Practical Problems Connected with Essential Oil Distillation

(d) Practical Problems Connected with Essential Oil Distillation.

At this point it is advisable to devote space to a few practical suggestions for the operation of essential oil stills. Many of the points brought out in the following paragraphs have already been mentioned, but it appears desirable to emphasize them, since failure to adhere to them may well represent the difference between successful and unsuccessful operation. The three general methods of conducting steam distillation will be considered in order.

Water Distillation. 

Let us first consider the operation of a water distillation system. In every method of plant distillation, whether steam distillation, water and steam distillation, or water distillation, only those quantities of the essential oil with which the steam comes in direct contact can be vaporized. Any oil held within the plant tissue must first be extracted from the glands and brought to the surface of the plant by osmosis. But the forces of hydrodiffusion work very slowly whenever the distances to be bridged are relatively long. Water distillation necessitates a thorough comminution of the plant material to the smallest possible size; in other words, the reduction must exceed that required for direct steam distillation or water and steam distillation. All interspaces between the plant particles, which in the case of water distillation are filled with water, must be penetrated continuously by rising steam.

The retort is charged with the plant material to be distilled, and sufficient water added to fully cover the entire charge, leaving, however, ample vapor space above the charge to avoid boiling over and carrying over of spray into the condenser. After the cover has been fastened tightly using a suitable gasket between cover and still to avoid loss of vapors at that point the retort is connected to the condenser, and the cooling water permitted to flow through the latter. The fire is then started, if a direct fired still is being used, or the steam line opened, if either a steam jacket or steam coils are used for heating. Once the charge has reached its boiling point at the particular pressure used, condensate will begin to issue from tjie open end of the condenser, and should be run directly into the separator, which was previously filled with water. The rate of distillation can be controlled by the intensity of the fire, by the pressure in the steam jacket, or by the rate of introduction of steam. With direct fire, special care must be taken to avoid overheating the plant material. As the water and 01' evaporate, part of the charge will soon cease to be covered with water, and hence no longer will be automatically protected from overheating. It may be advisable to add more water as the distillation proceeds, to prevent any part of the charge from becoming exposed to the full heat of the fire. When a steam jacket or closed steam heating coils are used, there is less danger of overheating unless the water level falls below the top steam coils. Here again the addition of sufficient water will prevent such an undesirable result. With open steam coils this danger is largely avoided, since for every pound of steam injected, a pound of steam condenses as distillation water in the condenser. However, care must be taken to prevent accumulation of condensed water within the retort, or the water level will rise gradually to the top. Therefore, the still should be well insulated and not exposed to draft or cold wind. Furthermore, the water charged into the retort at the beginning of the operation should be hot, as cold water would condense too much of the injected live steam.

The rate of distillation must be adjusted to suit the particular equipment and material being distilled. This rate should, of course, be maintained near the maximum in order to obtain the maximum production of oil. There are other, perhaps less apparent, reasons for maintaining a rapid rate of distillation when using water distillation. Principal among these is the fact that only by rapid distillation can the charge be maintained in a sufficiently loose condition to insure thorough penetration of the plant material by the rising steam. Steam which does not contact the charge, for example steam generated at the water surface as in a slow distillation, cannot carry any essential oil with it, and will be wasted. A lively conduct of distillation prevents to a large extent undesired agglomeration of the plant material, and brings about a more effective contact area between charge and steam. This in turn causes not only an increase in the rate of production, but also a better total yield of oil. It is commonly assumed that during water distillation all parts of the plant charge are kept in motion by boiling water. This, however, is only partly true. Steam bubbles form mainly along the closed steam coils, along the heated bottom and walls of the retort, and rise to the surface by the shortest way, avoiding any obstacles. Provided the distillation material is charged loosely and remains loose in the boiling water, the steam bubbles probably will contact all plant particles quite evenly and vaporize their volatile oil. This is the case especially with woody material, but flowers have a tendency to agglutinate under the influence of steam and form large lumps. True, the volatile oil diffuses quite readily from tenderwalled epidermis glands, but when leaves or flowers cling together diffusion is slowed down. Distillation must then be accelerated to a point where all particles of the plant charge are agitated and kept in continuous motion by rising and exploding steam bubbles. The degree of comminution, the weight of the plant charge and the construction of the still should be calculated accordingly. Plant material which contains an essential oil composed of high boiling constituents can be exhausted by water distillation only if comminuted to small particles.

Von Rechenberg pointed out that many years ago distillation was carried out almost exclusively over direct fire, water distillation then being the rule, steam distillation the exception. Experience had been that complete exhaustion of many plant materials could be effected only under great difficujties and after several days of distilling. Extraction of oil of cloves, for instance, seems to have caused a great deal of trouble. Directions dating back to the middle of the last century claim that cloves could be exhausted only by repeated distillation; in other words, the retort had to be opened from time to time, the content stirred, and the evaporated water replaced. In the case of cloves this was repeated from three to eight times. Very probably the plant charge relative to the size of the still was much too large and distillation had to be carried out much too slowly with a small fire; otherwise the cloves in the still would have foamed over into the condenser. Small-scale operators, especially field distillers employing directly fired retorts, still commit the mistake of not putting sufficient water into the retort. Ignorant of the simple rules underlying water distillation, they seem, to prefer a slow distillation, or they are handicapped by too small condensers, or by lack of water. Frequently they add to the plant material such a small quantity of water that only the still bottom, which is directly fired, remains covered with water at the end of the operation. This practice is faulty. Plant parts rising above the level of the boiling water in the course of water distillation tend to lump together, to become almost impenetrable for steam, and therefore not to yield their oil completely. For this reason, the retort should be only partly filled with plant material, which should remain fully immersed in water, even when distillation is completed. Only by following this precaution is it possible to exhaust the plant charge by water distillation, as far as this can be done at all.

Water distillation is still quite widely used with portable equipment in primitive countries. There, lack of roads and poor transport facilities prevent hauling of the plant material from outlying growing regions to centrally located distilleries. Therefore, the apparatus must be moved into the growing sections, or in other words, follow the plant material. Small stills, simple, sturdy and low priced, hence retain the favor of ira-iy native producers.

Aside from these purely practical conveniences, water distill aiion possesses one decided advantage. It permits processing of very finely powdered material (root, bark, and wood, etc.) or of piant parts which by contact with direct (live) steam would easily agglutinate and form lumps through which the steam cannot penetrate (e.g., roses or orange blossoms). From such an agglutinated mass, live steam vaporizes the oil only from the outside and not from the inside. Steam distillation, therefore, would remain incomplete. The nascent steam bubbles attack all parts of the plant charge only if the latter moves loosely and freely in boiling water. As a matter of fact, material which readily agglutinates can be processed only by water distillation.

On the other hand, water distillation suffers from several disadvantages. Whether comminuted or not, the plant material cannot always be completely exhausted. Furthermore, certain esters, linalyl acetate, for example, are partly hydrolyzed; other sensitive substances, such as aldehydes, tend to polymerize under the influence of boiling water, etc. Consequently, all other conditions being the same, the quality of product from a rapid distillation will be better, in general, than that of the product from a slow distillation. Water distillation requires a greater number of stills, more space, and more fuel. It demands considerable experience and familiarity with the method and its effect, in fact more experience and care than any other form of plant distillation ; otherwise the yield of oil will be affected and fall considerably below that obtained by water and steam distillation or by direct steam distillation. Water distillation is the least economical process, water and steam distillation giving, in general, better results in the case of field distillation.

Another peculiarity of water distillation lies in the fact that high boiling and somewhat water-soluble oil constituents cannot be completely vaporized from the large quantities of water which must cover the plant charge in the still, or they require so much steam that they can be recovered only partly from the distillation (condensed) water; therefore, the distilled oil will be deficient in regard to these constituents. In other words, distillation remains incomplete. Such compounds are high boiling alcohols (phenylethyl alcohol, cinnamyl alcohol, benzyl alcohol, etc.), phenols (eugenol, etc. ), certain nitrogenous substances, and some acids. A typical example is orange blossom oil : the methyl anthranilate present in the flowers cannot be completely recovered by distillation, extraction with volatile solvents giving better results. The case is similar with roses : distilled rose oil lacks the somewhat spicy note of the extracted product (concrete or absolute) and contains much less phcnylethyl alcohol, because eugenol and phenylethyl alcohol, as the author16 proved, remain in the residual still waters. How much of these high boiling, somewhat water-soluble compounds are actually carried over by distillation depends upon their boiling points, their degrees of solubility in water, and the quantity present in the plant material. If the plant charge, despite comminution, contains coarser particles, which during the boiling do not soften and, therefore, are not torn apart, these particles will retain high boiling, water-msoluble oil constituents, because diffusion through the greatly swollen tissue layers acts too slowly. These factors explain why essential oils obtained from the same plant material by water distillation or by steam distillation vary considerably in regard to yield, physical properties and chemical composition.

For all these reasons, water distillation is used today in essential oil factories and for large-scale production only in cases where the plant material by its very nature cannot be processed by water and steam distillation or, even better, by direct steam distillation.

For most efficient operation, a modern retort serving for water distillation should be flat and wide, thereby offering a large surface of evaporation. The plant material should be filled in evenly, not higher than 4 in. Water is then pumped into the still until it stands about 2 in. above the charge. Steam of at least 3 atmospheres absolute pressure, generated in a separate steam boiler, is injected into the steam jacket beneath the still, so that the water in the still is brought to lively boiling, and each particle of the plant charge thoroughly and continually agitated. The quantity of the plant charge does not necessarily depend upon the size of the still. A somewhat loose charge contains sufficient interspaces to permit an unhindered penetration by the steam bubbles rising from the still bottom; hence the charge can be higher than 4 in. If, in addition, the plant material does not agglutinate or lump while softening under the influence of heat, the charge may be considerably higher. However, complete exhaustion is not always assured ; in general, good results in the case of water distillation are obtained only if the charge is sufficiently low to permit the rising steam bubbles to overcome the weight of the plant charge. In other words, the steam should continually agitate the plant particles. In this case it is preferable to work without a perforated grid above the bottom of the retort. If, on the other hand, the charge is high, exercising a marked pressure upon the bottom, the insertion of a perforated grid is advisable. For certain types of plant materials e.g., roses, orange blossoms, and ylang ylang flowers which can be kept floating in the boiling water by lively steam development, much deeper or spherical stills may be employed. Heating coils (instead of a steam jacket) should be avoided in this case because plant particles easily attach themselves to the coils and may give trouble.

-------------------

18 Guenther and Garnier, "Bulgarian Rose Oil," Am. Perfumer 25 (1930), 621.

Very finely powdered material such as almond or apricot kernels has a tendency to "burn" in contact with the hot steam jacket; the water in the still should, therefore, be heated, not by indirect steam, but by direct steam, injected through a steam coil within the still. High, cylindrical stills arc better adapted to this purpose than wide, flat ones. In this case, the distillation water is collected separately, and not pumped back into the still during the process, because too much liquid would accumulate in the still by condensation of the injected live steam.

A general rule which applies to all methods of distillation is that each charge should be completed on the same day. The quantity of plant material charged and the rate of distillation must be calculated accordingly. It should be kept in mind that the shorter the distillation, the less the forces causing hydrolysis, decomposition and rcsinificntion will come into action. The loss of essential oil arising from these forces may amount to several per cent, as calculated upon the oil. In the case of water distillation, it is not always possible, however, to shorten distillation to a one-day operation.

Fig. 3.18 shows a still for water distillation, with automatic return of the distillation waters.

Water and titeam Distillation. 

Let us now consider some practical aspects of water and steam distillation, a method which in recent years has become quite popular among small producers using portable distillation equipment that can be moved from field to field, following the harvest. The smaller units are heated by direct fire, the larger ones by a steam jacket, a closed steam coil, or in rarer cases by open steam coils. When using direct fire, precaution must be taken to insure that only the bottom of the still, the section containing water below the grid which carries the plant charge, is heated. Otherwise, one of the major advantages of water and steam distillation over water distillation, namely, freedom from the danger of overheating the plant material, will be lost. As was stated previously, when this method of distillation is employed the plant charge itself is kept out of contact with boiling water. Hence, if the upper part of the still were exposed to direct fire, the plant material might be dangerously overheated. It is advisable, therefore, to use indirect steam as r, source of heat, but not direct fire.

In this type of distillation, observing the precautions mentioned in the last paragraph, steam alone contacts the charge, the steam either being generated from, or passing through, water in the still. Thus overheating or drying of the charge is avoided because the temperature cannot rise above that of saturated steam at the pressure prevailing in the still (at atmospheric pressure never above 100). Water and steam distillation, therefore, represents a typical case of distillation with saturated low pressure steam. 

 FIG. 3.18. Still for water distillation.

For this reason, the condensate contains fewer decomposition products than that obtained by direct steam distillation with live steam, and particularly with high pressure or superheated steam.

Preparation of the plant material is much more important in this method of distillation than in water distillation. Since the steam contacts the material only by rising througn it, the plant charge must be so disposed that all parts of it are uniformly contacted, if high yields of oil are to be maintained. This requires that the charge be homogeneous as to size and, furthermore, that the average size of the individual pieces be controlled within rather narrow limits. If, for example, the material is finely ground, it will tend to pack and offer strong resistance to the passage of steam. This in turn may develop steam pressure beneath the charge until such pressure is sufficient to penetrate it. Such penetration, however, will take place at only a few places, releasing the pressure and permitting the steam to escape through only a few passages or channels sometimes called "rat holes." Obviously, under these circumstances most of the plant material is never contacted by the steam, and the recovery of oil is incomplete. If, on the other hand, a charge consists of, say, whole stalks, leaves and flowers, there obviously will be some fairly large passages through the charge which offer little or no resistance to the passage of steam. Steam will then escape through these and again permit m,ost of the charge to remain unaffected by it. Therefore, in the case of water and steam distillation, the plant material should not be too finely ground ; nor should it contain excessively long stalks or large roots or pieces of bark. Granulation usually gives the best results. Experience alone can determine the optimum size to which the material should be reduced, and this will vary from plant to plant. At any rate, the preparation of the charge for water and steam distillation must always be given most careful attention.

Another problem to be considered in water and steam distillation arises from the fact that the charge is cold at the start, and that the first steam to enter it will condense, thus wetting the plant material. This wetting will continue until the entire charge reaches the boiling temperature of water at the operating pressure. With certain types of plant materials for example, leaves or ground seeds, bark, roots, etc.- excessive wetting may result in lumping or agglomeration of the charge and, therefore, in a subnormal oil yield. Such wetting, again, may cause channeling of the steam. If a charge tends to agglomerate when wet, it is sometimes advisable to add dried twigs or short small pieces of stalk, or any other loose but absolutely neutral material, in limited quantities, so that the charge may be kept porous. To avoid continuation of wetting due to loss of heat by radiation from the walls of the still, the upper part of the retort in other words, the section housing the charge should be insulated.

The rate of distillation in the case of water and steam distillation is not as important as in the case of water distillation. It affects only the rate of production but not always the quality or yield of oil. A lively pace of distillation recommends itself, however, in order to prevent excessive wetting of the plant charge and in order to increase the production rate. Regarding oil production per hour, water and steam distillation is less efficient than steam distillation; it approaches that of water distillation. 

evenly and completely by steam. Although the method of direct steam distillation serves for a variety of plant materials, water and steam distillation is suitable only for certain types. It is especially adapted to field distillation in small or medium sized stills.

Water and steam distillation can also be carried out under reduced or increased pressure. Indeed, in some cases, reduced pressure gives excellent results.

Fig. 3.19 shows a retort for water and steam distillation. 

 PLATE 4. Field distillation of rosemary in Tunis. The stills are directly fired.

The plant material is transported by camels to the distillation post.

Steam Distillation.

Live steam, usually of a pressure higher than atmospheric, is generated in a separate steam boiler, and injected into the plant charge within the retort. This type of distillation is referred to as direct steam distillation, or distillation with live steam, or dry steam distillation. Most aromatic plants are distilled today with direct live steam at atmospheric pressure.

The application of steam distillation is subject to exactly the same reservations mentioned in the discussion of water and steam distillation, plus one additional factor. When using steam distillation, it is always possible, after the initial period during which the charge in the retort is warming up and condensation taking p.lace, that the steam may be slightly superheated. Indeed, in some cases the steam may be purposely superheated, as already mentioned, in order to improve the oil to water ratio. In expanding from the much higher boiler pressure to the lower pressure prevailing within the retort, the steam tends to become superheated. Two factors then assume importance. First, the temperature of the charge will no longer be maintained at the boiling point of water, under the operating pressure, but will rise to the temperature of the superheated steam. The operator, therefore, must guard carefully against overheating. Second, superheated steam has a tendency to dry out the charge and reduce the rate of recovery of the essential oil. As was pointed out above, a good part of the oil is vaporized only after diffusing, as an aqueous solution, through the cell membranes to the outside of the plant particles. This diffusion, however, becomes possible only by the presence of a certain amount of hot water, and may be stopped altogether, or seriously slowed down, when the charge is completely dried. If, therefore, in the case of direct steam distillation, the flow of oil stops prematurely, it may be necessary to continue distillation with saturated (wet) steam for a time, until hydrodiffusion is re-established. After that slightly superheated steam may again be employed.

In general, it can be said that direct steam distillation excels water distillation, as well as water and steam distillation in regard to cost, rate of distillation, and capacity of production. As far as the condition of the plant material and the method of charging are concerned, the same principles apply here as to water and steam distillation. Special attention must be paid to the quality of the live steam. The higher the pressure of the steam, the higher is the temperature at which it enters the still ; but in this respect the moisture content of the steam plays an important role. Saturated & earn usually carries some water in the form of minute droplets, which are condensed by the expanding steam. Hence, the effect of superheating becomes noticeable only if saturated (but dry) steam, of markedly high pressure, is used. The higher the pressure of the steam in the steam boiler, the drier the plant material will remain during distillation. Only the portions of the charge touching the still walls will then become moist through condensation, despite insulation of the still against emanation of heat. In order to limit such loss of heat, and consequent excessive lowering of the temperature, the high-pressure steam, before entering the still, is sent through a water separator, and partly dried. In this connection, it should also be kept in mind that the different systems of steam boilers generate live steam, containing more or less moisture. In cases of prolonged distillation, considerable quantities of steam are condensed in the plant charge, and water accumulates on the bottom of the still. This may give trouble by wetting the lower part of the plant charge. Such condensed water must be drawn off, from time to time, through a stopcock in the still bottom.

Since high-pressure steam causes considerable decomposition, distillation is best started with steam of low pressure, followed by steam of higher pressure toward the end of the operation, when the oil content of the charge has decreased considerably, and when chiefly the high boiling constituents of the essential oil remain in the retort. No general rule can be laid down in this respect, as every type of plant material requires a different and specific method of preparation and also of distillation.

End of Distillation. 

As the distillation proceeds, and as the oil content of the charge decreases correspondingly, the ratio of water to oil in the condensate will increase, because the steam can no longer contact the oil in the charge efficiently, regardless of the rate of distillation, and also because the remaining constituents are mostly high boiling. The operator must then decide at what point it is no longer economical to continue the distillation. Several criteria can be applied here. From a knowledge of the size of the charge and the yield to be expected, and from experience or trial distillations in a pilot still, it can quickly be determined whether or not the charge has been nearly exhausted. If yield data on the particular material charged are not available, it usually will suffice to take a small sample of the condensate directly into a test tube or glass cylinder and estimate from this the rate at which oil is being distilled at any particular time. Then, knowing the amount of oil already distilled, and calculating the amount that will be distilled in any additional period of time, it can be decided whether distillation should be continued for that period, or whether it would be more economical to stop and begin a fresh charge. The value of the product also enters into consideration, since a very valuable oil can be run profitably to a much larger water to oil ratio than can a less valuable oil. Certain oils e.g., vetiver or angelica root oil contain their most valuable constituents in the last runs (highest boiling fractions), and in these cases distillation must be prolonged for hours even though almost no oil seems to distill over toward the end of the operation. Otherwise valuable, high boiling constituents will be lacking in the oil. This rule, by the way, applies to all types of distillation.

It should also be kept in mind that the oil to water ratio measured at any time during distillation will always be higher than during any succeeding period, since this ratio decreases as the distillation continues. Experience with the distillation of any particular plant material will enable the operator to evaluate these matters properly, so as to obtain a maximum yield, a maximum rate, and a high quality of oil.

Treatment of the Volatile Oil.

The handling of the condensed oil is worthy of brief comment since its quality may deteriorate, particularly if the oil must be stored for some time. Just as the condensed water (distillation water) is always saturated with oil, so the condensed oil will always be saturated with water. There remains also the probability of slow reaction between the oil and water, unless the latter is almost completely removed. The oil can be brightened (cleared of cloudy appearance) by filtering through kieselguhr or magnesium carbonate on filter paper. . This procedure removes all small droplets of water which cause the cloudiness, but it does not completely dry the oil. Larger quantities of oil may be filtered through mechanical filters, filter presses, or run through high-speed centrifuges. For further details see the section in the Appendix on "The Storage of Essential Oils."

Treatment of the Distillation Water.

The distillation water flowing off the oil separator (Florentine flask) contains some of the volatile oil in solution or suspension, the quantity depending upon the solubility and specific gravity of the various oil constituents. Considering that the distillate presents a mixture of condensed steam and oil vapors, it is evident that the water phase of the distillate actually represents an aqueous solution of oil, completely saturated at the prevailing temperature. Those oil constituents which are somewhat soluble in water will be partly dissolved in the distillation water, and the dissolved portion of this oil will be different in composition from that of the oil separated in the Florentine flask. The latter is usually called main or direct oil, the former water oil. The water-soluble constituents consist mostly of oxygenated compounds, and since these compounds possess a higher specific gravity than nonoxygenated compounds (terpenes, sesquiterpenes, etc.), the water oil usually has a higher specific gravity than the main oil. This difference, however, is not always pronounced, because the distillation water contains not only oil in actual solution, but also in suspended (minute droplets) and emulsified form. A more or less milky appearance of the distillation water thus indicates the presence of oil.

Such distillation water cannot be discarded, but must be submitted to further treatment to prevent loss of oil. In the case of water distillation or water and steam distillation, it may be automatically returned into the retort during distillation. For this purpose the Florentine flask must be installed at a sufficient height above the still so that the flow from the flask overcomes the pressure within the still. In the case of steam distillation (with live steam from a separate steam boiler) the distillation water should not be returned into the retort, as too much liquid would condense and accumulate within it and wet the plant charge. The distillation water therefore, is pumped or injected into a separate still for redistillation. The process of recovering the oil from the water by redistillation is commonly called cohobation, the stills serving for this purpose being known as cohobation stills. In its original and stricter sense, the term "cohobation" implies that the distillation water is used over and over for the distillation of a new plant charge (in the case of water distillation or water and steam distillation), but today cohobation simply means redistillation of the distillation waters.

The distillation waters are redistilled most efficiently in round stills provided with a steam jacket or a closed steam coil. Indirect heating is preferable, because the injection of live steam into the retort would cause too much water to accumulate within the retort, and hinder the vaporization of the oil from the water. In the case of many distillation waters only 10 to 15 per cent need be distilled off to recover most of the oil dissolved or suspended therein. The residual water may be discarded. Occasionally, however, it is necessary to distill off more than half of the quantity of water; in such case, a considerable portion of the oil distilled over will again be dissolved in the distillation water. To shorten the cohobation and increase the quantity of oil in the condensate, the water in the cohobation still is saturated with common salt (NaCl). This decreases the solubility of the volatile oil in water: the oil distills over more quickly, and with a smaller quantity of water. This procedure is recommended particularly where the distillation water contains slightly water-soluble constituents of high boiling point, which cannot be recovered by mere steam distillation. The separation of oil and water by cohobation is based upon the simple principle that a mixture of oil vapors and steam possesses a slightly lower boiling point than pure water vapors (steam), and that the vapor mixture arising contains more oil than the liquid phase. By a reduction of the speed of cohobation, the oil content of the distillate may be increased because the rising steam will be more thoroughly saturated with oil vapors.

The following figures cited by Folsch¹⁷ give an idea of the quantities of volatile oils which can be obtained by the cohobation of various distillation waters :

Another method of recovering the oil dissolved or suspended in the distillation water consists in saturating the latter first with salt and then extracting the solution with volatile solvents e.g., highly purified petroleum ether or benzene. This is usually done twice. The drawn off and united solvent solutions are then concentrated in a still by driving off the solvent, first at atmospheric pressure, and later in vacuo, until every trace of solvent is eliminated from the oil.

---------------------

17 "Die Fabrikation und Verarbeitung von atherischen Olen," Wien und Leipzig (1930),

Any distillation of aromatic plants, unless conducted at fairly low temperatures, gives rise to products of decomposition in the nonvolatile plant constituents. These products (methyl alcohol, formaldehyde, acetaldehyde, acetone, low fatty acids, nitrogenous compounds, phenols, etc.) are carried into the condensate and present objectionable impurities. Because of their water solubility, they dissolve mainly in the distillation water, since the quantity of water by far exceeds that of oil. Because of the presence of such decomposition products, the crude water oil obtained by cohobation or extraction will in most cases be of dark color, often of disagreeable odor. It should not be combined with the main oil, as it would spoil the odor and flavor of the latter. It is, therefore, advisable to rectify the crude water oil by fractionation in a good vacuum still.

In many cases the great water solubility of the aforementioned decomposition products serves for the purification of volatile oils : when rectifying (redistilling) an oil by hydrodistillation, the distillation water is then simply discarded.

Disposal of the Spent Plant Material.

The disposal of the spent plant material, which represents a rather large bulk, frequently offers an annoying problem. One very economical method of disposal consists in using it as fuel after air drying, of course, either in the sun or near the still in the case of direct fire stills, or near the boiler when a separate steam generator is used. Since the spent material has a rather low fuel value per unit volume, consideration must be given to the construction of a special fuel box. In many cases the spent material may be used effectively as fertilizer. Certain spent plants make an excellent cattle feed ; this is particularly true of seeds which contain a high percentage of protein and fatty oil. The drying is done in dehydrating apparatus or by air drying on shelves. When sweetened with molasses, some spent grasses, such as lemongrass, seem to be relished by cattle.

Trial Distillation.

No discussion of distillation as used in the essential oil industry would be complete without some consideration of the interpretation to be placed upon the results of laboratory distillations, or, as they are frequently called, trial distillations. Since the oil content of plant material to be distilled fluctuates rather widely with such variables as geographical origin, growing conditions, ambient temperature, rainfall, period of harvest, moisture content, etc., it is not usually possible to state any values for the oil content other than by upper and lower limits (which in some cases may be quite widely separated). As already pointed out, handling of the plant material after harvesting, and prior to distillation, also has a marked effect on the oil content. As knowledge of the efficiency with which a large scale distillation is being conducted can be obtained only by comparison of the actual yield with the possible yield, it becomes quite important that the latter value be known with some accuracy. The only means of determining this value is to conduct a laboratory distillation using a sample of the plant material to be distilled in the larger scale operation.

The aim of any commercial distillation is, of course, to recover as large a percentage of the valuable oils in as high a state of purity as possible. Only in the laboratory, on a small scale, and under carefully controlled conditions, can both of these conditions be met. Therefore, the results of such laboratory distillation may be considered as a standard which the largescale operation should approach as closely as practically possible.

There are two ways of carrying out such trial distillations : (a) on a very small scale in a glass flask, and (b) on a larger scale in a pilot still.

(a) Numerous methods of assaying the contents of essential oil in plant materials have been suggested. 

The literature offers many modifications of these methods, all of which aim at a quantitative yield of oil. The best and most commonly used method is that of Clevenger, which has found official recognition in "Methods of Analysis” published by the Association of Official Agricultural Chemists, Fifth Edition, 1940. For details of Clevenger's method see below, Chapter 4, on "The Examination and Analysis of Essential Oils, Synthetics and Isolates." This method permits assaying quantitatively the content of essential oil in a small amount (50 to 500 g.) of plant material. Although the amount of oil thus obtained is not sufficient to carry out a complete analysis, conclusions regarding its odor and flavor characteristics can be drawn from the small sample. Occasionally, the oil will have to be set aside for several days, until the slightly "burnt" or "still" odor of the freshly distilled oil has disappeared.

(b) A much more satisfactory method consists in distilling a sample of 20 to 50 Ib. of aromatic plant material in a regular "pilot" still. 

Such a still, made of tin-lined copper, should be constructed so as to embody all the characteristics of large stills. It should allow for water distillation, water and steam distillation, and direct steam distillation. It will thus be possible to find for each new plant material the most appropriate method of distillation, to study, as well as possible, the rate of distillation and the consumption of steam (by measuring the quantity of distillation water*), and to determine the maximum yield of oil. Interesting observations regarding the effects of hydrodiffusion can be made. In the case of direct steam distillation the use of high-pressure or superheated steam may be studied. The quantity of oil recovered will be sufficiently large to examine the oil analytically, even to fractionate it. The pilot still should be provided with several trays, in order to find out the most opportune way of charging, if seed materials are to be processed. A small crusher and hay cutter will permit trying out the effects of comminuting the plant material according to different sizes. Needless to say, the pilot still should be well insulated in other words it should resemble large stills in every possible way except size.

---------------------------

* This will be only approximately correct, since heat losses from the distillation system have not been considered.

For all-around operation the pilot still should also be equipped for automatic return of the distillation waters into the still, in the case of water distillation or water and steam distillation, if the return (cohobation) of these waters into the still during operation seems desirable. In the case of direct steam distillation, the distillation water or a small measured part of 

 FIG. 3.20. Sketch of an experimental still.

it is saturated with ordinary salt, and three times extracted with low boiling petroleum ether. The drawn off and united petroleum ether extracts are then carefully evaporated on a hot water bath, and the residue dried in a desiccator to constant weight. From this small quantity the oil content of the total distillation waters can be calculated. Obviously, the extraction of only a part of the distillation waters gives an exact result only where the total distillation waters, after completion of the distillation, have been bulked in a tank. The distillation water should always be processed right after distillation of the plant material, because when exposed to the air for some time it loses oil by evaporation. While a small part of the distillation water is extracted experimentally with a solvent, another part should be steam distilled (cohobated). If cohobation yields no oil, the distillation water will have to be extracted with solvents.

Figs. 3.20 and 3.21 show the construction of pilot stills which may have a capacity of about 50 gal.

Steam Consumption in Plant Distillation.

In the distillation of aromatic plants the distillate (condensed water and oil) usually contains much more water than if the isolated oil itself had been hydrodistilled.

The following table indicates the average oil content, by weight percentage, in the distillates of completed operations as established by von Rechenberg, 18 based upon years of experience with industrial distillation of aromatic plants.

Distillation of Plant Material; Average Content of Volatile Oil in the Distillate

18 "Theorie der Gewinnung und Trennung der atherischen Ole," Leipzig (1910), 362. 

 FIG. 3.21. Sketch of an experimental still with automatic cohobation.

2.22 to 3.04 per cent oil, whereas in the case of caraway oil distillation the condensate contains 8.80 to 10.11 per cent of oil. This obviously implies a much greater consumption of steam, in the first case, for the same quantity of oil, and longer hours of distillation.

The paucity of oil in the distillation of plant material, according to the same author, is caused by several factors :

1. Many aromatic plants contain a quantity of oil insufficient to saturate the relatively large quantities of steam blowing rapidly through the plant charge. On the other hand, it is not advisable to reduce the speed (rate) of distillation below a certain limit. A high steam velocity causes pressure differentials within the still, which prevent the steam from stagnating in the more densely packed parts of the plant filling. For this reason, and in order to increase the efficiency of a still, the operator is always tempted to inject into the still much more steam than is actually required. This results in a large volume of distillation water.

Example: Let us suppose that a charge of 2,000 kg. of plant material can be exhausted in 11 hours, if we inject 250 kg./hr. steam (250 kg. distillation water in 1 hour). If, instead, we inject twice the amount of steam, i.e., 500 kg./hr., the length of distillation will be shortened at best by one-third, and in most cases only by one-fourth; but not by one-half, as might be expected.

2. In the course of distillation the oil content of the plant charge decreases gradually and the vaporization of oil is not stopped abruptly toward the end of the operation. This does not even take place in hydrodistillation of volatile oils per se, and much less with plant material. Plainly, such a prolongation of the distillation greatly increases the steam consumption and depends also upon the individual operator.

3. While retained in the plant material, the volatile oil may be subjected also to forces of adhesion ; this seems true even if small quantities of oil are distributed over large surfaces of comminuted plant particles. Experiments to this effect were undertaken by Rodewald.²⁰

4. The volatile oil is enclosed within the plant tissue and cut off from direct contact with steam by several layers of membrane, often very tough. For this reason most plant materials must be comminuted prior to distillation. Where steam distillation is practiced, this process of comminuting (grinding, pounding, milling, crushing, rasping) should not be carried too far (certainly not to the point of reducing the material to the size of flour particles), because the interspaces within the plant charge would then become too small. The rising steam must have sufficient space to penetrate all parts of the charge uniformly. Very small interspaces necessitate a slow, ineffective distillation, because any increase in pressure would cause the steam to break channels ("rat holes") through the plant charge, or to hurl parts into the gooseneck and condenser. In other words, too finely powdered material is not penetrated evenly by steam and cannot be completely exhausted by steam distillation.

-----------

20 Z. physik. Chem. 24 (1897), 193.

If, on the other hand, the plant material is not powdered, but granulated, only a portion of the oil is freed, and another portion remains enclosed within the oil glands in the plant tissue. When crushing plant material, such as seed, a portion of the freed volatile oil will be covered again by crushed plant particles. The distillation of excessively crushed seed material, if not properly conducted, may, therefore, require longer hours than that of torn or slightly milled seed, provided the quantity of injected steam is the same.

If the plant material is distilled in uncomminuted condition as with herbs and leaves, and most flowers the oil remains enclosed within the plant tissue. Hacking with an axe or machete or cutting in a hay cutter offers an advantage only in that the material can be packed into the still more uniformly ; the steam then penetrates the charge more evenly, but very few oil glands will actually be broken up. Since the steam can vaporize only those volatile substances which it touches directly, and will not affect the oil enclosed within the plant tissue, the oil must first be dissolved by hot (liquid) water and carried, by diffusion, through the swollen cell walls toward the outside. Hydrodiffusion, however, requires much more time than vaporization, which takes place almost immediately, because all the enclosed volatile oil must be brought to the surface, and that is a slow process. This fact is primarily responsible for the paucity of oil in the condorisate, and for the relatively long duration of distillation in the case of uncomminuted leaves and herbs possessing a tough fiber.

5. If the plant material is comminuted prior to distillation, very high boiling or practically nonvolatile substances, such as resins, paraffins, waxes, fatty oils (contained in other cells or glands), mix with and dissolve in the freed volatile oil, thereby substantially lowering its vapor pressure, and reducing its rate of vaporization. This occurs particularly in the case of seeds, most of which contain large quantities of fatty oils. The Agricultural Experiment Station of Mockern, near Leipzig, Germany, reported the following content of fat and fatty oil (ether extract) in seeds from which the volatile oil had first been removed by steam distillation : 

Assuming that air-dried caraway seed, such as is used for distillation, contains about 15 per cent moisture and 5.5 per cent volatile oil, we arrive at a ratio of 5.5 per cent volatile oil to 12.8 per cent of fatty oil in the seed. For practical distillation, this implies that 5.5 parts of volatile oil must be vaporized from 12.8 parts of fatty, nonvolatile oil. In other words, it is necessary to distill a mixture of fatty oil and volatile oil, which holds 30 per cent of volatile substances in solution. Assuming a content of 5.0 per cent volatile oil in fennel seed, 3.0 per cent of volatile oil in anise seed, 3.5 per cent in ajowan seed, 2.5 per cent in celery seed and 1.0 per cent in coriander seed, we find that we would have to distill :

Ajowan oil mixture containing 11.5% volatile oil*

Anise oil mixture containing 16.0% volatile oil

Caraway oil mixture containing 30.0% volatile oil

Celery oil mixture containing 9.7% volatile oil

Coriander oil mixture containing 4.2% volatile oil

Fennel oil mixture containing 27% volatile oil

Such relatively large quantities of fatty, nonvolatile oils are well capable of reducing the vapor pressure, and thereby the rate of vaporization of the volatile oils dissolved in these fatty oils.

Other reasons aside, it is thus practically impossible, when distilling seed with steam or boiling water, to saturate the steam completely with oil vapors, even when packing the plant charge very high in the still. The oil vapor phase in this mixture will always remain unsaturated ; the more the content of fat in the seed exceeds thut of volatile oil, the less the steam will be saturated with volatile oil vapors. This theoretical consideration confirms practical experience in the case of seed distillation. In actual practice, therefore, distillation of seed material can seldom if ever be completed, because the fatty oil tends to retain small quantities of volatile oil. It becomes necessary, therefore, to halt distillation, since the small recovery of oil no longer warrants the increasing consumption of steam and labor.

6. The steam consumption is influenced further by the moisture content of the plant material, particularly in the case of herbs, grasses and roots, which are processed either in the fresh succulent, or scmidry, or dry, condition. When distilling peppermint herb with live steam, for instance, the following quantities of steam will be consumed, the steam consumption being measured by the quantity of distillation water in the condensate:

Fresh herb requires 250 to 350 kg. of steam per kilogram of oil.

Semidried herb requires 60 to 80 kg. of steam per kilogram of oil.

Air-dried herb requires 30 to 40 kg. of steam per kilogram of oil.

----------------------

* In all cases a 15% moisture content of the seed is assumed.

These figures, cited by Folsch,²¹ are evidently relative, as actual steam consumption depends upon the type of the still, the quality of the steam, the way of packing, and the experience of the operator.

Because of its high moisture content, fresh peppermint herb, when distilled, has a tendency to lump (agglutinate) and to prevent a uniform penetration by the steam. The volatile oil is, therefore, released from the fresh herb only very slowly. Taking the above figures of steam consumption as a basis, the steam/oil vapor mixture (in other words, the condensate) will contain the following quantities of volatile oil :

Fresh Herb ............0.3 to 0.4% peppermint oil

Semidried Herb ........ 1.2 to 1.6% peppermint oil

Air-dried Herb.......... 2.5 to 3.0% peppermint oil

The oil content in the condensate is not uniform from the beginning to the end of distillation, but amounts in the beginning to a multiple of the average oil content. In the case of air-dried peppermint herb, the condensate contains in the beginning about 8 per cent of oil, which decreases gradually toward the end until it amounts to only 0.004 per cent. For practical reasons distillation should then be stopped. As mentioned previously, certain plants contain volatile oils, the most valuable parts of which are very high boiling. When applying saturated steam of atmospheric pressure only, distillation must then be continued for very long periods, although only small quantities of oil are recovered toward the end. If this is not done the high boiling constituents are lacking in the oil, and the oil is of inferior quality. In such cases it will be advantageous to speed up and complete the operation by injecting slightly superheated steam toward the end.

Rate of Distillation.

According to Folsch, ²² the ratio between quantity of distillation (condensed) water and time (in other words, the quantity of water distilled over per hour) may be designated as rate (force or speed) of distillation. It must be regulated according to the diameter of the still, and the size of the interspaces within the plant charge (degree of comminution). If the velocity of the rising steam is too low, the steam will stagnate in the denser portions of the charge, and complete exhaustion by distillation will be impossible. If, on the other hand, the velocity is too high, the steam may break through the charge, form steam channels ("rat holes") and even hurl plant particles into the condenser, partly clogging it. By collecting the distillation water running off the condenser from time to time, and over a period of some minutes, and then weighing it, the rate of distillation can be controlled. For practical purposes the volatile oil may be ignored. The quantity of distillation water collected during these few minutes is calculated in terms of kilogram/hours per square meter. (See example below.)

--------------

21 "Die Fabrikation und Verarbeitung von atherischen Olen," Wien und Leipzig (1930), 40.

22 Ibid., 62.

This figure is then compared with the optimum rate of distillation, as established by trial distillation or by experience with the plant material in question (and taking into consideration its degree of comminution). The steam velocity in the actual operation may be regulated accordingly.

For example, if we obtain in one minute 8 kg. of distillation water, and if the smallest area covered by the charge on the perforated grid in a cylindrical still is 1.2 sq. m., the rate of distillation will be

8 x 60/1.2= 400 kg./hr. per sq. m.

Once the average oil content of the mixed vapors (steam plus oil vapors) has been established for a certain type of plant material and a certain degree of comminution, and once the most favorable rate of distillation is known, the amount of steam necessary for complete exhaustion of a plant charge can be calculated, and the steam supply adjusted accordingly. By weighing the quantity of distillation water from time to time, by converting the figures to the total length of distillation and by relating this to the quantity of oil expected, the operation may thus be regulated according to optimum conditions. Let us suppose, for example, that 1,000 kg. of coriander seed must be distilled, and that the seed, according to assay, contains 0.8 per cent of oil, in other words that the 1,000 kg. of coriander seed contain 8 kg. of oil. We know from experience or from trial distillations that the average oil content of the vapor mixture (condensate) in the case of coriander seed distillation is 0.5 per cent. Therefore, 1,600 kg. of steam are required to distill over 8 kg. of coriander seed oil. If we work with a distillation rate of 200 kg./hr., i.e., 200 kg. distillation water per hour, the charge should be exhausted in 8 hr. In order to shorten the time of distillation, the rate of distillation must be increased. However, in this case attention must be paid to the fact that, on increasing the speed of distillation, the average oil content of the vapor mixture decreases to a certain extent, because the quicker the steam penetrates the plant charge the less it has occasion to become saturated with oil vapors. In other words, much more steam will be consumed than is calculated theoretically.

Pressure Differential Within the Still.

The velocity of steam flow is caused by differences in pressure. In the case of plant distillation with live steam, which in the boiler is usually at a pressure above atmospheric, the plant charge in the retort prevents the injected steam from expanding immediately and completely. For this reason, the steam pressure cannot fall immediately to the level of the atmospheric pressure. Thus, there arises a certain excess pressure beneath the charge in the retort; but a gradual equalization with the atmospheric pressure takes place toward the top of the still. The degree of this excess pressure is a function of the force (speed) of distillation and of the interspaces within the plant charge. According to the height of the charge or the number of layers, this excess pressure can be increased by 0.3 atmospheres, and, in some cases, even more. But if the pressure exceeds a certain limit (which depends upon the type and height of the plant charge), the steam forms fine, often scarcely visible channels through a powdered charge, whereas coarser masses are torn apart, or even hurled into the gooseneck of the still. An excess pressure ("back pressure") within the retort may be caused also by a gooseneck or condenser pipes too narrow for the volume of steam injected into the still or by sharp bends in the pipes.

Irregular heating of the boiler and variations in the steam consumption (such as are occasioned by the turning on and turning off of neighboring stills) may cause the pressure in a steam generator to undergo continuous fluctuations. High-pressure steam has a tendency to blow into a still somewhat jerkily, giving rise to pressure variations even within the retort.

Such fluctuations, however, are by no means harmful ; on the contrary, they may exert a beneficial influence, as far as the yield of oil is concerned, by Forcing the injected steam to loosen and penetrate the more densely packed portions of the plant charge, where the steam would otherwise stagnate.

Pressure Differential Inside and Outside of the Oil Glands.

As it rises through the plant charge, the steam at first vaporizes all the freed volatile oil which by comminution of the plant material is within reach of the parsing steam. Saturated steam (not superheated!) will at the same time condense a certain quantity of water within the retort. Consequently, the temperature of high pressure steam will be reduced to that of saturated steam, in other words, to the boiling point of the water/oil mixture. It must be remembered that this boiling point is slightly lower than that of the saturated steam. As the volatile oil vaporizes from the plant material, the temperature of the steam rises again to that of pure saturated steam, at the pressure prevailing in the charge. If the plant charge is somewhat tightly packed, the temperature of the steam will show a certain range from the bottom to the top of the charge. This differential in temperature depends upon the force of distillation and the drop in the steam pressure from the lower to the upper section of the retort; in other words, the lowest part of the charge will have the highest, and the upper part the lowest, temperature. Gradually the temperature of the steam equalizes itself throughout the charge and, despite poor heat conduction, will prevail, even inside of all plant particles. As has been said, the boiling point of a water/oil mixture is somewhat lower than that of steam alone, the total vapor pressure a little higher. Since the temperature inside and outside of the plant particles has become equalized, a certain excess pressure will develop within those oil glands which still contain volatile oil and water enclosed. This pressure differential inside and outside of the oil glands probably has some influence upon the vaporization of the volatile oil through the cell walls. A sufficiently large pressure differential may well cause some cell membranes to burst (provided they are not too thick and strong) or at least to expand the cell walls, to enlarge the pores and to loosen agglomerated particles of the charge, thus opening new passages for the steam. The more the pressure differential is reduced, the more it loses its significance as a loosening agent ; but it remains important for the isolation of oil, in so far as it supports the forces of hydrodiffusion. The pressure differential inside and outside of the oil glands is more effective when first heating the retort and toward the end of distillation, provided temperature and pressure fluctuations actually occur inside of the still. A pressure differential, however, can be created only if water is present in liquid form, or by partial condensation of steam when first heating the retort : the water thus formed will penetrate the plant tissue and seep also into the oil glands, (von Rechenberg).

In the hydrodistillation of plant material at reduced pressure, the pressure differential inside and outside of the oil glands exerts itself to a marked degree only with low boiling substances. In the case of distillation above atmospheric pressure, however, the pressure differential assumes considerable importance.

Effect of Moisture and Heat upon the Plant Tissue.

Any plant material serving for distillation contains a certain quantity of moisture, even airdried material retaining 10 to 20 per cent of water. If saturated steam of atmospheric pressure is injected into the plant charge, condensation of steam will take place until the temperature of the still content has risen to that of the steam.

Heat in conjunction with moisture soon causes the plant tissues to swell, the cells and pores to enlarge, and the total volume to expand. Completely swollen seed material, for example, may have expanded by onefourth of its original volume. In actual distillation, this loosening of the plant material may, however, be partly counteracted by the weight and pressure of the softened plant charge.

An actual bursting of the plant membranes by the action of steam takes place probably to a limited extent only. The hot steam undoubtedly exerts a certain preparatory effect important for the vaporization of the enclosed volatile oil, but the action of steam per se is not sufficient to liberate that part of the oil which remains protected by resistant cell membranes.

Influence of the Distillation Method on the Quality of the Volatile Oils.

The quality, as well as the physicochemical properties, of a volatile oil are greatly influenced by the condition of the plant material (age, dried or fresh) and by the way distillation is carried out. Many factors enter the picture, viz., the method of distillation (water distillation, water and steam distillation, and steam distillation), the degree of comminution of the plant material, the quantity of the plant charge, the length of distillation, the pressure applied, the quality of the steam, the treatment of the distillation waters, whether the oil of cohobation is added to the main oil or not, etc.

The effects of water distillation and steam distillation differ considerably, in that high boiling constituents of the volatile oil are recovered only incompletely in the case of water distillation, if the plant material is insufficiently comminuted. Even leaf material yields volatile compounds of high boiling point only incompletely by water distillation. Von Rechenberg²⁸ reported that patchouli leaves yielded 3.27 per cent of volatile oil on steam distillation, and only 2.98 per cent on water distillation. The latter oil contained only a small quantity of the high boiling constituents, which incidentally possess also a high specific gravity and a high odor and fixation value. Oil constituents which are slightly soluble in water, phenols and certain alcohols and acids for example, are retained in the water, with the result that water distillation and steam distillation yield different types of oil, if the plant material contains only small quantities of oil.

General Difficulties in Distillation.

Essential oils consist of volatile compounds which are more or less sensitive to the influence of heat. It is doubtful, therefore, that all the volatile constituents present in the living plants can be isolated as such by distillation. In addition, distillation of certain plant materials is connected with difficulties of hydrodiffusion. The oil in part resinifies, and in part remains in the plant tissue. Hence every type of plant material requires a particular method of distillation.

Because of these difficulties, and because of the high cost of distillation in certain cases (through excessive steam and fuel consumption), it has been suggested that such materials be extracted with volatile solvents, and the concentrated extracts steam distilled. The oils obtained usually contain small quantities of resinous and waxy matter ; such oils may be soluble in a certain volume of dilute alcohol, but the solutions often become turbid when more of the dilute alcohol is added.

---------

28 "Theorie der Gewinnung und Trennung der atherischen Ole," Leipzig, (1910), 441.