Fast hydrogen purification through graphitic carbon nitride nanosheet membranes

Two-dimensional graphitic carbon nitride (g-C3N4) nanosheets are ideal candidates for membranes because of their intrinsic in-plane nanopores. However, non-selective defects formed by traditional top-down preparation and the unfavorable re-stacking hinder the application of these nanosheets in gas separation. Herein, we report lamellar g-C3N4 nanosheets as gas separation membranes with a disordered layer-stacking structure based on high quality g-C3N4 nanosheets through bottom-up synthesis. Thanks to fast and highly selective transport through the high-density sieving channels and the interlayer paths, the membranes, superior to state-of-the-art ones, exhibit high H2 permeance of 1.3 × 10−6 mol m−2 s−1 Pa−1 with excellent selectivity for multiple gas mixtures. Notably, these membranes show excellent stability under harsh practice-relevant environments, such as temperature swings, wet atmosphere and long-term operation of more than 200 days. Therefore, such lamellar membranes with high quality g-C3N4 nanosheets hold great promise for gas separation applications.

The peak heights of C1 and C2 carbon should be should be equal in 13 C NMR spectra, theoretically. However, the peak heights of C1 and C2 carbon are not equal, which is probably because the 13 C spectra obtained by the 1 H-13 C CP-MAS NMR technique cannot provide information regarding the relative abundance of carbon atoms in different chemical environments according to the signal strength, but rather record the diversity in CP efficiencies 1 . Actually, the signal intensity ratio varies due to different proximities to 1 H species in the 1 H-13 C cross polarization/MAS (CP-MAS) mode. The peak of C1 shows higher intensity, which because that C1 is closer to protons of nonpolymerized NH2 or partially polymerized NH species, where 13 C spin polarized NMR signals are greatly enhanced from neighboring 1 H species via a 1 H- 13  The higher zeta potential of the g-C3N4 suspension dispersed in isopropanol compared with those of g-C3N4 suspension dispersed in deionized water, demonstrates its more negatively charges, causing the stronger repulsive interactions to weaken the π-π interaction between adjacent nanosheets. membrane.
The C 1s spectrum of the g-C3N4 membrane could be divided into two peaks at 288.4 and 284.8 eV, respectively. The peak at 288.4 eV can be assigned to sp 2 C bonded to N in an aromatic ring, while the peak at 284.8 eV is corresponding to the standard reference carbon.
Besides, the N 1s spectrum of the g-C3N4 membrane can be divided into four peaks at 398. Errors bars indicate the standard deviation of three measurements. There is no obvious proportional relationship between the gas permeances and gas molecular weight.   where P is the gas permeance, A is the pre-exponential factor, Eact is the apparent activation energy, R is the ideal gas constant (8.314 J mol -1 K -1 ), and T is the absolute Kelvin temperature (K). ln(P) versus 1/T displays a straight line in which the slope was used to calculate Eact.
As shown in Supplementary Fig. 27, the act,H 2 is about 2.4 kJ mol -1 , and act,CO 2 is about 7.4 kJ mol -1 . The apparent activation energy is an association between diffusion activation energy and adsorption heat.
CO2 adsorption on g-C3N4 should release more heat than H2, considering the weak adsorption of H2 on g-C3N4. As a result, when diffusion through the g-C3N4 membrane, CO2 should also have a higher activation energy than H2. It is, therefore, inferred that CO2 diffusing through g- C3N4 membranes was a much more activated process than H2, and it was a tighter fit for CO2 in g-C3N4 flakes in high temperatures. The theory also explains why the H2/CO2 separation factor decreased with an increasing temperature. Because CO2 permeance rose faster with temperature than H2, CO2 diffusion was more activated than H2 in the g-C3N4 membrane.   Consequently, the DFT calculations predicted an excellent selectivity of H2/CO2 with 1.3 ×