0 Bewertungen0% fanden dieses Dokument nützlich (0 Abstimmungen)
13 Ansichten3 Seiten
Separation of methanol-hydrogen mixtures using inorganic membranes. Separation factors between 1. And 4 depending on the total pressure at 473 K. Results indicate that capillary condensation in the membrane pores is not the primary separation mechanism as had been previously reported.
Separation of methanol-hydrogen mixtures using inorganic membranes. Separation factors between 1. And 4 depending on the total pressure at 473 K. Results indicate that capillary condensation in the membrane pores is not the primary separation mechanism as had been previously reported.
Separation of methanol-hydrogen mixtures using inorganic membranes. Separation factors between 1. And 4 depending on the total pressure at 473 K. Results indicate that capillary condensation in the membrane pores is not the primary separation mechanism as had been previously reported.
123 Note Separation of methanol-hydrogen mixtures using inorganic membranes Richard D. Noble, J ohn L. Falconer, Meng-Dong J ia and Tracy W. Perkins University of Colorado, Chemical Engineering Department, Boulder, CO 80309-0424 (USA) Abstract In a previous paper, J. Membrane Sci., 60 (1991) 185-193, large separation factors (as high as 600) were reported for methanol/hydrogen mixtures. These reported results are not accurate. Repetition of the experiments gave separation factors between 1.5 and 4 depending on the total pressure at 473 K. The same experimental apparatus and alumina membranes were used in this study. These results indicate that capillary condensation in the membrane pores is not the primary separation mechanism as had been previously reported. Keywords: ceramic membranes; diffusion; gas and vapor permeation; gas separation; inorganic mem- branes; separation of hydrogen/methanol; alumina membranes Previously, large separation factors were re- ported for methanol/hydrogen mixtures using small diameter pore (4 nm) alumina mem- branes [ 11. I t was discovered that this data was fabricated by one of the authors (D.P. Sperry) and, thus, the results are not accurate. To cor- rect this situation, the experiments were re- peated with the same experimental apparatus and alumina membranes. The apparatus and experimental apparatus are described in the previous paper [ 11. The experimental conditions are listed on Fig. 1. Each experimental condition was re- peated 10 times. This figure shows the experi- mental results for the permeance of both meth- anol and hydrogen through the membrane. The total membrane thickness is also listed so that a permeability can be computed. The perme- ante is a better measure in this case since a thin perm-selective layer dictates the separation and the rest of the thickness is primarily for me- chanical support. I t should be noted that these experiments are difficult to perform mainly due to the maintenance of a negligible pressure dif- ference across the membrane. Even a small pressure difference across the membrane was very difficult to maintain due to the rapid per- meation of each solute. This is evident in the 0376-7388/93/$06.00 0 1993 Elsevier Science Publishers B.V. All rights reserved. 124 2 A I A A A 8 1 ) . A A A A @! I ! 5bo1&01doo2ci oo25003ooo35004ooo45005000 1 I Pressure (kPa) Mc~banol Vqor Flow Rate: 2.0I IO-6m3(S.T.P.)/s IMeoHAH2 Iiydrogcn Flow Rate: 3.0 II IO 6mYS.T.P.)/s TOM Membrane Thickness: 1.25mm Fig. 1. Permeance vs. total feed pressure ( T= 473 K ) . scatter in the data with feed pressure in Fig. 1. sure is shown in Fig. 2. The separation factor A representative error bar is shown on one data varies from 1.5 to 4. Again, a representative er- point to indicate the error associated with the ror bar is shown on one data point. These low average values reported on the Figure. separation factors indicate that capillary con- A plot of separation factor vs. total feed pres- densation of methanol is not the prima@ sep- 45 8 4 9 3.5. 5 n =_I m . m m 2 1 . m 1.5 0 560 1600 1500 2000 2500 3000 3500 4WO 4500 5 Pressure (kPa) Do Fig. 2. MeOH/H, separation factor vs. total feed pressure ( T=473 K) I 125 aration mechanism, as previously reported. The tors, which are 6 to 16 times the Knudsen value. ideal separation factor-is 0.25, based on Knud- sen diffusion. Surface flow of methanol proba- bly contributed to the increased separation fac- Reference 193. 1 D.P. Sperry, J.L. Falconer and R.D. Noble, Methanol- hydrogen separation by capillary condensation in in- organic membranes. J. Membrane Sci.. 60 (1991) 185- Editors Note Professor Noble contacted me as soon as he en- countered difficulty in reproducing the results cited here. Mr. Sperrys actions were referred to the University of Colorado Committee on Research Misconduct. He is no longer a stu- dent at the University of Colorado. I t is always unpleasant to confront such mat- ters, but I believe Professor Nobles and Pro- fessor Falconers actions in this situation were exemplary. W.J. KOROS Editor-in Chief
Macromolecular Microsymposium — 16: Main Lectures Presented at the Sixteenth Microsymposium on Macromolecules (Advances in Scattering Methods), Prague, 12 - 16 July 1976