HUNGARIAN JOURNAL OF
INDUSTRY AND CHEMISTRY

Vol. pp.45(1) pp. 5–8 (2017)

hjic.mk.uni-pannon.hu

DOI: 10.1515/hjic-2017-0002

SEPARATION OF GASES BY MEMBRANES:
THE EFFECTS OF POLLUTANTS ON THE STABILITY OF MEMBRANES

NÁNDOR NEMESTÓTHY*

Research Institute of Bioengineering, Membrane Technology and Energetics, University of
Pannonia, Egyetem u. 10, Veszprém, H-8200, HUNGARY

The long-term stability of membranes is determined mainly by their sensitivity to pollutants. Their stability was
tested using a novel, multichannel measuring system, which is based on pressure differences. This measuring
system is suitable to determine the changes in permeability of polymer membranes. The damaging effects of
H2S, BTX and n-dodecane were investigated in terms of polyimide gas separation membranes using nitrogen
gas.

Keywords: multichannel test equipment, pressure difference, hydrogen sulphide, BTX

1. Introduction

Previously, it was thought that the stability of memb-
ranes is determined by the mechanical stress (shear) and
the natural aging of polymers. Recently, however, it has
been confirmed that their stability is limited mainly by
the sensitivity of membranes to certain pollutants. These
aggressive compounds, pollutants, e.g. chlorine,
hydrogen sulphide, hydrocarbons, etc., may damage the
structure of the polymer, thus its physical and chemical
properties change, and consequently the permselectivity
of the membrane changes, as well.

It is known that polymeric reverse osmosis memb-
ranes are sensitive to strong oxidising agents, especially
chlorine compounds [1], therefore, intensive research
has been conducted to avoid or at least reduce any
damage [2-3]. Similar levels of membrane degradation
are observed in proton-exchange membrane (PEM) fuel
cells and batteries containing membranes, where oxi-
dising compounds are in contact with the membranes, as
well [4-5].

The stability of polymeric gas separation memb-
ranes has hardly been investigated. The long-term eff-
ects of H2S on inorganic membranes has been studied
by Australian researchers at low concentrations (50
ppm) [6]. However, H2S may not only cause long-term,
but immediate damage, mainly in the form of swelling,
which strongly influences the gas transport properties of
membranes. Koros and co-workers presented the effects
of extremely high H2S concentrations on a polymer
membrane (50,000 - 100,000 ppm) [7].

In the field of membrane technology, sensitivity
can be measured by a sort of effectiveness unit. The

*Correspondence: nemesn@almos.uni-pannon.hu

product of the concentration and time period yields a
value where the effectiveness of the membrane decr-
eases from 95 to 90 and then to 70% of its original
value (e.g. 1000 ppm*hour means that the membrane
was exposed to 1000 ppm of pollutant for 1 hour, or 0.1
ppm of pollutant for 10,000 hours during the tests).

According to the literature these types of
measurements have yet to be published for gas
separation membranes, thus the aim of this work was to
design, construct and operate a piece of test equipment
that conducts reliable laboratory tests.

For the determination of stability, direct and
indirect methods can be used to measure the gas
volumes passing through the membranes. Direct
methods are usually preferred, and – if the composition
of the gas is known – are more exact than indirect ones.
However, when the gas composition varies and small
amounts of gases need to be measured, indirect methods
are often more suitable. In this work an indirect method
based on a pressure differential technique was chosen,
where the pressure of a closed vessel is measured and
the varying pressure yields information about the
volume of the gas passing through the membrane.

During the investigation the effects of pollutants
on the permeability of nitrogen was to be studied. The
following pollutants were used:

 compounds containing sulphur at associated

gas
 BTX mixture (benzene, toluene, xylene)
 heavy hydrocarbons

In this research the aim was to determine

quantitatively the effects of pollutants on the
membranes to define the tolerance range of particular
membranes.

NEMESTÓTHY

Hungarian Journal of Industry and Chemistry

Figure 1. The small modules constructed for the tests

2. Experimental

For this series of measurements, polyimide gas
separation membranes (synthesised by UBE) were used.
They were taken from a hollow fibre module and can
accurately model the properties of industrial gas
separation membranes. From the hollow fibres small
modules containing 6 capillaries were constructed
(Fig.1) and their ends were closed, thus their tests were
carried out in a “sack” configuration.

In the design of the test system it was important
that several parallel measurements should be conducted
and the measuring channels combined with each other.
The scheme of a measuring channel can be seen in Fig.2

The gas was introduced into the measuring system
through valve V1 (which can be adjusted by valve V2 if
necessary). Before measurements were taken the
pressure of the vessel was checked by pressure
transducer PT1. To start the test the pressure was
adjusted by regulating valve PV1, which was checked
by pressure transducer PT2. Then the membrane was
installed into the thermostatic system in a way that
ensured its mobility was not restricted, thus the
permeation of gas could not influence the flux.

A photograph of the measuring system is shown in
Fig.3. For the permeability measurements nitrogen gas
from a cylinder was used (99.5%; Messer Hungarogáz
Kft., Hungary).

The permeability of the membrane was determined
from the pressure of the vessel and the transmembrane
pressure measured on-line during the experiments. In
the experiments, H2S, methyl mercaptan and ammonia
(compounds containing S or N), oleic acid, ethyl

Figure 2. The scheme of the test system

Figure 3. The test system

alcohol, moreover, a benzene-toluene-xylene mixture
(BTX) and n-dodecane (as a heavier hydrocarbon) were
used as pollutants.

For the stability experiments the small membrane
modules were put in a closed vessel (Fig.4) where the
headspace was saturated with the given pollutant. The
vessels were placed in a thermostatic incubator at 27 °C
usually for between 1 and 7 days.

Certain materials (e.g. BTX) damaged the epoxy
resin glue used to adhere the fibres of the membranes,
therefore, these experiments were repeated using poly-
ether-sulfone glue instead.

3. Results

The nitrogen permeability of the membranes was
determined before and after the incubations.

In the preliminary experiments, the ammonia
solution and methyl mercaptan severely damaged the
surface of the membranes, thus no flux could be
measured. Oleic acid and ethyl alcohol hardly
influenced the flux, while BTX, H2S and n-dodecane
changed the permeability of nitrogen considerably.

For further investigation of the pollutants an
experimental design was constructed, using appropriate

Figure 4: The membranes in the closed vessel

THE EFFECTS OF POLLUTANTS ON THE STABILITY OF MEMBRANES

45(1) pp. 5–8 (2017)

Table 1. The parameters of the experimental design

pollu-
tant

cmin
ppm

cav
ppm

cmax
ppm

tmin
d

tav
d

tmax
d

H2S 100,000 300,000 500,000 1 3.5 7
BTX
mixt.

1,000 750 500 1 3.5 7

dode-
cane

1,000 5,500 10,000 1 3.5 7

statistical methods. The parameters selected were the
concentrations of the pollutants (minimum, maximum
and average) and the incubation time (minimum,
maximum and average). The Statistica 8 computer
program was applied to the design presented in Table 1.

Firstly the effect of H2S on the nitrogen
permeability of the membranes was measured. The
experimental results are presented in Fig.5.

It can be seen that the permeability of nitrogen
increased even when the concentrations of pollutants
were low and rose by using higher concentrations and
longer periods of exposure.

From the H2S concentration and the incubation time
it was possible to calculate a special parameter of
exposure with the unit of ppm*h. Permeability was
presented as a function of this parameter (Fig.6), where
an almost linear relationship was observed.

The results suggest that the process can be described
as a first-order reaction, which means that no safe limit
can be determined where H2S is regarded as harmless,
on the contrary, it should be considered at all times.

The effect of the BTX mixture was studied using a
similar methodology. The results are summarized in
Fig.7. This figure shows that the permeability of
nitrogen increased at low concentrations of the BTX
mixture. At higher concentrations and over longer
periods of time, no further significant changes were
observed.

The exposure parameter was also calculated in the
unit of ppm*h and permeability was presented as its
function (Fig.8). The diagram can be described as a
saturation-type curve.

2**(2-0) design; MS Residual=49886.27

> 2000
< 2000
< 1500
< 1000
< 500
< 0

Figure 5. The effect of H2S on the nitrogen
permeability

Exposure [M ppm * h]

0 20 40 60 80 100

P
e
rm

e
a
b
il
it
y
c

h
a
n
g
e
[
%

]

500

1000

1500

2000

2500

Figure 6. Permeability changes against exposure to H2S

2**(2-0) design; MS Residual=2675.592

> 400
< 360
< 310
< 260
< 210
< 160
< 110
< 60

Figure 7: The effect of the BTX mixture

In the last series of experiments, the effect of

exposure to n-dodecane was investigated experi-
mentally. The results are presented in Fig.9.

n-dodecane caused – unlike H2S and BTX – a
reduction in the permeability of nitrogen even at low
concentrations. The flux fell to zero at higher
concentrations and over longer incubation periods.

Permeability was investigated as a function of
exposure (Fig.10). The process can be described as a
first-order reaction, thus the effect of heavier
hydrocarbons, e.g. n-dodecane, should always be
considered.

Exposure [k ppm * h]

0 20 40 60 80 100 120 140 160 180

p
e

rm
e

a
b

ili
ty

c
h

a
n

g
e

[
%

]

100

200

300

400

Figure 8: Permeability changes against BTX exposure

NEMESTÓTHY

Hungarian Journal of Industry and Chemistry

2**(2-0) design; MS Residual=40.32141

> 100
< 96
< 76
< 56
< 36

Figure 9: The effect of n-dodecane

4. Conclusion

The long-term stability of polyimide gas separation
membranes was tested against various pollutants: H2S, a
BTX mixture and n-dodecane. These compounds
significantly affected the nitrogen permeability of the
membranes which were described by using a special
parameter of exposure. It was found that H2S and the
BTX mixture increased the permeability, while n-
dodecane reduced the permeability of the membranes.
Further investigations are planned to investigate the
effect of other pollutants, moreover, to determine the
permeability of additional gases, e.g. carbon dioxide,
methane, etc.

Acknowledgement

We acknowledge the financial support of Széchenyi
2020 under the EFOP-3.6.1-16-2016-00015 project.
This research was supported by the János Bolyai
Research Scholarship of the Hungarian Academy of
Sciences.

REFERENCES

[1] Glater, J.; Hong, S.K.; Elimelech, M.: The search
for a chlorine-resistant reverse osmosis membrane,
Desalination, 1994 95(3), 325-345 DOI:
10.1016/0011-9164(94)00068-9

Exposure [ k ppm * h]

0 50 100 150 200

p
e

rm
e

a
b

ili
ty

c
h
a

n
g

e
[

%
]

Figure 10. Change in permeability against exposure to
n-dodecane

[2] Zhang, Y.; Zhao, C.; Yan, H.; Pan, G.; Guo, M.;
Na, H.; Liu, Y.: Highly chlorine-resistant
multilayer reverse osmosis membranes based on
sulfonated poly(arylene ether sulfone) and
poly(vinyl alcohol), Desalination, 2014 336, 58-63
DOI: 10.1016/j.desal.2013.12.034

[3] Gohil, J.M.; Suresh, A.K.: Chlorine attack on
reverse osmosis membranes: Mechanisms and
mitigation strategies, J. Membr. Sci., 2017 541,
108-126 DOI: 10.1016/j.memsci.2017.06.092

[4] Huang, X.; Pu, Y.; Zhou, Y.; Zhang, Y.; Zhang,
H.: In-situ and ex-situ degradation of sulfonated
polyimide membrane for vanadium redox flow
battery application, J. Membr. Sci., 2017 526, 281-
292 DOI: 10.1016/j.memsci.2016.09.053

[5] Lapicque, F.; Belhadj, M.; Bonnet, C.; Pauchet, J.;
Thomas, Y.: A critical review on gas diffusion
micro and macroporous layers degradations for
improved membrane fuel cell durability, J. Power
Sources, 2016 336, 40-53 DOI:
10.1016/j.jpowsour.2016.10.037

[6] Uhlmann, D.; Smart, S.; Diniz da Costa, J.C.: H2S
stability and separation performance of cobalt
oxide silica membranes, J. Membr. Sci., 2011
380(1), 48-54 DOI: 1016/j.memsci.2011.06.025

[7] Kraftschik, B.; Koros, W.J.; Johnson, J.R.; Karvan,
O.: Dense film polyimide membranes for
aggressive sour gas feed separations, J. Membr.
Sci., 2013 428, 608-619 DOI:
10.1016/j.memsci.2012.10.025