Vinylidene chloride copolymer-based carbon molecular sieve adsorbent compositions and processes therefor

Abstract

Novel carbon molecular sieve (CMS) compositions comprising carbonized vinylidene chloride copolymer having micropores with an average micropore size ranging from 3.0 to 5.0. These materials offer capability in separations of gas mixtures including, for example, propane/propylene; nitrogen/methane; and ethane/ethylene. Such may be prepared by a process wherein vinylidene chloride copolymer beads, melt extruded film or fiber are pretreated to form a precursor that is finally carbonized at high temperature. Preselection or knowledge of precursor crystallinity and attained maximum pyrolysis temperature enables preselection or knowledge of a average micropore size, according to the equation ?=6.09+(0.0275?C)?(0.00233?T), wherein ? is the average micropore size in Angstroms, C is the crystallinity percentage and T is the attained maximum pyrolysis temperature in degrees Celsius, provided that crystallinity percentage ranges from 25 to 75 and temperature in degrees Celsius ranges from 800 to 1700. The beads, fibers or film may be ground, post-pyrolysis, and combined with a non-coating binder to form extruded pellets, or alternatively the fibers may be woven, either before or after pre-treatment, to form a woven fiber sheet which is thereafter pyrolyzed to form a woven fiber adsorbent.

Claims

1. A molecular sieve composition comprising unsupported carbonized polyvinylidene chloride copolymer prepared by a process comprising pyrolyzing an unsupported polyvinylidene chloride polymer precursor having a precursor crystallinity percentage at an attained maximum pyrolysis temperature, wherein the molecular sieve composition has micropores having an average micropore size ranging from 3.0 ? to 4.3 ?, wherein the average micropore size of the molecular sieve composition further characterized according to the equation
?=6.09+(0.0275?C)?(0.00233?T) wherein ? is the average micropore size in Angstroms, C is the precursor crystallinity percentage, and T is the attained maximum pyrolysis temperature, wherein the precursor has a crystallinity percentage, as measured by differential scanning calorimetry, ranging from 25 to 75%.

2. The molecular sieve composition of claim 1, wherein the attained maximum pyrolysis temperature in degrees Celsius is from 1000 to 1500.

3. The molecular sieve composition of claim 2 being further characterized as having an average micropore volume according to the equation
V=0.346+0.00208?C?0.000152?T wherein V is the average micropore volume in milliliters per gram.

4. The molecular sieve composition of claim 2 wherein the polyvinylidene chloride polymer precursor is pre-treated, prior to pyrolysis, to at least 10 percent dehydrochlorinate it, to form a pre-treated precursor.

5. The molecular sieve composition of claim 4 wherein the polyvinylidene chloride copolymer precursor is prepared by polymerization or melt-extrusion to form a precursor bead, a precursor film or a precursor fiber, the polyvinylidene chloride copolymer precursor optionally comprising vinylidene chloride and at least one additional monomer selected from a vinyl monomer, a vinyl chloride monomer, an acrylate monomer, a methacrylate monomer, a styrenic monomer, acrylonitrile, methacrylonitrile, itaconic acid, chlorotrifluoroethylene, and mixtures thereof, and the melt-extrusion being optionally carried out at a stretch ratio from 1 to 8, the precursor bead, the precursor film or the precursor fiber optionally having an average thickness or average cross-sectional diameter or average width, as applicable, ranging from 10 micrometers to 1000 micrometers.

6. The molecular sieve composition of claim 4 wherein the pre-treatment is carried out at a temperature ranging from 100? C. to 180? C. for a time ranging from 1 hour to 48 hours, and optionally, in the case of the precursor film or the precursor fiber, applying tension, simultaneously with the pre-treatment, in an amount from 0.01 megapascals to 10 megapascals.

7. The molecular sieve composition of claim 5 further comprising weaving the precursor fiber to form a precursor woven sheet, prior to pyrolysis; and pre-treating the precursor woven sheet either before or after weaving; wherein the pyrolysis of the precursor woven sheet forms a woven fiber adsorbent being characterized as having voids representing from 10 percent to 50 percent, as measured by mercury porosimetry, of the total volume thereof.

8. The molecular sieve composition of claim 1 further comprising grinding the pyrolyzed molecular sieve composition to form particles having an average size ranging from 10 micrometers to 1000 micrometers; combining the particles with at least a non-coating binder and water to form a pellet precursor material; and pelletizing the pellet precursor material to form a structured pellet adsorbent.

9. The molecular sieve composition of 1 being used to separate a mixture selected from (a) propane and propylene; (b) nitrogen and methane; (c) ethane and ethylene; (d) carbon dioxide and nitrogen; and (e) n-butane and iso-butane; the molecular sieve composition having been prepared such that it has an average micropore size that is suitable to enable separation of the selected mixture.

10. A process for separating two gases in a mixture thereof, comprising contacting a mixture of two gases, wherein at least one gas has a representative molecular diameter ranging from 3.0 ? to 5.0 ?, and the molecular sieve composition of claim 1, under conditions suitable to adsorb, in the micropores of the molecular sieve composition, at least 5 weight percent of the at least one gas, under conditions such that the at least 5 weight percent of the at least one gas is separated from the other gas; and then desorbing the at least one gas.

Description

EXAMPLES 1-17 AND COMPARATIVE EXAMPLES 1-16

(1) A series of example (ES) and comparative (CS) samples are prepared from polyvinylidene chloride that is copolymerized with a monomer selected from methyl acrylate (MA), ethyl acrylate (EA), or butyl acrylate (BA), the monomer being present in each case in an amount shown in TABLE 2. In each case the copolymerization is accomplished by suspension polymerization. In general this includes mixing the selected monomers according to their weight ratio with a polymerization initiator and then carrying out the polymerization reaction in a water dispersion. The copolymer powder is then dried to remove water and any unreacted monomers. The powder is then sieved and the 30-50 U.S. mesh portion thereof is selected, to ensure uniformity, for the CMS preparation.

(2) The precursor powder is then dehydrochlorinated to pre-treat it, at a temperature of 130? C. for 24 hr, followed by 150? C. for 24 hr, in an oven purged by 2 liters per minute (L/min) of air.

(3) Following the dehydrochlorination pre-treatment, the pre-treated powder is then pyrolyzed in a three-stage pyrolysis procedure. The first stage includes loading 300 gram (g) samples of vinylidene chloride resin (copolymer) into a low temperature oven. A scrubber connected to this oven contains a 10 wt % sodium hydroxide aqueous solution. The loaded oven is first heated at a ramp rate of 1? C./min to 130? C. and held for 24 hr, then heated at a ramp rate of 1? C./min to 150? C. and held for 24 hr under 2 L/min of air purge, before cooling to an ambient temperature.

(4) Next, the second stage of pyrolysis includes loading the precursor powder into a cubic foot retort furnace. A scrubber connected to this furnace contains a 10 wt % sodium hydroxide aqueous solution. The loaded furnace is heated to 650? C. at a ramp rate of 5? C./min and held for 15 min, under 2 L/min of nitrogen, before cooling to an ambient temperature.

(5) A third stage of pyrolysis is then carried out in a graphite furnace. Samples of the precursor powder (10 g each) from the second stage processing in the retort furnace are loaded, in turn, in a graphite boat measuring 4 inches by 4 inches by 0.5 inch (4?4?0.5). The boat containing each sample is heated according to the conditions shown for the inventive and comparative samples in TABLE 2, with a 10 L/min of nitrogen purge (one volume turnover every 12 min). After completion of the third stage of pyrolysis for each, the furnace is cooled at a ramp rate of 10? C./min to 450? C., below which the furnace is allowed to cool to ambient temperature at a slower rate due to the heat transfer limitations.

(6) TABLE 2 also shows the properties of the CMS compositions formed based upon six process variables: 1) Attained maximum pyrolysis temperature; 2) Hold time at attained maximum pyrolysis temperature; 3) Ramp rate to attained maximum pyrolysis temperature; 4) Comonomer type; 5) Comonomer content; and 6) Precursor crystallinity. The overall micropore volume is measured using a N.sub.2 BET t-plot method, which is typically used in the art.

(7) The average micropore size, alternatively termed the effective micropore size or the average effective micrropore size, is also measured, using a kinetic adsorption method using multiple probe molecules. To estimate the effective micropore size of each CMS adsorbent, first, all pairs of the probe gases with selectivity higher than 10 are determined for each CMS adsorbent. For each gas pair having a selectivity greater than or equal to 10, the smallest molecule rejected and the largest molecule adsorbed are selected as the defining molecule pair. Then, the average of this defining molecule pair's representative molecular diameters is taken to be the effective micropore size of that particular CMS adsorbent.

(8) For example, the smallest and the largest gas molecules that are rejected and accepted by the Example 1 (EX 1) adsorbent are C.sub.3H.sub.8 (4.3 ?) and C.sub.2H.sub.6 (4.1 ?), respectively. Consequently, the effective micropore size of the EX 1 adsorbent is estimated and understood to be 4.2 ?.

(9) TABLE-US-00002 TABLE 2 CMS preparation parameters and properties Comon- Sample Hold omer Crystal- Micropore Effective (EX or Temp. Time** Ramp rate Comon- content linity*** volume micropore CS #) [? C.] [min] [? C./-min] omer [mol %] [%] [mL/g] size[?] CS 1 1075 30 6 MA 8.4 53 0.309 5.8 CS 2 950 30 6 EA 6.7 63 0.342 5.8 CS 3 700 30 2 BA 5 70 0.334 5.8 CS 4 950 60 10 BA 5 70 0.338 5.8 CS 5 1075 60 2 EA 5 71 0.313 5.8 CS 6 950 0 2 MA 8.4 53 0.260 5.8 CS 7 1200 30 10 MA 5 71 0.383 5.8 CS 8 700 60 10 EA 8.4 46 0.340 5.8 CS 9 1200 0 2 EA 6.7 63 0.318 5.2 CS 10 825 30 10 BA 8.4 37 0.271 5.2 EX 1 1200 60 6 BA 8.4 37 0.246 4.2 CS 11 825 60 2 MA 6.7 65 0.363 5.8 CS 12 1075 0 10 BA 6.7 45 0.293 5.2 CS 13 700 0 6 MA 6.7 65 0.371 6.2 CS 14 825 0 6 EA 5 71 0.367 5.8 ES 2 1250 30 6 EA 8.4 46 0.292 4.6 ES 3 1025 30 6 BA 8.4 37 0.276 4.2 ES 4 1200 30 6 BA 8.4 37 **** 4.2 ES 5 1575 30 6 EA 6.7 63 **** 4.2 ES 6 1450 30 6 EA 6.7 63 0.242 4.2 CS 15 1200 30 6 MA 6.7 65 0.324 5.2 ES 7 1450 30 6 BA 8.4 37 **** 3.4 ES 8 1575 30 6 MA 8.4 53 0.204 4.2 ES 9 1450 30 6 MA 5 71 0.300 4.6 ES 10 1700 30 6 BA 6.7 45 **** 3.5 ES 11 1200 30 6 BA 8.4 37 0.204 4.0 ES 12 1700 30 6 MA 5 71 0.176 4.2 ES 13 1325 30 6 BA 6.7 45 0.249 4.2 ES 14 1700 30 6 EA 8.4 46 **** 3.3 ES 15 1325 30 6 MA 8.4 53 0.312 4.6 ES 16 1575 30 6 BA 5 70 0.182 3.7 CS 16 1200 30 6 EA 5 71 0.346 5.2 ES 17 1325 30 6 EA 5 71 0.317 4.6 *Attained maximum pyrolysis temperature **Hold time at attained maximum pyrolysis temperature *** Crystallinity of precursor (i.e., of pre-pyrolysis composition) ****Micropore volume too low to measure by N.sub.2 BET method.

EXAMPLE 18

(10) Four (4) example samples (EX), denoted as ES 1, ES 6, ES 8 and ES 13, shown in TABLE 2 as having an effective micropore size of 4.2 ?, are used for an experiment to compare high throughput kinetic adsorption in separations of propylene and propane. To calculate selectivity, the formula shown in Equation (3) is used.

(11) $\begin{array}{cc}\mathrm{Alpha}\text{-}\mathrm{PD}=\frac{t_{0.5C3H8}}{t_{0.5C3H6}}?\frac{?P_{C3H6}}{?P_{C3H8}}& \left(\mathrm{Equation}3\right)\end{array}$
In the equation ?P represents the pressure drop (from the 45 psi, 0.31 MPa, starting pressure to the equilibrium pressure) due to adsorption, which is proportional to the amount of adsorption according to the ideal gas law. The half time adsorption (t0.5) represents the time at which 50% of the pressure drop (adsorption) happens, which corresponds to the diffusion speed. The selectivity (Alpha-PD) is defined in the following equation to take into consideration both the equilibrium and kinetic selectivities. Results for the four example samples tested are shown in TABLE 3.

(12) TABLE-US-00003 TABLE 3 C.sub.2H.sub.4/C.sub.2H.sub.6 kinetic adsorption summary of CMS samples: ES 1, 6, 8 and 13 ? P t0.5 ? P t 0.5 Micropore C3H6 C3H6 C3H8 C3H8 Alpha- volume Sample [psi*] [min] [psi*] [min] PD [cm.sup.3/g] ES 1 33.1 4.4 5.3 100 142 0.246 ES 6 27.3 4.9 12.7 53 23 0.242 ES 8 27.8 6.3 5.8 95 72 0.204 ES 13 27 5.4 7.3 60 41 0.249 *1 psi = approximately 0.007 MPa

(13) It is noted that researchers have previously believed that Zeolite 4A, with an effective micropore size of about 4.2 ?, offered the best potential for use in propylene/propane separations. However, the micropore volume of Zeolite 4A is known to be 0.20 mL/g. See, for example, Da Silva F. A., Rodrigues A. E., Adsorption Equilibria and Kinetics for Propylene and Propane over 13? and 4 A Zeolite Pellets, Ind. Eng. Chem. Res. (1999) 38, 2051-2057, which is incorporated herein by reference in its entirety. Thus, certain embodiments of the present invention may offer significant increases in micropore volume, and higher volume generally results in higher throughput. It is also noted that effective micropore size refers to pore sizes that are effective to result in separation, but such may offer greater or lesser rates of diffusion dependent in part upon geometry.

EXAMPLE 19

(14) An ethylene/ethane selectivity measurement is carried out using ES 11 in a high throughput kinetic adsorption. The results of this separation are shown in TABLE 4. These results show that, not only ethane adsorbs to a much lesser extent in the inventive CMS than ethylene, due to the pressure drop resulting from the adsorption, it also adsorbs much more slowly, by a factor of about 10. Thus, the two molecules can be easily and effectively separated using the inventive compositions.

(15) TABLE-US-00004 TABLE 4 C.sub.2H.sub.4/C.sub.2H.sub.6 kinetic adsorption summary of CMS sample: ES 11 ? P C2H4 t0.5 C2H4 ? P C2H6 t 0.5 C2H6 [psi*] [min] [psi*] [min] Alpha-PD[?] 31.3 3.4 16.6 27.8 15 *1 psi = approximately 0.007 MPa

EXAMPLE 20

(16) A nitrogen/methane selectivity measurement is carried out using ES 16 in a high throughput kinetic adsorption. The results of this separation are shown in TABLE 5. These results show that, although the nitrogen adsorbs to a lesser extent than methane, due to the pressure drop resulting from the adsorption, it adsorbs almost 40 times faster than the methane. Thus, the inventive CMS composition provides an effective kinetic separation of these two molecules.

(17) TABLE-US-00005 TABLE 4 N.sub.2/CH.sub.4 kinetic adsorption summary of CMS sample: ES 16 ? P N2 t0.5 N2 ? P CH4 t 0.5 CH4 Alpha-PD [psi*] [min] [psi*] [min] [?] 9.4 0.5 15.9 18.3 22 *1 psi = approximately 0.007 MPa

EXAMPLE 21

(18) A selectivity measurement of propane (representative of a linear chain paraffin) and iso-butane (representative of a branched chain paraffin) is carried out using ES 15 in a high throughput kinetic adsorption. The results of this separation are shown in TABLE 6. These results show that the propane adsorbs both to a greater extent and also almost 4 times faster than the iso-butane. Thus, the inventive CMS composition provides an effective separation of these two molecules. Furthermore, it is noted that, because the micropore volume of this CMS (i.e., 0.312 mL/g) is significantly higher than that of Zeolite 5A (i.e., 0.198 mL/g), currently being utilized for certain commercial linear/branched chain paraffin separations, such as n-butane/i-butane separations, the inventive CMS may offer a comparative and significant advantage.

(19) TABLE-US-00006 TABLE 6 C.sub.3H.sub.8/i-C.sub.4H.sub.10 kinetic adsorption summary of CMS sample: ES 15 Micropore ? P C3H8 t0.5 C3H8 ? P i-C4H10 t 0.5 i-C4H10 Alpha-PD volume [psi*] [min] [psi*] [min] [?] [cm.sup.3/g] 25.5 5.3 0.8 28 168 0.312 *1 psi = approximately 0.007 MPa

EXAMPLES 22-24

(20) Three exemplary melt extruded copolymer tapes, designated as ES 22, ES 23, and ES 24, comprising vinylidene chloride monomer and, as comonomers therewith, methyl acrylate (MA) 4.8 wt %, MA 8.5 wt % or vinyl chloride (VC) 17.6 wt %, respectively, are prepared. Approximately 5 g of PVDC tapes are laid down on aluminum (Al) pans and allowed to shrink freely during the pre-treatment step. The PVDC films are kept in a one-cubic-foot oven purged by approximately 10 L/min of air. The oven temperature is raised to 130? C. with a 1? C./min ramp and held for 24 hr, then raised to 150? C. with a 1? C./min ramp rate and held for another 24 hr, before cooling to an ambient temperature. The crystallinity of each of the tapes is shown in TABLE 7. As compared to the crystallinity of the as-polymerized resins in TABLE 2, the crystallinity in the tapes after melt extrusion is reduced by an amount ranging from 10% to 30%.

(21) TABLE-US-00007 TABLE 7 Crystallinity of various PVDC precursors Precursor Crystallinity # Copolymer (morphology) [%] ES 22 MA 4.8 wt % (melt extruded tape) 50 ES 23 MA 8.5 wt % (melt extruded tape) 38 ES 24 VC 17.6 wt % (melt extruded tape) 35

(22) Samples of approximately 2 g each of pre-treated films are loaded into a one-inch diameter quartz tube furnace. The tube furnace loaded with the resin samples is then raised from 550? C. to 1000? C. at a ramp rate of 5? C./min or 10? C./min, respectively, to finish the HCl releasing reaction. The carbonized film obtained from the first step of pyrolysis is then put into an ASTRO? furnace (ASTRO is a trademark of Astro Thermal Tec Ltd.) with electrical heating, water cooling, and argon (Ar) purging. The temperature is raised from 1000? C. to the final temperature (i.e., attained maximum temperature) of 1500? C. at a 10? C./min ramp, and held at the final temperature for 15 min. The pyrolysis conditions are shown in the names of samples as follows: ramp-final temperature-hold time. Thus, for example, the notation 5C-1000-15 min defines a 5? C./min ramp, final temperature of 1000? C., and hold time of 15 min during pyrolysis.

(23) The high throughput kinetic adsorption is performed in a high throughput reactor (HTR) system installed in a triple dry box. Adsorbate gases (propylene C.sub.3H.sub.6 and propane C.sub.3H.sub.8) can be injected into each cell at a controlled pressure and temperature. The kinetic adsorption measurements are performed in the following sequence: (1) Load approximately 0.5 g of CMS sample into the 14 cm.sup.3 high throughput cells; (2) de-gas at 140? C. for 4 hr by a semi-continuous N.sub.2 purge; (3) introduce the C.sub.3H.sub.6 gas to a pressure of 45 pounds per square inch (psi, 0.31 MPa) and monitor the pressure drop for 8 hr at 35? C.; (4) de-gas at 140? C. for 24 hr by N.sub.2 purge; and (5) introduce the C.sub.3H.sub.8 gas to 45 psi (0.31 MPa) pressure and monitor the pressure drop for 96 hr at 35? C.

(24) TABLE 8 shows the high throughput kinetic adsorption results on the pyrolyzed carbon tapes. Because each of the adsorption cells contains (1) the same amount of CMS adsorbent (0.5 g); (2) the same volume (14 milliliters, mL); and (3) is pressurized to the same 45 pounds per square inch (psi, 0.31 MPa) initial pressure; the pressure drop (?P) is, therefore, an indicator of the amount of gas adsorbed by the adsorbent. The results show that, for the CMS adsorbents prepared from the same precursor type, ?P of C.sub.3H.sub.6 and C.sub.3H.sub.8 is approximately the same at low pyrolysis temperature. ?P of both adsorbate gases increases slightly and then decreases as the pyrolysis temperature increases. When pyrolyzed at 1500? C., the CMS film from the MA 4.8 wt % precursor shows C.sub.3H.sub.6/C.sub.3H.sub.8 selectivity as high as 50. Also, when pyrolyzed at above 1000? C., the CMS films from both MA 8.5 wt % and VC 17.6 wt % precursor show some C.sub.3H.sub.6/C.sub.3H.sub.8 selectivity.

(25) From the above results, then, it may be concluded that propylene/propane selective materials may be more conveniently prepared in the form of melt-extruded tapes than in the form of as-polymerized resins because the melt-extruded tapes may be pyrolyzed at a lower temperature than the resins to achieve comparable selectivity. The final pyrolysis temperature for the MA 4.8 wt % resin decreases from approximately 1700? C. for as-polymerized resin to approximately 1500? C. in the melt-extruded tape. Similarly, the final pyrolysis temperature for the MA 8.6 wt % resin decreases from approximately 1300? C. for as-polymerized resin to approximately 1000? C. in the melt extruded tape. This shows that melt extrusion reduces crystallinity of PVDC copolymers, which in turn enables the formation of propylene/propane selective micropores at lower pyrolysis temperatures.

(26) TABLE-US-00008 TABLE 8 C.sub.3H.sub.6/C.sub.3H.sub.8 kinetic adsorption summary of CMS films Tape-Composition-Ramp rate-Pyrolysis ? P C3H6 t0.5 C3H6 ? P C3H8 t 0.5 C3H8 Alpha-PD temperature-Pyrolysis time [psi**] [min] [psi**] [min] [?] SARAN* Tape-MA 4.8 wt %-5? C.-850-15 min 17 7 16 8 1 SARAN Tape-MA 4.8 wt %-10? C.-1000-15 min 18 5 17 5 1 SARAN Tape-MA 4.8 wt %-10? C.-1200-15 min 20 3 19 5 2 SARAN Tape-MA 4.8 wt %-10? C.-1500-15 min 18 19 6 300 50 SARAN Tape-MA 8.5 wt %-5? C.-800-15 min 16 5 14 4 1 SARAN Tape-MA 8.5 wt %-5? C.-900-15 min 17 4 14 4 1 SARAN Tape-MA 8.5 wt %-5? C.-1000-15 min 17 3 14 38 18 SARAN Tape-MA 8.5 wt %-5? C.-1000-60 min 16 3 13 36 18 SARAN Tape-VC 17.6 wt %-10? C.-1000-15 min 14 3 13 3 1 SARAN Tape-VC 17.6 wt %-10? C.-1200-15 min 6 67 3 246 7 SARAN Tape-VC 17.6 wt %-10? C.-1500-15 min 2 125 1 450 6 *SARAN is a trademark of The Dow Chemical Company **1 psi = approximately 0.007 MPa

EXAMPLES 25-41

(27) Two types of 0.28 millimeter (mm) diameter CMS fibers (obtained from SATTI?, Germany, and denoted as Doll Hair due to the fact that the largest application for the fiber is hair used in children's toy dolls) are pretreated according to two methods: Method A, wherein the fiber is maintained at constant length during the pre-treatment step, and Method B, wherein the fiber is allowed to shrink freely during the pre-treatment step. Both methods are carried out under the same temperature profiles as shown in Example 1. It is noted that the fiber always breaks in the middle of pre-treatment when attempts are made to maintain constant length. The broken fibers then shrink similarly to those left to shrink freely. The crystallinity of the precursor fibers is shown in TABLE 9.

(28) TABLE-US-00009 TABLE 9 Crystallinity of the 0.28 mm PVDC fiber Crystallinity Copolymer (morphology) [%] MA 4.8 wt % (0.28 mm fiber) 47

(29) The fibers are then subjected to pyrolysis at different temperatures. TABLE 10 shows the kinetic adsorption results of the fibers corresponding to pyrolysis at each temperature. There is a pyrolysis temperature window ranging from approximately 850? C. to 1000? C. for both types of CMS fibers to reach an optimal C.sub.3H.sub.6/C.sub.3H.sub.8 selectivity. The temperature window is significantly lower than the 1500? C. pyrolysis temperature that is required for the CMS film, despite the fact that both fibers and film are prepared from the same MA 4.8 wt % precursor. The major difference between the film and fibers is in their respective crystalline morphologies, which depend heavily upon the melt extrusion process employed. The 4.8 wt % MA film is extruded with a negligible amount of stretch applied and left to crystallize slowly over an approximately two-day timeframe. In contrast, when a stretch ratio as high as 4 is used to melt spin fibers, the crystallization process happens in the timeframe of a few seconds. It is noted that the crystallites in the stretched fiber are generally much smaller and more highly aligned than those in films, even though the crystallinity levels in the film and fiber precursors are approximately the same, as shown in TABLE 7 and TABLE 9. Without wishing to be bound by any theory, it is speculated that the increased alignment of crystallites in the precursor, due to stretching, may reduce crosslinking and, upon pyrolysis, increase graphitization, which may in turn result in the need for, or tolerance of, a lower pyrolysis temperature. The lower pyrolysis temperature for CMS fiber generation also results in a more economical pyrolysis production process for scale-up purposes.

(30) TABLE-US-00010 TABLE 10 C.sub.3H.sub.6/C.sub.3H.sub.8 high throughput kinetic adsorption summary for CMS fibers Method-Morphology- Atmosphere-Tension/no ? P t0.5 ? P t 0.5 Ex- tension-Ramp rate- C3H6 C3H6 C3H8 C3H8 Alpha ample temperature-Pyrolysis time [psi] [min] [psi] [min] [?] ES 25 A-Doll Hair-Air-tension- 17 3 14 7 3 5 C.-550-15 min ES 26 A-Doll Hair-Air-tension- 17 4 14 6 2 5 C.-700-15 min ES 27 A-Doll Hair-Air-tension- 19 5 12 55 19 5 C.-850-15 min ES 28 A-Doll Hair-Air-tension- 19 8 12 738 138 5 C.-900-15 min ES 29 A-Doll Hair-Air-tension- 19 7 13 593 134 5 C.-950-15 min ES 30 A-Doll Hair-Air-tension- 16 20 11 483 36 5 C.-1000-15 min ES 31 A-Doll Hair-Air-tension- 19 12 4 105 38 10 C.-1100-15 min ES 32 A-Doll Hair-Air-tension- 18 26 6 228 26 10 C.-1200-15 min ES 33 A-Doll Hair-Air-tension- 3 92 2 294 4 10 C.-1500-15 min ES 34 B-Doll Hair-Air-no tension- 17 4 15 5 1 5 C.-550-15 min ES 35 B-Doll Hair-Air-no tension- 19 4 16 7 2 5 C.-700-15 min ES 36 B-Doll Hair-Air-no tension- 19 13 11 692 88 5 C.-850-15 min ES 37 B-Doll Hair-Air-no tension- 20 8 14 332 55 5 C.-900-15 min ES 38 B-Doll Hair-Air-no tension- 20 13 9 733 124 5 C.-950-15 min ES 39 B-Doll Hair-Air-no tension- 17 20 10 486 41 5 C.-1000-15 min ES 40 B-Doll Hair-Air-no tension- 18 5 9 19 7 10 C.-1100-15 min ES 41 B-Doll Hair-Air-no tension- 18 5 11 67 21 10 C.-1200-15 min

EXAMPLE 42

(31) CMS fiber is obtained by pyrolyzing a 170 ?m outside diameter (OD) vinylidene chloride/methyl acrylate copolymer fiber (SARAN?, obtained from FUGAFIL? in Germany), using the two step process described hereinbelow. Composition and crystallinity of the precursor fiber are shown in TABLE 11.

(32) TABLE-US-00011 TABLE 11 Crystallinity of the 0.17 mm PVDC fiber Crystallinity Copolymer (morphology) [%] MA 4.8 wt % (0.17 mm fiber) 51

(33) In a pre-treatment step, 100 g of PVDC fiber is preheated in a convection oven in air at 130? C. for one day and at 150? C. for another day, to form stabilized CMS precursor. The pre-treated fiber is then pyrolyzed to 550? C. under an N.sub.2 purge (5? C./min ramp, hold for 15 min at 550? C.) to complete the chemical decomposition. The CMS fiber is denoted as 0.17 mm CF, with CF meaning carbon fiber. The CMS fiber obtained is then ground to an average particle size ranging from 30 ?m to 200 ?m.

EXAMPLE 43

(34) A paste is made by manually mixing three components: 10 g of ground CMS fiber (Example 42), 1 g of METHOCEL? A4M (METHOCEL is a trademark of The Dow Chemical Company), and 10 g of deionized (DI) water. The paste is extruded through a 5 mm die and cut into approximately 5 mm short cylindrical pellets. The pellets are dried in an N.sub.2-purged oven at 50? C. overnight. The pellet is denoted as 10% Methocel-0.17 mmCF, with % representing wt %.

EXAMPLE 44

(35) CMS fiber is prepared using the protocol of Example 42, except that the paste components include 5 g of METHOCEL? A4M. Further processing is continued as in Example 43, with the final pellet sample being denoted as 50% Methocel-0.17 mmCF.

(36) The gravimetric transient adsorptions of CMS fiber and the pellet adsorbents prepared in Example 42, 43, 44 are carried out in a modified thermogravimetric analysis (TGA) instrument. Approximately 200 milligrams (mg) of CMS material is loaded into a TGA pan and heated to 90? C. (at 10? C./min ramp and hold time of 30 min) under 25 standard cubic centimeters per minute (sccm) of helium (He) purge. The He purge gas is then switched to 25 sccm of mixture gas containing 50 mole percent (mol %) of He and 50 mol % of C.sub.3H.sub.6. The weight gain of the samples (due to the adsorption of C.sub.3H.sub.6) over time is recorded. The C.sub.3H.sub.6 capacity is recorded as the percentage of weight gain at equilibrium compared to the fresh CMS samples. The half time adsorption is the time required to reach 50% of the equilibrium weight change, which is used as a parameter for the rate of kinetic adsorption. The results are shown in TABLE 12.

(37) TABLE-US-00012 TABLE 12 Propylene capacity and half time adsorption into CMS fiber and pellets Propylene Half time capacity adsorption Example Carbon Fiber (CF) composition [wt %] [min] ES 42 017 mm CF (non-pellet) 8.2 0.8 ES 43 10% Methocel-0.17 mm CF (pellet) 8.2 1.3 ES 44 50% Methocel-0.17 mm CF (pellet) 6.1 1.0

(38) Results show that a significant variation in the amount of cellulose ether (10 wt % versus 50 wt % METHOCEL?) does not significantly reduce the diffusion speed of C.sub.3H.sub.6. In fact, the diffusion into the pellet is closely similar to diffusion into the 0.17 mm CMS fiber that has not been pelletized at all. The fact that mass transport speed does not appear to be significantly affected by either the fact of pelletization, or by the proportion of binder when the CMS adsorbent has been pelletized, is unexpected. Without wishing to be bound by theory, it is speculated that the highly hydrophilic nature of the cellulose ether, i.e., the fact it is non-coating when employed in combination with the hydrophobic CMS fiber or film, may result in preservation of a void-filled interface that ensures effective mass transport.

EXAMPLES 46-47

(39) One type of SARAN? woven cloth (obtained from FUGAFIL?, Germany, denoted as SARAN cloth) is pre-treated under similar temperature profiles as described in Example 34-41. The SARAN? cloth is left to shrink freely during the pre-treatment and the subsequent pyrolysis step according to the maximum attained pyrolysis temperatures shown in TABLE 13.

(40) Testing includes TGA kinetic adsorption testing using C.sub.3H.sub.6, and C.sub.3H.sub.8. For the kinetic adsorption testing, first, approximately 200 mg of the cloth is loaded into a TGA pan and heated to 90? C. (at 10? C./min ramp and hold time of 30 min) under 25 sccm of He purge. The He purge gas is then switched to 25 sccm of mixture gas containing 50 mol % of He and 50 mol % of C.sub.3H.sub.6. Then 200 mg of the fresh sample is loaded into a TGA pan and heated to 90? C. (at 10? C./min ramp and hold time of 30 min) under 25 sccm of He purge. The He purge gas is then switched to 25 sccm of mixture gas containing 50 mol % of He and 50 mol % of C.sub.3H.sub.8. For separation based on kinetic adsorption, a cycle time to reach 50% equilibrium of the faster gas (C.sub.3H.sub.6) is desirable to maximize the selectivity. Therefore, selectivity in this gravimetric method (Alpha-G) is defined as the ratio of weight gain for C.sub.3H.sub.6 and C.sub.3H.sub.8 at the half time for adsorption of C.sub.3H.sub.6.

(41) TABLE 13 shows the kinetic adsorption results for cloth samples that have been carbonized by pyrolysis to different temperatures. There is a temperature window ranging from 1100? C. to 1200? C. to form CMS with propylene/propane selective pores, which is similar to that found in the SARAN fiber derived CMS.

(42) TABLE-US-00013 TABLE 13 C.sub.3H.sub.6/C.sub.3H.sub.8 high throughput kinetic adsorption summary of CMS Cloth Pyrolysis Propylene t0.5 temperature capacity propylene Example [? C.] [wt %] [min] Alpha-G [?] 46 1100 9.6 11.3 13 47 1200 2 15.8 100