Iron-containing defect-rich mixed metal oxides for Friedel-Crafts alkylation

Article abstract of DOI:10.1016/j.apcata.2012.07.025

Several stable iron-containing spinel oxides with the composition CuFe ₂O₄ or FeCr₂O₄ were prepared by different methods including thermal nitrate decomposition, combustion synthesis, and mechano-chemical synthesis. Materials were compared to commercially available spinels produced by conventional ceramic route with respect to their structural properties and catalytic activity. Phase compositions and microstructures of the catalysts were studied in detail. The catalysts were tested in the Friedel-Crafts alkylation of various aromatic compounds with different alkyl halides as alkylating agents. Microstructural properties (microstrains) of the synthesized binary oxides have been identified as a reason for the increased reactivity. Even at low reaction temperatures of 50-70 °C catalysts maintained their high activity resulting in TOFs of up to 3000 h ^-1 and excellent selectivity with total heterogeneous behavior in catalysis.

Full text of DOI:10.1016/j.apcata.2012.07.025

Applied Catalysis A: General 443–444 (2012) 87–95
Contents lists available at SciVerse ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Iron-containing defect-rich mixed metal oxides for Friedel–Crafts alkylation
Benjamin Jägera, Antje Wermanna, Peter Scholza, Matthias Müllerb, Udo Reislöhnerc,
Achim Stollea, Bernd Ondruschkaa,∗
a Institute for Technical Chemistry and Environmental Chemistry, Friedrich-Schiller University Jena, Lessingstr. 12, D-07743 Jena, Germany
b Otto-Schott Institute, Friedrich-Schiller University Jena, Fraunhoferstr. 6, D-07743 Jena, Germany
c Institute for Solid State Physics, Friedrich-Schiller University Jena, Max-Wien-Platz 1, D-07743 Jena, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
Several stable iron-containing spinel oxides with the composition CuFe2O4 or FeCr2O4 were pre-
pared by different methods including thermal nitrate decomposition, combustion synthesis, and
mechano-chemical synthesis. Materials were compared to commercially available spinels produced by
conventional ceramic route with respect to their structural properties and catalytic activity. Phase com-
positions and microstructures of the catalysts were studied in detail. The catalysts were tested in the
Friedel–Crafts alkylation of various aromatic compounds with different alkyl halides as alkylating agents.
Microstructural properties (microstrains) of the synthesized binary oxides have been identified as a rea-
son for the increased reactivity. Even at low reaction temperatures of 50–70 ◦C catalysts maintained
their high activity resulting in TOFs of up to 3000 h−1 and excellent selectivity with total heterogeneous
behavior in catalysis.
Received 19 April 2012
Received in revised form 10 July 2012
Accepted 19 July 2012
Available online 28 July 2012
Keywords:
Combustion synthesis
Crystallite size
Friedel–Crafts alkylation
Liquid-phase catalysis
Mechano-chemical synthesis
Microstrains
© 2012 Elsevier B.V. All rights reserved.
Spinels
1. Introduction
short time, whereby high temperatures are generated by a redox
reaction from an oxidant and a fuel component. The fast pro-
The Friedel–Crafts alkylation is a well-known widely-used
organic reaction for C C bond formation between aromatics and
alkyl halides, alcohols, or esters. Bulk chemicals like ethylben-
zene from ethylene and benzene or pharmaceutical compounds
are produced using this methodology. The traditional reaction is
carried out in the liquid homogeneous phase using Lewis acids in
either (over)stoichiometric (e.g. AlCl3) or catalytic amounts (e.g.
FeCl3) [1]. Heterogeneously catalyzed Friedel–Crafts alkylations
are still an important field of research [2–7]. Among those cata-
lysts iron-containing compounds are favored [2,4,5,8–11]. Besides
mesoporous materials [2–4,11,12] and zeolites [13,14] the appli-
cation of mono- or mixed metal oxides has been described also
[5–7,9,15].
For the synthesis of oxidic materials like spinels several meth-
ods are known from the literature. Among these the ceramic route
as high temperature reaction between solids (>1000 ◦C) is the clas-
sic way. Thermal decompositions of metal precursors, i.e. salts like
nitrates, can be realized at lower temperatures but it still several
hours are required for the reaction. Furthermore, the combus-
tion synthesis allows the preparation of spinels [16–18] in very
gression and the evolution of large amounts of gases lead high
cooling rates resulting in materials with smaller crystallites and
increased number of defects [16]. Spinels with similar structural
characteristics [19,20] can be obtained following a mechano-
chemical route with the advantage that low bulk temperatures
(≈50 ◦C) are reached due to the milling process [21–23]. The struc-
tural factors should have an impact on the existence of Lewis
acidic sites and therefore they should influence the catalytic activ-
ity in the Friedel–Crafts alkylation. Additionally the Fe-content
can influence the acidity shown for materials with the formula
CuCr2−xFexO4. Lewis acidity increased along with a rise of the iron
content [9].
This work is focused on the comparison of the activity and struc-
tural properties of oxidic materials prepared by different synthetic
procedures: namely, ceramic route (CR), thermal nitrate decom-
position in a muffle furnace (MF), combustion synthesis (CS1),
scaled-up combustion synthesis (CS10), and mechano-chemical
synthesis (MS). The Friedel–Crafts alkylation is used as a model
reaction proving the formation of defects due to their action
as Lewis acidic sites. Especially from combustion synthesis and
mechano-chemical synthesis highly defective materials and there-
fore active catalysts are expected. Furthermore, it is anticipated that
the spinel phases can be synthesized in a faster reaction and/or at
lower reaction temperatures.
∗
Corresponding author. Tel.: +49 3641 948400; fax: +49 3641 948402.
E-mail address: Bernd.Ondruschka@uni-jena.de (B. Ondruschka).
0926-860X/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.apcata.2012.07.025

88
B. Jäger et al. / Applied Catalysis A: General 443–444 (2012) 87–95
2. Experimental
Particle size measurements were performed using a Coulter LS 230
laser diffraction particle size analyzer working in the range from
40 nm to 2 mm.
Reagents, solvents, and metal precursors were purchased from
commercial suppliers (Sigma–Aldrich, Merck) and were employed
without further purification. Weights of metal precursors used for
catalyst preparation were corrected by means of the purity of the
respective metal salts. Commercially available spinels (denoted as
CA) were purchased from Ferro GmbH and Sigma–Aldrich. After
their synthesis catalysts were ground to free-flowing powders prior
to application.
2.3. Catalytic tests
Reactions were carried out in batch mode in a round bot-
tom flask equipped with reflux condenser connected to a washing
flask for trapping the evolved HCl gas. CuFe2O4 was applied with
0.16 mol% (4 mg) referred to benzyl chloride (10.69 mmol) and
related to the formal stoichiometry of the spinel CuFe2O4. The aro-
matic compound o-xylene (106.9 mmol) was used as reagent and
solvent in a ratio 10:1 referred to the alkylating agent for sup-
pressing overalkylation. The CuFe2O4 catalyst was suspended in
o–xylene followed by application of benzyl chloride. The reaction
was carried out for 10 min at 80 ◦C using oil bath for heating.
The reusability of the catalysts was tested with FeCr2O4 CS10 in
the alkylation of o-xylene for both alkylating agents benzyl chlo-
ride and cyclohexyl chloride. The extreme small amount of catalyst
makes it impossible to carry out a recycling by filtration. There-
fore, the reaction system was cooled down, analyzed and reused
completely.
2.1. Catalyst preparation
2.1.1. Thermal decomposition (MF series)
Thermal nitrate decomposition was carried out with the metal
nitrates in molar ratio 1:2 for achieving MˇıMˇıˇı2O4 following a mod-
ified literature procedure [24]. The respective amounts of metal
salts for 1 g oxide material were diluted in distilled water (20 ml).
The solution was evaporated to dryness and afterwards heated in
a muffle furnace for 6 h at 300 ◦C and for additional 4 h at 800 ◦C
under static air in each step.
2.1.2. Combustion synthesis (CS series)
2.4. Product analysis
Glycine and metal nitrates were used as fuel and oxidant,
respectively. Glycine and nitrate were set to a 1:1 molar ratio.
The respective amount of metal salts for 1 g oxide material and
glycine were dissolved in distilled water (20 ml). Excess water was
removed by heating to 100 ◦C under atmospheric pressure for 5 h
until gel-formation. Afterwards, the mixture was heated to the
combustion temperature of 200–250 ◦C on a heat plate with max-
imum power (415 W). After the gas evolution stopped (ca. 10 s),
metal oxide foams with a low material density were achieved (CS1
series, cf. Supplementary Information video CS 1g).
The procedure was scaled up to 10 g oxide material as prod-
uct (CS10 series). The respective amounts of metal nitrates and
glycine were dissolved in distilled water (100 ml) and the excess
water was removed by heating to 100 ◦C under atmospheric pres-
sure for approximately 5–7 h until gel-formation. The ignition also
took place at 200–250 ◦C on a heat plate and the gas evolution was
finished after 20–30 s.
Analysis of the reaction mixtures was carried out using a GC-FID
(HP 6890) with n-decane as internal standard. Exact measure-
ment conditions are available in the Supplementary Information.
Calibrations were made by dilution of a stock solution of the alky-
lating agent covering the expected range of concentrations with
six points. The reaction products were identified by GC–MS Agi-
lent 6890N and in the case of cyclohexyl-dimethyl-benzenes by
NMR spectroscopy (selective NOESY, Bruker Avance) additionally
for identifying the isomers (cf. Supplementary Information).
3. Results and discussion
The combustion synthesis is much faster than thermal nitrate
decomposition. After gel formation the reaction is finished within
seconds (as shown in Supplementary Information Video), often
with bright glowing and under evolution of large amounts of gases.
Comparably, 10 h are required for classical thermal synthesis by
decomposition of the nitrate precursors. For the mechano-chemical
synthesis the longest reaction time of 16 h was applied, but it can
be carried out near room temperature without any need of external
heating. A calcination step is avoided thus, preventing the healing of
lattice defects [25]. In comparison with other “mechano-chemical”
routes the active phases are formed during the comminution of the
reagents and not until after additional thermal treatment.
2.1.3. Mechano-chemical synthesis (MS series)
The materials were synthesized in a ball mill “Pulverisette 7
classic line” (Fritsch GmbH) by using the respective oxides FeO
and Cr2O3 or CuO and Fe2O3 in stoichiometric amounts to receive
FeCr2O4 or CuFe2O4, respectively. Magnesia-stabilized zirconium
oxide was used as material for the grinding beaker (45 ml) and the
grinding balls (8 mm × 10 mm, ball to powder mass ratio of 9:1).
The reaction was carried out at 500 min−1 with 13 min milling time
followed by 2 min pause and in reverse mode. This program was
applied 64 times, resulting in a total reaction time of 16 h.
3.1. Catalyst characterization
2.2. Catalyst characterization
The TGA/DTA investigations of the catalysts (cf. Supplementary
Information Fig. S1) revealed stable oxide phases for all materi-
als without any relevant changes in the sample mass except for
CuFe2O4 MS. Residual CuO leads to a weight loss due to the forma-
tion of Cu2O during the heating process. In conclusion, the thermal
nitrate decomposition as well as the combustion synthesis leads
to complete conversion of the precursors and no metal nitrate
remained in the final catalysts.
X-ray powder diffraction (XRPD) was used for investigation of
the phase composition (cf. Supplementary Information Fig. S2 and
S3). Furthermore, diffractograms allowed the determination of the
crystallite sizes D (integral breadth) and microstrain ε (Table 1). The
details of the fitting procedure are described elsewhere [26]. For
CuFe2O4 metastable cubic spinel phase was detected for samples
Thermogravimetric analyses were carried out using a Shimadzu
DTG 60 with platinum crucible under flowing air atmosphere of
30 ml min−1, heating rate 10 K min−1 from room temperature up to
900 ◦C. X-ray powder diffractograms were obtained on a Siemens
D5000 diffractometer using Cu K␣ radiation. The microstructural
properties of the spinel phases were calculated by whole pattern fit-
ting with the software TOPAS (Bruker AXS GmbH) using the Pawley
method. The space groups and additional phases were determined
by searching the JCPDS database. BET surface was determined by
N2 sorption using Autosorb-1 (Quantachrome GmbH). Mößbauer
spectra were obtained at room temperature with a velocity-
scanning spectrometer using 57Co in a rhodium matrix as source.

B. Jäger et al. / Applied Catalysis A: General 443–444 (2012) 87–95
89
Table 1
Microstructural investigation, detected phases, BET surface and mean particle size of the catalysts.a
Sample
D (nm)
ε (4 × 103)
Space group
Detected phases
BET (m2 g−1
)
d50 (␮m)
13.7
214
70.1
120
CuFe2O4 CA
11.87 (45)
134 (67)
37.2 (25)
99 (27)
24.1 (46)
11.3 (10)
77.5 (30)
78.3 (50)
7.08 (26)
14.3 (20)
–
0.0(13)
129 (21)
159.7 (91)
97 (21)
278 (54)
282 (44)
74.5 (37)
92.5 (55)
181 (15)
310 (55)
–
Fd-3m
I41/amd
Fd-3m
Fd-3m
I41/amd
Fd-3m
R-3c
CuFe2O4
CuFe2O4; Fe2O3; CuO
CuFe2O4; CuO; Fe2O3; CuFeO2
CuFe2O4 cub.; CuFe2O4 tetr.; CuO; Fe2O3
29.51
0.93
6.42
6.32
CuFe2O4 MF
CuFe2O4 CS1
CuFe2O4 CS10 cubic
Tetragonal
CuFe2O4 MS
Fe(Fe,Cr)2O4 CA
FeCr2O4 MF
FeCr2O4 CS1
FeCr2O4 CS10
FeCr2O4 MS
CuFe2O4; Fe2O3; CuO
Cr1.3Fe0.7O3; Fe2O3; Cr2O3
Cr1.3Fe0.7O3; Fe2O3; Cr2O3
FeCr2O4
FeCr2O4; Cr2O3; (Fe2O3)
Cr2O3; Fe2O3; FeO
6.28
3.94
8.27
126.90
59.70
5.34
8.48
0.887
0.52
85.5
70
3.68
R-3c
Fd-3m
Fd-3m
R-3c
a
The values in parentheses are referred to the error of the given value. It indicates the error of the last two given decimals. For example 11.87 (45) represents 11.87 0.45.
Italicized values indicate results, where the calculated error exceeds the calculated value.
888
884
880
876
prepared by the CS1 and the MS method indicating a quench-
ing process during preparation [27]. Otherwise, the MF method
resulted in the formation of the tetragonal spinel phase. During up-
scaled combustion synthesis CS10 both the cubic and the tetragonal
phase of CuFe2O4 were formed. Analysis of mechano-chemically
synthesized CuFe2O4 MS not indicated the presence of a ZrO2 phase
resulting from abrasion processes of the milling tools. Only CuFe2O4
MS led to a spinel phase. Due to the extreme hardness of Cr2O3 no
spinel phase was formed during the mechano-chemical treatment
of FeO and Cr2O3 since no particle refinement took place, which
would form reactive species. The only observed reaction was oxi-
dation of FeO to Fe2O3. Diffractograms show that in most samples
additional phases are present and that formation of such phases
takes place in nearly all used synthetic procedures. Catalyst FeCr2O4
CS1 contains of a phase-pure cubic spinel (Fd-3m), while the MF
method as well as the CA led to a product exhibiting the trigonal
space group R-3c.
(a)
-8
-4
0
4
8
velocity [mm s-1]
(b)
In Table 1 the results of microstructural investigation, measured
BET surfaces, and the mean particle sizes d50 are listed (exam-
ple histogram for the particle size distribution cf. Supplementary
Information Fig. S4). The crystallite size D and the amount of
microstrain ε were calculated by whole pattern fitting of the diffrac-
tograms.
440
438
436
Fe-containing FeCr2O4 prepared by the thermal decomposition
(MF) and by combustion synthesis (CS) as well as the commercially
available material (CA) were subjected to Mößbauer spectroscopy
(Fig. 1) in order to determine the structure around the Fe centers.
This would allow affirmation to the structural determination from
XRPD studies and location of Fe in the spinel crystal structure.
The material FeCr2O4 CS10 (Fig. 1c) exhibits an isomer shift of
0.17 mm s−1 relative to the standard 57Fe/Rh and a quadruple split-
ting of 0.89 mm s−1 (Table 2). Due to the cubic structure of FeCr2O4
CS10 (Table 1) and the fact that it is a normal spinel iron should
be located on tetrahedral sites with a cubic symmetry. Therefore,
one single line is expected [28] and confirmed for FeCr2O4 in some
studies [29,30]. Regarding FeCr2O4 CS10 a quadruple splitting is
observed leading to the consequence of a distorted structure due
to local defects around the iron centers. Similar quadruple splitting
was found in an oxidized FeCr2O4 spinel containing Fe3+ leading to a
cation deficient structure [31] and underlining the results obtained
by XRPD (high ε, cf. Table 1). The Mößbauer spectra of FeCr2O4 CA
and MF (Fig. 1a and b) show a different structure. Both materials
exhibit a magnetic splitting nearly similar to the spectrum of Fe2O3
[32,33]. Regarding the XRPD results this confirms the finding of a
trigonal space group belonging to Cr1.3Fe0.7O3 which can be consid-
ered as Cr-substituted iron oxide. Last mentioned spectra consist
of two individual sextets confirming the presence of Cr1.3Fe0.7O3
and residual Fe2O3 as additional phase. The central peak pair near
0 mm s−1 for the less intensive sextet cannot be resolved and is
fitted as one broader peak due to its low intensity.
-8
-4
0
4
8
velocity [mm s-1]
(c)
146
145
144
143
-2
-1
0
1
2
velocity [mm s-1]
Fig. 1. Mößbauer spectra of FeCr2O4 CA (a), MF (b) and CS10 (c) at room temperature
fitted by Lorentzian profiles. (For interpretation of the references to color in this
figure legend, the reader is referred to the web version of the article.).

90
B. Jäger et al. / Applied Catalysis A: General 443–444 (2012) 87–95
Table 2
Results of the Mößbauer spectra fitting for FeCr2O4 CA, MF and CS10.
Sample
CA
Signal
Isomer shift ꢀ (mm s−1
)
Magnetic splittinga ı (mm s−1
)
Line width ꢁ (mm s−1
)
1st sextet
2nd sextet
1st sextet
2nd sextet
Doublet
0.310
0.135
0.358
0.141
0.169
16.648
15.721
14.867
13.557
0.879b
0.28
0.75
0.44
1.31
0.63
MF
CS10
a
Magnetic splitting is referred to the distance of the outer lines.
In this case only a quadruple splitting is observed.
b
3500
3000
2500
2000
1500
1000
500
Cl
+
cat, ΔT
- HCl
Cl
CuFe2O4 CA
CuFe2O4 MF
CuFe2O4 CS1
CuFe2O4 CS10
CuFe2O4 MS
FeCr2O4 CA
FeCr2O4 MF
FeCr2O4 CS1
FeCr2O4 CS10
FeCr2O4 MS
+
Scheme 1. Catalytic Friedel–Crafts alkylations used as model reactions.
3.2. Catalytic activity
The Friedel–Crafts alkylation of o-xylene (Scheme 1) as an acti-
vated aromatic compound and benzene as non-activated aromatic
compound as well as other arenes were used for investigation of the
effect of catalyst preparation on the catalytic properties. Addition-
ally, the reactivity in a reaction using cyclohexyl chloride, which
leads to a less stabilized cation/complex than benzyl chloride, was
tested.
0
50
60
70
80
90
100
110
A blank test with the most reactive system o-xylene and
benzyl chloride for 6 h at 140 ◦C resulted in a conversion of
benzyl chloride <1%. Results obtained for the Friedel–Crafts
alkylation of o-xylene with benzyl chloride are summarized in
Table 3. The selectivity S refers to the main product 1,2-dimethyl-
4-(phenylmethyl)benzene. Alkylation in the ortho-position to
1,2-dimethyl-3-(phenylmethyl)benzene took place to a negligi-
ble extent only. Furthermore, consecutive alkylation products are
detected.
Reaction temperature [ºC]
Fig. 2. Comparison of turn-over frequencies (TOFs) in dependency of the required
reaction temperature for the alkylation of o-xylene with benzyl chloride using the
catalysts of the different synthetic routes (reaction conditions: 106.9 mmol o-xylene,
10.69 mmol benzyl chloride, 0.16 mol% catalyst, 10 min).
chloride are summarized. The further difference in selectivity to
the main product corresponds to a consecutive alkylation of cyclo-
hexyldimethylbenzene to dialkylated products.
The reactions proceeded very fast and with the more active
catalysts they are completed within 10 min. The required reac-
tion temperature correlates with the reactivity of the catalysts
allowing working at lower temperatures for more active catalysts.
Differences in the reactivity are visible, even if the same mate-
rial is used resulting from different synthetic routes. Regarding
CuFe2O4 the differences are less pronounced than for FeCr2O4.
Using CuFe2O4 the reaction takes place in the range from 50–80 ◦C,
whereas with prepared FeCr2O4 and the commercial product on-set
temperatures of 60 and 110 ◦C are required to see any conversion,
respectively. Comparison of the reached TOFs depending on the
necessary reaction temperature (Fig. 2) shows the differences in the
activity of the materials obtained by the different synthetic routes
(TOFs were calculated by dividing the amount of main product by
the amount of catalyst and the reaction time [34]). Obviously CA-
series catalysts need the highest temperature of 110 ◦C for reaching
comparable TOFs. With the MF or CS1 catalysts high TOFs can be
achieved working at lower temperatures of 70–80 ◦C. When using
CS10 catalysts TOFs of 500–1000 h−1 are reached at 60 ◦C. Highest
activity is detected for catalysts from the MS series (e.g. 2000 h−1
at 50 ◦C).
In the alkylating reaction with cyclohexyl chloride only FeCr2O4
CS catalysts show activity in which CS10 (78% conversion at 140 ◦C
after 60 min) is more active than CS1 (50% conversion). Similar to
the other reaction the FeCr2O4 MS catalyst is less reactive than
CS1 or CS10 leading to 49% conversion at 140 ◦C and 120 min.
With CuFe2O4 catalysts no alkylation reaction can be observed at
increased reaction temperatures and times. Just a small amount
of cyclohexyl chloride is converted. Instead a side reaction leading
to oxidation of one methyl group of o-xylene in <1% takes place
(identified with GC–MS). The resulting cation is less stabilized and
therefore more reactive than the cation formed by reaction of ben-
zyl chloride. This leads to the formation of a higher amount of the
ortho-product compared to reaction of benzyl chloride. The amount
of alkylation in the ortho-position is nearly independent from the
achieved conversion or the used reaction temperature and seems
to be characteristic for the alkylating agent cyclohexyl chloride in
combination with the catalysts tested.
The lower reactivity of cyclohexyl chloride can be checked
by using benzyl chloride and cyclohexyl chloride as alkylating
agents in a parallel reaction. The results are summarized in
Fig. 3. At the used reaction conditions benzyl chloride reaches
total conversion whereas the conversion of cyclohexyl chlo-
ride is incomplete (Fig. 3a). The product distributions (Fig. 3b)
confirm that benzyl chloride is the more active compound
The alkylation of o-xylene with cyclohexyl chloride yields two
isomers in relevant amounts. Due to lower stability of the cyclo-
hexyl cation it is more reactive and attacks both available positions
in o-xylene. In Table 4 the results of the alkylation with cyclohexyl

B. Jäger et al. / Applied Catalysis A: General 443–444 (2012) 87–95
91
Table 3
Catalytic activity at the specified reaction conditions for the Friedel–Crafts alkylation of o-xylene with benzyl chloride.a
Catalyst
T (◦C)
X (%)
S (%)
TOF (h−1
)
p/ob
CuFe2O4 CA
CuFe2O4 CA
CuFe2O4 MF
CuFe2O4 MF
CuFe2O4 CS1
CuFe2O4 CS1
CuFe2O4 CS10
CuFe2O4 CS10
CuFe2O4 MS
80
70
70
60
80
70
80
70
60
50
110
100
90
70
70
70
60
60
70
70
60
60
60
>99
0
>99
0
>99
0
>99
11
>99
58
>99
15
0
88
–
88
–
88
–
3300
–
3300
–
3300
–
>99:1
–
>99:1
–
>99:1
–
87
94
88
92
87
>99
–
>99
87
92
92
97
88
96
96
>99
>99
3263
388
3300
2001
3263
563
–
188
1631
1829
983
655
3201
1872
900
600
38
>99:1
>99:1
>99:1
>99:1
>99:1
>99:1
–
>99:1
>99:1
>99:1
>99:1
>99:1
>99:1
>99:1
>99:1
>99:1
>99:1
CuFe2O4 MS
Fe(Fe,Cr)2O4 CA
Fe(Fe,Cr)2O4 CA
FeCr2O4 MF
FeCr2O4 CS1
FeCr2O4 CS10c
FeCr2O4 CS10c,d
FeCr2O4 CS10c
FeCr2O4 CS10c,d
FeCr2O4 CS10
FeCr2O4 CS10d
FeCr2O4 CS10
FeCr2O4 CS10d
FeCr2O4 MS
5
>99
53
57
18
97
26
25
8
1
a
Reaction conditions: 106.9 mmol o-xylene, 10.69 mmol benzyl chloride, 0.16 mol% catalyst, 10 min.
Ratio of 1,2-dimethyl-4-(phenylmethyl)benzene to 1,2-dimethyl-3-(phenylmethyl)benzene.
mcat was increased to 0.32 mol%.
b
c
d
Reaction time was decreased to 5 min.
Table 4
Catalytic activity at the specified reaction conditions for the Friedel–Crafts alkylation of o-xylene with cyclohexyl chloride.a
Catalyst
T (◦C)
t (min)
X (%)
S (%)
TOF (h−1
)
p/ob
CuFe2O4 CS1
CuFe2O4 CS10
CuFe2O4 MS
FeCr2O4 CS1
FeCr2O4 CS10
FeCr2O4 CS10
FeCr2O4 CS10
FeCr2O4 CS10
FeCr2O4 CS10
FeCr2O4 CS10
FeCr2O4 CS10
FeCr2O4 CS10
FeCr2O4 MS
140
140
140
140
140
140
140
140
120
120
120
110
140
120
120
120
60
120
90
60
10
60
30
1
3
1
–
–
–
–
–
–
–
–
–
50
95
86
78
31
32
15
0
95b
91
85
90c
297
270
305
439
1163
196
171
–
79:21
80:20
78:22
79:21
78:22
79:21
78:22
–
>99
98c
91d
–
10
60
120
0
49
–
–
132
–
86d
78:22
a
Reaction conditions: 106.9 mmol o-xylene, 10.69 mmol cyclohexyl chloride and 0.16 mol% catalyst.
Ratio of 4-cyclohexyl-1,2-dimethylbenzene to 3-cyclohexyl-1,2-dimethylbenzene.
Side product cyclohexene from elimination reaction of cyclohexyl chloride detected (<2%).
Cyclohexene is the only detectable side product.
b
c
d
and therefore at first 1,2-dimethyl-4-(phenylmethyl)benzene is
formed followed by the alkylation of further o-xylene with
cyclohexyl chloride forming 4-cyclohexyl-1,2-dimethylbenzene
and 3-cyclohexyl-1,2-dimethylbenzene. The overall selectivity for
monoalkylated products and the ratio of isomers formed by alky-
lation with cyclohexyl chloride is comparable to the experiments
with just one alkylating agent. Therefore the two reaction systems
do not seem to interact with each other.
In Table 5 the alkylation results of other aromatic compounds
(Scheme 2) with benzyl chloride are summarized. Due to the
fact that benzene is a non-activated aromatic compound, higher
temperatures and longer reaction times are needed compared
to the alkylation of o-xylene and other arenes. Additionally, the
alkylation of toluene was carried out at higher temperatures but
medium reaction time of 30 min is possible. Activated arenes
like o-xylene, p-xylene, and mesitylene can be alkylated at lower
temperatures and short reaction times of 10 min. The substrate
durene needs higher temperatures for melting allowing a homo-
geneous reaction solution. The selectivities in Table 5 referred to
benzene
mesitylene
toluene
p-xylene
durene
Scheme 2. Other aromatic compounds used in the Friedel–Crafts alkylation with
benzyl chloride.
the respective monoalkylated products. Difference to 100% selec-
tivity is attributed to consecutive alkylations affording dialkylated
products and in small amounts intramolecular alkylation of benzyl
chloride.
For all substrates the alkylation with benzyl chloride can be car-
ried out successfully. The selectivity to monoalkylated products is
in the same range as observed for the reaction with o-xylene. Only
if benzene is converted the selectivity is significantly lower due

92
B. Jäger et al. / Applied Catalysis A: General 443–444 (2012) 87–95
Fig. 3. Catalytic activity at specified reaction conditions for the Friedel–Crafts alkylation of o-xylene with benzyl chloride and cyclohexyl chloride. Conversion of the alkylating
agents (a), selectivity to the specified products (b) (reaction conditions: 106.9 mmol o-xylene, 10.69 mmol benzyl chloride and cyclohexyl chloride, 0.16 mol% FeCr2O4 CS10,
140 ◦C).
Table 5
Catalytic activity at the specified reaction conditions for the Friedel–Crafts alkylation of other aromatic compounds with benzyl chloride.a
Catalyst
mcat (mol%)
Educt
T (◦C)
t (min)
X (%)
S (%)
TOF (h−1
)
FeCr2O4 CS10
FeCr2O4 CS10
FeCr2O4 MS
2.50
1.25
1.25
0.63
0.63
0.01
0.16
0.16
0.16
0.16
0.16
0.16
0.16
0.16
Benzene
Benzene
Benzene
Benzene
Benzene
Benzene
Toluene
o-Xylene
o-Xylene
p-Xylene
p-Xylene
Mesitylene
Mesitylene
Durene
80
80
80
80
80
80
90
70
60
70
60
70
60
100
120
120
120
120
120
120
30
10
10
10
10
>99
78
>99
90
>99
>99
99
97
25
97
34
75
77
74
75
74
73
90
88
96
87
97
95
15
24
30
54
59
3673
1106
3201
900
3174
1239
3437
416
3461
FeCr2O4 MS
CuFe2O4 MS
CuFe2O4 MS
FeCr2O4 CS10
FeCr2O4 CS10
FeCr2O4 CS10
FeCr2O4 CS10
FeCr2O4 CS10
FeCr2O4 CS10
FeCr2O4 CS10
FeCr2O4 CS10
10
10
10
96
11
>99
>99
92
a
Reaction conditions: 106.9 mmol arene, 10.69 mmol benzyl chloride.
to the effect that Friedel–Crafts alkylation yields a product with
higher activity toward additional alkylation compared with ben-
zene. Lower activity is manifested in the calculated TOFs also. Just
with CuFe2O4 MS a TOF of >3600 h−1 could be realized with the
unsubstituted arene.
hot-filtration step, which indicates a rather homogeneous catalytic
reaction. With CuO and Cr2O3 nearly no reaction takes place. This
fact allows the assumption that Fe in the spinel is the active element
and neither Cu nor Cr contributes to catalysis under the reaction
conditions investigated. But this arrangement allows the formation
of stable heterogeneous catalysts in contrast to Fe2O3.
In Fig. 5 the conversion-selectivity plot for the alkylation of o-
xylene with benzyl chloride is presented. It can be seen that the
selectivity depends on the conversion and is nearly independent
from the used catalyst. In contrast the conversion reveals a depen-
dency on the used catalyst in combination with the applied reaction
temperature. At low conversion the selectivity reaches maximal
values of >99%. But also at full conversion the selectivity does not
drop below 87% (e.g. catalysts FeCr2O4 CS1, CuFe2O4 MS, CuFe2O4
CS10).
For investigation of important parameters on the conversion
of benzyl chloride, reaction time, reaction temperature, and cata-
lyst amount were varied using 23 factorial plan. The response data
used for statistical evaluation are summarized in Table 6. Results
4. Discussion
Besides the pure activity of the catalysts, the heterogeneity of
the reaction is a very important factor. Therefore, heterogeneity
tests were carried out with the model reaction system o-xylene
and benzyl chloride by stopping the reaction after 10 min and sep-
aration of the catalyst by hot filtration. Afterwards the reaction
mixture was allowed to react for further 20 min. As shown in Fig. 4,
no further reaction takes place in case of FeCr2O4 CS10. There-
fore, the reaction can be considered as completely heterogeneous
without any metal traces present in solution after removal of the
catalyst. If Fe2O3 was used, the reaction rate is not affected by the

B. Jäger et al. / Applied Catalysis A: General 443–444 (2012) 87–95
93
Table 6
Used data for statistical analysis.a
Catalyst
mcat (mol%)
T (◦C)
t (min)
X (%)
S (%)
FeCr2O4 CS10
FeCr2O4 CS10
FeCr2O4 CS10
FeCr2O4 CS10
FeCr2O4 CS10
FeCr2O4 CS10
FeCr2O4 CS10
FeCr2O4 CS10
0.16
0.16
0.16
0.16
0.32
0.32
0.32
0.32
60
60
70
70
60
60
70
70
5
10
5
10
5
10
5
10
8
34
26
97
18
57
53
>99
>99
96
>99
88
>99
92
92
87
a
Reaction conditions: 106.9 mmol o-xylene, 10.69 mmol benzyl chloride.
100%
90%
80%
70%
60%
50%
40%
FeCr2O4 CS10
Fe2O3
CuO
30%
Cr2O3
20%
10%
0%
FeCr2O4 CS1
FeCr2O4 CS10
FeCr2O4 MS
0%
20%
40%
60%
80%
100%
-10
-5
0
5
10
15
20
Conversion
Reactiontime [min]
Fig. 6. Conversion-selectivity plots for the alkylation of o-xylene with cyclohexyl
chloride for two FeCr2O4 catalysts (cf. Table 4) (reaction conditions: 106.9 mmol
o-xylene, 10.69 mmol cyclohexyl chloride, 0.16 mol% catalyst).
Fig. 4. Heterogeneity tests for FeCr2O4 CS10, Fe2O3, CuO, and Cr2O3. Negative reac-
tion times indicate reaction in presence of catalysts, which have been removed at
t = 0 min (reaction conditions: 106.9 mmol o-xylene, 10.69 mmol benzyl chloride,
0.16 mol% catalyst, 60 ◦C).
indicated that the reaction time has the major influence on the
reaction (52%) followed by the reaction temperature (37%). Inter-
estingly, variation of the catalyst amount contributes only to an
extent of 8% as already demonstrated by using of very small cata-
lyst amounts for a successful reaction. Statistical analysis revealed
that parameter interactions are not significant.
100%
90%
Fig. 6 shows the conversion-selectivity plot for the alkylation of
o-xylene with cyclohexyl chloride. Also in this case the selectivity
is quite stable (85–99%) over the complete range of conversion. But
the selectivity does not reach its maximum at low conversions due
to a side reaction of the alkylating agent, namely the elimination of
HCl leading to cyclohexene.
80%
CuFe2O4 CA
CuFe2O4 MF
CuFe2O4 CS1
70%
In Table 7 a comparison of the used catalysts with data avail-
able in the literature is done. Also with benzene as educt, the TOF
for FeCr2O4 CS10 is higher than the material used by Ghorpade
et al. [9]. In the alkylation of o-xylene the TOFs are much higher
than the values found in the literature. The TOF for FeCr2O4 CS10
reaches 3200 h−1 at 70 ◦C or 983 h−1 at a lower reaction temper-
ature of 60 ◦C. The particle sizes, crystallite sizes, or BET surfaces
are not able to explain the high activity of the CS or MS cata-
lysts only. Therefore, the amount of microstrains seems to be the
determinant. The most active catalyst FeCr2O4 CS10 contains the
highest amount of microstrains. Within the studied catalysts of the
CuFe2O4 type the presence of the tetragonal phase leads to more
active catalysts compared to cubic single phase CuFe2O4 resulting
in the best performance when using the pure tetragonal phase cat-
alyst CuFe2O4 MF. Regarding the cubic phase CuFe2O4 catalysts this
CuFe2O4 CS10
CuFe2O4 MS
FeCr2O4 CA
60%
FeCr2O4 CS1
FeCr2O4 CS10
50%
0%
20%
40%
Conversion
60%
80%
100%
Fig. 5. Conversion-selectivity plots for the alkylation of o-xylene with benzyl chlo-
ride for different catalysts (cf. Table 3) (reaction conditions: 106.9 mmol o-xylene,
10.69 mmol benzyl chloride, 0.16 mol% catalyst, reaction time 5 or 10 min).

94
B. Jäger et al. / Applied Catalysis A: General 443–444 (2012) 87–95
Selectivity
TON
Selectivity
TON
(a)
100%
(b)
100%
250
200
150
100
50
600
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
90%
80%
500
70%
400
60%
50%
300
40%
200
100
0
30%
20%
10%
0%
0
1
2
3
4
5
1
2
3
4
5
Cycle
Cycle
Fig. 7. Recycling tests for FeCr2O4 CS10 used in the alkylation of o-xylene with benzyl chloride (a) and cyclohexyl chloride (b) (reaction conditions: 106.9 mmol o-xylene,
10.69 mmol alkylating agent, 0.16 mol% catalyst at 70 ◦C and 10 min (a) and at 120 ◦C and 60 min (b)). For interpretation of the references to color in this figure legend, the
reader is referred to the web version of the article).
Table 7
drop due to enrichment of the alkylated products which are more
prone to consecutive alkylation.
Comparison of calculated TOFs for some heterogeneously catalyzed Friedel–Crafts
alkylations with benzyl chloride (cf. Tables 3–5).
Catalyst
Educt
TOF (h−1
)
Reference
5. Conclusion
CuFeCrO4
Benzene
Benzene
Benzene
Benzene
Benzene
Toluene
6 (80 ◦C)
24 (80 ◦C)
[9]
FeCr2O4 CS10
FeCr2O4 MS
CuFe2O4 MS
Fe-JLU-15 (70)
ZnFe2O4
FeCr2O4 CS10
ZnTiO3
FAPO4
FeCr2O4 CS10
FeCr2O4 CS10
Fe/Al-MCM-41
FeCr2O4 CS10
FeCr2O4 CS10
CuFe2O4 MF
CuFe2O4 MS
CuFe2O4 MS
FeCr2O4 CS10
FeCr2O4 CS10
FeCr2O4 CS10
This work
This work
This work
[4]
[15]
This work
[7]
Several very active iron containing binary oxides for the
Friedel–Crafts alkylation have been synthesized by different syn-
thetic routes. Differences in the catalytic activity of the materials
obtained from the respective synthetic procedures were found.
The amount of microstrains as a measure of the defect con-
centration is identified as the major characteristic parameter of
the catalyst. Neither the crystallite or particle size nor the BET
surfaces are responsible for the observed differences in the cat-
alytic activity. Therefore, both the CS (combustion synthesis)
and MS (mechano-chemical synthesis) methods are able to pro-
duce structures with significant increased numbers of defects
than other synthetic routes. Thus, those procedures yield more
active catalysts for the Friedel–Crafts alkylation. The combustion
synthesized materials exhibited higher activity than the com-
mercially available oxides, wherein the CS10 catalysts obtained
from the scaled-up combustion synthesis revealed higher activity
than the CS1-type. With respect to the achieved TOFs two rows
of activity were found depending on the reaction temperature.
For CuFe2O4 the activity behaves like CA ≈ CS1 < CS10 < MF < MS
and for FeCr2O4 CA ≈ MF < CS1 ≤ MS < CS10. The mechano-chemical
synthesis, without heating the materials to the usually needed
high temperatures for calcination or phase formation, resulted in
active catalysts also. Therefore, the mechano-chemical activation
in the synthesis of CuFe2O4 MS leads to the phase formation of
a spinel just by grinding the individual oxides. The TOFs for the
Friedel–Crafts alkylation of o-xylene are much higher than the val-
ues found in literature even at lower temperature (50–70 ◦C). Also
less reactive aromatic compounds or sterical hindered arenes can
be alkylated successfully with comparably high TOFs at mild reac-
tion conditions and less reactive alkylating agents like cyclohexyl
chloride. Employing cyclohexyl chloride a higher tendency for the
formation of both possible isomers is observed. Also with this
54 (80 ◦C)
3673 (80 ◦C)
492 (70 ◦C)
289 (70 ◦C)
1106 (90 ◦C)
<1 (70 ◦C)
Toluene
p-Xylene
p-Xylene
p-Xylene
p-Xylene
o-Xylene
o-Xylene
o-Xylene
o-Xylene
o-Xylene
o-Xylene
Mesitylene
Mesitylene
Durene
3 (80 ◦C)
[10]
1239 (60 ◦C)
3174 (70 ◦C)
10 (140 ◦C)
983 (60 ◦C)
3201 (70 ◦C)
3300 (70 ◦C)
3300 (60 ◦C)
2001 (50 ◦C)
416 (60 ◦C)
3467 (70 ◦C)
3461 (100 ◦C)
This work
This work
[11]
This work
This work
This work
This work
This work
This work
This work
This work
result is exceeded by the application of CuFe2O4 MS containing only
the cubic spinel phase but with the highest amount of microstrains
in the whole row of the respective materials. Also at a reaction
temperature of 50 ◦C the TOF still reaches 2001 h−1
.
The reusability of the catalysts was tested with FeCr2O4 CS10 in
the alkylation of o-xylene for both alkylating agents benzyl chloride
and cyclohexyl chloride (Fig. 7). While the TON (turnover number,
calculated as mole converted alkylating agent per mole catalyst)
for benzyl chloride (Fig. 7a) keeps quite stable (TON = 625–619) the
TON for cyclohexyl chloride (Fig. 7b) decreases, especially between
the third and fourth reaction cycle. This indicates a significant deac-
tivation of the catalyst when using cyclohexyl chloride as alkylating
agent. The selectivities to the monoalkylated products in both cases

B. Jäger et al. / Applied Catalysis A: General 443–444 (2012) 87–95
95
reaction system TOFs > 1000 h−1 can be achieved. Overall, the mate-
rial CuFe2O4 MS is the best catalyst for achieving high TOFs in the
alkylation with benzyl chloride at low temperatures. FeCr2O4 CS10
shows the best performance with a broad range of educts and is also
able to convert cyclohexyl chloride whereby the CuFe2O4 catalysts
show hardly any conversion.
[6] C. Tagusagawa, A. Takagaki, A. Iguchi, K. Takanabe, J.N. Kondo, K. Ebitani, T.
Tatsumi, K. Domen, Catal. Today 164 (2011) 358–363.
[7] N. Pal, M. Paul, A. Bhaumik, Appl. Catal. A 393 (2011) 153–160.
[8] F. Adam, A.E. Ahmed, Chem. Eng. J. 145 (2008) 328–334.
[9] S.P. Ghorpade, V.S. Darshane, S.G. Dixit, Appl. Catal. A 166 (1998) 135–142.
[10] H. Hentit, K. Bachari, M.S. Ouali, M. Womes, B. Benaichouba, J.C. Jumas, J. Mol.
Catal. A: Chem. 275 (2007) 158–166.
[11] M.E. Preethi, S. Revathi, T. Sivakumar, E-J. Chem. 5 (2008) 467–472.
[12] K. Bachari, O. Cherifi, J. Mol. Catal. A: Chem. 253 (2006) 187–191.
[13] D. Mravec, M. Michvocík, M. Hronec, P. Moreau, A. Finiels, P. Geneste, Catal.
Lett. 38 (1996) 267–270.
Acknowledgments
[14] P.G. Smirniotis, E. Ruckenstein, Ind. Eng. Chem. Res. 34 (1995) 1517–1528.
[15] M.M. Shinde, M.R. Sawant, J. Chin. Chem. Soc. 50 (2003) 1221–1226.
[16] U. Zavyalova, B. Nigrovski, K. Pollok, F. Langenhorst, B. Müller, P. Scholz, B.
Ondruschka, Appl. Catal. B 83 (2008) 221–228.
[17] L.H. Ai, J. Jiang, Powder Technol. 195 (2009) 11–14.
[18] N.M. Deraz, J. Anal. Appl. Pyrol. 88 (2010) 103–109.
[19] T. Verdier, V. Nivoix, M. Jean, B. Hannoyer, J. Mater. Sci. 39 (2004) 5151–5154.
[20] E. Manova, D. Paneva, B. Kunev, C. Estournès, E. Rivière, K. Tenchev, A. Léaustic,
I. Mitov, J. Alloys Compd. 485 (2009) 356–361.
For carrying out the BET measurements and NMR spectroscopy
the support of the Institute for Physical Chemistry (A. Schmidt)
and the Institute for Inorganic and Analytical Chemistry (Dr. M.
Friedrich), respectively of the Friedrich-Schiller University Jena are
gratefully acknowledged.
Appendix A. Supplementary data
[21] G. García-Pacheco, J.G. Caban˜as-Moreno, H. Yee-Madeira, F. Cruz-Gandarilla,
Nanotechnology 17 (2006) 2528.
[22] S. Bid, S.K. Pradhan, Mater. Chem. Phys. 82 (2003) 27–37.
[23] V.G. Harris, D.J. Fatemi, J.O. Cross, E.E. Carpenter, V.M. Browning, J.P. Kirkland,
A. Mohan, G.J. Long, J. Appl. Phys. 94 (2003) 496–501.
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.apcata.2012.07.025.
[24] W.M. Shaheen, A.A. Ali, Mater. Res. Bull. 36 (2001) 1703–1716.
[25] M. Casas-Cabanas, G. Binotto, D. Larcher, A. Lecup, V. Giordani, J.M. Tarascon,
Chem. Mater. 21 (2009) 1939–1947.
References
[26] B. Jäger, A. Stolle, P. Scholz, M. Müller, B. Ondruschka, Appl. Catal. A 403 (2011)
152–160.
[27] J.E. Tasca, C.E. Quincoces, A. Lavat, A.M. Alvarez, M.G. González, Ceram. Int. 37
(2011) 803–812.
[28] N.N. Greenwood, Angew. Chem. 83 (1971) 746–755.
[29] H.B. Mathur, A.P.B. Sinha, C.M. Yagnik, Solid State Commun. 3 (1965) 401–403.
[30] M.J. Rossiter, J. Phys. Chem. Solids 26 (1965) 775–779.
[31] E. Schmidbauer, Phys. Chem. Miner. 14 (1987) 533–541.
[32] Y.R. Uhm, W.W. Kim, S.J. Kim, C.S. Kim, C.K. Rhee, J. Appl. Phys. 93 (2003)
7196–7198.
[1] I. Iovel, K. Mertins, J. Kischel, A. Zapf, M. Beller, Angew. Chem. Int. Ed. 44 (2005)
3913–3917.
[2] C. Anand, P. Srinivasu, S. Alam, V.V. Balasubramanian, D.P. Sawant, M.
Palanichamy, V. Murugesan, A. Vinu, Microporous Mesoporous Mater. 111
(2008) 72–79.
[3] C. Anand, B. Sathyaseelan, L. Samie, A. Beitollahi, R.P. Kumar, M. Palanichamy,
V. Murugesan, E.R. Kenawy, S.S. Al-Deyab, A. Vinu, Microporous Mesoporous
Mater. 134 (2010) 87–92.
[4] Y. Du, S. Liu, Y. Ji, Y. Zhang, F. Liu, Q. Gao, F.S. Xiao, Catal. Today 131 (2008)
70–75.
[33] M. Ristic, S. Music, M. Godec, J. Alloys Compd. 417 (2006) 292–299.
[34] T. Choudhury, N.M. Misra, Bull. Mater. Sci. 34 (2011) 1273–1279.
[5] V. Rajpara, S. Banerjee, G. Sereda, Synthesis (2010) 2835–2840.