Supercritical Fluid Engineering Science - American Chemical Society

---

Chapter 28

Production of Mesophase Pitch by Supercritical Fluid Extraction 1

2

Downloaded by UNIV OF ARIZONA on April 5, 2017 | http://pubs.acs.org Publication Date: December 17, 1992 | doi: 10.1021/bk-1992-0514.ch028

T. Hochgeschurtz, K. W. Hutchenson , J . R. Roebers , G.-Z. Liu, J. C. Mullins, and M . C. Thies Department of Chemical Engineering, Center for Advanced Engineering Fibers, Clemson University, Clemson, SC 29634

Supercritical fluid extraction is being investigated for the production of the mesophase pitch used to make high-performance carbonfibers.A heat-soaked, isotropic petroleum pitch has been fractionated with supercritical toluene in a region of liquid-liquid equilibrium at temperatures and pressures to 656 Κ and 140 bar, respectively. Since both phases are black and opaque at the conditions of interest, an AC impedance bridge technique was used to detect the liquid-liquid interface level in the equilibrium cell. For all tested experimental conditions, the bottom-phase fraction contained a high percentage of the mesophase necessary for producing anisotropic carbon fibers. Generalized correlations have been developed for the characteristic constants of the Peng-Robinson equation applicable to high molecular weight, aromatic hydrocarbons.

Petroleum pitch is a high molecular weight, carbonaceous material which can be produced from the by-products of the thermal and catalytic cracking of crude oil distillates. These by-products are generally considered to be waste materials by refiners because upgrading into a more useful form, such as lubricating oil, is relatively expensive. Thus, the isotropic pitch which can be produced is relatively inexpensive and readily available. Isotropic pitch has considerable potential as a raw material for the production of economical high-performance carbon fibers, which are used in the manufacture of advanced composites (7). However, the current methods for treating isotropic pitch to produce the fiber precursor, or mesophase, have deficiencies that result in high production costs and less than ideal properties (2). 1

Current address: Ε. I. du Pont de Nemours and Company, P.O. Box 80304, Wilmington, DE 19898 Current address: Bayer AG, Ingenieurbereich Anlagenplanung, D-5090 Leverkusen, Bayerwerk, Germany

2

0097-6156/93/0514-0338$06.00/0 © 1993 American Chemical Society

Kiran and Brennecke; Supercritical Fluid Engineering Science ACS Symposium Series; American Chemical Society: Washington, DC, 1992.

348

SUPERCRITICAL FLUID ENGINEERING SCIENCE

Downloaded by UNIV OF ARIZONA on April 5, 2017 | http://pubs.acs.org Publication Date: December 17, 1992 | doi: 10.1021/bk-1992-0514.ch028

Supercritical fluid (SCF) extraction is being investigated by our group for the production of mesophase pitch fractions from isotropic petroleum pitch. Previous work (3,4) focused on thefractionationof isotropic pitch with supercritical toluene in a region of vapor-liquid equilibrium at temperatures and pressures to 673 Κ and 76 bar, respectively. Although the isotropic pitch was successfully separated into narrow molecular weight fractions in this region, no more than 20% of the pitch could be extracted into the vapor phase. Since a significant fraction of the low molecular weight species in isotropic pitch must be removed before mesophase can form (5), no pitch fractions containing mesophase were isolated. In this study we report on the fractionation of a heat-soaked, isotropic pitch with supercritical toluene in a region of liquid-liquid equilibrium, which was found to exist at high pressures. Experimental Apparatus and Procedure A continuous-flow apparatus has been constructed that can be used to measure both vapor-liquid and liquid-liquid equilibrium compositions at temperatures to 673 Κ and pressures to 350 bar. The apparatus is conceptually similar to that reported by previous workers (6,7), but is capable of operating with materials which have melting points as high as 573 K. A flow apparatus is necessary for our work to produce large enough samples of pitch fractions for subsequent analysis and for spinning into carbon fibers. Such an apparatus is also needed to eliminate thermal polymerization reactions of the pitch by reducing residence times at elevated temperatures to only a few minutes. Several hours at 673 Κ are known to be required before any appreciable extent of polymerization can occur (8). Figure 1 shows a simplified schematic of the experimental apparatus, and a brief description is given below. Most of the details of the experimental apparatus and procedure have been described elsewhere (6,9,10). However, several important modifications have been made since our last paper, and these are discussed in detail. The isotropic pitch is pumped indirectly using a high-pressure cylinder equipped with an internal floating piston; the piston is displaced by the regulated flow of a working fluid (toluene, in this case). In previous work, the pitch was pumped in the molten state by maintaining the cylinder at 423 K. In this study, a 50/50 mixture by weight of pitch and toluene is charged to the high pressure cylinder. This mixture is a viscous, homogeneous liquid at ambient temperatures, so no heating of the cylinder is required. The solvent and working fluid are delivered at flow rates of approximately 290 and 250 mL/h, respectively, using LDC/Milton Roy Model 396 metering pumps. The weight ratio of solvent to pitch for this study was maintained constant at 3:1. The two streams are preheated, combined in a impingement mixer, and then further preheated before passing through a Kenics-type static mixer into the equilibrium cell, which functions as a phase separator. The pitch/solvent mixture enters the cell in the middle of the inner chamber. At the operating temperatures, pressures, and solvent-to-feed (S/F) ratio investigated in this study the mixture separates into two phases: a solvent-rich, lighter liquid phase and a pitch-rich, heavier liquid phase. Each phase is drawn off independently through top and bottom sample ports in the cell. Micrometering valves (Autoclave Engineers, 6 0 V R M M 4882-GY) are located on both lines exiting the cell. The top-phase valve is used primarily for pressure control and the bottom-phase valve for level control. After expansion to atmospheric pressure, both phases pass through on/off valves (Autoclave Engineers 60VM4071-GY) and are collected.

Kiran and Brennecke; Supercritical Fluid Engineering Science ACS Symposium Series; American Chemical Society: Washington, DC, 1992.

Kiran and Brennecke; Supercritical Fluid Engineering Science ACS Symposium Series; American Chemical Society: Washington, DC, 1992.

-

Figure 1.

T"

MICROMETERING VALVE

Simplified schematic of the experimental apparatus.

HIGH-PRESSURE CYLINDER

•ISOTROPIC PITCH

ISOTHERMAL BATH

STATIC MIXER

PREHEATING SECTION

Downloaded by UNIV OF ARIZONA on April 5, 2017 | http://pubs.acs.org Publication Date: December 17, 1992 | doi: 10.1021/bk-1992-0514.ch028

BOTTOM-PHASE SAMPLE

TOP-PHASE SAMPLE

Downloaded by UNIV OF ARIZONA on April 5, 2017 | http://pubs.acs.org Publication Date: December 17, 1992 | doi: 10.1021/bk-1992-0514.ch028

350

SUPERCRITICAL FLUID ENGINEERING SCIENCE

The sample collection scheme has been substantially modified from that previously reported. A l l sample lines and valves exiting the nitrogen bath are wrapped with electrical heating tape and insulated, and can be maintained at temperatures to 673 Κ if necessary. In this study, the solvent-rich top phase contained the lower molecular weight components of pitch and was maintained at temperatures from 373-523 K . However, the pitch-rich, bottom liquid phase contained the highest molecular weight components of pitch and had to be maintained above its melting point, i.e., at approximately 600 K. Several changes have been made to reduce the plugging of lines with high-melting solids: (1) The sample tubing between each micrometering and on/off valve is now larger (6.35 mm o.d. χ 2.11 mm i.d.); (2) the sample tubing exiting each on/off valve has an o.d. of 9.5 mm, an i.d. of 3.2 mm, and is made of aluminum to facilitate heat transfer to the tip of the tubing; (3) the aluminum tubing is inserted approximately halfway into insulated Pyrex reaction vessels. As before, samples of each phase are collected in 250-mL sample bottles placed inside the vessels. The equilibrium cell was originally designed by Roebers (10,11) for visual determination of the interface level between two fluid phases. However, for our measurements in the liquid-liquid region both phases were found to be black and opaque. After evaluating several possible methods of nonvisual interface detection, we chose an A C impedance bridge technique, which is described below, for the detection of the interface level (see Figure 2). An electrode made of 304 stainless steel was fabricated and inserted into the slot of the equilibrium cell chamber, and served as one electrode of a parallel-plate capacitor. The body of the cell served as the other electrode. The capacitor, then, consisted of these two electrodes and the two liquid phases between them. The internal electrode was insulated from the cell body with spacers made of a machinable ceramic (Macor; Corning Glass) and was connected to the external circuit with an 18-gauge electrical wire. The wire was insulated from the cell wall and sealed against pressure with a soapstone cone, compression-type electrical connector (Newport Scientific), which is rated for operation to 700 Κ and 1300 bar. The internal volume of the cell is reduced to approximately 30 ml with the electrode insert. The cell capacitor serves as one leg of an A C impedance bridge and indicates the differences in dielectric constant and electrical resistivity which exist between these two phases. The basic concept of the technique is that a change in the impedance occurs with a changing interface level and is converted to a voltage output. Details of the bridge design are presented elsewhere (72). To control the level of the liquid-liquid interface during an experimental run, the system is first brought to steady state at the desired operating conditions of temperature, pressure, and S/F ratio. The bottom-phase micrometering valve is kept sufficiently open so that there is no accumulation of the pitch-rich bottom phase in the cell (i.e., the cell only contains the solvent-rich, top liquid phase). The impedance bridge is then nulled to 0.0 ± 0 . 1 volts by adjusting the capacitors and resistors on the variable leg of the bridge. After the bridge is nulled, the bottom-phase micrometering valve is closed and the level of bottom phase is allowed to build until an output voltage of 4-10 volts is obtained on the Fluke D M M . This micrometering valve is then controlled throughout an experimental run to maintain a constant voltage output, and thus a constant liquidliquid interface level. Since the properties of the pitch-rich bottom phase (e.g., the percentage of mesophase) affect the voltage output obtained, the interface level is not precisely known. However, we have experienced little difficulty in operating the apparatus in this manner and have had no problems with cross-contamination of phases.

Kiran and Brennecke; Supercritical Fluid Engineering Science ACS Symposium Series; American Chemical Society: Washington, DC, 1992.

HOCHGESCHURTZ ET AL.

351

Production of Mesophase Pitch

Downloaded by UNIV OF ARIZONA on April 5, 2017 | http://pubs.acs.org Publication Date: December 17, 1992 | doi: 10.1021/bk-1992-0514.ch028

28.

<

8 73

Ο

I §·

1 ε

.

«

ο

ri

c

(S .1

Kiran and Brennecke; Supercritical Fluid Engineering Science ACS Symposium Series; American Chemical Society: Washington, DC, 1992.

SUPERCRITICAL FLUID ENGINEERING SCIENCE

Downloaded by UNIV OF ARIZONA on April 5, 2017 | http://pubs.acs.org Publication Date: December 17, 1992 | doi: 10.1021/bk-1992-0514.ch028

352

Although both phases in the cell are black and opaque, their appearances during sample collection are distincdy different. The solvent-rich top phase, which was found to contain only isotropic pitch, is collected at near ambient temperatures and has a viscosity similar to that of liquid toluene. Upon exiting the micrometering valve, the pitch-rich bottom phase flashes, removing most of the toluene, and the pitch is collected as a porous solid at approximately 473 K . A number of safety features were incorporated into the design of the apparatus and are required for working with flammable supercritical solvents. A n inert atmosphere of less than 1 mol % oxygen is maintained inside the isothermal bath to prevent a deflagration should a solvent leak occur. A polycarbonate shield is mounted over the observation window of the bath. A n independent temperature controller serves as a high-temperature shutoff. Rupture disks are used to protect both the equilibrium cell and the high-pressure cylinder from overpressure. Finally, power to the bath heaters is automatically shut off if the system pressure drops below some preset value.

Experimental Measurements The cell pressure was measured with a Heise Bourdon-tube gauge (0-2500 psi range) which had been calibrated with a Budenberg dead weight pressure gauge tester to 0.05% of the indicated pressure. The system pressure in the liquid-liquid region was controlled to within ± 50 psi, giving an experimental uncertainty of ± 50 psi. The equilibrium temperature of the combined feed stream entering the cell and of the top and bottom phases in the cell were measured with Type Κ thermocouples referenced to an aluminum block located in the isothermal bath. The temperature of the reference block was measured with a secondary standard RTD (Burns Engineering). Both the thermocouple voltages and RTD resistance were measured with a Keithley Model 191 D M M . The RTD and D M M were previously calibrated as a unit to ±0.05 Κ using a Rosemount 162CE primary temperature standard and a Mueller resistance bridge. The reported values present average temperatures measured during the course of an experiment and are believed to be accurate to ±0.5 K. Multiple samples of each phase were collected in 250-mL glass jars for a time period of 20-30 min each. The weight percent of toluene in the top phase was determined by drying the collected samples in a vacuum oven (National Appliance, Model 5851) for 12-14 h at approximately 450 K . Subsequent analysis of the dried samples by gel permeation chromatography (GPC) confirmed that no residual toluene was present. Since the top-phase samples were collected at near ambient temperatures, no correction for toluene losses was necessary. Determining the weight percent toluene in the bottom-phase samples was more difficult. Since the bottom-phase sampling lines had to be maintained at temperatures of approximately 600 K , most toluene in the phase flashed after passing through the micrometering valve. Compositions were therefore calculated by subtracting the topphase flow rates of pitch and toluene from the corresponding feed flow rates. Since the bottom-phase flow rate is relatively small, this calculational method produced highly variable results for the bottom-phase compositions. To calculate extraction yields, the weight of pitch collected in each phase had to be determined. The top-phase samples were dried to remove all toluene in the manner described above. The bottom-phase samples were dried in a similar fashion. The residual samples were then dissolved in 1,2,4-trichlorobenzene and analyzed by gel permeation chromatography to correct for residual toluene (if any) that was still present.

Kiran and Brennecke; Supercritical Fluid Engineering Science ACS Symposium Series; American Chemical Society: Washington, DC, 1992.

Downloaded by UNIV OF ARIZONA on April 5, 2017 | http://pubs.acs.org Publication Date: December 17, 1992 | doi: 10.1021/bk-1992-0514.ch028

28. HOCHGESCHURTZ ET AL.

Production of Mesophase Pitch

353

Gel permeation chromatography was performed with a Waters 150-C A L C / G P C chromatograph equipped with a refractive index detector. HPLC-grade 1,2,4trichlorobenzene was used as the mobile phase at a flow rate of 1 mL/min. Details of the G P C techniques used are presented elsewhere (3 JO). Since the purpose of this study is to produce mesophase pitch by SCF extraction, bottom-phase samples were analyzed to determine the percentage of mesophase present. First, the samples were dried under vacuum at 630 Κ for 30 min. with a Vacuum/Atmospheres Co. TS-4000 Dri-Lab system. Using this system, samples were exposed to elevated temperatures for a maximum of two hours, and the elevatedtemperature environment was maintained at