E.V. Starokon et al. / Molecular Catalysis 443 (2017) 43–51
45
reflection scale to the Kubelka-Munk scale, F(R) = (1–R)2/2R, and
normalized to the intensity of absorption band 2260 cm−1 (vibra-
tions of the zeolite Si-O-Si fragment).
tion to ethylene, which proceeds on the surface of Na-FeZSM-5
zeolite, was observed in our earlier studies [48,52].
The extract contains also 8.6% of methoxypropanols (1-
formed most likely as a result of the interaction of the produced
PO with methanol:
2.4. Product extraction and analysis
In this work, we used an extraction procedure similar to that
described for the case of ethylene oxidation by ␣-oxygen [48]. After
termination of the reaction and evacuation of alkene as described
above, the sample was rapidly poured from the reactor to a vial with
the extractant having a temperature of 0 ◦C. The vial was closed
and shaken for 1 min and then placed in a centrifuge for 30 s to
separate the catalyst. For convenience of chromatographic analysis,
aqueous methanol (10% H2O) was used as the extractant in the case
of propylene, and aqueous acetonitrile (10% H2O) in the case of 1-
butene. Special experiments showed that the solvent nature exerts
only a weak effect on the composition of extracted products (see
Table S1 in Supplementary data).
(3)
It seems interesting that the product of hydration of PO, 1,2-
propanediol, is absent in the extract. The extract contains also a
minor (∼4%) amount of non-identified products, which are denoted
as “Others”. They are represented by a set of low-intensity chro-
matographic peaks which are difficult to identify. The total yield
of extracted products referred to the initial amount of ␣-oxygen
(a mole of product per a mole of O␣) is 51%. The rest 49% include
compounds that cannot be extracted from the zeolite. Their origin
will be considered later.
Several techniques were used to identify the extracted products.
The gas chromatographic (GC) analysis was performed on a Kristall
2000 M instrument equipped with an automatic liquid sampler and
a capillary column HP Plot U designed for low-boiling oxygen-
containing organic substances. The gas chromatographic–mass
spectrometric (GC–MS) analysis was carried out using Agilent Tech-
nologies 7000 GC/MS Triple Quad and GC System 7890A equipped
with a HP-5MS column. The 1H spectra of nuclear magnetic reso-
nance (NMR) were recorded at 400.13 MHz on a Bruker Avance-400
NMR spectrometer.
3.2. The effect of Na content in the zeolite
To reveal the effect of sodium modification of FeZSM-5 zeolite
on the composition of products, a series of experiments was per-
formed with the samples having different Na content. One can see
that raising the Na content of the zeolite from 0.2 to 0.4% leads to
only a small increase in the selectivity for PO, up to 64% (Table 1,
exp. 2) This occurs due to some decrease in the content of both
methoxypropanols and products with the composition C6H12O in
the extract.
In the case of FeZSM-5 sample not modified with sodium, the
PO fraction in the products was substantially smaller (31.3%) and
the total yield of the oxidation products strongly decreased, to 24%.
The fraction of methoxypropanols in the extract increased to 33.3%,
Analysis results were used to calculate the composition (selec-
tivity) of extracted products and their total yield with respect to
␣-oxygen:
Amount of products extracted(ꢀmol/g)
Yield =
Amount of␣-oxygen deposited(ꢀmol/g)
whereas the fraction of products with the composition C6H12
O
remained virtually unchanged. The AA content increased several-
fold, to 4.0%. Thus, the presence of Na ions results in a significant
suppression of side reactions. Most likely, this is due to a decrease
in the number of Brönsted acid sites after addition of sodium salt.
All the next experiments were carried out with the 0.2Na-FeZSM-5
The quantum chemical computations were carried out using
Density Functional Theory (DFT) implemented in GAUSSIAN 09
software [49]. In the present study, a M06-2X functional and
6–311g(df,p) basis set was used [50]. This method was employed in
work [51] to study the CH4 + O␣ reaction mechanism. The FeII(OH)2
structure, which was earlier shown to possess all typical properties
of ␣-site [51], was chosen as a model. The calculation procedure and
the electronic structure of ␣-sites were considered in detail in [51].
3.3. Effect of temperature on the C3H6 + O˛ reaction
was performed where the reaction temperature was varied from
−60 to 25 ◦C while the other conditions (temperature of sample
evacuation, sample before extraction, and extractant) remained
constant (Table 2).
3. Results and discussion
It is seen that elevation of the reaction temperature from −60
to −25 ◦C does not affect, within the measurement accuracy, the
amount and composition of extracted products. A further raising of
the temperature to 0 and 25 ◦C produces significant changes in the
composition of products. According to expectations, the content
of PO decreases, and that of PA increases. When the temperature
reaches 25 ◦C, PA becomes one of the main products of the reac-
tion. Most likely PA is formed due to the secondary reaction of PO
isomerization:
Our first experiments on the oxidation of propylene by ␣-oxygen
were carried out with the 0.2Na-FeZSM-5 sample, which earlier
showed a high selectivity toward ethylene epoxidation [48]. It is
seen (Table 1, exp. 1) that at −60 ◦C the main product of propylene
oxidation extracted from the surface is PO (59.4%). PO was iden-
tified by gas chromatography, 1H NMR (see Fig. S1) and GC–MS,
particularly with the use of labeled 18O␣ (see Fig. S2a-S2c). Allyl
alcohol was also detected in the extract, but its amount was very
low, 0.6% of the total amount of extracted products.
In addition to PO, the products having the composition C6H12O,
were detected: DMTHF, carbonyl compounds (2-methylpentanal
and 2-hexanone), and unsaturated alcohols (2,3-dimethyl-3-
buten-1-ol, 4-methyl-4-penten-2-ol and 5-hexen-2-ol). These
products are results of the secondary reaction of PO addition to
propylene (see Scheme S1). Similar reaction of ethylene oxide addi-
Noteworthy is the fact that at low temperatures (−60 ÷ −25 ◦C)
such isomerization product of PO as acetone is hardly detectable
(∼0.1%), while at 0 and 25 ◦C its content does not exceed 0.6%. In
the study of PO isomerization on zeolite catalysts [53], it was con-
cluded that the isomerization reaction pathway depends on the