the two
metals are separated by a linear and rigid spacer, a good prerequisite in order to build wheel-
and-axle (waa) compounds. These are morphologically irregular molecules based on a linear
spacer connecting two relatively bulky groups. The irregular shape of waa compounds
frustrates the close packing principle and their solid-state chemistry is consequently
[11]
characterized by the tendency to crystallize as solvates or clathrates. In general, the flexible
crystalline framework which characterizes waa compounds allows the introduction of guest
species by heterogeneous solid/gas uptake processes into the waa host framework, as
2
ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
demonstrated by us with several organometallic waa complexes containing [(p-cymene)RuCl ]
2
[12-15]
wheels.
These inclusion processes occur with structural rearrangements of the host
framework necessary in order to accommodate the incoming guest molecules, which are
possible only when the host-guest interactions can compete with the host-host ones. The
analysis of the X-ray structures of several PBIs reveals extended stacking between the large -
[16,17]
surface of the perylene cores.
These strong intermolecular interactions are expected to
inhibit the insertion of guest species through heterogeneous gas-uptake processes. However,
the large -surface of the perylene core could be exploited in order to trigger on -
interactions with aromatic volatile organic compounds (VOCs), such as toluene or the p-xylene
isomers, giving rise to host-guest species. This might be facilitated by the introduction of the
organometallic wheels, since they are expected to upset the high stacking between the
[18]
perylene platforms, making the packing more flexible and then more reactive towards VOCs.
Monitoring of aromatic VOCs has important environmental concerns, and then molecular solids
able to absorb them are highly desirable for their detection and/or capture. The possibility of
clathrating aromatic compounds by metal complexes is a well- known phenomenon. Octahedral
metal complexes with the general formula ML
4
X
2
(M = transition metal, L = py ligand, X =
anion), the so-called Werner type complexes, form clathrates with different aromatics from the
[19,20]
liquid phase, as reported by different authors in 1957.
More recently, Werner type
complexes were described to be able to form clathrates with xylene isomers also by solid-gas
[21,22]
heterogeneous reactions, in some cases with high/moderate selectivity,
although the
initial phases are not porous. This ability was tentatively attributed to conformational flexibility
3
ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
[23]
of the host complexes. The development of well-defined molecular solids able to clathrate
VOCs has potential applications in several important fields, such as pollution control, separation
techniques and heterogeneous catalysis. Thus, the ligands L1 and L2 and their ruthenium
complexes 1 and 2 were subjected to vapor absorption experiments of toluene and the three
xylene isomers, under identical experimental conditions. The use of ligand L2 and its complex 2
allows testing the effect that the perturbation of the linearity and rigidity of the spacer has onto
the uptake properties of the corresponding complex, keeping constant the affinity towards
aromatic VOC dictated by the perylenic core. The comparison of the reactivity observed for the
free ligands and the corresponding complexes will elucidate the importance of the introduction
of the organometallic units [(p-cymene)RuCl ] on the sorption properties of these PBI-based
2
solids. An isothermal kinetic analysis of the uptake reactions will also be shown.
Results and Discussion
The synthesis of the ligands L1 and L2 was accomplished by following literature reported
[24]
methods describing the synthesis of PBI ligands (Figure 1). These consider a prolonged reflux
of a DMF solution containing perylene-3,4,9,10-tetracarboxydianhydride and an excess of the
appropriate amine. The progress of the reactions could be conveniently monitored by FTIR
analysis conducted on small samples of the reactant mixtures withdrawn at regular intervals of
1
time, monitoring the disappearance of the anhydride C = O stretching band (1772 cm ) in favor
1
of the imide C = O stretching bands of the final products (1708 and 1666 cm for L1, 1691 and
1
1
656 cm for L2, respectively). The red solids isolated after the workup contained solvent
4
ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
molecules, as established by TGA and elemental analysis, with weight losses percentages of
11.5 and 11.9%, for L1 and L2, respectively, at temperatures higher than 140°C. Desolvation
was accomplished by thermal treatment at 140°C (oven, overnight). The purity of the samples
1
3
was further confirmed by C CPMAS SSNMR spectroscopy (the spectra indicated the absence
of traces of unreacted dianhydride (Figs S1 and S2 in the Supporting Information). Dissolution of
1
the free ligands in a mixture of CF
3
COOD and CDCl
3
allowed the recording of their H NMR
spectra in solution. In both cases the pyridine protons gave rise to two distinct doublets. All the
signals are in the expected chemical shift regions. For the synthesis of the Ru-complexes 1 and
2
two different protocols were followed (Figure 2): a) addition of solid L1 or L2 to a
dichloromethane solution of [(p-cymene)RuCl and subsequent stirring of the mixture for 24
hours at room temperature, or b) heating at 100°C of a cyclohexanone solution of [(p-
cymene)RuCl containing solid L1 or L2 for 24 hours. The workup considers a prolonged
]
2 2
]
2 2
washing of the isolated products with dichloromethane where the Ru-dimer precursor is well
soluble. This ensure to have the final products free from traces of the unreacted organometallic
precursor. The FTIR spectra of the final red solids did not show appreciable changes with
respect to the spectra of the starting ligands, as expected for the binding of the Ru-
organometallic fragments to the pyridine rings. No differences were observed in the spectra of
the complexes prepared in the two different solvents. The same behavior was observed for the
decomposition temperatures, which started for all the complexes around 250°C. As seen for the
free ligands, also the Ru-complexes are highly insoluble in the common organic solvents. For
1
the recording of the H NMR spectra clear solutions were obtained by dissolving the solids in a
5
ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
few drops of CF
3
COOD and diluting with CDCl . The spectra of 1 and 2 showed the signals of the
3
p-cymene fragment in addition to the signals of the starting ligands, in the expected 2:1 ratio.
The elemental analyses of the products were conducted on thermally evacuated samples (oven
dried at 140°C for 10 hours) and they were in agreement with the proposed stoichiometries. As
regards the reactions conducted in cyclohexanone, complex 1 was initially isolated as
1
cyclohexanone solvate, as indicated by H NMR analysis (CDCl
3
/CF COOD, signals belonging to
3
cyclohexanone at 1.81, 1.97 and 2.48 ppm, see Supporting Information). Cyclohexanone was
not detected by FTIR spectroscopy, probably due to overlapping of the C = O stretching band of
the ketone with signals belonging to the PBI ligand. Integration of the NMR signals indicated a
rough 6 to 1 ratio between cyclohexanone and the complex. However, after storing the solid
sample at room temperature the signals belonging to cyclohexanone disappeared giving 1 back.
1
No signals belonging to the solvent were observed in the H NMR spectrum of 2 synthesized in
1
3
cyclohexanone. C CPMAS SSNMR spectra of the complexes were characterized by broad
signals, indicative of a rather low crystallinity. However, the absence of signals belonging to
unreacted reagents assured high degree of purity for the four isolated complexes (Figs S1 and
S2 in the Supporting Information). Unfortunately, owing to the extremely low solubility it was
not possible to structurally characterize the Ru-complexes, although many attempts were made
in order to isolate X-ray quality single crystals. The wheel-and-axle geometry of the complexes 1
and 2 prompted us to undertake a study on their sorption properties towards environmentally
important VOCs, such as toluene, o-, m- and p-xylene, in comparison with the ones shown by
the free ligands L1 and L2. Weighed amounts of the solids were thus kept in contact with the
6
ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
vapors of the chosen VOC at room temperature for different time intervals in a closed H-cell
reactor (see Fig. S4 in the Supporting Information). Initially, the amount of adsorbed VOC was
tentatively determined by weighing the sample before and after the adsorption, but it soon
appeared that the host-guest systems formed were highly unstable in the absence of guest-
vapors, as evidenced by the continuous and rapid weight loss observed on the balance plate.
This high instability prevented TGA analysis. The attempts made to monitor the uptake
1
processes by H NMR analysis conducted on freshly prepared samples was again hampered by
the fast guest release and led to not fully reproducible data. This can be inferred by Figure 3
1
which shows the H NMR spectra quickly recorded after having exposed complex 1 to p-xylene
vapors for one week (spectrum 1), and after having stored the same sample at room
temperature in the absence of the aromatic compound for 40 minutes (spectrum 2) and 2 days
(spectrum 3), respectively. Nevertheless, the NMR analysis gave evidences of the absorbing
capacity of the complexes towards VOCs as well as of the inertness of the free ligands. Due to
the high instability of the final materials an exact quantification of the amount of absorbed
VOCs was impossible. However, based on the NMR integrals a rough complex/VOC ratio of 1 to
2.5 was estimated for both complexes. During the uptake processes no color changes of the
solids were observed. In order to reliably quantify the amount of absorbed VOCs and establish
the sorption kinetic profiles of the uptakes, isothermal kinetic studies were undertaken. On the
basis of the similar behavior shown by the four complexes during the preliminary uptake
experiments we initially restricted the sorption kinetic profile study to complexes 1 (synthesized
in dichloromethane) and 2 (synthesized in cyclohexanone), using toluene as VOC, comparing
7
ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
the results with those collected with the free ligands L1 and L2. The uptake reactions were now
carried out using a microbalance which monitors the weight change as a function of time under
controlled conditions of temperature and pressure. The apparatus employed is reported in
[25]
Figure 4. Figure 5 shows the absorption profiles observed at 35°C for the toluene sorption.
The sorption profiles clearly indicate that the introduction of the organometallic moieties
activates the PBI systems towards the uptake of toluene, since the two free ligands are
completely inert. With both complexes the uptake occurs speedily and within 50 minutes the
sorption curves reach a plateau corresponding to about 25% weight of absorbed toluene.
Prolonged exposures did not bring to significative variations. In light of these results, the study
was extended to the three xylene isomers. Table 1 collects the amount of VOCs absorbed by
the two complexes at 35°C. The results indicate a similar behavior for 1 and 2, irrespective to
the VOC used, with weight percentages of absorbed guest ranging from about 18 to 24.5%,
corresponding to 3 molecules of guest for each molecule of dinuclear complex. Noteworthy is
the fact that the introduction of the organometallic wheels makes the complexes reactive
towards VOCs, whereas the introduction of a methylene spacer between the imide nitrogen
and the pyridine ring, as in ligand L2, is not sufficient to overcome the high stacking of the
[18]
perylene platforms. Regrettably, the lack of crystallographic information impedes to highlight
the structural differences responsible of the different reactivity found, although it can be
concluded that the responsiveness of the complexes to VOCs is dictated by the combination
between the waa geometry and the perylene spacer. The sorption curves were finally
[28]
converted into α-time curves (see Fig. S5 and S6 in the Supporting Information) and fitted
8
ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
into standard kinetic models using the α range of 0.1 to 0.95 to determine the model of
2
sorption. Good fits (r > 0.993) of the experimental data were obtained for several models, as
indicated in Tables 2 and 3. However, it can be concluded that R3 model showed dominance
over the other models with the exception of the system 1/m-xylene which was better fitted
with the R2 model. Hence, it can be concluded that the VOCs uptake by complexes 1 and 2 is
[28]
governed by a geometrical model. For complex 2 and its free ligand L2 a BET analysis was
carried out in order to measure the surface areas and then the porosity of the samples (see
Table S2 and Figure S7 in the Supporting Information). In both cases the values of the
2
1
2 1
experimentally determined surface areas were very low (8.68±0.08 m g and 4.60±0.07 m g ,
respectively) pointing to the absence of intrinsic porosity. Because of the similar uptake
behavior, BET analyses were not carried out for the couple 2/L2. The ability shown by the
[29]
complexes to absorb VOCs must then be imputed to the so called “porosity without pores”,
to say the ability of the crystalline framework to rearrange once exposed to suitable guests in
order to allow its entrance.
In conclusion, two dinuclear half-sandwich Ru(II) complexes containing two different PBI ligands
with a wheel-and-axle geometry have been synthesized and characterized. These complexes,
differently from their free ligands, are able to absorb aromatic volatile organic compounds,
such as toluene and the three xylene isomers, through heterogeneous solid-gas reactions. This
ability must be ascribed to the introduction of the half-sandwich organometallic units, which
possibly work by reducing the stacking of the perylene moieties thus making the corresponding
solids more responsive. The isothermal kinetic profiles of the uptakes indicate a geometrical
9
ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
model but, currently, the lack of structural information hampers the definition of a plausible
sorption mechanism. These results are nonetheless stimulating for the development of other
PBI based organometallic systems with responsive character towards VOCs which could find
application in all the fields where the VOC monitoring is essential.
Experimental
All the manipulations have been carried out under an atmosphere of dry nitrogen, using oven
dried Schlenk vessels and freshly distilled solvents. Perylene-3,4,9,10-tetracarboxylic
dianhydride, 4-aminopyridine and 4-aminomethylpyridine were purchased from Aldrich and
[30] 1
used as received. [(p-cymene)RuCl
2
]
2
was synthesized as reported in literature.
H NMR
spectra were recorded on an Avance-300 or Avance-400 Bruker spectrophotometers at 25°C
and the chemical shift values are referred to TMS (for SSNMR measurements see Supporting
Information). IR (ATR) spectra were collected by means of a Nicolet-Nexus spectrophotometer
1
in the range 4000-600 cm by using a diamond crystal plate, or by KBr plates on a Thermo
1
Avatar spectrophotometer, in the 4000-400 cm range. Elemental analyses were performed by
using a FlashEA 1112 Series CHNS-O analyzer (ThermoFisher) with gas-chromatographic
separation. TGA analyses were conducted by means of a Perkin Elmer TGA7 apparatus, with a
1
heating rate of 5°C min from 25°C to 250°C, under an atmosphere of dry nitrogen.
10
ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
Synthesis of ligands L1 and L2
Perylene tetracarboxydianhydride (0.5 g, 1.27 mmol) and the appropriate amine were placed in
a Schlenk vessel under a flow of dry nitrogen, together with a magnetic bar. 15 mL of anhydrous
DMF saturated with nitrogen were introduced by syringe and the reactor was sealed with a
teflon stopcock. The mixture was heated at 130°C monitoring the progress of the reaction by
FTIR spectroscopy (disappearance of the C = O stretching bands of the starting anhydride in
favor of the C = O stretching bands of the imide groups). The FTIR monitoring was performed as
follows: about 0.5 ml of the reaction mixture was withdrawn at regular time intervals by a
pipette, the sample was diluted with dichloromethane and the solid collected by centrifugation.
After washing with diethyl ether and vacuum drying the solid was analyzed by FTIR (ATR).
Typically, the synthesis of L1 reached completion after 48-72 hours, while the synthesis of L2
reached completion within 24 hours. The final purple solids undergone the same workup
described for the FTIR samples. In order to completely remove DMF L1 and L2 were oven dried
at 140°C overnight.
N,N-bis(4-pyridine)perylene-3,4,9,10-bis-dicarboxydiimide (L1) Yield: 86%. M.p.: > 350°C. Anal.
Calcd. for C34
H
16
N
4
O
4
∙½DMF (581.07): C, 73.38; H, 2.59; N, 7.40. Found: C, 73.20; H, 3.00; N,
1
3
7
.84. H NMR (CF
3
COOD/CDCl
3
): 9.18 (d, 4H, py, JHH = 6Hz), 8.95 (s, 8H, perylene), 8.27 (d, 4H,
3
1
py, JHH = 6Hz). FTIR (KBr, cm ): 1708 (C = O), 1666 (C = O), 1590, 1576, 1361, 1346, 709, 734,
39. TGA (from 25°C to 250°C, 10°C/min, under nitrogen atmosphere): loss of 11.5% in the T-
interval 140-190°C.
6
11
ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
2.1.2. N-N-bis(4aminomethylpyridine)perylene-3,4,9,10-bis-dicarboxydiimide (L2) As for L1 but
using 4-aminomethyl pyridine instead of 4-aminopyridine. Yield (95%). M.p.: >350°C. Elemental
analysis calcd. for C36
H
20
N
4
O
4
∙1/2DMF∙1/2H O (618.135): C, 72.86; H, 3.99; N, 10.19. Found: C,
2
1
7
2.73; H, 3.38; N, 10.09. H NMR (CF
3
COOD/CDCl ): 8.85 (m, 12H, py+perylene), 8.17 (d, 4H,
3
3
1
py, JHH = 6.6Hz), 5.75 (s, 4H, CH
2
). FTIR(KBr, cm ): 1691 (C = O), 1656 (C = O), 1594, 1578, 1341,
811, 749, 644. TGA (from 25°C to 250°C, 10°C/min, under nitrogen atmosphere): loss of 11.9%
in the T-interval 140-190°C.
Synthesis of the Ru-complexes 1 and 2
Synthesis in dichloromethane
[
(p-cymene)RuCl
2
2
] (0.1 g, 0.163 mmol) was placed in a Schlenk tube equipped with a stirring
bar and dissolved in 30 ml of anhydrous dichloromethane under a dry nitrogen atmosphere.
The appropriate ligand was introduced as solid and the deep red mixture was stirred at room
temperature for 24 hours. Then, the deep red solid was collected by centrifugation, plenty
washed with dichloromethane in order to remove possible traces of unreacted Ru-dimer and
then washed with diethyl ether and finally vacuum dried at 100°C.
{
(
(
[(p-cymene)RuCl
2
]
2
(L1)} (1) Yield (90%). M.p.: 251.6°C. Anal. Calcd. for C54
H
44Cl
4
N
4
O
4
Ru
2
1
1156.9): C, 56.06; H, 3.83; N, 4.84. Found: C, 56.21; H, 3.70; N, 5.01. . H NMR
3
3
CF
3
COOD/CDCl
3
): 9.13 (d, 4H, py, JHH = 6.6Hz), 8.89 (s, 8H, perylene), 8.22 (d, 4H, py, JHH =
3
3
5
.1Hz), 5.61 (d, 4H, cymene, JHH = 6Hz), 5.42 (d, 4H, cymene, JHH = 6Hz), 2.78 (m, 2H,
12
ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
3
1
CH(CH
3
)
2
), 2.22 (s, 6H, CH
3
), 1.30 (d, 12H, CH(CH
3
) , JHH = 6.9Hz). FTIR (KBr, cm ): 3055, 2962,
2
1701 (C = O), 1668 (C = O), 1592, 1576, 1357, 809, 746, 654.
{
(
(
[(p-cymene)RuCl
2
]
2
(L2)} (2) Yield (94%). M.p.: 247.4°C. Anal. Calcd. for C56
H
48Cl
4
N
4
O
4
Ru
2
1
1184.95): C, 56.76; H, 4.08; N, 4.73. Found: C, 56.39; H, 4.06; N, 4.74. H NMR
3
CF
3
COOD/CDCl
3
): 8.82 (m, 12H, perylene+py), 8.12 (d, 4H, py, JHH = 5.4Hz), 5.71 (s, 4H, CH
2
),
), 2.22
, JHH = 6.9Hz). FTIR (ATR, cm ): 1694 (C = O), 1657 (C = O),
3
3
5
.61 (d, 4H, cymene, JHH = 6Hz), 5.42 (d, 4H, cymene, JHH = 6Hz), 2.77 (m, 2H, CH(CH
)
3 2
3
1
(s, 6H, CH
3
), 1.29 (d, 12H, CH(CH
)
3 2
1593, 1343, 810, 748.
Synthesis in cyclohexanone
The reactions were carried out at 100°C in a teflon stopcock sealed rotaflow reactor, using
cyclohexanone as solvent. The amount of reagents and the concentrations of the reactant
solutions were as for the reactions conducted in dichloromethane. Complex 1 was initially
1
isolated as 1∙6C
6
H
10O, as indicated by the H-NMR analysis (CDCl
3
/CF COOD, signals belonging
3
to cyclohexanone at 1.81, 1.97 and 2.48 ppm). However, at room temperature 1∙6C
6
10
H O
1
desolvates giving 1 back within 24-36 hours, as established by H-NMR. A complete desolvation
could be accomplished by heating the solvate complex at 140°C under vacuum for 10 hours.
1
Complex 2 was isolated as an unsolvated species, whose characterization (IR, H-NMR and
elemental analysis) was equivalent to that of complex 2.
13
ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
Sorption kinetics of volatile organic compounds by ruthenium complexes
Both 1 and 2 were dried under a vacuum for 5hrs at 95°C before being exposed to solvent
vapors of o-xylene, p-xylene, m-xylene and toluene. Kinetic studies were carried out to analyze
the uptake of volatile organic compounds by the activated samples of 1 and 2 at 35°C. Samples
weighing between 80 and 95 mg with particle size in the range of 38-45 µm were used. A
balance, placed in a thermostatically controlled oven (precision ca. 0.5°C), is connected to a
computer which tracks the experiment. A sample holder is contained within a brass cylinder
which has six solvent chambers evenly distributed around the central sample container. The
brass container fits over the sample on the balance and can be sealed. The solvent was placed
in the six solvent chambers, the container sealed and the vapor pressure allowed to equilibrate
for 30 minutes. The scale was zeroed, the seal between sample and solvent vapor removed and
the monitoring computer program (PB303S31) started. The increase in mass with time was
recorded and then fitted to analyze the extent of reaction with time.
14
ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
References
[1] Huang, C.; Barlow, S.; Marder, S. R. J. Org. Chem. 2011, 76, 2386.
[2] Ruiz Delgado, M. C.; Kim, E. G.; da Silva Filho, D. A.; Bredas, J. L. J. Am. Chem. Soc. 2010, 132,
3375.
[
[
[
[
[
[
[
[
[
[
3] Geng, Y.; Li, H. B.; Wu, S. X.; Su, Z. M. J. Mater. Chem. 2012, 22, 20840.
4] Yan, L.; Ye, Z.; Peng, C.; Zhang, S. Tetrahedron 2012, 68, 2725.
5] Dobrawa, R.; Würtner, F. Chem. Commun. 2002, 1878.
6] Würtner, F.; Sautter, A. Chem. Commun. 2000, 445.
7] Würtner, F.; Sautter, A.; Schmid, D.; Weber, P. J. A. Chem. Eur. J. 2001, 7, 894.
8] Addicott, C.; Oesterling, I.; Yamamoto, T.; Müllen, K.; Stang, P. J. J. Org. Chem. 2005, 70, 797.
9] You, C.C.; Würtner, F. J. Am. Chem. Soc. 2003, 125, 9716.
10] Castellano, F. N. Dalton Trans. 2012, 41, 8493.
11] Soldatov, D. V. J. Chem. Cryst. 2006, 36, 747.
12] Bacchi, A.; Cantoni, G.; Granelli, M.; Mazza, S.; Pelagatti, P.; Rispoli, G. Cryst. Growth Des.
2011, 11, 5039.
[13] Bacchi, A.; Cantoni, G.; Mezzadri, F.; Pelagatti, P. Cryst. Growth Des. 2012, 12, 4240.
15
ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
[14] Bacchi, A.; Carcelli, M.; Pelagatti, P. Crystallogr. Rev., 2012, 18, 253.
[15] Bacchi, A.; Bourne, S.; Cantoni, G.; Cavallone, S. A. M.; Mazza, S.; Mehlana, G.; Pelagatti, P.;
Righi, L. Cryst. Growth Des. 2015, 15, 1876.
[16] Würthner, F. Chem. Commun. 2004, 1564.
[17] Hino, K.; Sato, K.; Takahashi, H.; Suzuki, S.; Mizuguchi, J. Acta Crystallogr. Sec. E, 2005, E61,
o440.
[18] Hino, K.; Mizuguchi, J. Acta Crystallogr. Sec. E, 2005, E61, o672.
[19] Schaeffer, W. D.; Dorsey, W. S.; Skinner, D. A.; Christian, C. G. J. Am. Chem. Soc. 1957, 79,
5870.
[20] Williams, F. V. J. Am. Chem. Soc. 1957, 79, 5876.
[21] Lusi, M.; Barbour, L. J. Angew. Chem. Int. Ed. 2012, 51, 3928.
[22] Batisai, E.; Lusi, M.; Jacobs, T.; Barbour, L. J. Chem Commun. 2012, 48, 12171.
[23] Nassimbeni, L. R.; Niven, M. L.; Suckling, A. P. Inorg. Chim. Acta 1989, 159, 209.
[24] Biedermann, F.; Elmalem, E.; Ghosh, I.; Nau, M.; Scherman, O. A. Angew. Chem. Int. Ed.
2012, 51, 7739.
[25] Bourne, S. A.; Oom, B. M.; Toda, F. J. Chem. Soc., Perkin Trans.2 1997, 585-587.
[26] Barton, B.; Caira, M. R.; McCleland, C. W.; Taljaard, B. J. Chem. Soc., Perkin Trans. 2 2000,
865-869.
16
ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
[
27] Mehlana, G.; Bourne, S. A.; Ramon, G. Dalton Trans. 2012, 41, 4224-4231.
28] Brown, M. E. Introduction to Thermal Analysis – Techniques and Application (2 Edition),
nd
[
Kluwer Academic Publishers, 2001.
[29] Barbour, L. J. Chem. Commun. 2006, 1163.
[30] Bennet, M. A.; Huang, T. N.; Matheson, T. W.; Smith, A. K. Inorg. Synth. 1984, 74.
17
ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
Table 1 Amounts of absorbed VOCs by complexes 1 and 2 during the uptake experiments
conducted at 35°C.
Sample VOC Amount
Amount
absorbe
d/mg
%weight
absorb
G/H
ID
use
d/
b
ed
mg
1
1
1
1
2
2
2
2
m-x
o-x
p-x
tol
m-x
o-x
p-x
tol
80
21.6
21.26
23.22
22.45
20.83
24.54
21.41
18.03
20.83
2.9
3.3
3.2
3.3
3.6
3.0
2.5
3.4
80
95
95
95
80
80
95
24.2
27.5
25
30.9
21.8
17.6
25
b
Calculated as {[mass VOC]/[mass of (H+VOC)]}x100, where VOC =
adsorbed VOC, H = host (i.e. complex 1 or 2)
18
ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
Table 2
Analysis of the uptake of VOCs by 1 at 35°C.
2
Solvent
Kinetic model
R
Rate constant, k
1
(min )
1
/3
ortho-xylene
(R3) 1 - (1 - α)
0.997 5.36E-2
(
D2) (1 - α) ln (1 - α) + α
0.994 6.25E-2
0.996 2.51E-2
1
/2
meta-xylene
para-xylene
Toluene
(R2) 1- (1- α)
1
/3
(R3) 1 - (1 - α)
0.997 2.10E-2
0.999 1.05E-2
0.996 8.55E-3
0.995 1.95E-2
1
/2
(R2) 1- (1- α)
1
/3
(R3) 1 - (1 - α)
1
/3
(R3) 1 - (1 - α)
(
D2) (1 - α) ln (1 - α) + α
0.996 2.67E-2
0.999 1.75E-2
2
/3
(D4) [1 - 2α/3] - (1 - α)
19
ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
Table 3 Analysis of the uptake of VOCs by 2 at 35°C.
Rate
2
Solvent
Kinetic model
R
constant,

1
k (min )
ortho-
xylene
1
/3
(
(
(
(
(
R3) 1 - (1 - α)
0.999 2.55E-02
0.996 2.18E-02
0.995 3.63E-02
0.983 3.01E-02
0.993 1.79E-02
1
/2
R2) 1- (1- α)
R2) 1- (1- α)
meta-
xylene
1
/2
1
/3
R3) 1 - (1 - α)
para-
xylene
1
/2
R2) 1- (1- α)
1
/3
(
R3) 1 - (1 - α)
0.996 1.59E-02
0.998 2.24E-02
(
(
α)
D2)(1 - α) ln (1 - α) + α
D4) [1 - 2α/3] - (1 -
0.996 6.86E-03
2
/3
1
/2
Toluene
(R2) 1- (1- α)
0.999 3.04E-02
0.999 2.31E-02
1
/3
(
R3) 1 - (1 - α)
20
ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
Figure 1 Synthesis of the PBI ligands L1 and L2
21
ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
Figure 2 Synthesis of the Ru-complexes 1 and 2
22
ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
Figure 3 1H-NMR spectrum (aliphatic region) of 1/p-xylene (px) at different time intervals: immediately
after the uptake reaction (spectrum 1), after 40 minutes at rt in the absence of px vapors (spectrum 2),
after 2 days at rt in the absence of px vapors (spectrum 3).
23
ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
Figure 4 Apparatus used for the sorption kinetics of VOCs. After charging with the solid host sample and
liquid VOC the brass cylinder is sealed (cover not shown in the figure) and the vapor pressure allowed to
equilibrate for 30 minutes at the desired temperature.
24
ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
Figure 5 Profiles of the uptakes of toluene at 35 °C for 1 (a), 2 (b), L1 (c) and L2 (d).
25
ACCEPTED MANUSCRIPT