J . Cireni. Soc'.. Fui*aduj. Trans. I , 1989, 85(9), 2675-2682 Solubility of H,S, CO, and CH, in N-Formyl Morpholinet F-Y. Jou, R. D. Deshmukh, F. D. Otto and A. E. Mather" University of Alberta, Edmonton, Alberta, Canada T6G 2G6 The solubility of H,S, CO, and CH, in N-formyl morpholine has been measured at five temperatures between 298.15 and 403. I5 K at pressures up to ca. 7 MPa (1 3.6 MPa in the case of methane). The experimental data were correlated by the Peng-Robinson equation of state. Parameters for the N-formyl morpholine were estimated from liquid molar volumes and vapour pressures. Values of the binary interaction parameter were obtained from the experimental data. Using the expressions connecting the values of the binary interaction parameter with the parameters of the Krichevsky- Ilinskaya equation, Henry's coefficients for the three solutes were obtained.Most physical solvents that are used for the removal of the acid gases (H,S and CO,) from gas streams are polar organic compounds which have a strong affinity for hydrogen sulphide and/or carbon dioxide. At low pressures, the solubility of gases in physical solvents is usually proportional to the partial pressure of the gas at a given temperature. The enthalpy of solution is generally small (< 5 kJ mol-') and it is often possible to regenerate the solvent by flashing to a lower pressure. Hence processes using physical solvents require smaller amounts of energy than the traditional processes using chemical solvents such as alkanolamines. N-Formyl morpholine is a physical solvent which has been proposed for the separation of the acid gases, H,S and CO,, from natural and synthesis gas streams.A knowledge of the solubility of these gases is necessary to decide if this solvent has any advantage over those solvents now in use. The solubility of methane is also important, since its magnitude is a measure of the loss of hydrocarbons in the solvent. To answer these questions, the solubility of these gases in N-formyl morpholine was measured at five temperatures between 298.15 and 403.15 K at pressures up to ca. 7 MPa. Experimental The experimental apparatus and procedure are similar to that outlined by Jou et al.' The liquid and vapour phases were brought to equilibrium in a windowed Jerguson cell. A 250 cm3 cylindrical reservoir was attached to the top of the cell to increase the volume of the vapour phase.The vapour from the reservoir was recirculated through the solvent by a magnetically driven piston pump. The cell and pump were enclosed in a 0.4 m3 air bath maintained at f 0.5 "C of the set-point temperature. The heater consisted of a series of electrically heated fins which allowed for measurements to 130 "C. The air within the bath was mixed with a blade fan. With the exception of the pump piston, all metallic materials in contact with the fluids were type 316 stainless steel. The pump piston was made from Carpenter 450 steel as it must be ferromagnetic. The pressure in the cell was measured with calibrated Heise gauges, which have an accuracy of f O . 1 YO of full-scale span.The temperature of the fluid in the cell was monitored by an iron-constantan thermocouple. A gas sample line extended from the reservoir to the sample loop of the gas chromatograph. The liquid sample line led from the base of the cell to a needle valve located outside of the air bath. University of Surrey, 23-26 August, 1988. f Paper presented at the Third International IUPAC Symposium on Solubility Phenomena, held at the 89 2675 F A R I2676 Solubilities in N-Formyl Morpholine Table 1. Solubility of CO, in NFM total pressure mole fraction / kPa" of co, S/kmol kg-' T = 298.15 K 6410 0.971 0.287 6378 0.958 0.183 6359 0.941 0.137 6298 0.910 0.087 5 6267 0.899 0.076 9 6216 0.854 0.050 8 6126 0.803 0.035 3 5906 0.720 0.022 3 2950 0.369 0.005 07 1120 0.145 0.001 48 379 0.052 4 0.000 48 1 550.8 (45.5) 0.007 05 0.000 06 1 7 540.4 (4.56) 0.000 663 0.000 005 76 487.4 (0.452) 0.000 0704 0.000 000 612 T = 313.15 K 7090 0.620 0.0142 5530 0.504 0.008 8 1 3240 0.294 0.003 63 1320 0.136 0.001 37 334 0.039 3 0,000 352 553.0 (38.2) 0.00422 0.000 036 8 503.2 (26.4) 0.002 99 0.000 026 1 506.8 (4.24) 0.000 464 0.000 004 04 45 1.2 (0.788) 0.000 094 3 0.000000819 T = 343.15 K (NFM vapour pressure = 0.13 kPa) 6590 0.389 0.005 52 5200 0.304 0.003 79 4660 0.285 0.003 46 3810 0.232 0.002 62 1930 0.1 18 0.001 16 560.4 (72.3) 0.004 86 0.000 042 5 506.4 (6.86) 0.000 4 1 9 0.000 003 64 502.3 (2.79) 0.000201 0.000001 75 489.3 (2.22) 0.000 136 0.000001 19 T = 373.15 K (NFM vapour pressure = 0.81 kPa) 6750 0.286 0.003 48 5440 0.232 0.002 62 3730 0.162 0.001 67 2210 0.094 8 0.000 9 10 87 1 0.038 3 0.000 346 264 0.012 3 0.000 108 554.3 (22.3) 0.001 03 0.000 008 98 T = 403.15 K (NFM vapour pressure = 3.25 kPa) 491.7 (5.31) 0.000 244 0.000 002 12 6270 0.20 1 0.002 19 4950 0.164 0.001 71 31 10 0.099 3 0.000 958 808 0.029 1 0.000 261 299 0.0106 0.000093 1 468.2 (22.3) 0.000 83 1 0.000 007 22 53 1.4 (4.74) 0.000 188 0.000 00 1 63 "The partial pressure of the solute is given in parentheses when nitrogen was added in order to raise the total pressure.F-Y. Jou, R.D. Deshmukh, F. D. Otto and A . E. Mather Table 2. Solubility of H,S in NFM total pressure mole fraction /kPa' of H,S S/kmol kg T = 298.15 K 1900 0.968 0.264 1500 0.799 0.0344 942 0.563 0.01 12 464.2 (1 32.0) 0.124 0.001 23 471.4 (93.4) 0.095 3 0.00091 5 449.8 (1 8.4) 0.022 2 0.000 198 481.3 (4.85) 0.006 31 0.000055 1 464.6 (0.670) 0.001 17 0.000010 1 468.7 (0.0797) 0.000 236 0.000 002 06 T = 313.15 K 278 1 0.984 0.529 1760 0.684 0.0184 1370 0.583 0.012 1 1030 0.480 0.008 01 753 0.366 0.005 03 555 0.304 0.003 79 458.7 (197.0) 0.121 0.001 20 496.4 (44.5) 0.032 1 0.000 288 445.3 (18.1) 0.0144 0.000 127 339.4 (5.89) 0.005 34 0.000 046 7 455.3 (0.657) 0.000 852 0.000 007 4 1 393.2 (0.169) 0.000 258 0.000 002 54 T = 343.15 K (NFM vapour pressure = 0.13 kPa) 5230 0.970 0.280 3140 0.670 0.0177 1720 0.434 0.006 65 655 0.199 0.002 16 473.4 (1 65.0) 0.057 9 0.000 534 468.4 (29.8) 0.0120 0.000 106 500.6 (5.27) 0.002 92 0.000 025 4 469.0 (1.13) 0.000 744 0.000 006 47 T = 373.15 K (NFM vapour pressure = 0.8 1 kPa) 472.3 (0.0905) 0.000 116 0.000001 01 6500 0.812 0.0374 4840 0.668 0.0175 3650 0.540 0.0102 2040 0.347 0.004 6 1 933 0.180 0.001 91 254 0.059 1 0.000 545 493.8 (53.1) 0.0148 0.000 13 1 464.3 (6.23) 0.002 33 0.000 020 3 414.2 (0.796) 0.000457 0.000 003 97 462.5 (0.237) 0.000 168 0.000 00 1 46 T = 403.15 K (NFM vapour pressure = 3.25 kPa) 6750 0.636 0.0152 4950 0.506 0.008 90 3060 0.350 0.00468 1230 0.161 0.001 67 364 0.054 1 0.000497 495.6 (34.1) 0.007 26 0.000 063 6 440.3 (3.62) 0.001 11 0.000009 65 444.3 (1.23) 0.000430 0.000 003 74 470.4 (0.100) 0.000 055 0 0.000 000 478 2677 "The partial pressure of the solute is given in parentheses when nitrogen was added in order to raise the total pressure.89-22678 Solubilities in N-Formyl Morpholine Table 3. Solubility of CH, in NFM total total /kPa 103X /kPa 103X pressure 1 05S/kmol kg-l pressure 1 05S/ kmol kg-' I3 450 10 030 7 360 4 840 2410 1230 452 93 13 620 1 1 650 7 200 3 800 2 020 988 295 13 470 10 800 6610 4 420 2 230 847 198 T = 298.15 K 55.3 50.9 43.8 39.8 35.2 31.7 24.7 22.0 13.1 11.5 7.06 6.17 2.60 2.26 0.555 0.482 T = 313.15 K 56.9 52.4 50.5 46.2 34.9 31.4 20.0 17.7 11.2 9.82 5.48 4.78 I .75 1.52 T = 343.15 K 61.4 56.8 51.0 46.7 34.3 30.8 23.8 21.2 12.5 11.0 4.79 4.18 1.20 I .05 13 080 9 720 6 660 3 550 1200 576 26 1 1 3 090 9 190 6 490 4110 2 760 1300 629 185 T = 373.15 K 64.2 59.6 50.6 46.3 36.1 32.5 20.2 17.9 6.96 6.09 3.45 3 .OO 1.60 0.92 1 T = 403.15 K 67.7 63.1 50.6 46.3 37.6 33.9 24.9 22.2 16.7 14.7 7.92 6.93 3.90 3.40 1.22 1.06 Table 4.Equation of state constants for NFM for the Peng-Robinson equation T/K u , , Pa m'; molP h, cm:' mol-' ~ ~~ ~- 298.15 6.7909 93.203 313.15 6.57 19 93 -404 343.15 6.2 157 93.718 373.15 5.9336 93.909 403.15 5.6995 93.965 ~~ - ~~ Table 5. Correlation parameters (a,,) for d,, ~~ ~- ~~ ~ co, H,S CH, ~~ -0.0174 - 0.06 1 7 0.1 182 ~~~ - Prior to introduction of the fluids, the apparatus was brought to the desired temperature and purged with nitrogen to remove traces of oxygen. Ca.50cm3 of the solvent was fed by gravity to the equilibrium cell. Hydrogen sulphide or carbon dioxide was added to the cell in an amount indicated by the pressure. The pump was started and the vapour recirculated through the solvent. Additional amounts of acid gas were addedF-Y. Jou, R . D. Deshmukh, F. D.Otto and A . E. Mather 2679 8 7 6 5 --. $ 4 !?i. v1 n 3 2 1 0 I I I I P 1 0 0.2 0.4 0.6 0.8 1 mole fraction CO, Fig. 1. Solubility of CO, in N-formyl morpholine. T/K: @, 298.15; 0, 323.15; V, 348.15; 0, 373.15; ., 398.15. until the desired partial pressure had been approximately obtained. When necessary, nitrogen was added to maintain the system pressure above 200 kPa. At equilibrium, as established by a constant cell pressure, the pump was stopped and the phases analysed. A portion of the vapour was released to the sample loop of a gas chromatograph. A sampling valve was used to inject the gas into a 3 m long, 6.35 mm OD column packed with Chromosorb 104. The liquid sample was withdrawn from the equilibrium cell into a vessel containing 1 mol dm-3 NaOH, thus converting free dissolved acid gas into the involatile ionic species.A 40 cm3 high-pressure sample bomb containing the sodium hydroxide was used for sampling at loadings of CO, where the reaction rate was not great enough to prevent pressure build-up over the basic solution. A 50 cm3 Erlenmeyer flask fitted with a rubber septum served as a collection vessel for the sampling of solutions of CO, at partial pressures < 1000 kPa and for the sampling of solutions containing H,S. The procedure was somewhat different when no acid gases were present. The liquid sample was passed into a 50 cm3 weighed sample bomb while constant pressure was maintained in the equilibrium cell by gas addition. The bomb was reweighed to determine the amount of sample and then attached to a mercury-filled burette and the pressure brought to atmospheric.The gas which evolved from the liquid was collected in the calibrated 50 cm3 burette. The amount (in mol) collected was calculated from the P-V-T data after subtracting the vapour pressure of the liquid. In addition, it was necessary to account for the small amount of gas, equivalent to the solubility at atmospheric pressure, which remained in the liquid sample. Gas chromatography was used to measure these residual solubilities.2680 7 6 5 2 4 2 : Solubilities in N-Formyl Morpholine I I I I 0 0.2 0.4 0.6 0.8 1 mole fraction H2S Fig. 2. Solubility of H,S in N-formyl morpholine. Symbols as for fig. 1. The CO, content in an aliquot of the liquid sample was determined by adding excess 0.05 mol dm-, BaCl, to precipitate the carbonate as BaCO,.The precipate was washed and titrated with standardized 0.1 mol dm-, HCl using methyl orange-xylene cyanol indicator to a grey-green endpoint. The H,S content in an aliquot of the sample was determined by reacting the liquid with a solution of acidified 0.05 mol dmP3 I,. The unreacted I, was back-titrated with 0.05 mol dm-, Na,S,O, using starch indicator. The experimental error in the liquid-phase concentration is estimated to be +2-3 % in the range studied. Results and Discussion The solubility of H,S and CO, in N-formyl morpholine was determined at 298.15, 313.15, 343.15, 373.15 and 403.15 K at partial pressures of the acid gases up to ca. 7 MPa. The data are presented in tables 1 and 2 for CO, and H,S, respectively.The solubility of CH, in N-formyl morpholine was measured at the same five temperatures listed above. Partial pressures ranged up to 13.6 MPa. The data for methane are given in table 3. At a given partial pressure there is only a small effect of temperature on the solubility of methane. Both the mole fraction of solute in the liquid phase and the solubility S (kmol of solute per kg of solvent) are reported. For the acid gases, the partial pressure, the product of the mole fraction solute in the vapour phase and the total pressure, is reported when nitrogen was used to maintain the total pressure above the atmospheric pressure. The experimental data were correlated using the Peng-Robinson2 equation of state. The procedure followed was analogous to that described by Jou et aL3 The parameters for the pure solvent (NFM) were obtained from the liquid density and vapour pressure obtained from the fragmentary points of Cinelli et al.* and Vetere et al.5F-Y.Jou, R. D. Deshmukh, F. D. Otto and A . E. Mather 15 10 2 2 [ 5 I I I I I i / 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 mole fraction CH4 Fig. 3. Solubility of CH, in N-formyl morpholine. Symbols as for fig. 1. Table 6. Parameters for the Krichevsky-Ilinskaya equation H2,/MPa z?;/cm3 mol-' A / R T 298.15 313.15 343.15 373.15 403.15 298.15 313.15 343.15 373.15 403.15 298.15 313.15 343.15 373.15 403.15 6.89 8.92 13.8 19.6 25.9 1.15 1.62 2.93 4.70 6.92 187 178 167 160 155 co, 35.55 36.60 38.93 41.60 44.69 35.07 35.97 37.97 40.23 42.83 H2S 37.32 CH, 38.56 41.29 44.39 47.94 0.099 0.105 0.121 0.142 0.168 -0.151 -0.137 -0.1 10 - 0.084 -0.059 0.736 0.702 0.656 0.632 0.623 268 12682 Solubilities in N-Formyl Morpholine I 275 325 375 42 5 TIK Fig.4. Effect of temperature on the Henry’s constant: (----) CH,, (--- -) co,, (..-. .) H,S. The parameters are given in table 4. The experimental data were used to obtain the binary interaction parameter which appears in the mixing rule of the equation of state. The values of aij were found to be independent of temperature. The values are presented in table 5 for the three solutes. The results of the correlation are compared with the experimental data in fig. 1-3. In general, the correlated values are in good agreement with the experimental data over the wide ranges of temperature and pressure involved.Bender et aL6 have shown the connection between the binary interaction parameter and the three parameters in the Krichevsky-Ilinskaya7 equation : the Henry’s constant, the partial molar volume of the solute at infinite dilution, and the Margules parameter. Their equations (corrected) have been used to obtain these parameters, which are presented in table 6. The results for the Henry’s constants are shown in fig. 4. No comparison with other data are possible as only Zawacki et al.8 have measured data in these systems. The authors are grateful to the Alberta/Canada Energy Resources Research Fund for financial support of this research. References 1 F-Y. Jou, A. E. Mather and F. D. Otto, Ind. Eng. Chem. Process Des. Dev., 1982, 21, 539. 2 D-Y. Peng and D. B. Robinson, Ind. Eng. Chem. Fundum., 1976, 15, 59. 3 F-Y. Jou, R. D. Deshmukh, A. E. Mather and F. D. Otto, Fluid Phase Equilibria, 1987, 36, 121. 4 E. Cinelli, S. Noe and G. Paret, Hydrocurbon Process., 1972, 51(4), 141. 5 A. Vetere, R. De Simone and A. Ginnasi, Ind. Eng. Chem. Process Des. Dev., 1975, 14, 141. 6 E. Bender, U. Klein, W. P. Schmitt and J. M. Prausnitz, Fluid Phase Equilibria, 1984, 15, 241. 7 I. Kritchevsky and A. Iliinskaya, Acta Physicochim. U.R.S.S., 1945, 20, 327. 8 T. S. Zawacki, D. A. Duncan and R. A. Macriss, Hyclrocurhon Process., 1981, 60(4), 143. Paper 8104463E ; Received 4th Nocember, 1988