Microporous and Mesoporous Materials 284 (2019) 258–264

Contents lists available at ScienceDirect

Microporous and Mesoporous Materials
journal homepage: www.elsevier.com/locate/micromeso

Ultra-thin MFI membranes with different Si/Al ratios for CO2/CH4
separation

T

Liang Yu∗, Shahpar Fouladvand, Mattias Grahn, Jonas Hedlund
Chemical Technology, Luleå University of Technology, SE-971 87, Luleå, Sweden

A R T I C LE I N FO

A B S T R A C T

Keywords:
MFI zeolite membranes
Si/Al ratios
CO2/CH4 separation
High permeance
Natural gas
Biogas

Ultra-thin MFI zeolite membranes with different Si/Al ratios (152, 47 and 26) were prepared on graded αalumina supports in the presence of organic template molecules and evaluated for separation of equimolar CO2/
CH4 mixtures at temperatures from 315 to 249 K. The thicknesses of all membranes were less than 500 nm and
permporometry showed that the number and size of defects were low for the two membranes with the highest
Si/Al ratio (152 and 47). The membrane with the lowest Si/Al ratio (26) also had low amounts of defects in the

mesopore range, but did have a few macropore defects. All membranes showed very high CO2 permeances in the
entire temperature range studied and the permeances increased with increasing temperature. The CO2 permances were also correlated to the Si/Al ratio of the membranes. The higher permeances was observed for
membranes with higher Si/Al ratio. The highest observed CO2 permeance was 142 × 10−7 mol s−1 m−2 Pa−1 at
room temperature for the membrane with Si/Al = 152. The separation factor, on the other hand, increased with
decreasing temperature for the two membranes with the highest Si/Al ratio (152 and 47), but for the membrane
with a Si/Al ratio of 26, the separation factor went through a maximum at ca. 270 K. The highest separation
factor observed was 7.1 at 249 K for the membrane with Si/Al = 47. These observations are consistent with an
adsorption controlled separation mechanism.

1. Introduction
Natural gas and biogas containing mainly CH4 and CO2 are environmentally friendly fuels and feed stocks. However, CO2 reduces the
heating value, may cause pipe corrosion and acts like a ballast.
Consequently, developing technologies for efficient removal of CO2
from biogas and natural gas is a necessary and timely research topic
[1].
Removal of CO2 from CH4 can be achieved by e.g., pressure swing
adsorption, cryogenic separation, amine absorption or membrane separation [2]. Absorption or adsorption based processes have been used
widely for CO2 separation in industry, but these techniques are costly.
Consequently, the application of membrane technology for gas separation has recently attracted much attention [1]. Despite much progress in the development of polymer membranes, the industrial use is
still limited due to the low permeance and limited robustness of these
organic materials [3]. Inorganic zeolite membranes with porous framework and high stability have therefore drawn much interest. The
well-defined porous zeolite framework may potentially give both high
permeance, which reduces the required membrane area and also high
selectivity [4].

∗

Several types of zeolite membranes have been investigated for CO2
separation from CH4. Small-pore zeolite membranes, like CHA, DDR
with high selectivity, e.g. 100–200, but very low CO2 permeance,

namely 0.4–5 × 10−7 mol s−1 m−2 Pa−1, have been reported [5–9].
Zeolite membranes with larger pore size, for example MFI and FAU,
have also been evaluated for CO2/CH4 separation, showing higher
permeance (7.5–9 × 10−7 mol s−1 m−2 Pa−1) and separation selectivities in the range of 5–40 [10–12]. Our group has developed a masking
technique, which enables preparation of ultra-thin MFI zeolite membranes with limited invasion of zeolite into the support on open graded
α-alumina discs resulting in membranes with exceptionally high permeances [13]. The membranes have a high Si/Al ratio of about 139 and
display a high CO2 permeance of 45 × 10−7 mol s−1 m−2 Pa−1 and
good CO2/CH4 separation factor of 4.5 at 10 bar feed pressure and
277 K [14].
MFI zeolite can be prepared with different Si/Al ratios. The amount
of aluminium included in the zeolite framework has a profound effect
on the properties of the zeolite, e.g. the polarity of the framework,
which increases with increasing aluminium content [15]. Many systematic studies on the influence of the Si/Al ratio on the performance of
zeolites have been reported. As an example, the adsorption of CO2 in

Corresponding author.;
E-mail address: (L. Yu).

/>Received 13 November 2018; Received in revised form 29 March 2019; Accepted 19 April 2019
Available online 19 April 2019
1387-1811/ © 2019 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license
( />

Microporous and Mesoporous Materials 284 (2019) 258–264

L. Yu, et al.

top surface of the support with a monolayer of MFI seed crystals and 3)
growth of the seed layer into a continuous zeolite film. The detailed
procedure has been described in detail in a patent application [25].

50 nm silicalite-1 crystals were used as seeds. For the aim of obtaining a zeolite membrane free of aluminium, an aluminium free
synthesis
mixture
with
a
molar
composition
of
3TPAOH:25SiO2:1450H2O:100EtOH was used. The growth was carried
out in an oil bath under reflux at 88 °C for 55 h. For the aim of a
membrane with a Si/Al ratio of 50, aluminium isopropoxide (≥98.0%,
Aldrich) was used as aluminium source and sodium hydroxide (NaOH,
≥99.0, Merk) was used to increase the alkalinity of the solution. The
final molar composition of the synthesis mixture with an Si/Al ratio of
50 was 3TPAOH:0.25Al2O3:Na2O:25SiO2:1600H2O:100EtOH and the
zeolite films were grown in an oil bath at 100 °C for 22 h under reflux. A
synthesis solution with a Si/Al ratio of 25 with a molar composition of
3TPAOH:0.5Al2O3:Na2O:25SiO2:1600H2O:100EtOH was used for preparation of a membrane with a target Si/Al ratio of 25. The hydrothermal treatment was carried out at 150 °C for 14 h in an autoclave.
After synthesis, the membranes were rinsed in a 0.1 M NH3 solution and
then calcined at 500 °C for 6 h using a heating rate of 0.2 °C min−1 and
a cooling rate of 0.3 °C min−1.
The membranes were characterized by n-hexane/helium adsorption-branch permporometry to evaluate the quality of the membrane
and determine the distribution of flow-through defects [26]. A detailed
description of the experimental and data evaluation procedures are
given in our previous work [27]. Scanning electron microscopy (SEM,
FEI Magellan 400 field emission XHR-SEM) was used for investigating
the morphology, microstructure and thickness of the membranes. Energy dispersive spectroscopy (EDS) was used to estimate the Si/Al ratio
of the zeolite film obtained from the aluminium free synthesis mixture.
Before measurement, the membranes were ion-exchanged to Cs+ form
as described in our previous report and thereafter coated with carbon

[24].

MFI zeolites with different Si/Al ratios was studied by Tuanny et al.
[16]. The CO2 adsorption/desorption isotherms showed that the MFI
zeolites with lower Si/Al ratio, i.e. higher aluminium content, showed
higher affinity for CO2. A strong interaction between these polar sites
associated with aluminium and the large quadrupole moment and polarizability of CO2 have been pointed out as the factors responsible for
the relatively high affinity of CO2 for zeolites with low Si/Al ratio. Our
group has also reported a detailed investigation of both single component and binary adsorption of carbon dioxide and methane in Na-ZSM-5
[17,18]. The adsorption selectivity is dependent on the composition,
and selectivities in the range of 15–31 were observed at 308 K. Furthermore, the adsorption selectivity was increasing with decreasing
temperature and for an equimolar mixture at atmospheric pressure,
adsorption selectivivities of about 8 and 15 were observed at 393 and
308 K, respectively. Modelling work reported by Krishana et al. showed
an adsorption selectivity of about 3 for an equimolar CO2/CH4 mixture
at 300 K and 10 bar pressure for pure silica MFI membranes [19].
Systematic investigations on how the performance of zeolite membranes is affected by the Si/Al ratio have also been reported. MFI
membranes with Si/Al ratios of 57 and ∞ were prepared by Noack et al.
[15]. The membranes were characterized by equimolar H2/SF6 mixture
permeation measurements under various conditions. The ZSM-5 membrane with a Si/Al ratio of ∞ was H2 selective with a separation factor
of 42 and showed a H2 permeance of 2.36 × 10−7 mol s−1 m−2 Pa−1.
Meanwhile, it was also observed that both the H2 and SF6 permeance
decreased with increasing Al content. For the membrane with a Si/Al
ratio of 57, the H2/SF6 separation factor increased to 52, whereas the
H2 permeance decreased to 0.89 × 10−7 mol s−1 m−2 Pa−1. The decreased permances in the membrane with higher aluminium content
was ascribed by the presence of charge balancing cations reducing the
effective pore diameter somewhat. It was also found that the number of
intracrystalline defects increased with increasing Al content in the
synthesis mixture. Therefore, high quality Al-rich MFI membranes are
difficult to prepare at least when organic template molecules are used

[20]. However, in the absence of organic template molecules, we
showed that it even possible to prepare high quality membranes with a
Si/Al ratio as low as about 11.5. The high quality was for instance
demonstrated with a N2/SF6 separation factor as high as 110. However,
the permeance of these membranes was low as it was very difficult to
remove water completely from the pores of the zeolite. We have previously reported results on the influence of the Si/Al ratio of MFI
membranes prepared in the presence of organic template molecules on
the separation of hexane isomers, permeation of H2 and SF6 and separation of methanol and ethanol from synthesis gas [21–24]. Typically, the membranes were selective towards the more polar component
in the feed and higher permeance was observed for membranes with
higher Si/Al ratio. The observed differences were assigned to differences in adsorption properties of the membranes, as well as the narrower effective pore diameter due to the existence of counter-cations,
normally sodium ions, in the membranes with lower Si/Al ratio.
In the present work, ultra-thin MFI membranes with different Si/Al
ratios (the Si/Al ratio in the membrane synthesis mixture are ∞, 50 and
25) were prepared using organic template molecules on porous graded
α-alumina discs and, for the first time, the separation was evaluated for
an equimolar CO2/CH4 mixture as a function of temperature.

2.2. Gas separation experiments
Before separation experiments, the membrane was mounted in a
Wicke –Kallenbach type steel cell and subsequently dried in a flow of
helium at 300 °C for 6 h with a heating rate of 1 °C min−1 followed by
natural cooling. A thermostated bath with silicone oil was used to
achieve sub-ambient membrane temperatures by submerging the
membrane cell in the oil. Permeation of gas mixtures was performed in
a continuous flow mode using an equimolar mixture of CO2 and CH4
introduced to the cell by digital mass flow controllers. The retentate
pressure was controlled by a back pressure regulator. The pressure on
both sides of the membrane was monitored by pressure gauges. The
feed mixture was fed to the membrane at a total pressure of 700 kPa at a
flow rate of 8 Ndm3/min, and the permeate was kept at atmospheric

pressure. No sweep gas was used during measurements. The gas flow in
the permeate stream was measured using a bubble flow meter, and the
composition of the retentate and permeate streams were analysed using
a GC (490 Micro GC, Agilent).
The separation performance of the membrane was described using
separation factor β, flux J, permeance π and separation selectivity α as
follows.
For the equimolar binary CO2 and CH4 mixture separation, the separation factor β CO2/CH4 is defined as the ratio of the permeate concentrations of components i and j:

2. Experimental
2.1. Membrane preparation

β
Porous graded α-alumina discs (Fraunhofer IKTS, Germany) were
used as supports. The discs have a diameter of 25 mm with a 30 μm
thick top layer and a 3 mm thick base layer. The pore size of the topand base layer are 100 nm and 3 μm, respectively [13]. The general
procedure for membrane growth consist of; 1) masking the support to
prevent growth of zeolite into the pores of the support, 2) coating the

CO2 /CH 4

=

CCO2
CCH4

(1)
−2

The flux Ji (kg m


F
Ji = i
A
where Fi (kg h
259

h

−1

) for component i is defined by:
(2)

−1

) is the permeate mass flow rate of component i and A


Microporous and Mesoporous Materials 284 (2019) 258–264

L. Yu, et al.

Fig. 1. Top view (a) and cross-sectional (b) SEM images of an as-synthesised MFI membrane with a Si/Al ratio of 152.

permeance) as high as 110 × 10−7 mol s−1 m−2 Pa−1. When n-hexane
was added to the feed (p/p0 = 0.025), the zeolite pores and micro pore
defects smaller than about 1.25 nm, were blocked [27] and the helium
permeance was dramatically reduced to 0.53 × 10−7 mol s−1 m−2
Pa−1, i.e. a decrease by 99.5%. When the relative pressure of n-hexane

was further increased, the helium permeance decreased gradually to
0.02 × 10−7 mol s−1 m−2 Pa−1 at a relative pressure of n-hexane in the
feed of 0.91. This indicates that the membrane is of very high quality
with no large defects in line with the SEM observations. The Si/Al ratio
of the Cs+ exchanged membrane was estimated by EDS to 152. This is
somewhat higher than the value 139 reported previously MFI membranes prepared using a different masking method [13].
Fig. 3 illustrates the separation results for a feed comprised of an
equimolar mixture of CO2/CH4 as a function of temperature for the
same membrane. The CO2 permeance through the membrane was as
high as 142 × 10−7 mol s−1 m−2 Pa−1 at room temperature and decreased only slightly to 106 × 10−7 mol s−1 m−2 Pa−1 at the lowest
temperature (249 K). The thin zeolite film in combination with an open
graded support is the main explanation for the high permeance. This
permeance is quite similar to our previous reports for slightly thicker
MFI membranes with slightly lower Si/Al ratios of about 139 [Fig. 2 in
reference 14] and much higher than MFI membranes with similar Si/Al
ratio reported by other groups [29–31]. Consequently, the observed
CO2 flux was also very high although the total pressure difference
across the membrane was kept relatively low at 6 bar. The CO2 flux was
as high as 686 kg m−2 h−1 at room temperature and 495 kg m−2 h−1 at
249 K. The CO2 and CH4 permeances and fluxes are generally lower at
lower temperatures.
The separation factor and separation selectivity are very similar and
increase with decreasing temperature. The separation factor was
around 1.4 at room temperature increasing to a maximum of 2.5 at the
lowest investigated temperature, i.e. 249 K. As the permporometry results showed that the membrane was of high quality, the low separation
factor is probably a result of a low CO2/CH4 adsorption selectivity as
reported by Krishna et al. Table 1 summarizes the properties of the
membranes.
The membrane prepared from a synthesis mixture with a Si/Al ratio
of 50 shows a very similar morphology as the membrane prepared from

an aluminium free synthesis mixture as observed from the top-view and
cross-section SEM images shown in Fig. 4. The zeolite film is continuous
without any apparent crack or pinholes and has almost the same
thickness (i.e. 400 nm) as the membrane prepared from an aluminium
free synthesis mixture. Again, the pores of the support are completely
open and no invasion was observed by SEM. The Si/Al ratio of the film
was measured to 47 by EDS. The single gas helium permeance was
98 × 10−7 mol s−1 m−2 Pa−1, i.e. slightly lower than for the membrane with a Si/Al ratio of 152. During permporometry, the helium
permeance decreased by 99.5% when the relative pressure of n-hexane
(p/p0) in the feed increased from 0 to 0.025 as shown in Fig. 5,

(m2) is the membrane area.
The permeance πi (mol m−2 s−1 Pa−1) for component i is defined
by:

πi =

Ji
Mi ΔPi × 3600

(3)

where Mi (kg mol−1) is the molar mass of component i, ΔPi (Pa) is the
partial pressure difference of component i through the membrane.
The separation selectivity α CO2/CH4 is defined as the ratio of the
permeances of the components CO2 and CH4 through the membrane:

α

CO2 /CH 4


=

πCO2
πCH4

(4)

3. Results and discussion
Fig. 1 shows top view and cross-section SEM images of an as-synthesised MFI membrane prepared from an aluminium free synthesis
solution. The zeolite film is continuous and comprised of well-intergrown crystals. Grain boundaries are of course observed, but no defects
can be observed. The film is even with a thickness of around 350 nm
and no invasion, i.e. growth of zeolite or deposition of solid material
from the synthesis mixture, in the support could be observed by SEM.
The
single
gas
helium
permeance
was
as
high
as
110 × 10−7 mol s−1 m−2 Pa−1. This very high permeance is the result
of a thin film on an open, non-invaded graded support. This permeance
is also correspondingly higher than the 81 × 10−7 mol s−1 m−2 Pa−1
that we have reported previously for 500 nm thick MFI membranes
[28]. Fig. 2 shows the permporometry pattern of the membrane with
the initial helium permeance (p/p0 = 0, i.e. single gas helium


Fig. 2. Permporometry pattern of an as-synthesised MFI membrane with a Si/Al
ratio of 152.
260


Microporous and Mesoporous Materials 284 (2019) 258–264

L. Yu, et al.

Fig. 3. Separation factor, permeance, thickness normalized permeance and flux as a function of temperature for separation of an equimolar CO2/CH4 mixture by an
MFI membrane with a Si/Al ratio of 152.

Table 1
Summary of the properties of the membranes with different Si/Al ratios.
Properties of the membranes

Thickness of zeolite film (nm)
Single gas He permeance
(10−7 mol m−2 s−1 Pa−1)
Thickness normalized single gas He
permeance (mol m−2 s−1 Pa−1 m−1)
Decrease in helium permeance at nhexane p/p0 = 0.025 (%)
Maximum CO2/CH4 separation factor

Si/Al ratios
152

47

26


350
110

400
98

750
66

31

25

9

99.5

99.5

99.1

2.5 (249K)

7.1 (249K)

3.3 (271K)

indicating a high quality membrane.
Fig. 6 shows the separation results for an equimolar CO2/CH4 gas

mixture for the membrane with a Si/Al ratio of 47. The CO2 permeance
was slightly lower than that of the membrane with a Si/Al ratio of 152
but still very high, viz. 98 × 10−7 mol s−1 m−2 Pa−1 at room temperature. The lower CO2 permeance for membranes with lower Si/Al
ratio is consistent with previous reports [20,23,32]. At room temperature, the separation factor and separation selectivity were 2, which is

Fig. 5. Permporometry pattern for the MFI membrane with Si/Al = 47.

somewhat higher than for the membrane with a Si/Al ratio of 152 as
described above. When the temperature was decreased to 249 K, the
separation factor increased to 7.1, which was around 3 times higher
than for the membrane with a Si/Al ratio of 152. The higher separation

Fig. 4. SEM images of an as-synthesised MFI membrane with a Si/Al ratio of 47, (a) top view; (b) cross-section.
261


Microporous and Mesoporous Materials 284 (2019) 258–264

L. Yu, et al.

Fig. 6. Separation factor, permeance, thickness normalized permeance and flux as a function of temperature for separation of an equimolar CO2/CH4 mixture by an
MFI membrane with a Si/Al ratio of 47.

factor compared with the membrane with a Si/Al ratio of 152 membrane is probably a result from the higher density of polar sites where
CO2 is preferentially adsorbed [33–35], and consequently higher CO2/
CH4 adsorption selectivity.
Representative SEM images for a membrane grown from a synthesis
mixture with a Si/Al ratio of 25 are shown in Fig. 7. The membrane is
continuous and the zeolite film is even with a thickness of around
750 nm, i.e. about two times thicker than the membranes with higher

Si/Al ratios as described above. The morphology of this membrane was
somewhat different from the membranes with higher Si/Al ratios described above; the zeolite crystals were larger and more well-defined,
and consequently the grain boundaries appear more clearly. It has
previously been reported that pinholes form more frequently in films
with well-developed crystals and grain boundaries [36]. The Si/Al ratio
of the film was measured to 26 by EDS. The single helium gas permeance was 66 × 10−7 mol s−1 m−2 Pa−1, i.e. lower than for the
thinner films, as expected.
The permporometry pattern in Fig. 8 shows that the membrane has
reasonably good quality. When n-hexane was added to the feed such
that the p/p0 was 0.025, the helium permeance decreased by 99.1% as
compared to 99.5% for the two membranes discussed above. Also, the
permeance measured at the highest relative pressure of 0.9 was as high
as 0.3 × 10−7 mol s−1 m−2 Pa−1, which indicates the existence of a

Fig. 8. Permporometry pattern for an MFI membrane with Si/Al = 26.

few larger defects, most likely pinholes in the membrane. These defects
should be larger than 48 nm as estimated from the Harkins–Jura and
Kelvin equations for a relative pressure of n-hexane of 0.9 [26]. Our
results with higher density of defects in the membrane with lower Si/Al

Fig. 7. SEM images of an as-synthesised MFI membrane with Si/Al = 26, (a) top view; (b) cross-section.
262


Microporous and Mesoporous Materials 284 (2019) 258–264

L. Yu, et al.

Fig. 9. Separation factor, permeance, thickness normalized permeance and flux as a function of temperature for separation of an equimolar CO2/CH4 mixture by an

MFI membrane with a Si/Al ratio of 26.

the flux of CO2 decreased faster than the decrease in CH4 flux.
Our previous work has showed that the performances of our high
flux membranes may be affected by both concentration polarization
and pressure drop over the support [43]. For example in the work by Yu
et al., high flux MFI membranes were evaluated for separation of olefin/
nitrogen (20% olefin in nitrogen) at different temperature. At 298 K, the
C3H6 permeance and selectivity were about 22 × 10−7 mol m−2 s−1
Pa−1 and 45, respectively. At these conditions, the relative pressure
drop over the support was only 8% and the concentration polarization
index i.e. the ratio between concentration of the main permeating
specie at the membrane surface to the concentration in the feed gas
bulk was 0.83. However, the relative pressure drop over the support
increased to 22% and the concentration polarization index decreased to
0.75 for separation of C2H4/N2 with C2H4 permeance of
70 × 10−7 mol m−2 s−1 Pa−1 and a selectivity about 4 at room temperature. For the gas separation in this work, the permeance is even
higher. Therefore, pressure drop over the support likely decreased both
the selectivity and the permeance. However, as the feed here was a
50:50 mixture and the selectivities were modest, the effect of concentration polarization should be minor.
At room temperature, the separation factor increased and the CO2
permeance decreased with decreasing Si/Al ratio. The differences in the
separation performance of these membranes can be explained by the
dependence of the adsorption and transport properties on the zeolite
framework composition. As sodium was used in the synthesis mixtures
for preparing the membranes with the lower Si/Al ratios, sodium is the
cation associated to aluminium. These ions will both provide strong
adsorption sites, which may lead to strong correlation effects [42] and
probably also reduce the accessibility of the pore space, which may be a
factor for the observed decrease in CO2 transport with increasing aluminium content of the zeolite [15]. For better understanding, it would

probably be useful to apply an effective medium approximation model
[42] to the experimental data, which however is beyond the scope of
the present work. In addition, the thickness normalized permeance as
shown in Figs. 3, 6 and 9 as well as Table 1 also indicated that the
aluminium and sodium content in the zeolite pores affect the permeance. Thus, for the lower Si/Al ratio membranes, ion exchange to H+
as counter ion could probably increase the pore accessibility, promoting
a higher permeance [44].

ratio is in line with previous reports from other groups [18].
The permeation data from the separation experiments for the
membrane with a Si/Al ratio of 26 are shown in Fig. 9a. The CO2
permeance was 71 × 10−7 mol s−1 m−2 Pa−1 at room temperature
(corresponding to a flux of 302 kg m−2 h−1), which was lower than for
the membranes with Si/Al ratios of 152 and 47. However, this membrane was about twice as thick as the two membranes with higher Si/Al
ratio, which most likely is the main cause for the lower permeance.
Still, the permeance is much higher than previously reported for MFI
zeolite membranes with similar or low Si/Al ratio and much higher
than that for polymeric membranes [37,38].
The separation factor was 3.1 at room temperature, i.e. higher than
the separation factors of 1.4 and 2 observed for the membranes with Si/
Al ratios of 152 and 47, respectively. The separation factor of this
membrane showed a different trend with decreasing temperature as
compared to the other membranes with higher Si/Al ratios. As the
temperature was decreased to 271 K, the separation factor increased
similarly as for the membranes with higher Si/Al ratios. However, as
the temperature was decreased further, the separation factor decreased
again, probably due to further increase in adsorbed concentration in
combination with decreased diffusion or desorption rate. We expect
that the effect of temperature on the transport through the pores should
be greater for this membrane compared to the one with higher Si/Al

ratios where the densities of strong adsorption sites are ca 2 and 0.6 per
unit cell for the membranes with Si/Al of 46 and 152, respectively. It is
well known that transport through silicalilte-1 is correlated i.e. a
stronger adsorbing but slower diffusing species (carbon dioxide) may
slow down the transport of a weaker adsorbing but faster diffusing
specie (methane) and this can be described by the Maxwell-Stefan approach [39–41]. However, in the more heterogeneous aluminum containing ZSM-5, the correlation effects can be even larger and the
Maxwell-Stefan approach may be inaccurate. In this case, an effective
medium approximation may be more successful when the differences in
relative adsorption strengths of the species are very high [42]. On the
other hand, the CO2/CH4 adsorption selectivity increases with decreasing temperature which should result in increased membrane selectivity with decreasing temperature as well [17]. Our results clearly
suggest that the most favourable separation conditions for the membrane with Si/Al = 26 occurred at 271 K and below this temperature
263


Microporous and Mesoporous Materials 284 (2019) 258–264

L. Yu, et al.

4. Conclusions

[18] L. Ohlin, M. Grahn, Detailed investigation of the binary adsorption of carbon dioxide and methane in zeolite Na-ZSM-5 studied using in situ ATR-FTIR spectroscopy, J. Phys. Chem. C 118 (2014) 6207–6213.
[19] R. Krishna, J.M. van Baten, In silico screening of zeolite membranes for CO2 capture, J. Membr. Sci. 360 (2010) 323–333.
[20] M. Noack, G.T.P. Mabande, J. Caro, G. Georgi, W. Schwieger, P. Kölsch, A. Avhale,
Influence of Si/Al ratio, pre-treatment and measurement conditions on permeation
properties of MFI membranes on metallic and ceramic supports, Microporous
Mesoporous Mater. 82 (2005) 147–157.
[21] J. Hedlund, M. Noack, P. KoÈlsch, D. Creaser, J. Caro, J. Sterte, ZSM-5 membranes
synthesized without organic templates using a seeding technique, J. Membr. Sci.
159 (1999) 263–273.
[22] F. Jareman, J. Hedlund, J. Sterte, Effects of aluminum content on the separation

properties of MFI membranes, Separ. Purif. Technol. 32 (2003) 159–163.
[23] S.A.S. Rezai, J. Lindmark, C. Andersson, F. Jareman, K. Möller, J. Hedlund, Water/
hydrogen/hexane multicomponent selectivity of thin MFI membranes with different
Si/Al ratios, Microporous Mesoporous Mater. 108 (2008) 136–142.
[24] L. Sandström, J. Lindmark, J. Hedlund, Separation of methanol and ethanol from
synthesis gas using MFI membranes, J. Membr. Sci. 360 (2010) 265–275.
[25] J. Hedlund, A. Holmgren, L. Yu, Methods for preparing supported zeolite films, GB
patent application 6 (Sep 2017) GB1714269.
[26] J. Hedlund, D. Korelskiy, L. Sandström, J. Lindmark, Permporometry analysis of
zeolite membranes, J. Membr. Sci. 345 (2009) 276–287.
[27] D. Korelskiy, M. Grahn, J. Mouzon, J. Hedlund, Characterization of flow-through
micropores in MFI membranes by permporometry, J. Membr. Sci. 417–418 (2012)
183–192.
[28] J. Hedlund, J. Sterte, M. Anthonis, A.J. Bons, B. Carstensen, N. Corcoran, D. Cox,
H. Deckman, W. deGijnst, P.P. deMoor, F. Lai, J. McHenry, W. Mortier, J. Reinoso,
High-flux MFI membranes, Microporous Mesoporous Mater. 52 (2002) 179–189.
[29] C. Algieri, P. Bernardo, G. Golemme, G. Barbieri, E. Drioli, Permeation properties of
a thin silicalite-1 (MFI) membrane, J. Membr. Sci. 222 (2003) 181.
[30] M.C. Lovallo, A. Gouzinis, M. Tsapatsis, Synthesis and characterization of oriented
MFI membranes prepared by secondary growth, AIChE J. 44 (1998) 1903.
[31] F. Bonhomme, M.E. Welk, T.M. Nenoff, CO2 selectivity and lifetimes of high silica
ZSM-5 membranes, Microporous Mesoporous Mater. 66 (2003) 181–188.
[32] D. Newsome, M.-O. Coppens, Molecular dynamics as a tool to study heterogeneity
in zeolites – effect of Na+ cations on diffusion of CO2 and N2 in Na-ZSM-5, Chem.
Eng. Sci. 121 (2015) 300–312.
[33] J.A. Dunne, R. Mariwala, M. Rao, S. Sircar, R.J. Gorte, A.L. Myers, Calorimetric
heats of adsorption and adsorption isotherms. 1. O2, N2, Ar, CO2, CH4, C2H6 and SF6
on Silicalite, Langmuir 12 (1996) 5888–5895.
[34] J.A. Dunne, M. Rao, S. Sircar, R.J. Gorte, A.L. Myers, Calorimetric heats of adsorption and adsorption isotherms. 2. O2, N2, Ar, CO2, CH4, C2H6, and SF6 on NaX,
H-ZSM-5, and Na-ZSM-5 zeolites, Langmuir 12 (1996) 5896–5904.

[35] D. Newsome, S. Gunawan, G. Baron, J. Denayer, M.-O. Coppens, Adsorption of CO2
and N2 in Na–ZSM-5: effects of Na+ and Al content studied by Grand Canonical
Monte Carlo simulations and experiments, Adsorption 20 (2014) 157–171.
[36] J.C. Jansen, J.H. Koegler, H. van Bekkum, H.P.A. Calis, C.M. van den Bleek,
F. Kapteijn, J.A. Moulijn, E.R. Geus, N. van der Puil, Zeolitic coatings and their
potential use in catalysis, Microporous Mesoporous Mater. 21 (1998) 213–226.
[37] H. Lin, M. Yavari, Upper bound of polymeric membranes for mixed-gas CO2/CH4
separations, J. Membr. Sci. 475 (2015) 101–109.
[38] R.W. Baker, B.T. Low, Gas separation membrane materials: a perspective,
Macromolecules 47 (2014) 6999–7013.
[39] R. Krishna, J.M. van Baten, Insights into diffusion of gases in zeolites gained from
molecular dynamics simulations, Microporous Mesoporous Mater. 109 (2008)
91–108.
[40] L.J.P. van den Broeke, W.J.W. Bakker, F. Kapteijn, J.A. Moulijn, Binary permeation
through a silicalite-1 membrane, AIChE J. 45 (1999) 976–985.
[41] R. Krishna, J.M. van Baten, Maxwell-Stefan modeling of slowing-down effects in
mixed gas permeation across porous membranes, J. Membr. Sci. 383 (2011)
289–300.
[42] V. Iyengar, M.-O. Coppens, Dynamic Monte-Carlo simulations of binary self-diffusion in zeolites: effect of strong and weak adsorption sites, Chem. Eng. Sci. 59
(2004) 4747–4753.
[43] L. Yu, M. Grahn, P. Ye, J. Hedlund, Ultra-thin MFI membranes for olefin/nitrogen
separation, J. Membr. Sci. 524 (2017) 428–435.
[44] K. Aoki, V.A. Tuan, J.L. Falconer, R.D. Noble, Gas permeation properties of ionexchanged ZSM-5 zeolite membranes, Microporous Mesoporous Mater. 39 (2000)
485–492.

Ultra-thin MFI zeolite membranes with Si/Al ratios of 152, 47 and
26, were grown on porous, graded alumina supports. The membranes
were evaluated by SEM and permporometry and for separation of
equimolar CO2/CH4 mixtures at different temperatures. All membranes
showed high CO2 permeances, which decreased with decreasing temperature and decreasing Si/Al ratio of the membranes. The separation

factor on the other hand increased with decreasing temperature except
for the membrane with the lowest Si/Al ratio where the separation
factor went through a maximum as a function of temperature. The increased separation factor is probably a result of higher polarity of
membranes with low Si/Al ratio of zeolite framework, resulting in increased affinity to CO2.
Acknowledgements
The Swedish Energy Agency, Bio4Energy and Swedish Research
Council for Environment, Agricultural Sciences and Spatial Planning
Formas are gratefully acknowledged for financially supporting this
work.
References
[1] S. Basu, A.L. Khan, A. Cano-Odena, C. Liu, I.F.J. Vankelecom, Membrane-based
technologies for biogas separations, Chem. Soc. Rev. 39 (2010) 750–768.
[2] X.Y. Chen, H. Vinh-Thang, A.A. Ramirez, D. Rodrigue, S. Kaliaguine, Membrane gas
separation technologies for biogas upgrading, RSC Adv. 5 (2015) 24399–24448.
[3] R.W. Baker, K. Lokhandwala, Natural gas processing with membranes: an
Overview, Ind. Eng. Chem. Res. 47 (2008) 2109–2121.
[4] N. Kosinov, J. Gascon, F. Kapteijn, E.J.M. Hensen, Recent developments in zeolite
membranes for gas separation, J. Membr. Sci. 499 (2016) 65–79.
[5] T. Tomita, K. Nakayama, H. Sakai, Gas separation characteristics of DDR type
zeolite membrane, Microporous Mesoporous Mater. 68 (2004) 71–75.
[6] Y. Cui, H. Kita, K.I. Okamoto, Preparation and gas separation performance of zeolite
T membrane, J. Mater. Chem. 14 (2004) 924–932.
[7] S. Li, M.A. Carreon, Y. Zhang, H.H. Funke, R.D. Noble, J.L. Falconer, Scale-up of
SAPO-34 membranes for CO2/CH4 separation, J. Membr. Sci. 352 (2010) 7–13.
[8] S. Li, J.G. Martinek, J.L. Falconer, R.D. Noble, T.Q. Gardner, High-pressure CO2/
CH4 separation using SAPO-34 membranes, Ind. Eng. Chem. Res. 44 (2005)
3220–3228.
[9] S. Li, J.L. Falconer, R.D. Noble, SAPO-34 membranes for CO2/CH4 separations:
effect of Si/Al ratio, Microporous Mesoporous Mater. 110 (2008) 310–317.
[10] Y. Hasegawa, T. Tanaka, K. Watanabe, B.H. Jeong, K. Kusakabe, S. Morooka,

Separation of CO2–CH4 and CO2–N2 systems using ion-exchanged FAU-type zeolite
membranes with different Si/Al ratios, Kor. J. Chem. Eng. 19 (2002) 309–313.
[11] J.C. Poshusta, R.D. Noble, J.L. Falconer, Temperature and pressure effects on CO2
and CH4 permeation through MFI zeolite membranes, J. Membr. Sci. 160 (1999)
115–125.
[12] J. Lindmark, J. Hedlund, Modification of MFI membranes with amine groups for
enhanced CO2 selectivity, J. Mater. Chem. 20 (2010) 2219–2225.
[13] J. Hedlund, F. Jareman, A.J. Bons, M. Anthonis, A masking technique for high
quality MFI membranes, J. Membr. Sci. 222 (2003) 163–179.
[14] L. Sandström, E. Sjöberg, J. Hedlund, Very high flux MFI membrane for CO2 separation, J. Membr. Sci. 380 (2011) 232–240.
[15] M. Noack, P. Kölsch, V. Seefeld, P. Toussaint, G. Georgi, J. Caro, Influence of the Si/
Al-ratio on the permeation properties of MFI-membranes, Microporous Mesoporous
Mater. 79 (2005) 329–337.
[16] T.S. Frantz, W.A. Ruiz, C.A. da Rosa, V.B. Mortola, Synthesis of ZSM-5 with high
sodium content for CO2 adsorption, Microporous Mesoporous Mater. 222 (2016)
209–217.
[17] L. Ohlin, P. Bazin, F. Thibault-Starzyk, J. Hedlund, M. Grahn, Adsorption of CO2,
CH4, and H2O in zeolite ZSM-5 studied using in situ ATR-FTIR spectroscopy, J.
Phys. Chem. C 117 (2013) 16972–16982.

264