Refining Processes

The discussion of the refining processes from the standpoint of the chemistry involved will be taken up first according to the general commercial process of complete cooling of distillate as soon as it leaves the retort, resulting in the separation of the non-condensible gas from the liquid products, followed by settling the liquid products to separate the pyroligineous acid from the tar. Then other more complicated systems will be discussed.

The condensation of vapors of volatile liquids when mixed with non-condensible gases is a more difficult problem than when the vapors occur alone. This is for the reason that the uncondensed gases tend to carry a part of the condensible products with them as they leave the condenser outlet. The condensible products carried in the vapors are in two distinct forms and require different means for their recovery. The first and most obvious form in which these products are carried in the gas is as a fog or mist, that is, in fine droplets of liquid so small that they settle out only very slowly, if at all. The exact cause of this fog formation is not known and it is a question whether different conditions of condensation may vary the amount of fog. It is certain, however, that it is not due to inefficient condensation, since the particles are in liquid form and may be carried by a perfectly cool gas. These small particles can be stopped only by some mechanical means such as bafRe plates, a scrubbing tower, or an electrostatic precipitator. The slow and incomplete settling noticed in the gas mains is due to the larger particles only, and even greater retardation of the speed of the gas flow, as in a gas tank, would probably not settle the finer particles in any reasonable length of time. Baffle plates in large gas mains would probably stop some more of the fog by the contact of the particles with the wet plates, but a scrubbing tower filled with coke or provided with water sprays would be more efficient.

It is doubtful, however, whether any of these means of stopping the fog would give a complete separation, The problem is similar to 73

that in other industries where liquid particles like S03 fog or solid particles like cement dust are carried in gases. In these cases complete and rapid settling can be obtained only by an electrostatic precipitator.

It has been reported 1 that a Cottrell precipitator was installed on the outlet from the condenser of a wood distillation oven with complete settling and recovery of the fog. The amount of liquid recovered was about 1.7 per cent of that coming from the condenser in liquid form, and the composition was about the same except that there was not quite so high a proportion of the more volatile products, such as alcohol.

The second form in which condensible products may be carried away from the condenser by the gas is as a true vapor. This is due to the fact that the condensed liquid has a slight vapor pressure even when well cooled and the gas is, therefore, saturated with the vapors, the proportion of vapors in the gas depending entirely on the vapor pressure and the vapor pressure varying with the temperature. If, for instance, the liquid condensate has a vapor pressure of

17.4 mm, the gas will consist of or 2.3 per cent of vapors by volume. This vapor pressure of 17.4 mm is actually that of water at 68° F., which is an indication of the possibilities of loss in this way. The vapor pressures also increase rapidly with increasing temperature, so that, for instance in the case of water, the vapor pressure is more than doubled between 68° F. and 90° F. I11 fact, the loss of vapors would naturally be greater than this since the more volatile constituents of the pyroligneous acid, methyl alcohol and methyl acetate, have much higher vapor pressures than water (88.7 mm and 169.8 mm, respectively, at 2d0 C.) and the vapor pressure of the pyroligneous acid would probably be higher than that of pure water. In this case also the composition of the lost vapors would be different from that of the pyroligneous since the more volatile constituents would be lost in greater proportions. It should also be noted in this connection that in this unneutralized pyroligneous acid a fairly large proportion of the methyl alcohol exists in the form of methyl acetate which is subsequently hydrolized to methyl alcohol by the neutralization. The methyl acetate having a higher vapor pressure than methyl alcohol goes into the gas in greater quantities than pure methyl alcohol would.

Klar2 says that losses in the gas of 5 per cent of the total acetic acid and methyl alcohol have been observed by him.. Lawrence8 says that 5 per cent of acetic acid and 45 per cent of the alcohol may be lost m the gas but such a high loss of alcohol seems improbable

1 .Unpublished report of the Research Corporation.

il if i i i 1 ! i except in case of very poor cooling. In the case of the kiln process where the proportion of gases is higher, due to the admission of air to the kiln and the partial combustion of the charge, a loss of wood alcohol constituents amounting to 35 to 40 per cent of the total has been observed.4

The vapors of alcohol, methyl acetate, and acetone can be partly recovered by running the gases up through a scrubbing tower in which they are washed by a descending stream or spray of water, since all of these vapors are soluble in water and their vapor pressure is diminished by solution in water. Probably a non-volatile or high-boiling oil such as is used in toluene recovery from coal gas could also be used efficiently for scrubbing a wood gas.

Liquid Products

The liquid products from the condensers on standing separate into two liquid layers, the tar and the pyroligneous acid, the tar on account of its higher specific gravity settling to the bottom. On account of mutual solubility, however, the separation is not complete, a part of the oils and of the non-volatile constituents of the tar remaining dissolved in the pyroligneous acid and part of the water, acid and alcohol, of the pyroligneous acid remaining dissolved in the tar. According to Klar D European hardwood tars contain about 18 per cent water, 2 per cent acetic acid and 0.7 per cent wood alcohol. The tars produced in the typical American plants will often contain 3.5 to 4.0 per cent acid, and hardwood tars produced in the laboratory have run as high as 7 per cent acid. There are probably several reasons for these variations but one of them is certainly the difference in the acid concentration of the pyroligneous acid. It would b£ expected that a tar in contact with a pyroligneous acid containing a high concentration of acetic acid would also contain a high concentration of acetic acid.

The amount of "soluble tar'3 and oils in the pyroligneous acid is also variable. Klar6 gives "7 per cent and more" for the soluble tar in pyroligneous acid from air-dry beech wood. In American practice tfie.soluble tar will run from 6 to 10 per cent of the pyroligneous acid; in laboratopr distillations of hardwood pyroligneous acid is frequently obtained with 12 to 15 per cent of soluble tar. It might be expected that the amount of soluble tar would be greater when the concentration of acid or alcohol, or both, in the pyroligneous acid was greater, but there is no general rule to this effect. It is not even known whether the concentration of dissolved tar is limited by the amount of water-soluble pitch present or by the composition of the pyroligneous acid, although this could be indicated by the simple experiment of de* Hawley, Chem., & Met. Eng. 25, 198 (xgai). ■ hoc. cit., p. 57. *Loe. eitr, p. sa.

termining the solubility of settled tar in distilled or "tar free" pyro-ligneous acid.

Tar Distillation

On account of this solubility of alcohol and acid in the settled tar, the latter must be distilled in order to recover the valuable products contained in it. In cases where no special attempt is made to complete the distillation of the tar with the production of heavy oils and pitch, but where only the recovery of the alcohol and acetic acid is required, the settled tar is distilled with steam. When the tar is distilled with steam for this purpose the light oils are commonly distilled completely and also a varying amount of heavy oils, depending on the amount of steam used. The theory underlying this process is difficult to develop, since we are distilling with steam not merely an oil completely insoluble in water, which would be -a simple case, but an oil containing in solution a substance, acetic acid, also soluble in water, and it is this acid soluble in both the oil and water which we are interested in distilling and separating. Apparently, however, the solubility of the steam in the acetic acid ¡prevents the distillation of the latter along with the oils of the same boiling point, because although acetic has a boiling point of n8° C., yet it is not all removed from the tar even when all the oils boiling below 2000 C. have been removed by the steam distillation.

Whatever may be the theory of this distillation the best conditions for carrying it out efficiently have been determined.7 It has been found that when other conditions are constant a distillate more concentrated in acetic acid and a higher percentage recovery of acid are obtained by slower distillation. This is probably entirely a matter of obtaining equilibrium between the steam and the volatile constituents of the tar and might not be apparent in a tall and narrow still or below certain limits of speed. In a still containing 510 gallons of tar in a forty-inch layer and with speeds of steam averaging 25 and 50 gallons per hour, respectively, in two runs the slower speed gave 149 pounds of acetic acid in 219 gallons of water in 9 hours, 45 minutes, while the faster speed gave 139 pounds of acid in 273 gallons of water in 5 hours, 45 minutes.

The pressure of steam in the closed coils of the still or, in other words, the temperature of the tar while the steam is being blown through also has a very definite effect. In small laboratory apparatus, two distillations were carried on under exactly the same conditions except that in one the temperature of the tar corresponded to 10 pounds steam pressure, in the other to 50 pounds. In the distillation at lower temperature a total volume of 446 cc distillate contained only 8.6 gr.

THawley and Calderwood, Jour. Jnd. Eng. Chem. it, 684 (1930).

acid, while the higher temperature gave in 272 cc distillate 10.3 gr. acid.

There is no special problem in recovering the alcohol contained in the tar since on account of its low boiling point it is probably removed long before all the acid is recovered. The very likely possibility that the acid last removed from the tar by steam distillation consists largely of higher homologues of acetic acid has never been studied, although Fraps 8 identified some of the higher acids in the heavy tar oils.

Few details are available on the distillation of tar by direct fire heat for the production of light and heavy oils and pitch. As with other oils containing water in solution the first part of the distillation, while the water is distilling with the oils, must be carried on slowly and carefully to prevent foaming and "boiling over." After the water is all removed the distillation can be hastened without danger until near the finish, when, as the last of the heavy oils are removed, there is again danger from too rapid heating or carrying the distillation too far, since the pitch may be "coked." This "coking" of the pitch is apparently an exothermic reaction and may readily become uncontrollable. If after the distillation is finished the pitch is drawn off and exposed- to the air while still very hot, there is also danger from its spontaneous ignition. Judd and Acree 0 have shown that in small scale distillation of tar, agitation of the charge with steam makes it possible to carry the distillation to the "hard pitch" stage without danger of coking the pitch.

When the tar is distilled by direct fire without the use of steam, the water which comes off during the first part of the distillation contains a large proportion of acetic acid but the higher fractions of oil, even those boiling well above 200° C., still contain some acid probably mostly acetic. Since there is no water collected with these oils the acid remains dissolved in them and can be recovered only by washing with water or a solution of alkali. In one of Judd and Acree's distillations with agitation by a jet of steam, where nearly one and one-half times as much water as oil was collected with the heavy oil, there was still 0.7 per cent acid in the oil.

The light oils have a dark color, especially after standing in the air, and a very pungent disagreeable odor. On redistilling, the color is lightened but not permanently. Neutralization with sodium carbonate before redistilling makes the color of the distilled product .lighter and more permanent and also improves the odor. Preliminary experiments 10 have indicated that hydrogénation of the light oils gives an almost colorless product with a pleasant ketone odor, but no details are available on this treatment.

"Forest Products Laboratory "technical note," Chem. &r Met. En$, 16, 708 <1917).

The heavy oils direct from the still are frequently the final market product, but they may require redistillation or neutralization or both. The fraction with boiiing points between 195° and 230° C., is especially valuable for the production of "beechwood creosote" which is a specially refined part of the alkali soluble constituents boiling between 200° and 220° C., and probably consists mostly of guaia'col, creosol, homopyrocatechin dimethyl ether, and the xylenols, since these are the main known constituents of the oil having boiling points between .these limits. The first three of these constituents probably occur in largest proportions in a well refined product, since they have the highest gravities and the refining process 11 is partly for the purpose of obtaining a high-gravity product.

The pitch usually requires no special refining process to make it a finished product. Its hardness is controlled by the point to which the separation of the oil is carried in the distillation. Judd and Acree 12 note that when a hard pitch is prepared, the last small fraction of distillate solidified "to a jelly-like mass on account of the presence of paraffin."'

Pyroligneous Acid Refining

The pyroligneous acid, partly separated from the tar by settling but still containing in solution certain non-volatile tar constituents, may be completely separated from tar by distillation. This distillation is practically nothing more than the separation of volatile from nonvolatile material and no special problem of fractionation is encountered. A very small amount of oil distills over in this process and is separated from the aqueous distillate, but the "soluble tar" left behind in the still after all, the aqueous distillate has been removed is practically free from oils. The soluble tar, however, contains in solution 7 to 10 per cent acetic acid which can be separated by the same process as is described for distilling settled tar for recovery of acetic acid (seep. 73).

The tar-free pyroligneous acid is next neutralized with lime for the formation of a non-volatile product of the acetic acid, so that the volatile wood alcohol can be separated by distillation. This neutralization, however, accomplishes much more than the formation of acetate of lime, since it is the cause of several complex reactions which affect the quality of the products subsequently separated. The details of these reactions due to neutralization have never been studied and it is only known that commercial practice has very definite but somewhat variable rules for deciding when the neutralization has been properly accomplished. The addition of lime is accompanied by such

0 0

Post a comment