FIGURE
9 A) Series of gas permeability through the tubular
MXene/SS2.5 membrane. B) Effect of electrophoresis time
on the H2/CO2 separation performance of
the membranes. Tests condition: the feeding mixed gas (50 ml: 50 ml of
H2/CO2) at 1 atm and 25oC.
In addition, the effect of the operation parameters on the gas
permeance, such as temperature and feed gas humidity, had also been
investigated. The temperature-dependent
H2/CO2 separation performance of the
tubular MXene/SS2.5 membrane was depicted in Figure 10A,
where three temperature cycles included heating and cooling processes.
When the temperature was raised from 25 oC to 125oC, the hydrogen permeance gradually was raised from
1283 GPU to 1880 GPU, while the H2/CO2selectivity decreased from 55 to 19. During the cooling stage, the
hydrogen permeance fell from 1880 GPU to 1634 GPU, and the selectivity
was raised from 19 to 50. Furthermore, the performance of the
MXene/SS2.5 membrane could still recover well after
three heating and cooling cycles, with
H2/CO2 selectivity of 51 and
hydrogen permeance of 1568 GPU ,
demonstrating the tubular MXene/SS2.5 membrane had good
thermal resistance. The apparent activation energy of the system was
illustrated in Figure 10B via Arrhenius equation. The apparent
activation energies Eact (H2) and Eact
(CO2) of the two gases were 1.61 kJ/mol and 11.04 kJ/mol, respectively.
It could be found that the apparent activation energies were all
positive, which indicated that gas diffusion rather than adsorption was
primarily responsible for controlling the gas separation process, which
was consistent with the literatures.66 Moreover, the
synthesis gas produced by the methane steam reforming process always
contains water vapor (usually < 3 vol%), so a water
vapor-containing test on the separation system was required. Different
saturated salt solutions were used to produce water vapor with different
relative humidity to explore gas permeability.4 As
shown in Figure 10C, as the relative humidity increased from 0% to
92%, the hydrogen permeance decreased from 1244 to 1168 GPU, while the
CO2 permeance slightly increased from 21 to 22 GPU, thus
the corresponding H2/CO2 selectivity
decreased from 57 to 51. Obviously, as the relative humidity of the feed
gas increased, the condensable water molecules in the feed gas condensed
in the pores or interlayer spacing of the membrane, obstructing the
transmission of non-condensable small molecular gas of
H2. Because CO2 is more soluble in
condensed water than H2, it diffused through the
membrane more quickly, reducing the selectivity of
H2/CO2, which was consistent with the
phenomenon of layered graphene at various humidity
conditions.4