Evaporation of the Ethanol and Water Components Comprising a Binary Liquid Mixture.код для вставкиСкачать
Evaporation of the Ethanol and Water Components Comprising a Binary Liquid Mixture K.D. O'Hare', P.L. Spedding2' and J. Grimshaw3 ' British Gas plc, Midlands Research Station, Solihull, West Midland, 691 2JW, UK * Department of Chemical Engineering, The Queen's University of Belfast, Belfast BT9 5AH, UK Department of Applied Chemistry, The Queen's University of Belfast, BT9 5AH, UK Evaporation studies were conducted on the ethanol-water system. The evaporation rates of the ethanol and water components were determined experimentally, over a range of compositions. The rate of change in the composition of the evaporating mixture was monitored, and the effect of concentration on the ethanol component evaporation rate (per unit mass of ethanol remaining) was investigated. Preferential adsorption of alcohol on the interjace in low ethanol concentration solutions allowed ethanol to evaporate more rapidly in these solutions. Introduction The evaporation of pure liquids has been well documented in the literature.'-" By comparison very little research has been conducted on the evaporation of binary liquid mixtures. This can be attributed to the difficulties involved with the experimental measurement of the component evaporation rates, and the calculation of mass transfer rates considering that the mass transfer resistance can be in both the gas and liquid phases. The aim of this work was to investigate experimentally the evaporation of ethanol and water from an ethanol-water mixture (from a plane surface of a mixed liquid, under conditions of turbulent air flow) and to determine whether: (i) the composition of the evaporating mixture changed as evaporation proceeded; (ii) the component evaporation rates were concentration dependant; (iii) the component evaporation rates were greater than for the pure liquids alone. Literature Review for Binary-Liquid Evaporation HoffmanI2 conducted one of the earliest investigations into the evaporation of binary mixtures. He concluded that binary mixtures could be divided into two * To whom all correspondence should be addressed. Developments in Chemical Engineering and Mineral Processing, Vol. 1, No 2/3,page 118 Evaporation of the Ethanol and Water Components 119 groups: (i) mixtures that showed ideal behaviour; and (ii) constant evaporating mixtures. He stated that the first group consisted of mixtures from the same homologous series, like toluene and xylene, where both solvents evaporated simultaneously at different rates and had no effect upon each other. When the faster-evaporating component was gone, the remainder of the other component evaporated at a constant rate. The second group consisted of mixtures of two dissimilar compounds such as an alcohol and an aliphatic hydrocarbon. HoffmanI2 stated that these evaporated at a constant rate without a change in composition, the evaporation rate being faster than for either component alone. Lewis et al. l 3 studied the evaporation of mixed solvents in relation to their use for protective coatings, such as lacquers. They concluded that the evaporation process was more complicated than was suggested by Hoffman.12They found that for the evaporation of a methanol-benzene mixture, the composition of the liquid mixture changed continuously from the initial value to a mixture containing no detectable methanol, which was the more volatile component. S a r n ~ t s k y ’investigated ~ the evaporation rate of mixed solvents in relation to paint films. He found that the evaporation of alcohol from alcohol-hydrocarbon mixtures was concentration dependant and that the alcohol component evaporated faster than was calculated assuming ideal behaviour. He also found that the evaporation rate of a toluene-isopropanol mixture was greater than for either of the two components alone. Richardson” studied the evaporation of binary mixtures using the Winkelmann16 method, which involved the passing of air above an evaporation tube containing the liquid mixture. He studied the evaporation of (i) acetone and dibutyl phthalate; and (ii) carbon tetrachloride and dibutyl phthalate. Dibutyl phthalate was virtually a non-volatile liquid. Richardson” found that in the first case, the more volatile component evaporated faster leaving a residual liquid, which being of greater density, sank into the bulk of the liquid causing considerable mixing. However, in the second case, because the dibutyl phthalate was less dense than the more volatile component, the residue tended to concentrate in the upper layers of the mixture and therefore the volatile liquid had to diffuse through this region before it could evaporate. Clark and Judson King17 measured the evaporation rates of n-pentane, cyclopentane and ethyl ether from n-tridecane into flowing nitrogen. They found that cellular convection, which was concentration-gradient induced and surface tension driven, served to increase liquid-phase mass transfer coefficients substantially. Mackay and MatsuguI8 studied the evaporation rates of liquid hydrocarbon spills on land and water, and stated that two limiting conditions existed for the evaporation of binary mixtures. First, assuming that there was perfect mixing in the liquid, the vapour leaving the surface would be in equilibrium with the bulk liquid and generally had a composition different from that of the liquid. The liquid composition was constant with depth, but changed with time. The more volatile material evaporated first, and as a result the rate of evaporation fell as mass transfer to the gas proceeded. The second limiting condition existed when there was assumed to be no mixing in the liquid phase. The vapour composition was equal to that of the liquid, and the liquid composition remained constant with time and depth providing the evaporation rate remained constant. Physically this process was envisaged as successive layers of liquid ‘peeling off’ the surface to form vapour. In practice a situation intermediate between these two conditions existed, and was controlled by the magnitudes of the evaporation rate and the liquid diffusion rate. 120 K.D.O'Hare, P.L. Spea'ding and J. Grimshaw Experimental Details In order to measure the evaporation rates of the ethanol and water components from a mixture thereof, it was necessary to determine both the total amount of liquid evaporated and the change in composition with time. With a contained body of liquid, evaporation causes the level of the evaporating surface to decline. Hinchley and Himus' and Powell and Griffiths4 observed that the presence of a 'rim' affected the evaporation rate of liquid. Therefore, in this work it was considered desirable to maintain the evaporating liquid surface flush with the top of the container. For pure liquids, a satisfactory procedure would be to add make-up liquid during an experimental run, but for a binary liquid mixture this would alter the liquid composition. The best solution appeared to be to incorporate a positive-displacement device in the base of the container which could compensate for the change in volume on evaporation, and thus maintain the position of the liquid surface. The experimental system is shown schematically in Figure 1. The main features of the system were: (i) a wind tunnel; (ii) a novel evaporation cell; (iii) a density meter. The evaporation studies were conducted in a horizontal Perspex wind tunnel, 2 m long and 88 mm by 98 mm cross-section. Air was blown through the wind tunnel by a positive-displacement blower, the air flow rate being set by use of a bypass system. For flow rates below 80 m3 h-', measurements were taken using rotameters, while for higher flow rates, a calibrated orifice plate was used. Air flow rates up to 100 m3 h-' were used in the experiments. The air temperature and relative humidity entering the wind tunnel were monitored using a digital thermohygrometer (Solomat 355). All other temperature measurements were obtained using E-type thermocouples in conjunction with a digital voltmeter. The air pressure was monitored using a digital micrometer (Air Ltd., model MP3KDU), while the atmospheric pressure was measured using a Fartin-type barometer. An evaporation cell was machined from a cylinder of solid nylon as detailed in Figure 2. The evaporating liquid within the cell was moved in a circular pattern by pumping externally through a peristaltic pump to a density meter (Stanton-Redcroft, model DMA 602 HT). The bulk temperature of the evaporating liquid was maintained at approximately the same temperature as the air flowing through the wind tunnel (+l"C), by using an air bypass system to heat the evaporation cell and the peristaltic pump. Temperature, pressure and velocity profiles were recorded to check that the conditions at the gas-liquid interface were constant, in terms of hydrodynamics and temperature. The ideal experimental situation was to have a plane liquid surface at constant temperature over the duration of a run, above which existed a turbulent sublayer, bounded by a turbulent stream of flowing gas. The bulk liquid was mixed so as to ensure constant boundary layer conditions at the upper liquid surface. A fixed reference level was installed at the air-liquid interface in the form of a stainless steel hypodermic needle. The required liquid level was achieved by raising the floor of the cell until the liquid interface was just pierced by the needle point. The movable cell floor was sealed using rubber O-rings. Upward movement of the floor was obtained and measured using a micrometer screw gauge. In any time interval, the total amount of liquid evaporated was measured by the micrometer screw gauge, while the change i n mixture composition was determined by continuous measurement of the liquid mixture density using the RH - v - I P T - H - G - D E P - c - B - A - Vent A I T T v v Warm Air Enclosure Peristaltic Pump Density Meter Water Bath Rotameters Orifice Plate Pressure Measurement Temperature Measurement Point Velocity Measurement Relative Humidity Measurement Wind tunnel I v > T I I D T )-T PPPP P P P P PPP F c I P -T + ---t ' P T I I G I R H P T 4 122 K.D. O’Hare, P.L. Spedding and J. Grimshaw n U v h el v h U N Figure 2 The evaporation cell. Evaporation of the Ethanol and Water Components 123 density meter. When the vertical movement was translated into an amount of liquid evaporated, the overall accuracy was within f0.5%. However, the composition was obtained to f0.05%. The evaporation cell contained approximately 200 cm3 of liquid at the beginning of a run and this diminished by approximately 50% over about 18 h. The bulk liquid temperature of the evaporating mixture was maintained at a constant value within f2"C during an experiment, while the average temperature varied for different experimental conditions. Further experimental details are given in O'Hare and Spedding." Results and Discussion Figure 3 shows the concentration profiles for three typical runs. While the temperature varied less than 1°C during a run, there was a temperature variation of approximately 3°C between the three runs; and the atmospheric pressure varied between runs but remained constant during any given run. Atmospheric pressure had its largest variation between the first and subsequent runs, resulting in the curves not coinciding between the conclusion of run 1 and the start of run 2. However, the profiles in Figure 3 show that the average bulk composition of an evaporating mixture changed continuously as evaporation proceeded. As the bulk ethanol concentration decreased during a run, the rate of evaporation of the ethanol component also decreased. For example, the ethanol evaporation rate decreased from 1.82 to 1.26 g h-' during run 1. Figure 4 illustrates that the average component evaporation rates were concentration dependant; the ethanol 1( 10 - RUN 3 0c \ K v z 0 6- z W V z u 0 -I 0 z 4- z 2- L I 2.5 5.0 7.5 10.0 TIME (hours) 12.5 15.0 17.5 20.0 Figure 3 Average li uid ethanol concentration (% w h ) against time at an air-jlow rate of 40 h-'. Ethanol component evaporation rate (start to pnish): Run I : 1.82-1.26 g h-'; Run 2: 1.09-0.70g h-'; Run 3: 0.50-0.33 g h-I. 2 K.D. 0 ’Hare, P.L. Spedding and J. Grimshaw 124 17.5 20.0 I 0 % ETHANOL COMPONENT WATER COMPONENT 0 I 00 ETHANOL CONCENTRATION (% W/W) Figure 4 Average evaporation rate (g h-I) against ethanol concentration (% w h ) for both the ethanol and water components of the solution, at an airflow rate through the wind tunnel of 40 m3 h-I. component evaporation rate increased markedly with increasing ethanol concentration. However, the individual evaporation rates of both the ethanol and water components were less than for the pure liquids alone. The trends shown in Figure 4 were explained in terms of vapour pressure driving forces as discussed in O’Hare and Spedding.’’ Figure 5 shows the evaporation rate of the ethanol component per unit mass of ethanol remaining in the evaporating mixture as a function of ethanol concentration. Again there was a slight temperature effect and a more pronounced effect of total pressure particularly affecting run 1. However the data clearly showed that for a contained body of liquid, the ethanol component evaporation rate (glh per g ethanol remaining) increased significantly as the ethanol concentration decreased below about 4% wlw. On Figure 5 , the average evaporation rate for pure ethanol (obtained for nine experimental runs) is shown as a reference line in order to illustrate the magnitude of the increased relative ethanol evaporation. It is important to note that at concentrations less than approximately 4% wlw ethanol, the ethanol component evaporation rate was significantly greater than for the pure ethanol alone. The results obtained were also experimentally significant being well within the experimental accuracy range of the apparatus. Detailed calculations of activity coefficients using the method suggested by Reid et aLZogave a +150% variation in ethanol mass transfer coefficient for run 3, suggesting that the data obtained in Figure 5 cannot be attributed to the effect of activity coefficients alone but must be caused by some other effect. Sanatsky14 studied the evaporation of alcohols from dilute hydrocarbon solutions and found that the alcohol component evaporation rate was as much as 50 times greater than expected. He postulated that the evaporation of a pure Evaporation of the Ethanol and Water Components 0.5 0.4 I25 ii Pure Ethanol ( 3 1 ° C ) I I 2 4 6 ETHANOL CONCENTRATION (% w / w ) 8 10 Figure 5 Ethanol component evaporation rate (g h-' per g of ethanol) against ethanol concentration (% w h ) at an airjlow rate through the wind tunnel of 40 m3 h-'. The horizontal line indicates the corresponding evaporation rate for pure ethanol at 100% w h for a comparison only. alcohol was hindered by the molecular attraction due to hydrogen bonding. However, when the alcohol was mixed with a non-hydrogen bonding compound, for example toluene, hydrogen bonding of the alcohol molecules was diminished due to their physical separation, causing the alcohol component to evaporate more quickly. It follows that the greater the fraction of the alcohol molecules that were separated through increasing dilution, the faster the effective evaporation rate of the alcohol became. In an ethanol-water mixture, there are three types of hydrogen bonding; ethanol-ethanol, ethanol-water and water-water, listed in order of increasing strength. In ethanol-water mixtures of low ethanol concentration the results obtained would be explained if the water-water hydrogen bond became dominant such that the ethanol-water hydrogen bonds were weakened in comparison, thus allowing the ethanol component to evaporate more quickly. Infrared spectroscopy studies failed to identify any valid spectral changes over the concentration range of interest, even under the most favourable experimental conditions. This is in agreement with the work of Franks and Ives21 who suggested that the complex interactions occurring in the alcohol-water system covered a wide continuous range of energies, and resulted in broad adsorption bands that were virtually unresolveable. Further, SymonsZ2 concluded from studies on bonding in this system that IR and NMR studies were unlikely to lead (by themselves) to the identification of valid mechanisms to explain any unusual experimental phenomena Various data for the system (detailed as follows) have shown that abnormalities occurred in the low alcohol concentration ran e. The density and partial molar volume;23 sound velocity and compressibility;2Fheat capacity, heats of mixing and temperature of maximum density;21 all showed evidence of the 126 K.D. O'Hare, P.L. Spedding and J. Grimshaw alcohol molecule being accommodated in existing cavities in the three-dimensional hydrogen-bonded low-density arrangement of water. The hydroxyl ion is hydrogen bonded to the enclosing water molecules such that the hydrophobic alkyl group of the alcohol is masked in a fluctuating clathrate 17H20), thus reinforcing t h e over all hydrate structure ( C~HSOH. a ~ ~ a n g e m e n tThe . ~ ~effect * ~ ~is?observed ~~ to reach a maximum around 13% wlw alcohol where presumably the cavity sites in the water are saturated with alcohol. Such structural arrangements, while being interesting in their own right, act at bulk concentrations well removed from the enhanced evaporation region reported in this work, and would in general be expected to minimise rather than enhance evaporative mass transfer. Butler and Wightman26 showed that while the surface tension data for low ethanol-water concentrations superficially appeared to be regular,27 the Gibbs adsorption equation predicted that the excess alcohol surface concentration rose significantly in the low alcohol bulk concentration region. Gurney,28and Snavely and Schmidt29 justified the validity of the equation theoretically and experimentally, respectively. Later Breitenbach and Edelha~ser,~' and Scatchard31 confirmed the earlier work of Butler and Wightman,26 while Vergara and L e ~ p i n a s s eextended ~~ the approach to enable the excess surface alcohol concentration to be calculated at infinite dilution. These latter workers concluded that the alcohol molecule was oriented perpendicularly to the interface and that surface conditions extended only a few molecules depth into the liquid. These based conclusions were in agreement with earlier suggestions made by upon extensive experimental work, and were later reinforced by theoretical work.34-37 For example, Gubbins and Thompson34 concluded that with the perpendicular orientation of the alcohol molecule at the liquid surface, the larger more strongly interacting atoms were upwards towards the gas phase. There was also some evidence that on the gas side the molecules lay parallel to the interface. Such theoretical studies have been used to accurately predict the surface tension for simple fluids,35 and therefore can be extended with confidence to provide insight into surface orientation. It is postulated, therefore, that the increased evaporation reported in this work for <4% wlw alcohol solutions is caused by excess adsorption of alcohol at the interface. The authors calculated values of r for alcohol in the surface layer (where r is Gibbs excess surface adsorption, usually in mol m2, but for constant volume conditions in A, see ref.31), using the Gibbs adsorption equation, and showed that the excess alcohol surface concentration rose initially in a consistent manner up to an inflection point at about 4% wlw alcohol, at which stage about 40% of the surface was occupied by alcohol molecules. Thus the alcohol preferentially passed to the surface layer, while in the bulk liquid it was being incorporated into the cavities within the water structure so that very little change occurred in the molar volumes. In such a structural arrangement, the alcohol at the interface would be readily available for evaporative mass transfer. Above 4% wlw alcohol, the rate of increase of r decreased and the evaporation of alcohol became constant. The calculated excess alcohol concentration in the lower layers of the bulk liquid under the surface did not become positive until the water cavities in the bulk phase were completely saturated with alcohol, occurring at about 1,3% wlw concentration. The partial molar volume measurements of Franks and Johnson23 for the aqueous ethanol system, indicated an inflection in the data at about 4-5% wlw alcohol. This corresponds to the calculated decrease in the rate of change of r Evaporation of the Ethanol and Water Components 127 with bulk concentration mentioned above. These changes indicated some mechanistic alteration of the orientation of the molecules at the liquid surface. It is possible that the alcohol molecules began to interact with the surface orientation if there were insufficient water molecules present to provide adequate separation. Conclusion For the ethanol-water system, it was experimentally determined that: (1) the component evaporation rates were concentration dependant; (2) the composition of the evaporating mixtures changed continuously as evaporation proceeded; (3) the rate of change of composition decreased as ethanol concentration decreased; (4) the ethanol component evaporation rate (per unit mass of ethanol remaining) increased as the ethanol concentration decreased; ( 5 ) at low ethanol concentrations ( ~ 4 %w/w), the ethanol component evaporation rate (per unit mass of ethanol remaining) was significantly greater than the pure ethanol evaporation; (6) the enhanced evaporation was caused by the preferential adsorption of alcohol % The course of the on the interface at these lower ethanol concentrations ~ 4 w/w. adsorption can be predicted using the Gibbs adsorption equation. References 1 Carrier, W.H. 1921. The theory of atmospheric evaporation with special reference to compartment driers. Ind. Eng. Chem., 13, 432-438. 2 Hinchley, J.W. and Himus, G.W. 1924. Evaporation in currents of air. Trans. Inst. Chem. Eng., 2, 57-64. 3 Hine, J.B. 1924. The rate of evaporation of liquids in a current of air. Phy. Rev., 24, 79-9 1. 4 Powell, R.W. and Griffiths, E. 1935. The evaporation of water from plane and cylindrical surfaces. Trans. Inst. Chem. Eng., 13, 175-191. 5 Lurie, M. and Michailoff, N. 1936. Evaporation from free water surfaces. Ind. Eng. Chem., 28,345-349. 6 Wade, S.H. 1942. Evaporation of liquids in currents of air. Trans. Inst. Chem. Eng., 20, 1-13. 7 O'Brien, L.J. and Stutzman, L.F. 1950. Mass transfer of pure liquids from a plane, free surface. Ind. Eng. Chem., 42, 1181-1187. 8 Maisel, D.S.and Shenvood, J.K. 1950. Evaporation of liquids into turbulent gas streams. Chem. Eng. Prog., 46, 131-138. 9 Severson, D.E., Madden, A.J. and Piret, E.J. 1959. Evaporation rates of liquids to flowing gas streams. AIChEJ, 5,413-418. 10 Smolsky, B.M. and Sergeyes, G.T. 1962. Heat and mass transfer with liquid evaporation. Int. J. Heat Mass Transfer, 5, 101l"21. 11 Asano, K. and Fujita, S . 1971. Mass transfer for a wide range of driving force evaporation of pure liquids. Chem. Eng. Sci., 26, 1187-1 194. 12 Hoffman, H.E. 1931. Evaporation rate of organic liquids. Ind. Eng. Chem., 24,135-140. 13 Lewis, W.K., Lombard, S. and Sanders, C.E. 1935. Evaporation of lacquer solvents. Ind. Eng. Chem., 27, 1395-1399. 14 Samotsky, A.A. 1969. Evaporation of solvents from paint films. J. Painr Technol., 41, 492-500. 128 K.D. O’Hare, P.L. Spedding and J. Grimshaw 15 Richardson, J.F. 1959. The evaporation of two component liquid mixtures. Chem. Eng. Sci., 10, 234-240. 16 Winkelmann, A. 1884. About the diffusion of gases and vapours. Ann. Phys., 1, 152-154. 17 Clark, M.W. and Judson King, C. 1970. Evaporation rates of volatile liquids in a laminar flow system. AIChEJ, 16, 64-75. 18 Mackay, D. and Matsugu, R.S. 1973. Evaporation rate of liquid hydrocarbon spills on land and water. Can. J. Chem. Eng., 5 1 , 4 3 M O . 19 O’Hare, K. and Spedding, P.L. 1992. Evaporation of a binary liquid mixture. Chem. Eng. J., 48, 1-9. 20 Reid, R.C., Prausnitz, J.M. and Poling, B.E. 1987. The Properties of Gases and Liquids, 4th edn, p.291. McGraw-Hill, New York. 21 Franks, F. and Ives, D.J.G. 1966. The structural properties of alcohol-water mixtures. Quart. Rev., 20, 1 4 4 . 22 Symons, M.C.R. 1989. Liquid water - the story unfolds. Chem. Brit., 491-494,498. 23 Franks, F. and Johnson, H.H. 1962. Accurate evaluation of partial molar properties. Trans. Faraday SOC., 58, 656-661. 24 Onori, G. 1986. Isentropic compressibility behaviour of dilute solutions of ethyl alcohol in water. Nuovo Cimento, 8,465-480. 25 Glew, D.N. 1962. Solution hydrates of aqueous non-electrolytes. Nature, 195, 698-699. 26 Butler, J.A.V. and Wightman, A. 1932. Adsorption at the surface of solutions Part I. The surface composition of water-alcohol solutions. J. Chem. Soc., 2089-2097. 27 Valentiner, S. and Hohls, H.W. 1938. Surface tension of alcohol water mixtures. Z Physik, 108, 101-106. 28 Gurney, C. 1947. Surface tension in liquids. Nature, 160(4057), 166-167. 29 Snavely, E.S.,Schmidt, G.M. and Hurd, R.M. 1962. Simple experimental method for verification of the Gibbs adsorption equation. Nature, 194(4827), 439-441. 30 Breitenbach, J.W. and Edelhauser, H. 1953. Surface tension and excess surface concentration in binary liquid mixture. Monatsh, 84, 384-391. 31 Scatchard, G. 1962. The Gibbs adsorption isotherm. J. Phys. Chem., 66,618-620. 32 Vergara, F. and Lespinasse, B. 1971. Calculation of the area occupied by molecules in the interfacial boundary layer. Estimation of the molar area of water and ethanol at infinite dilution from surface tension. Comptes Rendus, 272, 149-152. 33 Danicz, B.L. 1965. Discontinuities in the surface structure of alcohol water mixture. AD 622835 1,49 pp. 34 Gubbins, K.E. and Thompson, S.M. 1981. Molecular orientation at a gas-liquid interface. Faraday SOC. Symp. No. 16, Structure of the Interfacial Region, 59-70. 35 Kim, S.W., Eyring, H. and Lee, Y. 1969. Significant structure theory applied to surface tension of binary liquid mixtures. J. Chem. Phys., 51(9), 3967-3971. 36 Eckert, C.A. and Prausnitz, J.M. 1964. Statistical surface thermodynamics of simple liquid mixtures. AIChEJ, 10, 677-683. 37 Matsumoto, M. and Katuoka, Y. 1988. Study on liquid-vapour interface of water I. Simulational results of thermodynamic properties and orientational structure. J. Chem. Phys., 88(5), 3233-3245. Received: 2 December 1991; accepted: 10 September 1992.