ISOPRENE SYNTHESIS FROM FORMALDEHYDE AND ISOBUTYLENE
943
of Al–BEA, Zr–BEA, Sn–BEA, and Nb–BEA zeo- helium stream for 1 h, and then cooled to 60°C.
lite catalysts synthesized by isomorphous substitution Ammonia saturation was implemented in a stream of a
methods was studied earlier [13]; it was found that the dried NH3/N2 (1 : 10) mixture for 15 min. Physically
most effective isoprene synthesis catalysts are Al- and
Nb-containing BEA zeolites. The isoprene productiv-
ity increased in the following order: Zr–BEA < Sn–
BEA < Nb–BEA < Al–BEA, which correlated with
the content of Brønsted acid sites in the samples,
whereas the formation of the major byproduct—car-
bon monoxide resulting from formaldehyde decom-
position—increases with an increase in the number of
Lewis acid sites. However, some authors have other
opinions on the nature of active sites responsible for
the formation of isoprene. Thus, the authors of [1]
believe that the isoprene synthesis reaction is catalyzed
by weak Lewis acid sites, whereas medium-strength
Lewis acid sites lead to the formation of byproducts; at
the same time, Krzywicki et al. [6] believe that iso-
prene is formed on strong Lewis acid sites.
This study is focused on a single-stage gas-phase
synthesis of isoprene from formaldehyde and isobuty-
lene in the presence of Al–BEA and Nb–BEA zeolite
catalysts containing different amounts of aluminum
and niobium and the effect of the metal content on the
physicochemical and catalytic properties of the syn-
thesized samples.
adsorbed ammonia was removed at 100°C in a dry
helium stream for 1 h. After that, the temperature in
the reactor was linearly increased to 800°C at a rate of
8 deg/min. Changes in the thermal conductivity of the
stream were recorded using a thermal-conductivity
detector.
Infrared (IR) spectra were recorded on a Nicolet
Protege 460 Fourier-transform IR spectrometer
equipped with an MCT detector in the range of 4000–
400 cm−1 at a resolution of 4 cm−1. A 20-mg catalyst
sample was compressed into a disc with a diameter of
2 cm2. Water was removed from the samples on a vac-
uum unit equipped with absolute pressure sensors at a
working vacuum of 5 × 10−4 Pa. A pellet of the sample
was placed in an IR cell, heated to 450°C for 2 h, and
held at 450°C for 1 h. Carbon monoxide adsorption
was implemented at the liquid nitrogen temperature
(77 K) by dosing the gas until complete saturation.
The recorded IR spectra were processed using the
OMNIC ESP version 7.3 software.
The reaction of isobutylene with formaldehyde was
studied in a fixed-bed catalytic flow reactor system at
a temperature of 300°C; a feed space velocity of isobu-
tylene and formalin of 3.85 and 1.09 g/(g h), respec-
tively; and an isobutylene : formaldehyde molar ratio
of 5 : 1. Before testing, the catalysts were heated to
350°C in a helium stream; after that, the temperature
was decreased to the reaction temperature. Formalin
containing 37 wt % of formaldehyde, 3 wt % of meth-
anol, and 60 wt % of water was used without further
purification.
The chromatographic analysis of liquid and gas-
eous reaction products was conducted on a Chro-
matec Analytic Kristall 2000M gas–liquid chromato-
graph equipped with a flame ionization detector and a
40 m × 0.32 mm capillary column coated with the SE-
30 nonpolar phase. Dioxane and methane were added
as an internal standard to the liquid and gaseous sam-
ples, respectively.
EXPERIMENTAL
The original Al–BEA samples were zeolites with
Si/Al = 12.5, 25, 75, and 150 purchased from Zeolyst.
All samples were calcined in a dry air stream at 550°C
for 6 h.
The BEA zeolite with Si/Al = 75 was dealuminated
by three runs of treatment with concentrated nitric
acid at 80°С under stirring for 12 h. The resulting sam-
ple was washed with distilled water, dried, and cal-
cined at a temperature of 550°С for 6 h.
Modification with niobium was conducted by
impregnation of the dealuminated BEA zeolite with
an NbCl5 solution in isopropyl alcohol at 90°С for
12 h. The resulting catalyst was washed with distilled
water, dried, and calcined at 550°С for 6 h.
To determine the amount of CO, the gaseous sam-
ples were also analyzed on a Kristall 2000M chro-
matograph equipped with a thermal conductivity
detector and a 3-m column packed with Porapak-Q.
Formaldehyde concentration in the liquid samples
was determined by titration with Na2SO3.
Total formaldehyde conversion (F), product selec-
tivity (Si), and isoprene productivity (productivity
with respect to isoprene, P) were calculated by the fol-
The elemental composition of all the studied sam-
ples was determined by X-ray fluorescence analysis on
a ThermoScientific ARL PERFORM’X WDXRF
instrument equipped with a 2.5-kW rhodium-anode
X-ray tube.
Low-temperature nitrogen adsorption isotherms
were recorded on a Micromeritics ASAP2000 auto-
matic porosimeter. Micropore volume was deter-
mined by the t-plot method.
The acidic properties of the samples were studied lowing formulas:
by temperature-programmed desorption of ammonia
F = m
– munreacted formaldehyde
fed formaldehyde
(NH3-TPD) on a USGA-101 universal sorption gas
analyzer. A weighed portion of the sample was placed
in a quartz reactor, heated in a helium stream to a tem-
perature of 500°C, calcined at this temperature in a
× 100/mfed formaldehyde, %,
Si = n ×100/Σn, %,
i
i
PETROLEUM CHEMISTRY Vol. 60 No. 8 2020