High pressure phase equilibria of the system phenol-m-cresol-carbon dioxide

Fluid Phase Equilibria, Elsevier Science B.V.

94 (1994) 313-327

313

HIGH PRESSURE PHASE EQUILIBRIA OF THE SYSTEM PHENOL-m-CRESOL-CARBON DIOXIDE G. Di Giacomo, S. Brandani, V. Brandani and G. Del Re Dipartimento di Chimica Ingegneria Chimica e Materiali, Universit& di L’Aquila I-67040 Monteluco di Roio, L’Aquila (Italy) Keywords: phenols.

experiments,

data, solid-liquid,

VLE high pressure,

carbon dioxide,

(Received November 17,1992; accepted in final form January 14,1994)

ABSTRACT The ternary system composed of phenol, m-cresol and supercritical carbon dioxide is characterized by unusual phase behaviour. In fact the ternary diagram shows five distinct regions: a three phase region of solid-liquid-gas, two two phase regions liquid-solid and liquid-gas, and two single phase regions liquid and gas. In addition when pressure was increased and temperatures were below the temperature of fusion of phenol, it was observed that the carbon dioxide causes recrystallization of phenol from the liquid phase. To understand and to describe all these phenomena, several experimental measurements were performed and are reported in this paper. Finally, using all the available data for the pure compounds as well as for the three binary sub-systems, a modified Peng-Robinson Equation of State was parametrized and then used to describe, at least qualitatively, the multicomponent multiphase behaviour in the whole range of composition under different values of pressure and temperature.

INTRODUCTION Bio-oils obtained by pyrolysis and/or liquefaction of lignocellulosic biomasses, are complex mixtures of many organic compounds greatly varying in size, molecular structure and polarity. Depending on the type of pyrolysis, bio-oils are usually characterized by a high percentage (from 10 to 60%) of phenol and phenol homologous (Churin et al., 1987). When these compounds are not recovered as chemicals they are usually selectively destroyed during the fuel up grading process (Churin and Maggi,1989) since bio-oils seldom meet the standards for commercial fuels.

037%3812/94/$07.00 01994 Elsevier Science B.V. All rights reserved SSDZ 0378-3812 (94) 02483-H

314

G. Dt Giacomo et al. IFluld Phase Equilibria 94 (1994) 313327

Supercritical fluid extraction (SFE) process for the up-grading of different bio-oils derived from wood pyrolysis, is based on the separation and recovery of phenols which are by far the most valuable products of wood pyrolysis (Brandani et al., 1989). In comparison to other conventional separation techniques like distillation and liquid-liquid extraction, SFE has the advantages of higher efficiency, ease of recovery of the solvent and improved transport properties. Among the possible supercritical solvents carbon dioxide has many good characteristics: low critical temperature and pressure (304 K and 73.8 bar respectively), it is no toxic, inexpensive and environmentally acceptable. A prerequisite to the study of the feasibility of such a process is the knowledge of the phase behaviour of the carbon dioxide-bio oils systems and of their subsystems. In this study we focused our attention on the ternary system composed of phenol, mcresol and supercritical carbon dioxide under different temperature and pressure conditions, with the aim of understanding and of describing its phase behaviour, which, so far, has not been reported in the literature. We made several experimental measurements which were used along with data from the literature related to pure components and to the binary systems carbon dioxide-m-cresol and carbon dioxidephenol to parametrize a modified Peng-Robinson Equation of State based model (Panagiotopoulos and Reid, 1987). Finally, the predictive capability of this model, which contains pure components and binary interaction parameters only, is briefly discussed and experimental and calculated results at 80 bar and 308.2 K are compared. EXPERIMENTAL Three different laboratory experimental apparatuses were used in this study for measuring all the phase equilibrium data reported in this section. A single pass flow apparatus, which can operate up to 1000 bar and 473 K, was used for measuring the composition of the carbon dioxide rich phase of the ternary system carbon dioxide-mcresol-phenol at 308.2 K under different pressure and for a 50% by weight of the liquid phase composition on solvent free basis. The raw experimental data are reported in Table 1 where y indicates the mole fraction in the light phase. Some experimental solubilities of m-cresol in dense carbon dioxide were also reported in a previous paper (Brandani et al., 1990). A detailed description of this apparatus and its operation mode have been reported in a previous work (Di Giacomo at al., 1989). An optical flow type apparatus schematically shown in Figure 1 was used for measuring the equilibrium composition of the liquid phase for the binary system mcresol-carbon dioxide as well as for the ternary system carbon dioxide-m-cresolphenol at 308.2 K under different pressure conditions. The raw experimental data are reported in Table 2 and in Table 3 respectively, where x indicates the mole fraction the liquid phase. A detailed description of this apparatus, its calibration and its operating mode were reported in a previous work (Di Giacomo et al.,1993). One of the optical cells used in this study, a stainless steel liquid level gauge Jerguson 17-T5 1, can operate up to 640 bar and 473 K. In addition, by using the optical cell of the

G. Di Giacomo et al. IFluid Phase Equilibria 94 (1994) 313327

315

Fig. 1 - Simplified flow sheet of the optical flow type apparatus. SCl, SC2: Sample collectors; PC: Carbon Dioxide Cylinder; R: Reservoir; Pl,P2: HP Pumps; AB: Air bath; OC: Optical cell; MP: Mixer-preheater.

experimental apparatus shown in Figure 1, a stainless steel liquid level gauge Jerguson 17-T-40, along with a synthetic method it was possible to visualize the solid-liquid-gas three phase region at 308.2 K and 80 bar and to locate the merging point of this region with the liquid phase region and with both the solid-liquid and liquid-gas regions on the triangular phase diagram at approximately 0.4 mole fraction of carbon dioxide and 0.09 mole fraction of m-cresol. For this purpose the optical cell was first charged with a predetermined amount of a liquid mixture of phenol and m-cresol of fixed composition. Then it was pressurized very slowly with carbon dioxide until a very small quantity of solid phase appeared. By assuming that under the above mentioned conditions the liquid composition on a solvent free basis is practically the same as the liquid mixture initially charged in the cell and, by volumetrically measuring the amount of carbon dioxide coming from the cell which after several hours, was depressurized, it was possible, after many trials, to locate the vertex of the three phase region. It is worth pointing out that the line connecting the pure carbon dioxide point with the saturation point at atmospheric pressure on the phenol-m-cresol line lies below the above mentioned vertex. An all glass apparatus was used for measuring the solubilities of solid phenol in liquid m-cresol at atmospheric pressure and at temperatures ranging from 288.2 to 308.2 K. The raw experimental data are reported in Table 4 where x indicates the mole fraction in the liquid phase at saturation. A detailed description of this apparatus, its calibration and its operation mode were reported in a previous work (Di Giacomo et al., 1992). When using the single pass flow apparatus, the mass flow rate of the solvent is

G. Di Giacomo et al. /Fluid Phase Equilibria

316

94 (1994) 313327

TABLE 1 Experimental solubilities of 50% by weight m-Cresol-Phenol mixtures in dense Carbon Dioxide at 308.2 K. P (bar)

(C&J

(Phlnol)

Y (m-Cresol)

P (bar)

(C’o,)

(Phlnol)

(m-C:esol)

80.

0.9977

0.0013

0.0010

160.

0.9800

0.0104

0.0096

80.

0.9986

0.0080

0.0006

200.

0.9705

0.0152

0.0143

100.

0.9899

0.0054

0.0047

200.

0.9732

0.0130

0.0138

100.

0.9893

0.0056

0.0051

200.

0.9745

0.0137

0.0118

100.

0.9898

0.0053

0.0049

200.

0.9726

0.0140

0.0134

120.

0.9864

0.0072

0.0064

300.

0.9643

0.0184

0.0173

120.

0.9852

0.0078

0.0070

300.

0.9646

0.0182

0.0172

120.

0.9868

0.0068

0.0064

400.

0.9607

0.0197

0.0196

160.

0.9793

0.0107

0.0100

400.

0.9630

0.0186

0.0184

TABLE 2 Experimental Solubilities of Carbon Dioxide in m-Cresol at 308.2 K. T (K)

P (bar)

x (CO,)

T (K)

P (bar)

x (CO,)

308.25

22.90

0.124

308.05

120.0

0.430

308.25

22.50

0.122

308.15

123.3

0.440

308.25

40.50

0.200

308.15

158.6

0.450

308.25

40.90

0.219

308.15

160.0

0.460

308.25

61.50

0.310

308.05

250.0

0.503

308.25

61.90

0.289

308.15

251.8

0.499

308.05

80.10

0.390

308.05

304.5

0.545

308.05

80.40

0.406

308.05

305.8

0.544

308.15

100.0

0.430

308.05

405.0

0.552

308.05

101.7

0.43

308.15

406.2

0.562

continuously

monitored by a mass flow meter (Micro Motion mod. D6) which also acts as totalizer. The accuracy is +0.4%. In all the high pressure experiments the pressure was measured by a pressure transducer located at the top of the saturator or at the top of the optical cell. Temperature is measured by a type J thermocouple previously calibrated against an NBS mercury-in-glass thermometer. The thermocouple

G. Di Giacomo et al. /Fluid Phase Equilibria

317

94 (1994) 313-327

TABLE 3 Experimental Solubilities of Carbon Dioxide in 50% by weight m-Cresol-Phenol mixtures at 308.2 K.

c-r:)

(&

(Phtnol)

(m-&sol)

&)

(&

(Phznol,

(m-&sol)

80.

0.4000

0.3000

0.3000

160.

0.5692

0.2269

0.2039

100.

0.4639

0.2793

0.2568

200.

0.6273

0.1960

0.1767

100.

0.4862

0.2614

0.2524

200.

0.6231

0.1952

0.1817

120.

0.4893

0.2669

0.2438

260.

0.7081

0.1512

0.1407

120.

0.4832

0.2743

0.2425

260.

0.7339

0.1376

0.1285

TABLE 4 Experimental Solubilities of Solid Phenol in liquid m-Cresol at atmosphere pressure. T (K)

x (Phenol)

T o + y1

(16) (17)

a,

(18)

z,

= x3 ( 1 -

k txi - yi

EL, -

y

)

+

(1%

y3 Q, + y

(20)

) = 0

i=l

where:

at =

v L+V+S

.

’

s y=

L+V+S

For each set of values of T, P and zl, z;?, 5 the system of equations from (16) to (20) can be solved to find CY~ and y in addition to the liquid composition, Xi, and to the gas composition , yi. The fugacity coefficients of component i, $, are calculated from the following exact thermodynamic relationship:

G. Di Giacomo et al. /Fluid Phase Equilibria

94 (1994) 313327

321

TABLE 5 Pure component adjustable parameters of the modified Peng Robinson Equation of State. Pure Component

0;;

BOT,,

AAPD

No. of data noints

Carbon dioxide

1.16489

1.2695

2.18

150

m-Cresol

0.71061

-0.30349

1.35

23

1.3501

0.29396

13.5

33

Phenol

TABLE 6 Values of X,, h,, bjT and XjF for the three binary sub-systems.

CO,-m-Cresol

-0.21785

0.97878 lo5

-0.15412

0.68479 1O-3

CO,-Phenol

0.23219

-0.29612 10”

0.24416

-0.42450 1O-3

m-Cresol-Phenol

0.26001

-0.11615 1O-3

0.26001

-0.11615 1O-3

TABLE 7 Average Absolute Percent Deviations between the experimental data and calculated values for the three binary sub-systems. System

No. of data points

AAPD(x)

AAPD(y)

Carbon dioxide-m-Cresol

27

0.893

0.201

Carbon dioxide-Phenol

30

m-Cresol-Phenol

5

7.76 5.202

where A is the reduced residual Helmhotz energy and Z is the compressibility factor. The expression for A and Z are reported in APPENDIX. By combining this equation with the Peng Robinson Equation of State, equation (1) to (6), the following relationship for calculating the fugacity coeffkients is obtained:

an8

-lnz2flb,RT

a( n2Qm) a ni

n %I

--

bi

bm _

v

+

b,,, ( 1 + fl)

In

v + bm (1

- p)

1

(22)

322

G. Di Giacomo et al. /Fluid Phase Equilibria

94 (1994) 313327

The expression for the partial derivatives which appear in equation (22) are reported in the APPENDIX. The numerical values of the carbon dioxide pure component adjustable parameters, 8r1 and BOT,,, were found by fitting the PvT data at temperature from 308.2 K to 328.2 K and at pressure from 1 bar to 150 bar (Casci et al., 1972). The numerical values of the m-cresol pure component adjustable parameters, & and BOT,,, were determined by fitting the available experimental vapour pressure at temperatures from 325.2 K to 475.95 K (Boublik at al., 1973; Perry and Chilton, 1973). The numerical values of the phenol pure component adjustable parameters, & and BOT,,, were determined by fitting the available vapour pressure data at temperature from 384.45 K to 473.15 K (Perry and Chilton, 1973; TRC, 1962) along with two sets of isothermal experimental solid-gas and gas-liquid equilibria data for the system carbon dioxide-phenol at 309.2 K and 333.2 K respectively (Van Leer and Paulaitis, 1980). As result of this fitting we also obtained the two binary interaction parameters kv and kji for the aforesaid system. All the values of the pure component parameters are reported in Table 5 together with the numerical values of the Average Absolute Percent Deviations (AAPD) between predicted and experimental density for carbon dioxide and the AAPD between predicted and experimental vapour pressure for m-cresol and phenol. The numerical values of the two adjustable parameters kii and k,, for the binary system carbon dioxide-m-Cresol were found by fitting the experimental solubilities reported in Table 2 along with the literature isothermal vapour-liquid equilibrium data at 308.2 K, 318.2 K and 328.2 K and at pressure from 20 bar to 240 bar (Lee and Chao, 1988). The two binary interaction parameters for the system m-cresol-phenol were determined by fitting the experimental solubilities of solid phenol in liquid mCresol reported in Table 4. The temperature dependence of ki, and kji for the three binary systems carbon dioxide-phenol, carbon dioxide-m-cresol and m-cresol-phenol is expressed by: kv = Xv+A;T

(24)

Given the structural similarity of the two components of the last above mentioned binary system we set the value of k, equal to k,,.309.8 K and 80.9 bar. Table 6 gives the numerical values of X,, Xji, Xi,Tand hjiT for the three binary subsystems while the numerical values of AAPD between experimental and calculated values of the solubility are reported in Table 7. The above described EOS based model along with the numerical values of the adjustable parameters reported in Tables from 5 to 7, can be used to predict the phase behaviour for the Carbon dioxide-m-Cresol-Phenol ternary system. As an example the results of these calculations at 308.2 K and at 80 bar are shown in Fig.

G. Di Giacomo et al. /Fluid Phase Equilibria

94 (1994) 313-327

323

m-CRESOL 1.0 0.9 0.6

0.5 0.4 0.3 0.2 0.1

L+s+v

0.0 0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.S

PHENOL

0.9

1.0

CO2

2 - Phase diagram of the Carbon dioxide-m-Cresol-Phenol of 80 bar. 0 This work; A Lee and Chao, 1988.

ternary system at 308.2 K and pressure

m-CRESOL 0.006

0.006

0.060 0.992

PHENOL

0.996

0.996

1.600

CO2

Fig. 3 - Enlarged phase diagram of Carbon Dioxide-m-Cresol-Phenol ternary system at 308.2 K and pressure of 80 bar. 0 This work; A Lee and Chao, 1988; 0 Van Leer and Paulaitis, 1980 at

324

G. Di Giacomo et al. IFluid Phase Equilibria

94 (1994) 313327

Figure 2 and in Figure 3 together with the available binary and ternary experimental data obtained under the same pressure and temperature conditions. As can be seen the qualitative behaviour is correctly described in the whole range of composition. The three phase liquid-solid-gas region visualized by the optical experimental apparatus is also predicted and the agreement between the experimental and calculated intersection of liquid-solid and liquid-gas boundary lines is reasonably good. A similar agreement between calculated and experimental results has been verified under different pressure and temperature conditions. However, we have to point out that the model does not predict recrystallization of phenol in contrast to our experimental observation. CONCLUSIONS As a result of this work it has been found that the ternary system carbon dioxide-mcresol-phenol shows quite complex behaviour including a three phase solid-liquid-gas region in contact with a solid-liquid region, a gas liquid region and an all liquid and an all gas region. An EOS based model can be used to qualitatively describe this behaviour under different temperature and pressure conditions. In addition, recrystallization of solid phenol has been observed at 308.2 K and 80 bar when carbon dioxide is added to a liquid mixture of phenol and m-Cresol. However, we were not able to describe this phenomenon by the above mentioned model. ACKNOWLEDGEMENTS Our sincere experimental

thanks are due to Mr. Spagnoli for his help in carrying work, and to the CEC for financial support.

out the

APPENDIX A) Helmhotz free energy

,g=

am

A-Agzh nRT

2flb,RT

(Al) v+b,(l

+fl)

v+b,,,(l-\/-2-)

I

G. Di Giacomo et al. IFluid Phase Equilibria

94 (1994) 313-327

325

Z=pv RT

B) Expression

for the partial derivatives which appear in equation (22)

yz.o_)=@ .

i

(% + a&

)

zk +

e6 da, i-l

r

( k& - kh ) zk

zi

•t

k=l

(A3) - ~ ~

~~

( k~ - kjk )

Zk2 Zj

k=l j-l

LIST OF SYMBOLS A A ; BOT, kii L m n P Pit R s T TR V

V x Y ‘z

Helmhotz energy. reduced residual Helmhotz energy. model parameter given by equations (2), (4) and (5). model parameter given by equations (3) and (6). pure component adjustable model parameter. temperature dependent binary interaction adjustable model parameters. moles of liquid phase. ratio defined by equations (12) and (15). number of moles. pressure, bar. reduced pressure. gas constant. solid phase. temperature, K. reduced temperature. molar volume. moles of gaseous phase. mole fraction in the liquid phase. mole fraction in the light phase. mole fraction in the whole system. compressibility factor.

Greek symbols. v/(L+v+s). pure component adjustable model parameter. s/(L+v+s). temperature independent binary interaction adjustable model parameter.

G. Di Giacomo et al. f Fluid Phase Equilibrca

326 cp

94 (19941313327

fugacity coefficients.

Superscripts. id L s T

SU V 0

ideal property. liquid. solid. temperature coefficient in equations (26) and (27). sublimation. vapour. generalized property.

Subscripts. i, i, k m 1, 2, 3

components “i”, “j” and “k”. mixture. indicates carbon dioxide, m-cresol and phenol respectively.

REFERENCES Boublik, Y., Fried V. and Hala, E., Vapour Pressures of Pure Substances, 1st ed. Elsevier, Amsterdam 1973, p. 326. Brandani, V., Del Re, G., Di Giacomo, G. and Flammini, D., 1989. Treatment of Pyrolysis Oil with Compressed Carbon Dioxide. In: G. L. Ferrero, K. Mania& A. Buckeens, A. V. Bridgwater, (Eds.), Pyrolysis and Gasification, Elsevier, Amsterdam, pp. 430-434. Brandani, V., Del Re, G., Di Giacomo, G., Flammini, D. and Mucciante, V., 1990. Supercritical Fluid Extraction of Pyrolytic Oil, In: G. Grassi, G. Gosse, G. DOSSantos, (Eds.), Biomass for Energy and Industry, Vol. 2, Conversion and Utilization of Biomass, Elsevier, Amsterdam, pp. 2.710-2.714. Casci, C., Macchi, E. and Angelino, G., Proprieta Termodinamiche dell’Anidride Carbonica, Tamburini, Milano, 1972. Churin, E., Maggi, R., Grange, P. and Delmond, B., Characterization and up grading of a pyrolytic oil. Pyrolysis as basic technology for Large Agro-Energy Projects, L’Aquila, Italy 15 16 October, 1987. Churin, E. and Maggi, R., 1989. What Can We Do with Pyrolysis Oils. In: G. L. Ferrero, K. Maniatis, A. Buckeens, A. V. Bridgwater, @is.), Pyrolysis and Gasification, Elsevier, Amsterdam, pp. 326-333. Di Giacomo, G., Brandani, V., Del Re, G. and Mucciante, V., 1989. Solubility of Essential Oil Components in Compressed Supercritical Carbon Dioxide, Fluid Phase Equilibria, 2: 405-4 11. Di Giacomo, G., Brandani, P., Brandani, V. and Del Re, G., 1992. Solubility of Boric Acid in Aqueous Solutions of Sulfate Salts, Desalination, 89: 185-202. Di Giacomo,G., Brandani, V., Del Re, G. and Martinez de la Ossa, E., 1993. Phase Equilibria and Process Simulation for High-Pressure Supercritical Extraction Processes: Experimental Investigation. In: P. A. Pilavachi, (Ed.) Energy Efficiency in Process Technology, Elsevier, Amsterdam pp. 412-420. Lee, R.J. and Chao, K.C., 1988. Extraction of 1-Metylnaphthalene and m-Cresol with Supercritical Carbon Dioxide and Ethane, Fluid Phase Equilibria, 43: 329-340. Panagiotopoulos, A. Z. and Reid, R. C., Supercritical Fluids. Squires, T. and Paulaitis, M. E.

G. Di Giacomo et al. /Fluid Phase Equilibria

94 (1994) 313-327

327

(4s.). Am. Chem. Sot., Washington, 1987, pp. 115. Michelsen, M. L. and Kistenmaker, H., 1990. On composition-dependent interaction coefficients. Fluid Phase Equilibria, s: 229-230. Perry, H.R. and Chilton, C.H., Chemical Engineers Handbook, 5th ed. McGraw-Hill, New York 1973. TRC Thermodynamic Tables-Nonhydrocarbons, TRC, TAMU, College Station, USA 1962 Van Leer, R.A. and Paulaitis, M.E., 1980. Solubilities of Phenol and Chlorinated Phenols in Supercritical Carbon Dioxide, J. Chem. Eng. Data, 25: 257-259