Two types of adsorption distillation, i.e. fixed-bed adsorption distillation (FAD) and suspension adsorption distillation (SAD), are discussed in this chapter. In FAD the separating agent is stagnantly supported in the column as the catalyst in FCD, while in SAD the separating agent is constantly moving in the column as the catalyst in SCD. So in SAD the separating agent has the similar behavior in tray efficiency and hydrodynamics as the catalyst of catalytic distillation.
1. FIXED-BED ADSORPTION DISTILLATION 1.1. Introduction
Fixed-bed adsorption distillation (FAD) has presently been proposed by Abu Al-Rub [1-3]. It involves replacement of the inert packing material in packed-bed distillation column by an active packing material. The active packing materials used by Abu Al-Rub for separating ethanol and water are 3 A or 4 A molecular sieves, which are thought to be able to alter the VLE of ethanol and water considerably.
On the other hand, we think that some ion-exchange resins also can be assumed to be composed of one anion which generally has large molecular weight and one cation which is generally a single metal ion. Since salt effect may plays a role in increasing the relative volatility of aqueous solution, ion-exchange resins can also be selected as the packing material in the adsorption distillation column.
Evidently, in FAD molecular sieves or ion-exchange resins as the separating agents can be used for the separation of the mixture with close boiling point or azeotropic point. Furthermore, FAD is an extremely environment-friendly process because no extra organic solution is introduced except the components to be separated, and thus there doesn't exist solvent loss, unlike azeotropic distillation and extractive distillation.
Only was the separation of the ethanol / water system by FAD investigated because ethanol is a basic chemical material and solvent used in the production of many chemicals and intermediates [1,2].
The experimental results obtained from the VLE of the ethanol / water system at 70 °C in the presence of 6.5 g of 3 and 4 A molecular sieves showed considerable change from those without the molecular sieves. The azeotropic point for this system was eliminated. Moreover, no further improvement in this separation was achieved by increasing the weight of the molecular sieves beyond the optimum value. The alteration of the VLE is a result of the external force field exerted by the molecular sieves on the mixture's components. These results prove the feasibility of using active packing materials to alter the VLE of binary mixtures.
To interpret the results obtained thermodynamically, we will use the excess Gibbs energy to express the nonideality of a system. Suppose that the system is composed of component 1 (the light component), component 2 (the heavy component) and the solid (separating agent), and it is related to the interactions of the system components:
All the solid interactions are included in AG'1. Thus,
where yu is the activity coefficient in the absence of molecular sieves, and yis is the activity coefficient due to the presence of external force fields created by the solid on the liquid. Therefore, from an experimental measurement of the activity coefficients in the absence and in the presence of external force fields, yix can be obtained.
For the ethanol (1) / water (2) system, it shows a positive deviation from ideality in the absence or presence of molecular sieves. However, due to the effect of external force fields, the deviation from ideality is reduced, i.e. y. < yif and yis < 1 under the same temperature and concentration.
Another factor that may affect the VLE of a binary mixture in porous media is the curvature of the gas-liquid interface inside the pore which may cause a reduction in the vapor pressure of the components to be separated , But when the 3 and 4 A molecular sieves are employed in FAD, ethanol maybe can't be adsorbed into the pore because of its relative large molecular diameter (>4 A). Therefore, it can be assumed that only water (diameter 2.75 A) is competitively adsorbed.
The vapor pressure of water (2) in the microporous molecular sieves is expressed by the well-known Kelvin equation [5-8]:
\Pu where a2 is surface tension (N m"), V, is molar volume of water (m mof ), r is a radius of curvature over a concave vapor-liquid interface, R is gas constant (8.314 J mof1 K"1) and T is temperature (K); or
y2/x2 P2T2 P2Y21Y2S rRT Yis rRT
2cr,F, which indicates that af2 > ai2 because y.H < 1 and exp(——-) > 1 . That is why the rRT
relative volatility of ethanol to water is improved in the presence of molecular sieves.
However, there is another explanation for the reason about improvement of relative volatility . It is thought that the idea of an external force field emanating in a liquid from 2 mm diameter molecular sieve beads is far removed from the VLE inside capillaries, which implies that the theory of external force field isn't valid in the liquid bulk. Rather, the addition of molecular sieves to the water / ethanol system will result in the preferential uptake of water by the sieves, thereby reducing the liquid phase concentration of water. The quantified vapor phase would then be in equilibrium with a liquid phase of lower water content, irrespectively of how capillary surfaces may alter VLE. So that is why the relative volatility of ethanol to water is improved in the presence of molecular sieves.
We think that, whatever the theory, it is the truth that molecular sieves can play a role in increasing the relative volatility and thus making it possible to separate the components with close boiling point or azeotropic point. In other words, it is anticipated that FAD, as a promising special distillation process, may replace azeotropic distillation and extractive distillation in some cases in the near future.
Comparison of adsorption distillation and extractive distillation for separating ethanol and water is very attractive because adsorption distillation is an attractive "new" technique and extractive distillation is an "old" technique. The most commonly used separating agents in the extractive distillation are ethylene glycol, the mixture of ethylene glycol and one salt, CaCl2 and N, N-dimethylformamide (DMF) [9, 10].
Comparison of azeotropic distillation and extractive distillation has been made in chapter 2. Today, azeortopic distillation is almost replaced by extractive distillation for separating ethanol and water. So it is unadvisable to use azeortopic distillation for separating ethanol and water.
Comparison of adsorption distillation and extractive distillation is made on the basis of process experiment and VLE experiment.
The experimental flow sheet of extractive distillation process with two columns (extractive distillation column and solvent recovery column) has been established in the laboratory and is shown in Fig. 1. The extractive distillation column was composed of three sections, ( I ) rectifying section of 30 mm (diameter) x 800 mm (height), (II) stripping section of 30 mm (diameter) * 400 mm (height) and (III) scrubbing section of 30 mm (diameter) * 150 mm (height). The solvent recovery column was composed of two sections, ( I ) rectifying section of 30 mm (diameter) x 650 mm (height) and ( II ) stripping section of
30 mm (diameter) x 300 mm (height). The two columns were packed by a type of ring-shape packing with the size of 4 mm (width) x 4 mm (height). The theoretical plates are determined by use of the system of n-heptane and methylcyclohexane at infinite reflux, having 18 theoretical plates (including reboiler and condenser) for extractive distillation column and 14 theoretical plates (including reboiler and condenser) for solvent recovery column.
In the extractive distillation process experiment, the mixture of ethylene glycol and CaCh was used as the separating agent. The reagents, ethylene glycol and CaCb, were of analytical purity and purchased from the Beijing Chemical Reagents Shop, Beijing, PRC. Composition analyses of the samples withdrawn from the top of extractive distillation column were done by a gas chromatography (type Shimadzu GC-14B) equipped with a thermal conductivity detector. Porapark Q was used as the fixed agent of the column packing, and hydrogen as the carrier gas. The column packing was 160 °C and the detector 160 °C. The data were dealt with by a SC 1100 workstation. In terms of peak area of the components, the sample compositions could be deduced.
The experimental results of extractive distillation column are given in Table 1, where the feed concentration is ethanol of 0.3576 mole fraction and water of 0.6424 mole fraction. It is shown that high-purity of ethanol (above 99.0%wt), up to 0.9835 mole fraction (almost 99.5%wt), was obtained under low solvent/feed volume ratio 2.0 and low reflux ratio 0.5, which means that the mixture of ethylene glycol and CaCl2 as the separating agent is very effective for the separation of ethanol and water by extractive distillation and has been governed in industry for a long time.
Extractive distillation column Solvent recovery column Fig.l. The double-column process for extractive distillation.
Experimental results of extractive distillation column
Experimental results of extractive distillation column
Solvent/feed volume ratio
Ttop ! K
The top composition x\ (ethanol) x2 (water)
Was this article helpful?