Synthesis, crystal structure, and physicochemical properties of the new metal-organic framework - The iron(III) complex with benzene- 1,3,5-tricarboxylate

Article abstract of DOI:10.1007/s11172-014-0522-x

A new iron(III) coordination compound with the benzene- 1,3,5-tricarboxylate anion of the composition [Fe^III₂(OH)_0.3(H₂O)_1.7(btc)_4/3]Cl_1.7· 2.8DMF·1.9H₂O (1) was synthesized. Its crystal structure was determined, and the M ? ssbauer spectra were analyzed. The adsorption of salicylaldehyde and nitromethane by compound 1 was studied. The adsorption kinetics of salicylaldehyde is described by the pseudo-second order equation. The diffusion of this aldehyde in particles of compound 1 is, most probably, the rate-limiting step in the Henry reaction.

Full text of DOI:10.1007/s11172-014-0522-x

862
Russian Chemical Bulletin, International Edition, Vol. 63, No. 4, pp. 862—869, April, 2014
Synthesis, crystal structure, and physicochemical properties of
the new metalꢀorganic framework —
the iron(III) complex with benzeneꢀ1,3,5ꢀtricarboxylate
a
a
b
b
c
S. A. Sotnik, S. V. Kolotilov, M. A. Kiskin, Zh. V. Dobrokhotova, K. S. Gavrilenko,
b
b
d
d
a
V. M. Novotortsev, I. L. Eremenko, V. K. Imshennik, Yu. V. Maksimov, and V. V. Pavlishchuk
aL. V. Pisarzhevskii Institute of Physical Chemistry, National Academy of Sciences of Ukraine,
3
1 prosp. Nauki, 03028 Kiev, Ukraine.
Fax: +38 (044) 525 6216. Eꢀmail: shchuk@inphyschemꢀnas.kiev.ua
N. S. Kurnakov Institute of General and Inorganic Chemistry, Russian Academy of Sciences,
b
3
1 Leninsky prosp., 119991 Moscow, Russian Federation.
Fax: +7 (495) 952 1279. Eꢀmail: ileremenko@yandex.ru
Chembio Center, National Taras Shevchenko University of Kyiv,
c
6
1 ul. Chervonotkattskaya, 03022 Kiev, Ukraine.
Fax: +38 (044) 502 4832. Eꢀmail: kgavrio@mail.ru
N. N. Semenov Institute of Chemical Physics, Russian Academy of Sciences,
ul. Kosygina, 119991 Moscow, Russian Federation.
Fax: +7 (499) 137 2951. Eꢀmail: maksimov@chph.ras.ru
d
4
A new iron(III) coordination compound with the benzeneꢀ1,3,5ꢀtricarboxylate anion of the
III
composition [Fe 2(OH)0.3(H O)1.7(btc)4/3]Cl1.7•2.8DMF•1.9H O (1) was synthesized. Its
2
2
crystal structure was determined, and the Mössbauer spectra were analyzed. The adsorption of
salicylaldehyde and nitromethane by compound 1 was studied. The adsorption kinetics of
salicylaldehyde is described by the pseudoꢀsecond order equation. The diffusion of this aldeꢀ
hyde in particles of compound 1 is, most probably, the rateꢀlimiting step in the Henry reaction.
Key words: metalꢀorganic framework, iron(III), benzeneꢀ1,3,5ꢀtricarboxylate, Xꢀray difꢀ
fraction, Mössbauer spectroscopy, sorption, sorption kinetics, Henry reaction.
Metalꢀorganic frameworks (MOF), porous coordinaꢀ
as opposed to the earlier described14 isostructural MOF
tion polymers, have attracted rapidly increasing interest
due to a unique combination of their functional properties
[Fe (H O) (btc)4/3]Cl (2) containing equal amounts of
2
2
2
2
+
3+
Fe and Fe ions. Complex 1 exhibits high catalytic
activity in the Henry reaction, viz., the condensation of
salicylaldehyde with nitromethane. Therefore, it is of inꢀ
terest to study in detail the molecular and electronic strucꢀ
ture of this coordination compound.
The aim of the present work is to determine the crystal
structure of complex 1, which we have synthesized, analyze
its Mössbauer spectra, and study the adsorption kinetics of
salicylaldehyde by this compound, which is important for
the determination of the rateꢀlimiting step in the Henry
reaction with nitromethane catalyzed by this MOF.
and the ease of varying the chemical composition and crysꢀ
tal structures of these compounds.1
—4
The adsorption,
magnetic, catalytic, and luminescence properties of such
materials are of particular interest. Due to these properꢀ
5
ties, MOF are considered as promising adsorbents (materꢀ
6
ials for gas storage or separation of mixtures of different
7
8
9
compounds, enantioselective adsorbents ), catalysts (inꢀ
1
0
cluding stereospecific catalysts ), and active compounds
for the design of sensitive elements of sensors11 and displays.
Even quite slight changes in the conditions of the synꢀ
thesis and isolation of MOF lead to the preparation of
coordination polymers, which differ in the composition,
structure, and dimensionality (oneꢀ, twoꢀ, and threeꢀdiꢀ
12
Results and Discussion
mensional coordination polymers).1 We have shown that
3
9
Synthesis and the crystal structure of compound 1. In
1
4
the reaction of 1,3,5ꢀbenzenetricarboxylic acid (H btc)
attempting to reproduce the earlier described procedure
3
with iron(III) chloride in dimethylformamide (DMF) in
for the synthesis of the MOF [Fe (H O) (btc)4/3]Cl (2)
2
2
2
the presence of a base affords MOF of the composition
by the reaction of anhydrous iron chloride with trimesic
acid in dimethylformamide in the presence of triethylenediꢀ
III
[
Fe 2(OH)0.3(H O)1.7(btc)4/3]Cl1.7•2.8DMF•1.9H O (1),
2
2
Published in Russian in Izvestiya Akademii Nauk. Seriya Khimicheskaya, No. 4, pp. 0862—0869, April, 2014.
066ꢀ5285/14/6304ꢀ0862 © 2014 Springer Science+Business Media, Inc.
1

Iron(III) complex with benzeneꢀ1,3,5ꢀtricarboxylate Russ.Chem.Bull., Int.Ed., Vol. 63, No. 4, April, 2014
863
amine (as a base), we isolated a new coordination compound
b
c
III
of the composition [Fe 2(OH)0.3(H O)1.7(btc)4/3]Cl1.7
•
2
*
*
0
•
2.8DMF•1.9H O (1). Compounds 1 and 2 were syntheꢀ
2
a
sized using the reagents and solvents in equal molar ratios.
The only difference is that, in the present study, we used
anhydrous triethylenediamine, whereas its hexahydrate
*
1
4
was added to the reaction mixture in the earlier study.
Probably, this difference has a decisive effect on the acidꢀ
base characteristics of the reaction mixture and interferes
*
C(2)
Fe(1´)
C(1)
O(1)
3
+
2+
with the reduction of Fe to Fe . Under the reaction
conditions described earlier,14 the insoluble reaction prodꢀ
*
*
uct (complex 2) contained Fe3+ and Fe ions in a ratio of
2+
O(1w)
1
: 1, which can be attributed to the partial reduction of
Fe(1)
*
*
iron(III). In compound 1 synthesized in the present study,
both iron ions in the binuclear unit have the oxidation
state +3, as evidenced mainly by the Mössbauer spectra,
as well as by elemental analysis and Xꢀray diffraction data.
The solvent composition of the complex 1•2.8DMF•
Fig. 1. Fragment of 3D coordination polymer 1. The hydrogen
atoms are omitted for clarity; the carbon atoms of the carboxy
groups, to which the phenyl rings of btc3– (not shown in the
figure) are attached, are marked with an asterisk.
•
1.9H O corresponds to the sample existing in equilibriꢀ
2
3
–
um in air. The solvent content of the single crystal may
differ from that given above; however, we failed to posiꢀ
tion solvent molecules in the crystal structure.
Complex 1 is a 3D coordination polymer isostructural
to compound 2 described earlier. The crystal lattice of
btc anions have the local symmetry D . Therefore, all
3h
3
+
Fe ions and oxygen atoms are in equivalent positions.
3
+
The Fe ions are coordinated by five O atoms and
have a squareꢀpyramidal coordination geometry with
the carboxyl oxygen atoms at the base of the pyramid
(d(Fe—Ocarbox) = 2.031(8) Å) and a water molecule in the
apical position (d(Fe—O ) = 2.33(2) Å). The Fe—O bond
lengths in compound 1 are somewhat smaller than those
in the complex [Fe (H O) (btc) ]Cl•4.5DMF (2.048(4)
and 2.358(7) Å, respectively, at the same temperature ),
which is consistent with the presence of only Fe ions in
compound 1, as opposed to [Fe (H O) (btc)4/3]Cl•
•4.5DMF containing both Fe and Fe ions (in the
latter compound, the Fe and Fe2+ ions also occupy
equivalent positions, and the Fe—O bond lengths are avꢀ
eraged for both oxidation states of iron). The shortest disꢀ
tance between the iron ions in complex 1 is 2.970(6) Å.
The crystal lattice of complex 1 contains rings comꢀ
posed of three and four {Fe (btc)} groups (see Fig. 1). In the
3D framework of this complex, there are channels (Fig. 2),
the volume of which is 53% of the crystal unit cell volume.
This corresponds to the pore volume of 0.51 cm g (calꢀ
culated for the probe molecule with r = 1.4 Å using the
PLATON program taking into account the fact that the
cavities are partially occupied by chloride ions, which canꢀ
not be removed). The total solventꢀaccessible pore volume
in complex 1 (including coordinated water molecules) is
3
+
compound 1 is formed by the binding of Fe ions via
3
–
btc anions involved in the complex in a 3 : 2 ratio. Thereꢀ
fore, the crystals are composed of paddlewheel structural
units, which form the threeꢀdimensional structure of the
MOF. The positive charge of the framework is compenꢀ
sated by chloride ions, which are located in the cavities
similar to those in the structure of complex 2, and coꢀ
ordinated hydroxide ions. The amount of chloride ions
was fixed at the level of 1.7 Cl ions per two Fe ions,
which is consistent with the elemental analysis data (the
localization of the Cl ions was described in detail in the
Experimental section). Despite the fact that the positions
of the chloride ions in the crystal lattice cannot be unamꢀ
biguously determined by Xꢀray diffraction, the results of
Mössbauer spectroscopy study of complex 1, as well as the
data published in the literature1 for isostructural complex 2,
can attest to the presence of these anions in the outer
coordination sphere of Fe3+ ions.
Since the hydrogen atoms cannot be located by Xꢀray
diffraction, it is impossible to make a reliable assignment
of the coordinated oxygen atoms. Thus, they can belong
to either hydroxide ions or water molecules. Based on the
principle of electroneutrality of the complex, the ratio of
hydroxide anions to water molecules in complex 1 is 0.3 : 1.7
ax
2
2
2
4/3
1
4
3
+
1
4
2
2
2
2+
3
+
–
3+
3+
–
2
4
3
–1
1
5
3
–1
about 60% (0.68 cm g ).
Thermogravimetric and differential thermal analysis,
Xꢀray powder diffraction study, and Mössbauer spectroꢀ
scopy for compound 1. Complex 1 and desolvated sample
1´ were studied by thermogravimetric and differential therꢀ
mal (TG/DTA) analysis with the massꢀspectrometric deꢀ
tection of thermal decomposition products, Xꢀray powder
diffraction, and Mössbauer spectroscopy in order to imꢀ
prove the accuracy of the solvent content determination
(
nonꢀcoordinated water is not taken into account).
In the crystal lattice, the carboxyl oxygen atoms of the
1,3,5ꢀbenzenetricarboxylate anion are coordinated to the
iron(III) ions to form the binuclear units Fe (CO ) (H O)
2
2
2 4
2
(
CO is the carboxy group of 1,3,5ꢀbenzenetricarboxylate,
2
Fig. 1). These binuclear units have the local symmetry
D4h (the C axis coincides with the Fe—Fe bond), and the
4

8
64
Russ.Chem.Bull., Int.Ed., Vol. 63, No. 4, April, 2014
b
Sotnik et al.
Table 1. Parameters of the Mössbauer spectra of samples 1 and 1´a
0
c
b
Sample
State of Fe


mm s–1
+
1
1
1
Fe3 (paramag.)
0.43 0.68 0.60
0.47 0.68 0.59
0.43 0.62 0.60
3
+
´
Fe (paramag.)
3
+
´ (after the catalytic
Fe (paramag.)
experiment)
a
b
The accuracy of measurements is ±0.03 mm s–1
The linewidth.
.
in which Fe3+ ions are in the coordination environment
1
4,17
formed by oxygen donor atoms.
Apparently, the low
values of the quadrupole splitting are indicative of the
relatively highly symmetric electron density distribution
3
+
about the Fe ions and provide indirect evidence for the
III
presence of the chromophore of the composition Fe O5
Fig. 2. Fragment of the crystal lattice of complex 1. The atoms
are represented by van der Waals spheres to illustrate the cavities.
One position is shown for each of the disordered chloride ions.
Note. Fig. 2 is available in full color in the onꢀline version of the
journal (http://www.springerlink.com).
in compound 1 (as well as in its desolvated form 1´).
The Xꢀray powder diffraction patterns and the Mössꢀ
bauer spectra (see Fig. 3, Table 1) of desolvated sample 1´
are identical to those of the starting compound 1, as well
as to the corresponding characteristics of sample 1´ after
catalytic studies (the results of these studies are considꢀ
ered below). These observations suggest that the crystal
lattice of complex 1 is stable under conditions of activaꢀ
tion (desolvation) and the condensation of salicylaldeꢀ
hyde with nitromethane.
and the estimation of the oxidation state of the iron ions,
as well as to confirm the phase purity of the bulk sample
and the stability of the crystal lattice of this compound
after the desolvation.
During heating of complex 1, the weight loss started at
nearꢀroom temperature. At 250 C (and under atmospherꢀ
ic pressure), the weight loss is about 18%, which correꢀ
sponds to approximately 1.7 DMF molecules (or the loss
Catalytic activity of complex 1´. We studied the catalytꢀ
ic activity of sample 1´ in the Henry reaction, viz.,
the condensation of salicylaldehyde with nitromethane9
(Scheme 1). It is known that, depending on the condiꢀ
tions, this reaction can afford either nitro alcohols or unꢀ
of all nonꢀcoordinated water molecules, 1.9 H O, as well
2
1
8,19
as 1.3 DMF) per formula unit Fe . The subsequent heatꢀ
saturated nitro compounds.
2
ing leads to a sharp weight loss due apparently to the
decomposition of the sample.
Scheme 1
The mass spectrum of the thermal decomposition prodꢀ
ucts shows only a peak at m/z = 73 assigned to the molecular
ion of DMF. The peaks at m/z = 44—45, which could be
assigned to dimethylamine as the possible hydrolysis prodꢀ
uct of DMF in the reaction mixture, are absent. It should
be noted that the TG/DTA study of the samples obtained
at different periods of time gave somewhat different values
of the weight loss, as well as the different relative intensiꢀ
ties of the signal of water in the mass spectra of the thermal
decomposition products, which can be attributed to
a change in the solvent composition of complex 1 dependꢀ
ing on the atmospheric moisture and the period of the
storage of the sample. This behavior is typical of coordiꢀ
The reaction with the use of a suspension of sample 1´
in toluene gives exclusively transꢀnitrovinyl phenol durꢀ
ing the course of the reaction (this fact was confirmed by
H NMR spectroscopic data). It was found that the depenꢀ
dence of the reaction rate on the concentrations of the
reagents could be described by the equation of zero order
with respect to the concentration of nitromethane and the
equation of the first order with respect to the concentraꢀ
tions of salicylaldehyde and compound 1´:
9
1
1
6
nation compounds having porous crystal structures.
Figure 3 presents the Mössbauer spectra of the starting
complex 1 and desolvated sample 1´ measured at 300 K.
3
+
Both spectra show a doublet corresponding to Fe . The
–
1
isomer shifts  0.43—0.47 mm (Table 1), as well as the
–
1
quadrupole splitting of the doublets  0.62—0.68 mm
,
are in the ranges typical of highꢀspin iron(III) compounds,
v = k [SA][1´],
(1)
1

Iron(III) complex with benzeneꢀ1,3,5ꢀtricarboxylate Russ.Chem.Bull., Int.Ed., Vol. 63, No. 4, April, 2014
865
I (arb. units)
.00
a
b
1
0
0
.98
.96
–
10
–5
0
5
10 V/mm s–1
0
10
20
30
40 2
I (arb. units)
c
d
1
0
0
.00
.98
.96
–
10
–5
0
5
10 V/mm s–1
0
10
20
30
40 2
I (arb. units)
e
f
1
0
0
.00
.98
.96
–
10
–5
0
5
10 V/mm s–1
0
10
20
30
40 2
Fig. 3. Mössbauer spectra and Xꢀray powder diffraction patterns of compounds 1 (a, b), 1´ (c, d), and sample 1´ after the catalytic
experiment (e, f).
where SA is salicylaldehyde and k is the rate constant of
tion of nitromethane by compound 1´ does not have
1
the sorption/activation of SA by compound 1´, which is
the rateꢀlimiting step in the formation of the product.
Equation (1) can be written as follows:
a decisive role in the determination of the kinetics of the
Henry reaction, but the possibility of the filling of the pores
9
in sample 1´ by nitromethane is fundamentally important for
confirming the supposition that the catalytic reaction proꢀ
ceeds in pores of the crystal lattice of the MOF under study.
For the sorption from the gas phase, the adsorption
capacity of sample 1´ with respect to nitromethane at 20 C
v = k[SA],
(2)
where k = k [1´].
1
The rate constant k, as it is typical of the metal comꢀ
–
1
3
–1
plex catalysis, depends on the concentration of compound
is 1.6 mol mol (0.19 cm g at p/p = 0.9). Sample 1´
S
–
1
3 –1
1
´. At [1´] = 0.05 mol L (per formula unit Fe ), the
adsorbs 0.21 cm g of salicylaldehyde from the gas phase
2
constant k = 3.35•10 min–1 (see Ref. 9).
–
4
at p = p in the presence of air. Therefore, the accessible
S
Adsorption properties of complex 1´. The adsorption
isotherms of salicylaldehyde (Fig. 4, a) and nitromethane
pore volumes in the crystal lattice of sample 1´ are compaꢀ
rable for these substrates. These values are much smaller
than the accessible pore volume for the probe molecules
with r = 1.4 Å (see above), but are similar to the accessible
pore volume for molecules with r = 3.5 Å, which approxꢀ
imately corresponds to the kinetic radius of aromatic molecꢀ
(
Fig. 4, b) by compound 1´ can be described by the Langꢀ
muir equation. According to the Langmuir adsorption
model, the limiting adsorption capacities amax of comꢀ
–1
pound 1´ are 0.223(8) and 0.24(1) mol (mol Fe ) , and the
2
2
0
3 –1
adsorption constants K373 are 26(3) and 25(5) moles of
ules containing two substituents (0.21 cm g ; the estiꢀ
1
5
Fe per mole of the substrate for salicylaldehyde and nitroꢀ
mate was obtained using the PLATON program as deꢀ
scribed above). In the case of the adsorption from solutions
at 100 C, the adsorption capacity of MOF 1´ is lower (see
2
methane, respectively. Despite the similar values of amax
and K373 for salicylaldehyde and nitromethane, the adsorpꢀ

866
Russ.Chem.Bull., Int.Ed., Vol. 63, No. 4, April, 2014
Sotnik et al.
a/mol mol–1
a/mol mol–1
.25
0.20
a
0
.20
.15
.10
.05
3
0
2
1
0
0
0
0
.15
.10
.05
0
0
0
.1
0.2
0.3
[SA]/mol L–1
20
40
60
80
100 120 140 t/min
Fig. 5. Plot of the adsorption of salicylaldehyde (at the initial
concentration of 0.15 (1), 0.43 (2), and 0.69 (3) M) by comꢀ
pound 1´ versus the time. The curves are given for descriptive
purposes. The content of adsorbed aldehyde is given per mole of
a/mol mol–1
b
0
0
0
0
.20
.15
.10
.05
compound 1´ (the formula unit is Fe 100 C).
2
a/mol mol–1
However, the the relevant plots shown in Fig. 6 are
linear only in the initial regions, after which a sharp bend
and the second linear region are observed. Such patterns
are characteristic of the complex mechanism of the adꢀ
sorption, including the fast initial external diffusion step
followed by the prevailing slow internal diffusionꢀconꢀ
trolled adsorption in pores of MOF. This process can be
1
0
.2
.6
0
.3
0.6
0.9 p/ps
2
1
–
1
described by the following equation:
0
.1
0.2
[CH NO ]/mol L
3 2
Fig. 4. Isotherms of salicylaldehyde (a) and nitromethane (b)
adsorption from a toluene solution by compound 1´ (a is the
amount of adsorbed salicylaldehyde or nitromethane, [SA] and
a = k t + A,
1/2
(4)
t
d
where a is the amount of adsorbed salicylaldehyde per
t
[
CH NO ] are the equilibrium concentrations of salicylaldeꢀ
mole of compound 1´, k is the rate constant of internal
diffusion, t is the time, and A is the amount of aldehyde
adsorbed in the initial instant of time at t  0.
3
2
d
hyde and nitromethane, respectively, in the solution). The conꢀ
tent of adsorbed aldehyde is given per mole of compound 1´ (the
formula unit is Fe , 100 C). The inset shows the adsorption
isotherm of nitromethane by desolvated sample 1´ from the gas
2
The slow internal diffusion process, as the adsorption
reaches the steady state mode, may be evidenced by the
low slope of the corresponding region in the plot of the
amount of adsorbed aldehyde versus t0 (Fig. 7).
phase at 20 C.
.5
Fig. 4). This is due both to the higher temperature and the
possible competitive adsorption of the solvent (toluene).
In the present work, the adsorption of salicylaldehyde
by sample 1´ from a toluene solution was studied in detail.
This process can be the rateꢀlimiting step in the catalytic
reaction (as a necessary condition for the activation of
aldehyde). The kinetic curves of adsorption of salicylaldeꢀ
hyde (Fig. 5) are characterized by a sharp increase in the
initial region. The adsorption of the substrate reaches the
saturation plateau during approximately 20 min. Some
assumptions about the mechanism of the adsorption can
be made based on the analysis of the applicability of different
models to the interpretation of experimental data. If the
external diffusion is the rateꢀlimiting step of adsorption, the
–
ln(1 – F)
3
3
2
2
1
1
5
0
5
0
5
0
5
1
2
3
2
1
20
40
60
80
100 120 140 t/min
experimental data should be described by the equation:
Fig. 6. Plot of –ln(1 – F) versus the time for the adsorption of
salicylaldehyde (at the initial concentration of 0.15 (1), 0.43 (2),
and 0.69 (3) M).
ln(1 – F)= –kt,
(3)
where F is the rate, at which the saturation is reached.

Iron(III) complex with benzeneꢀ1,3,5ꢀtricarboxylate Russ.Chem.Bull., Int.Ed., Vol. 63, No. 4, April, 2014
867
a/mol mol–1
.25
t•a(SA)t–1/min mol(Fe ) mol(SA)
–1
2
0
0
0
800
.20
.15
6
4
00
00
1
2
3
1
2
3
2
4
6
8
10
12 t0.5/min0.5
200
Fig. 7. Plot of the amount of adsorbed salicylaldehyde versus t0.5
(
at the initial concentration of 0.15 (1), 0.43 (2), and 0.69 (3) M).
4
0
80
120
160
200 t/min
Fig. 9. Plot of t/a(SA) versus the time (at the initial salicylaldeꢀ
hyde concentration of 0.15 (1), 0.43 (2), and 0.69 (3) M, 100 C).
The estimated rate constant of the internal diffusion k
t
d
–
0.5
is ~0.005±0.0005 mol min
per mole of compound 1´
0.5
(
1.2 g min– per kg of compound 1´). This value falls in the
range of the constants k found for the diffusion of organic
dyes into adsorbents of the polysaccharide nature.
d
the pseudoꢀfirstꢀorder rate constant. The constant k =
2
2
2
–1
–1
=
4.3±0.3 mol(Fe ) mol(SA) min was estimated from
2
Presumably, it is the internal diffusion step that conꢀ
trols the rate of the Henry reaction of salicylaldehyde with
nitromethane. This conclusion is consistent with the fact
that the same changes in the salicylaldehyde concentraꢀ
tion are observed for the adsorption from a toluene soluꢀ
tion and from the reaction mixture containing aldehyde
and nitromethane (Fig. 8).
the corresponding linear regressions (Fig. 9) with a quite
satisfactory accuracy, and the correlation coefficient is
higher than 0.997 in all cases. The value of k is comparable
2
with the corresponding rate constants of adsorption kM of
divalent metal ions onto a sorbent based on sphagnum
23
moss peat (and its value is closer to the lower limit of the
values of the constant k , which is consistent with the
M
The adsorption associated with the chemisorption of
the substrate, the occurrence of which can be presumed in
the system under consideration, is often described using
difference in the size of the substrates). The adsorption of
salicylaldehyde is most likely controlled by the internal
diffusion coupled with the coordination of aldehyde to
iron(III) ions at the vacant coordination site, which deterꢀ
mines the total rate of the MOFꢀcatalyzed Henry reaction.
2
2,23
the following pseudoꢀsecond order equation:
2
t/a(SA) = 1/k a(SA) + t/a(SA)s,
t
2
s
3
+
The coordination of salicylaldehyde to Fe is eviꢀ
denced by a comparison of the diffuse reflectance spectra
of complex 1 and sample 1´, which was kept in a toluene
solution of salicylaldehyde (Fig. 10). The shifts of absorpꢀ
tion bands after the desolvation followed by the adsorpꢀ
where a(SA) and a(SA) are the current value of the
t
s
adsorption of the aldehyde and its adsorption after the
saturation, respectively (one mole of salicylaldehyde per
mole of compound 1´, the formula unit Fe ), and k is
2
2
C(SA)/mol L–1
R (%)
0
0
0
0
0
0
.152
60
.148
.144
.140
.136
.132
1
2
2
4
2
0
0
1
6
00
800
1000
1200 /nm
5
0
100
150
200
t/min
Fig. 10. Diffuse reflectance spectra of complex 1 (1) and the
sample obtained after keeping 1´ in a toluene solution of salicylꢀ
aldehyde (2). The maxima of the longꢀwavelength absorption
bands are indicated by arrows.
Fig. 8. Changes in the salicylaldehyde concentration during the
adsorption from a toluene solution (1) and from the reaction
mixture containing toluene and nitromethane (2) (100 C).

868
Russ.Chem.Bull., Int.Ed., Vol. 63, No. 4, April, 2014
Sotnik et al.
tion of salicylaldehyde (for example, the shift of the longꢀ
est wavelength band from 975 to 967 nm) can be due to the
replacement of a water molecule in the axial position of
Table 2. Crystallographic parameters and the structure refineꢀ
ment statistics for 1
the coordination environment of the Fe3+ ion by a salicylꢀ
aldehyde molecule similar to the change in the spectrum
Parameter
1
а
Molecular formula
Molecular weight/g mol
C H Cl Fe O
15
18
6
2,55
3
2
+
upon the coordination of benzaldehyde to Cu ions in
Cu (btc) described earlier.2
–1
720.62
4
2
4/3
Т/K
293(2)
Therefore, the following conclusions can be drawn.
Crystal system
Space group
a/Å
Cubic
Fm3 m
–
III
The new complex of the composition [Fe 2(OH)0.3
ꢀ
26.669(17)
18968(21)
16
1.009
1.089
25.08
32110
910
578
0.3504
39
1.362
0.1482
0.3869
(
H O)1.7(btc)4/3]Cl1.7•2,8DMF•1.9H O (1) belonging to
2
2
V/Å3
the family of metalꢀorganic frameworks was synthesized.
The crystal structure of 1 was determined. The presence of
iron ions in the oxidation state +3 in compound 1 was
established by Mössbauer spectra. The rate of diffusion of
salicylaldehyde in the desolvated complex (sample 1´) is
controlled by the internal diffusion and, as a consequence,
the rate of condensation of salicylaldehyde with nitroꢀ
methane catalyzed by sample 1´ is most likely determined
by the internal diffusionꢀcontrolled adsorption of salicylꢀ
aldehyde within pores of this MOF.
Z
dcalc/g cm–3

/mm–1
max/deg
Number of measured reflections
Number of unique reflections
Number of reflections with I > 2(I)
Rint
Number of parameter
GOOF
R (I > 2I))
1
wR (I > 2I))
2
Experimental
a
The hydrogen atoms at the coordinated oxygen atom were
not localized.
Salicylaldehyde (Merck) and DMF (Makrokhim) were disꢀ
tilled before use. Nitromethane and toluene were dehydrated
according to standard procedures.2 Other reactants and solꢀ
vents (reagent grade) were used without additional purification.
Desolvated complex 1 (referred to as sample 1´) was prepared by
heating complex 1 in vacuo (0.1 Torr) over P O at 150 C.
5
The elemental analysis was carried out on a Carlo Erba 1106
automated analyzer. The analysis for chlorine was performed as
described earlier. The Xꢀray powder diffraction patterns were
obtained on a Bruker D8 Advance diffractometer (CuꢀK radiaꢀ
tion,  = 0.154 nm).
The TG/DTA study of complex 1 was carried out on a LABꢀ
SYS ATD/DSC1600 instrument coupled with a Pfeiffer Omniꢀ
star mass spectrometer.
The thermal decomposition of compound 1 was studied by
differential scanning calorimetry (DSC) and thermogravimetric
analysis (TGA) on NETZSCH instruments. The thermograviꢀ
metric measurements were performed in a flow of artificial air
1
4
2
5
The singleꢀcrystal Xꢀray diffraction data set for compound 1
was collected on a Bruker SMART APEX II diffractometer
equipped with a CCD detector (MoꢀK radiation,  = 0.71073 Å,
2
6
graphite monochromator) according to a standard procedure.
The structure was solved and refined by the fullꢀmatrix leastꢀ
2
squares method based on F using the SHELXLꢀ97 program
package.27 All nonhydrogen atoms were found in the difference
Fourier map and refined with anisotropic displacement paraꢀ
meters. The hydrogen atoms were placed in idealized positions
according to their type and refined using a riding model. The
solvent molecules in the cavities of the crystal lattice were not
located and the corresponding electron density was removed
using the SQUEEZE routine in the PLATON software packꢀ
–
1
(O , 20.8%; CH , < 0.0001%) (20 mL min ) and argon (Ar,
2
4
> 99.998%; O , < 0.0002%; N , < 0.001%, water vapor,
2
2
–
1
< 0.0003%; CH , < 0.0001%) (20 mL min ) on a TG 209 F1
4
–1
instrument in alundum crucibles at a heating rate of 10 deg min
.
1
5
age. The low quality of the crystal did not allow us to unambigꢀ
uously determine the positions occupied by chloride ions, and
the possibility of their coordination to Fe3+ ions cannot be exꢀ
cluded based on the Xꢀray diffraction data (the R factor depends
only slightly on the choice of Cl– sites and the occupancies for
these sites). Chloride ions were located in two sites with the
same occupancies equal to 1.7/16 (the number of chloride ions
was fixed at the level corresponding to the results of chemical
analysis; both atoms are on fourfold axes). To retain the electroꢀ
static balance, the oxygen atoms coordinated to the iron ions
should be considered as constituents of water molecules or hydrꢀ
oxy groups, which are indistinguishable based on the Xꢀray difꢀ
fraction data. It should be noted that the assumption of the presꢀ
ence of Fe2 ions in complex 1 is in contradiction with the
Mössbauer spectroscopic data. The crystallographic data for comꢀ
pound 1 are given in Table 2.
The composition of the gas phase was studied on a QMS 403C
Aёolos massꢀspectrometric unit coupled with thermogravimetric
analysis. The ionizing electron energy was 70 eV; the maximum
determined mass number (the massꢀtoꢀcharge (Z) ratio of the
ion) was 300 amu. The weights of the samples used in thermoꢀ
gravimetric experiments were 0.5—3 mg. Differential scanning
calorimetry was performed in a flow of dry artificial air and
argon on a DSC 204 F1 calorimeter in aluminum cells at
a heating rate of 10 deg min– . The weights of the samples were
1—4 mg. The temperature of the thermobalance and the caloꢀ
rimeter was calibrated based on the phase transition points of the
reference compounds (C H , Hg, KNO , In, Sn, Bi, CsCl,
1
6
12
3
99.99% purity) according to the ISO/CD 11357ꢀ1 standard. The
samples used for the TGA and DSC experiments were weighed
using a Sartorius Research R 160P analytical balance with an
accuracy of 1•10–2 mg. The thermal analysis data were proꢀ
+

Iron(III) complex with benzeneꢀ1,3,5ꢀtricarboxylate Russ.Chem.Bull., Int.Ed., Vol. 63, No. 4, April, 2014
869
cessed using the NETZSCH Proteus Thermal Analysis software
according to the ISO 11357ꢀ1, ISO 11357ꢀ2, ISO 11358, and
ASTM E 1269ꢀ95 standards.
References
1
2
. H.ꢀC. Zhou, J. R. Long, O. M. Yaghi, Chem. Rev., 2012,
12, 673.
. E. Chelebaeva, J. Larionova, Y. Guari, R. A. S. Ferreira,
The 57Fe Mössbauer spectra were measured on a Wissel elecꢀ
trodynamic spectrometer in the temperature range of 16—300 K
using a Janis (CCSꢀ850) helium cryostat equipped with a Lake
Shore Cryotronics (332) temperature controller. The temperature
was maintained with an accuracy of at least 0.1 K. A 57Co(Rh)
radiation source with an activity of 1.1 GBq was used. The isomer
shifts were measured relative to the center of the magnetic hyperꢀ
fine structure of metallic iron. The Mössbauer spectra were proꢀ
cessed using the standard leastꢀsquares programs (LORENꢀICHPH
RAS, NORMOSꢀGermany) assuming that the lines are described
by Lorentzian profiles.
1
L. D. Carlos, F. A. A. Paz, A. Trifonov, C. Guérin, Inorg. Chem.,
2
008, 47, 775.
3
4
. S. V. Kolotilov, M. A. Kiskin, I. L. Eremenko, V. M. Novoꢀ
tortsev, Curr. Inorg. Chem., 2013, 3, 144.
. M.ꢀH. Xie, X.ꢀL. Yang, Y. He, J. Zhang, B. Chen, C.ꢀD.
Wu, Chem. Eur. J. 2013, 19, 14316.
. H. Wu, Q. Gong, D. H. Olson, J. Li, Chem. Rev., 2012, 112, 836.
. R. B. Getman, Y.ꢀS. Bae, C. E. Wilmer, R. Q. Snurr, Chem.
Rev., 2012, 112, 703.
. J.ꢀR. Li, J. Sculley, H.ꢀC. Zhou, Chem. Rev., 2012, 112, 869.
. D. N. Dybtsev, A. L. Nuzhdin, H. Chun, K. P. Bryliakov,
E. P. Talsi, V. P. Fedin, K. Kim, Angew. Chem., Int. Ed.
Engl., 2006, 45, 916.
. S. A. Sotnik, I. L. Eremenko, V. V. Pavlishchuk, K. S.
Gavrilenko, Teor. Eksp. Khim., 2013, 49, 94 [Theor. Exp.
Chem. (Engl. Transl.), 2013, 49].
0. M. Yoon, R. Srirambalaji, K. Kim, Chem. Rev., 2012, 112, 1196.
1. L. E. Kreno, K. Leong, O. K. Farha, M. Allendorf, R. P.
Van Duyne, J. T. Hupp, Chem. Rev., 2012, 112, 1105.
2. Y. Cui, Y. Yue, G. Qian, B. Chen, Chem. Rev., 2012, 112, 1126.
3. T. B. eliè, M. Rangus, K. Lázár, V. Kau i , N. Z. Logar,
Angew. Chem., Int. Ed. Engl., 2012, 51, 12490.
5
6
To study the sorption of salicylaldehyde or nitromethane at
7
8
1
00±1 C, a toluene solution of the corresponding substrate at
a specified concentration was placed in sealed vials containing
weighed samples of compound 1´. The effective concentration
–
1
of compound 1´ in the suspension was 0.05 mol L (the concenꢀ
tration of compound 1´ was determined as the ratio of the amount
of suspended 1´ to the volume of the solution). The suspension
was magnetically stirred for 2 h until the adsorption equilibriꢀ
um was reached. Then the concentration of the substrate in
solution was determined by spectrophotometry. The sorption
kinetics of salicylaldehyde in the initial concentration range of
9
1
1
1
1
0
.15—0.69 mol L–1 was studied in a similar way.
The sorption of nitromethane from the gas phase by a sample
of desolvated compound 1´ was determined gravimetrically at 20 C.
The sample was desolvated as described above and then was
additionally heated inside an apparatus for sorption measureꢀ
ments at 120 C in vacuo (10–3 Torr). The adsorption capacity
with respect to salicylaldehyde from the gas phase was deterꢀ
mined by measuring the increase in the weight of desolvated
sample 1´ in air saturated with salicylaldehyde vapor at 20 C.
The adsorption isotherm of salicylaldehyde from the gas phase at
1
1
1
4. L. Xie, S. Liu, C. Gao, R. Cao, J. Cao, C. Sun, Z. Su, Inorg.
Chem., 2007, 46, 7782.
5. A. L. Spek, PLATON, A Multipurpose Crystallographic Tool,
Utrecht University, Utrecht, 1998.
6. A. V. Pavlishchuk, S. V. Kolotilov, M. Zeller, L. K. Thompꢀ
son, I. O. Fritsky, A. W. Addison, A. D. Hunter, Eur. J.
Inorg. Chem., 2010, 4851.
7. D. S. Nesterov, E. N. Chygorin, V. N. Kokozay, V. V. Bon,
R. Boèa, Y. N. Kozlov, L. S. Shulpina, J. Jezierska, A. Ozarowꢀ
ski, A. J. L. Pombeiro, G. B. Shulpin, Inorg. Chem., 2012,
1
2
0 C was not measured because of the low saturated vapor presꢀ
sure of this substrate.
Procedure for the synthesis of complex 1. A mixture of FeCl3
5
1, 9110.
(
0.455 g, 2.8 mmol), H btc (0.588 g, 2.8 mmol), and triethyleneꢀ
3
1
1
2
2
8. D. Enders, M. R. M. Huttl, J. Runsink, G. Raabe, B. Wendt,
Angew. Chem., Int. Ed. Engl., 2007, 46, 467.
9. K. Motokura, M. Tada, Y. Iwasawa, Angew. Chem., 2008,
diamine (0.470 g, 4.2 mmol) in dimethylformamide (84 mL) was
placed in a 200ꢀmL Teflonꢀlined autoclave, heated at 150 C for
four days, and then gradually cooled to ~20 C. The resulting
brown crystals were washed with dimethylformamide and dried in
1
20, 9370.
0. J.ꢀR. Li, R. J. Kuppler, H.ꢀC. Zhou, Chem. Soc. Rev., 2009,
8, 1477.
1. R. M. Alosmanov, Vestn. Mosk. Univ., Ser. Khim., 2011, 52,
45 [Vestn. Mosk. Univ., Ser. Khim. (Engl. Transl.), 2011, 52].
air. The yield of the product was 90% (based on FeCl ). Found (%):
3
3
C, 35.11; H, 3.73; N, 5.96; Cl, 8.4. Fe C20.4H28.3Cl1.7N2.8O13
2
III
(
[Fe 2(OH)0.3(H O)1.7(btc)4/3]Cl1,7•2.8DMF•1.9H O). Calꢀ
2
2
1
culated (%): C, 35.13; H, 4.09; N, 5.62; Cl, 8.6.
2
2
2
2. F.ꢀC. Wui, R.ꢀL. Tseng, R.ꢀS. Juang, Wat. Res., 2001, 35, 613.
3. Y. S. Ho, G. McKay, Wat. Res., 2000, 34, 735.
4. K. Schlichte, T. Kratzke, S. Kaskel, Microp. Mesop. Mater.,
We thank R. A. Polunin for measuring the adsorption
of nitromethane from the gas phase.
The Xꢀray diffraction experiment was performed in the
Center of Collective Use of the N. S. Kurnakov Institute
of General and Inorganic Chemistry of the Russian Acadꢀ
emy of Sciences.
2
004, 73, 81.
2
5. A. Weissberger, E. S. Proskauer, J. A. Riddick, E. E. Toops,
Organic Solvents, Interscience Publishers Inc., New York,
1955, 552 pp.
26. SMART (Control) and SAINT (Integration) Software, Verꢀ
This study was financially supported by the Joint
Project of the National Academy of Sciences of Ukraine
and the Russian Foundation for Basic Research (Project
Nos 10ꢀ03ꢀ12(U), 10ꢀ03ꢀ13(U), 12ꢀ03ꢀ90418ꢀUkr_a, and
sion 5.0, Bruker AXS Inc., Madison, WI, 1997.
2
7. G. M. Sheldrick, SHELXSꢀ97 and SHELXL97, Program for
the Refinement of Crystal Structure, University of Göttingen,
Göttingen (Germany), 1997.
14ꢀ03ꢀ90423ꢀUkr_f) and the Russian Foundation for Basic
Received July 18, 2013;
Research (Project No. 12ꢀ03ꢀ00007ꢀa).
in revised form January 28, 2014