c = critical property Li, Y.-H., K. H. Dillard and L. Robinson, Jr.: /. Chem. M = mixture property Eng. Data, 26, 52 (1981). r = reduced property Newton, R. H.: Ind. Eng. Chem., 27, 302 (1935). rl, r2 = reference fluid Nishiumi, H. and S. Saito: /. Chem. Eng. Japan, 8, 356 1,2,i,j = component 1, 2, /,j (1975). Ohgaki, K. and T. Katayama: Fluid Phase Equilibria, 1,
SOLUBILITY OF SOLID NAPHTHALENE IN GASEOUS ETHYLENE AT HIGH PRESSURES
Hirokatsu MASUOKAand Masahiro YORIZANE Department of Chemical Engineering, Hiroshima University, Hiroshima 730
Solubility of naphthalene in compressed ethylene was measured with a single-pass, continuous- flow apparatus at temperatures of 35, 50 and 65°C respectively and for a pressure range of 8 to 62atm. A calculation method is presented for determining the solubility of a heavy solute in a super- critical gas solvent. The modified BWRequation of Lee and Kesler together with a pseudo- critical method was used to calculate fugacity coefficients of the solute in subcooled liquid and vapor phases. The utility of the method presented is illustrated by a calculation of the ethylene- naphthalene system at temperatures between 12 and 65°C and over a pressure range of 8 to 300 atm.
The supercritical gas extraction process is based on the Intro duction ability of a supercritical gas to enhance the volatility Fluid fuels and chemical feedstocks derived from of the heavy organic compounds present in coal, coal are becoming increasingly attractive as a sub- provided high pressure is maintained. stitute for oil. Supercritical gas* extraction of coal The solvent effect of compressed gases was recog- is promising in many liquefaction techniques12K * Each fluid has its owncritical temperature. At tempera- Received June 5, 1981. Correspondence concerning this afticle should be tures exceeding its critical temperature the fluid cannot be lique- addressed to H. Masuoka. fied and is referred to as a supercritical gas.
VOL. 15 NO. 1 1982 nized as far back as in 1879 by Hanay and Hogarth6}. librium for the solute is given by Fank3) reports the pronounced solvent power of super- /!=/;à" (i) critical water. Weale21) has presented an excellent f\ = (p\Pyi (2) discussion of solubility of solids in compressed gases. Hagenbach5) demonstrated enhanced solvent proper- where f°±s is the fugacity of pure solid, y1 is the mole ties of supercritical polar gases. Paul and Wise15) fraction of the solute in the solution, and
JOURNAL OF CHEMICAL ENGINEERING OF JAPAN Method In calculating the solubility according to Eq. (3), an equation of state capable of representing the thermodynamic properties of subcooled liquid is required. For this purpose the Lee and Kesler11} three-parameter corresponding-states correlation was chosen from considerations similar to those in a pre- vious paper13} on the solid solubility in liquid. The Lee and Kesler equation is an extended and im- proved version of the Pitzer-Curl tables based on a modified BWR equation of state. The compressi- bility factor of a real fluid is related to properties of a simple fluid (w=0) and to those of octane as a ref- erence fluid. Z=Z(0H-^7(Z(i2) -Z(0)) (8)
where a>(R] =0.3978. Other thermodynamic functions librium cell or sampling tube can cause serious experi- (fugacity coefficients and enthalpy departure) have mental errors, and it is difficult to minimize this ef- been derived from Eq. (8). fect. These circumstances are similar to measure- In applying the Lee-Kesler equation to mixtures, ment of vapor pressures of high boiling-point sub- we used the pseudocritical method8'13). According stances. We used a single-pass, continuous-flow (also re- to the three-parameter corresponding-states theory, ferred to as gas saturation) method1)7'10'17). The the fugacity coefficient of component i in the mixture can be written as follows: methodis especially suitable for low vapor pressures In f* _inf* , n /dT' of solute. t\ Hm-H* m 3. 1 Apparatus and experimental procedure The flow apparatus is shown schematically in Fig. 1. _.HJliL\ (z_n The gas from a high-pressure cylinder was passed sequentially through two coils immersed in a con- /^\ fdMuipy, stant-temperature bath, which was controlled to with- \OritJT,P,n3\ OQ)m JT' in it0.02oC. The coils were made of stainless steel rr,P' For the pseudocritical constants the van der Waals pipe (7mm I.D.). The first coil (1.0m length) one-fluid model was used. ensured that the gas stream was at the prescribed tem- perature while passing through the equilibrium coil i 3 (0.6 mlength) which contained the solute material. Tf-Y yv-V'V Tå å /V/ 1c-La LsyiJOYcxo-1cijlY c Here the solute material was evaporated into a gas i 3 stream. A sintered stainless steel filter (2 /mi pore Zci =Pei Vci/RTci size) was incorporated into the line on the down- stream side of the equilibrium coil to eliminate any i possible entrainment. After leaving the filter the P'C=Z'CRT!C\ V'C (10) saturated gas was led to a throttle valve. The stream pressure was reduced to approximately atmospheric i veti=(i -stixviit+ vi?Y/s pressure across the throttle valve, after which the -*ciQ==\\ kij)\lci-Lcj) stream flowed through the sampling loop of a gas chro- matograph. The throttle valve, sampler (1.37 cc) and where ki3- and si5 are parameters determined from tubing outside the bath to the gas chromatograph binary-phase equilibrium data. were heated with a Nichrome-wired heating system and insulated to prevent any condensation of solute. 3. Experimental The temperature of the line outside the bath was Various methods for determining the solubility of controlled at 180±5°C. solid in vapor are noted in the literature. Static Pressure in the coils was read from a calibrated methods, which are extensively used in phase equilib- Heise gage (0.5 Kg/cm2 minimum scale, 200 Kg/cm2 rium determinations, may not be suited for systems full scale). Temperature of the bath was read from of very low content of solute in vapor, for several a mercury thermometer (0.05°C minimumscale). reasons1\ One is that adsorption of solute on equi- To assure complete saturation of the gas, the gas
VOL. 15 NO. 1 1982 71 Table 1 Solubility of solid naphthalene in supercritical ethylene at 35, 50 and 65°C respectively '(yi : mole fraction of naphthalene) 35°C 50°C 65°C P[atm] y± P [atm] y1 P [atm] y± 8.0 5.44E-5 7.9 1.69E-4 12.0 3.89E-4 13.3 4.09E-5 12.7 1.17E-4 15.9 2.80E-4 20.0 3.92E-5 19.0 9.55E-5 21.5 2.46E-4 30.7 4.27E-5 28.6 1.14E-4 31.3 2.77E-4 34.0 5.25E-5 39.1 1.73E-4 37.0 3.09E-4 40.0 7.28E-5 47.5 2.17E-4 40.8 3.33E-4 48.2 9.84E-5 51.3 2.75E-4 45.3 3.85E-4 55.1 1.57E-4 56.9 3.37E-4 48.9 4.75E-4 59.1 2.29E-4 61.6 5.04E-4 56.2 6.54E-4 62. 1 3.72E-4
The flow rate was held in the range of 10 to 15ml/ min. 3. 2 Materials Naphthalene obtained from WakoPure Chemical Industries was of special-grade purity. Ethylene obtained from Seitetsu Chemical Company had a reported minimumpurity of 99.5 %. These materials wasused without further purification. The column packing was PEG20 Mand the length of the column was 1 m. The thermostat bath containing the column Fig. 2 Experimental and calculated solubilities of recorded 200°C at all measurements during the test. solid naphthalene in ethylene at 35°C Nitrogen was the carrier gas. 3. 3 Gas analysis To determine solute concentration, measurements were madewith a Shimazu gas chromatograph (Model GC-4BPTF) with hydrogen flame ionization detector and a Takeda digital integrator (Model TR-2215A). 3. 4 Calibration The flow method (also referred to as the direct injection method) was chosen for calibration of the flame ionization detector. Helium was chosen as carrier gas because it has very low solubility in naphthalene, and its volumetric prop- erties are accurately known over the temperature range involved. From Eq. (7) the enhancement factor of naphthalene for the helium-naphthalene system is calculated to be less than 1.003 at 1 atm and over temperature levels of 35 to 65°C. In the calculation the second virial cross-coefficients were estimated by a correla- tion ofTsonopoulos19}. Mole fraction ofnaphthalene in the sampler was calculated as nsat psat y^E/j^=^ (1 1) The calibration curve was fitted by a quadratic equa- Fig. 3 Experimental and calculated solubilities of tion. Arithmetic average deviation of the data from solid naphthalene in ethylene at 50°C the equation was 3.1 %. flow rate measured by a soap-film flow meter was 4. Results and Discussion varied in the range 10 to 50 m//min by way of a trial. It was found that any variation of flow rate within a Solubility data were obtained at temperature levels range less than 25 m//min did not affect the results. of 35, 50 and 65°C and in the pressure range of 8 to
JOURNAL OF CHEMICAL ENGINEERING OF JAPAN Fig. 5 Consistency test of experimental solid solubilities in vapor using enhancement factor at 35, 50 and 65°C respectively
Eig. 4 Experimental and calculated solubilities of solid naphthalene in ethylene at 65°C 62atm. The results are given in Table 1 and Figs. 1 to 4. Data of Tsekhanskaya et alm at 35°C anc Diepen and Scheffer2) at 50°C are also shown in these figures for comparison. At 35°C there is no appareni discrepancy between our data and those of Tsekhan- skaya et ah But at 50°C the data of Diepen and Scheffer appear to be too high at lower pressures compared with our data. To check the accuracy of experimental data of this work, a plot of the enhancement factor was tried, as shown in Fig, 5; it must be equal to unity at the vapor pressure for the temperature under consideration. Onthis basis, the data from the present work appear reasonable. The solubility of component 1 in component 2 was calculated by two different methods. The first method uses Eqs. (3), (4), (8), (9) and (10). Terms including Fig. 6 Experimental and calculated solubilities of ACPin Eq. (4) were neglected, because these terms are solid naphthalene in ethylene at 12 and 25°C re- much less significant than the others. The second spectively method uses Eqs. (5), (8), (9) and (10). Results ob- tained by both methods show little difference. The and s12 in this experimental range. It is interesting to difference ranges from 1% at 65°C to 3% at 35°C. note that the calculations reproduced fairly good re- The following discussion is limited to the former sults even with k12=s12=0. The calculations with k12 method. and s12 from the correlation of the solid-liquid equi- The solubility was calculated by a trial-and-error libria did not improve the results very much. This procedure. Temperature and pressure were selected may be due to the difference of phases in which fugacity as the independent, known variables; the program coefficients were calculated : vapor phase in this work, calculates the mole fraction in vapor phase. We liquid phase in the solid-liquid equilibria. calculated the solubility by three different approaches ; Moreover, to test the applicability of our method first, assuming k12=s12=0\ second, using k12 and sl2 the calculations were performed at 12 and 25°C, as from the correlation of solid-liquid equilibria of the shown in Fig. 6. Results of calculations at 12°C previous study13}; third, using empirical kn and s12 showthat even a small variation of k12 and s12 ap- which best reproduce the experimental solubility data. preciably affects calculated results. This may be due The results are given in Figs. 2 to 4. They show that to the fact that the system temperature of 12°C is the calculations are not too sensitive to values of k12 very near the critical temperature of ethylene (9.5°C).
VOL. 15 NO. 1 1982 m mixture Conclusions OR) reference fluid Solubility of naphthalene in compressed ethylene s solid sat saturated was measured at temperature levels of 35, 50 and 65°C t triple point and in the pressure range of 8 to 62 atm. v vapor phase Acalculation method was proposed to determine (O) simple fluid the solubility of a heavy solute in a supercritical o pure 1 solute solvent. It was shown that the method can correlate 2 solvent well the solubility of solid naphthalene in ethylene in pseudocritical property of mixture the temperature range between 12 and 65°C and over ideal gas pressure ranges of 8 to 300atm. The method uses the Lee-Kesler equation of state together with the Literature Cited pseudocritical method, analogously to the solubility 1) Carruth, G. F. and R. Kobayashi: /. Chem. Eng. Data, 18, 115 (1973). of solids in liquid phase which was reported previously. 2) Diepen, G. A. M. andF. E. C. Scheffer: /. Am. Chem. Soc, The correlation requires two adjustable parameters, 70, 4085 (1848); /. Phys. Chem., 57, 575 (1953). but even without these parameters the results seem 3) Fank, E. U.: Z. Phys. Chem. (Frankfurt am Main), 82, 92, fairly reasonable for the ethylene-naphthalene system. 107, 192 (1956). 4) Gangoli, N. and G. Thodos: Ind. Eng. Chem., Prod. Res. Acknowledgement Dev., 16, 208 (1977). The authors wish to express their thanks to Mr. H. Takeda, 5) Hagenbach, A.: Ann. Phys. (Leipzig), 5, 276 (1901); 8, 568 Mr. M. Takahashi and Mr. T. Hojyo for their work during the (1902). study. 6) Hannay, J. B. and J. Hogarth: Proc. Roy. Soc. London, Ser. A, 29, 324 (1879). Nomenclature 7) Hiza, M.J., C.K. Heck and A.J. Kidnay: Chem. Eng. = second virial coefficient [cc-mol"1] Progr., Symp. Ser., 64 (88), 57 (1968). 8) Joffe, J.: Ind. Eng. Chem., Fundam., 15, 298 (1976). = heat capacity at constant pressure [cal -mol-^K"1] 9) Kaul,B. K. andJ. M. Prausnitz: AIChEJ., 24, 223 (1978). E 10) Kuebler, G. P. and C. McKinley: "Advances in Cryogenic f enhancement factor [-] Engineering", Vol. 19, p. 320, Plenum Press, New York H fugacity [atm] (1974). [cal - mol"1] molar enthalpy ll) Lee,B. I. andM. G. Kesler: AIChEJ.,21, 510(1975). kti interaction parameter [-] P 12) Maddocks, R. R., J. Gibson and D. F. Williams: Chem. pres sure [atm] Eng. Progr., 75 (6), 49 (1979). R gas constant, 1.9872 [cal-mol-1 - *:-1] 13) Masuoka, H., R. Tawaraya and S. Saito: /. Chem. Eng. gas constant, 82.054 [cc- atm-mol-^ K-1] Japan. 12, 257 (1979). S molar entropy [cal -mol-^ K-1] Sij 14) Masuoka, H. and M. Yorizane: submitted to Chem. Eng. interaction parameter [-] T temperature [K] Sci. (Letter To The Editors). V molar volume [cc - mol"1] 15) Paul. P.F. M. and W.S. Wise: "The Principles of Gas y Extraction", M & B Monograph CE15, Mills & Boon, mole fraction in vapor [-] London (1971). z [-] compressibility factor 16) Prausnitz, J. M. : "Molecular Thermodynamics of Fluid- Phase Equilibria", Prentice-Hall, Inc. (1969). fugacity coefficient [-] [-3 17) Prausnitz, J. M. andP. R. Benson: AIChEJ.,5, 161 (1959). acentric factor 18) Tsekhanskaya, Y. V., M. B. Iomtev and E. V. Mushkina: Superscripts and Subscripts> Muss. J. Phys. Chem., 38, 1173 (1964). c = critical 19) Tsonopoulos, C.: AIChE J., 20, 263 (1974). / = fusion 20) Vetere, A.: Chem. Eng. Sci., 34, 1393 (1979). i,j, k = components in solution 21) Weale, K. E.: "Chemical Reactions at High Pressures", / = liquid E. & F. N. Spon Limited, London (1967).
10 JOURNAL OF CHEMICAL ENGINEERING OF JAPAN