The Sulfate Process
( Originally Published 1920 )
THE third chemical process for the preparation of fibers from wood was developed by Dahl in Germany in 1883. It is an alkaline process similar in many ways to the soda process, and it is reported that its development was due to the desire to replace the unavoidable loss of soda with some material cheaper than soda ash. Among other things, Dahl tried sodium sulfate in the form of salt cake, and found that the resulting fiber was dark in color, but made stronger paper than other wood fibers. From these facts it derives its two common names, the "sulfate" process from the use of sodium sulfate, and the "kraft" process from the high strength of its paper and from the German word for strength. At first the cooks were intentionally made incomplete, and the final disintegration of the partly softened chips was carried out by mechanical means, as by a kollergang or edge-runner. This procedure was soon abandoned for more complete cooking, which produced fibers that could be handled in ordinary beating equipment.
From the name "sulfate" it might be assumed that sodium sulfate plays a principal part in the chemical reactions during cooking. Actually it has little or no action on the wood, and it is only after it has been reduced to sodium sulfide during the recovery of the alkali from the black liquor that it is effective in the cooking process. A more appropriate name would be the "sulfide" process, but the old name is universally used and the term "sulfide" will probably never be generally accepted.
The underlying chemical principles of this process are practically the same as those of the soda process except for the complications introduced by the presence of sulfur compounds. Sodium sulfide probably acts very much like caustic soda in the digester, but it may well be that it reacts first with water to form sodium sulfhydrate, which then reacts with the wood very much as caustic soda. This reaction between sodium sulfide and water to form sulfhydrate and caustic soda is reversible and takes place only as the caustic soda is taken up by the wood. The chief function of the sulfide in the cooking liquor is, therefore, quite likely that of forming a reserve supply of alkali which can be used as needed, but which is not all immediately available at the beginning of the cook. This automatically controlled cooking condition differs from that of a soda cook in which all the available alkali is present at the start, thus causing the action to vary from drastic at first to rather mild at the end. It is probable that this restraining influence of the sulfide is responsible for the higher yields and strength of sulfate pulps.
The presence of sulfides in the cooking liquor also leads to undesirable side reactions, with results differing markedly from those of the soda process. In the soda process small amounts of unnoticeable methyl alcohol are formed by hydrolysis of the lignin, but in the sulfate process the corresponding compound is methyl mercaptan, which has a vile and persistent odor. Organic sulfides with bad odors are also formed, but these are not as obnoxious as the mercaptans. The odor of these compounds is first noticeable when their concentration in the air reaches 1 part in about a mil-lion by weight. The amount of mercaptan formed depends on the kind of wood used and the cooking conditions, and the amount which escapes to constitute a nuisance depends on the construction of the pulp mill and the precautions which are taken to destroy these compounds. Under the best of operating conditions a sulfate mill is not a welcome addition to any closely settled community.
The digesters used in the sulfate process are similar in all respects to those described in the section on soda pulp. The tumbling or rotary types are seldom used, and practically all new plants install vertical stationary digesters. These have gradually increased in size until those in modern mills have capacities up to 4300 cubic feet, and there seems to be no reason why even larger ones could not be used if they were fabricated at the plant site, as is done with sulfite digesters.
The use of indirect heating is very general in sulfate cooking and there are a number of different kinds of heat exchangers for this work. These vary in their construction and efficiency, but with a good heater it is possible to cook rapidly and thoroughly without the use of any direct steam, except that necessary to occasionally blow out the strainers through which the liquor reaches the circulating pump, and to shake up the contents of the digester to aid the penetration of the cooking liquor and to prepare the charge for blowing. The strainers mentioned are variously placed and constructed according to the circulating system employed. In one system the false bottom in the lower cone is the strainer, and the liquor is pumped up through the heater and returned to the digester near its top. In another system there is a special device in the blow line below the cone which permits removal of the liquor, and in a third system the liquor is taken out through a perforated belt around the middle of the digester and is returned at both the top and the bottom.
The cooking liquor for the sulfate process is essentially a mixture of caustic soda and sodium sulfide, but it also contains small amounts of sodium carbonate, sulfate, silicate, aluminate, sulfite and thiosulfate, all of which are relatively inactive. The ratio of sulfide to caustic soda varies over a rather wide range; it is common practice to have from 20 to 35 per cent of the total alkali as sulfide. Because of the presence of two chemicals used in varying proportions it is difficult to express in simple terms the ratio of alkali to wood used in cooking. This is still further complicated by the fact that the best ratio of alkali to wood also depends on the ratio of sulfide to caustic in the liquor used. Merely as an example, if the "sulfidity" of the liquor were 40 per cent, the liquor for satisfactory cooking should contain about 9.5 per cent sodium sulfide and 14.2 per cent caustic soda, but if the sulfidity were 76 per cent the figures would be about 24 per cent sulfide and 7.7 per cent caustic, all based on the weight of wood used.
As the liquor is usually made stronger than is desired for use in the digester it is often diluted to the proper strength with dilute black liquor, which would otherwise be considered too weak to evaporate for recovery. Though this saves a little alkali, it is not the best possible practice, for easier bleaching and stronger pulps can be obtained by diluting the cooking liquor with water.
As in the soda process, the cooking cycle consists of a period for raising pressure and allowing the liquor to penetrate the chips, and a second period at full pressure, which may be from 100 to 135 pounds per square inch. To these there is added in some mills a period for reducing the digester pressure before blowing, usually to about 60 pounds. This gives more time for heat recovery from the blow-off steam, and for recovery of the last traces of turpentine where resinous woods are being cooked. It is questionable whether these gains will offset the digester time lost, and many mills blow at full pressure in 10 to 20 minutes, as is done when cooking by the soda process.
The longer the period for reaching full pressure, the less is the degrading action of the alkali on the cellulose. In a rapidly steamed cook, a high temperature is reached while considerable alkali is present, and cellulose degradation can take place. Since the greater part of the delignification takes place at low temperatures, a long heating period produces an easier bleaching fiber and one with high initial tearing strength, but which will develop good bursting strength on beating. Rapid heating, if less alkali is used, produces a strong pulp with a relatively high lignin content, such as is used in kraft bags and wrappers.
The temperature in the digester charge during cooking may be from about 320 to 350 degrees F. Below 320 degrees more alkali is needed to produce a well cooked pulp, and above 345 degrees little is gained, except a slight decrease in the time required to cook. The total time of cooking may vary from 3 hours to about 73 hours according to the kind of wood used and the type of pulp being made. Of this total time 40 to 60 per cent may be taken up in bringing the digester to full pressure.
The steam consumption during cooking depends on the size of the digester, the effectiveness of its insulating covering, the concentration of the cooking liquor and the amount used, as well as the temperature of the cook, and whether direct or indirect steam is used. The greatest amount of steam is, of course, required during the time the charge is being brought up to pressure. For a 15-cord digester heated by direct steam, the steam required will be about 20,000 to 22,000 pounds per hour during the heating period, but only about 1,250 pounds during the full-pressure period, when only enough is needed to make up for condensation due to radiation. The average steam consumption per ton of pulp in alkaline cooking is about 4,135 pounds when saturated steam is used at 125 pounds pressure. These figures apply to cooks by both the sulfate and the soda processes.
It is obvious that the variables influencing the sulfate process are somewhat more numerous than those in the soda process. Both are affected by the ratio of chemical to wood, the concentration of the cooking chemical, the maximum temperature, and the time it is maintained. The sulfate process is more affected by the kind of wood, because a wider range of species is used. In addition the relief schedule has more influence in the sulfate process, and of still more importance is the composition of the cooking liquor. All things considered there is a much greater chance to adjust conditions in the sulfate process, and it is, therefore, possible to make a considerably wider range of products from any given wood. Because of the number of variables involved, the study of any one, uninfluenced by the others, has been so difficult that much information still remains to be worked out.
Relieving gas and pressure from the digesters is of especial importance in sulfate cooks because it is one of the chief sources of evil odors, and because turpentine can be recovered in sufficient amounts to make its recovery worth while, at least when cooking pine wood. At a cooking pressure somewhere between 40 and 60 pounds per square inch the sulfur compounds in the liquor begin to react with the lignin, forming sulfides and mercaptans. Relief, prior to this stage, lets out chiefly the air escaping from the chips, but above this point the organic sulfur compounds escape, and cause very objectionable odors. There are various chemical methods for reducing these odors, but they all have features which have prevented their extensive use. The most practical way seems to be to scrub the gases in a barometric spray condenser with aerated water, collect all of the incondensible gas, and pass it through the blow pipes in the recovery furnace, where it will burn. The condenser effluent may be mixed with a little bleach solution to eliminate its odor.
Other sources of odor are blowing down pressure prior to discharging the digester; the condensate from the evaporators; and occasionally unsatisfactory recovery furnace conditions, which permit unburned gases to escape. With satisfactory furnace operations all mercaptans and organic sulfides entering with the black liquor are destroyed.
The blowing of the digester charge at the end of the cook is through a valve and large pipe connecting the bottom of the digester with a separator, as in the soda process, or the pipe may connect directly with a diffuser which collects the black stock, and in which it is washed. A diffuser is a large closed tank with a perforated false bottom, and a vent at the top through which steam containing a little entrained fiber passes. This goes to a receiving tank, which serves to separate the steam from the small amount of fiber which was carried over. From this the steam goes to some form of heat exchanger to recover as much as possible of its heating value. The diffusers are of such a size that each will hold a full digester charge, which is blown in at full digester pressure, and distributed uniformly by a cone-shaped plate in the top of the diffuser just under the entrance opening. Washing in a diffuser is by flooding with black liquor and finally with water, as is described for the open wash pits of the soda process, except that with diffusers the wash liquor is supplied by a pump which fills the diffuser completely and maintains pressure during the washing period. At the completion of washing, a manhole in the side of the diffuser just above the false bottom is opened, high pressure cold water is turned on at the top and the contents quickly forced out into a stock chest. The entire operation of emptying and replacing the door should take less than 30 minutes.
Continuous rotary filters, as described under soda cooking, now tend to replace the diffusers. They are often used in units of three in tandem, and the major factors for successful operation are regulation of stock consistency and the formation of a uniformly thick and smooth sheet on the filter surface. When properly operated such filters should give a loss of not over 60 pounds of salt cake per ton of fiber, which is considerably less than that from diffusers.
The screening operation to remove knots, slivers and smaller shives has been described under soda pulp. The equipment is about the same for both the soda and sulfate processes, except for the size of the screen openings; the longer fibers from the coniferous woods generally used in the sulfate process naturally require larger screen slots than the short fibers from deciduous woods used in most soda cooks.
The recovery of soda is as necessary in the sulfate, as in the soda process, and again the success of the plant depends on the efficiency of the recovery operation. That is the reason why so much stress is laid on washing the black stock with a minimum of water, and on reducing soda losses in every possible way. The early methods of evaporation by disc evaporators, burning in rotary black ash burners, mixing the ash with salt cake and shoveling it into smelters, are no longer employed. Evaporation is carried out in large, multiple effect evaporators, with either vertical or horizontal tubes, followed by a final treatment in a forced circulation evaporator to bring it up to the desired strength. If the black liquor going to the evaporator contains 15 per cent of solids and that coming away from the forced circulation evaporator contains 75 per cent, it means that for each pound of solids 5.3 pounds of water must be evaporated. Because it depends on so many conditions it is unsafe to generalize on the amount of water which must be evaporated, but it may easily be as much as 8 tons per ton of pulp made.
One of the difficulties encountered in evaporating and storing black liquor produced in cooking resinous woods is the separation of soaps formed by reaction of resin and alkali. These are more or less insoluble in the black liquor, and tend to float on its surface. This soap is sometimes collected by flotation, or centrifugal action, and treated for the recovery of the resin and fatty acids, which are collectively known as "tallol," which is frequently corrupted to "tall oil." It is a black, viscous, sticky liquid which finds some use in making soap and as an emulsifying agent. The amount of such material may vary from about 100 pounds per ton of pulp for north-ern pines to 250 pounds for the more resinous southern woods.
The recovery furnaces have, in general, followed the line of development noted for soda mills, but their design and operation has been complicated by the necessity for reducing the sodium sulfate to sulfide. This was originally done in a smelting furnace, in which the partly burned mass from a rotary furnace, containing carbon and inorganic chemicals, was subjected to an air blast to burn away the carbon and at the same time form the desired sulfide. These smelting furnaces were similar to very large, deep crucilbles. Next came the spray-type stationary furnaces. In these the upper portion concentrated and partly burned the black liquor, which entered through some form of spray head, while the lower part served as the smelting furnace for the material dropping from the upper combustion chamber. Smelting furnaces are lined with a refractory material, usually soapstone, which resists the action of the molten alkali and also acts as an insulator to prevent the loss of heat.
In both these types of furnace the temperature of the smelt must be uniform and sufficiently high to induce the reduction of the sulfate. The supply of air is quite critical, and it must be very care-fully regulated; if it is insufficient the temperature will drop so low that no reduction will take place, and if too much is supplied reoxidation of sulfide to sulfite will take place. A reduction of 88 to 92 per cent of the sulfate to sulfide is about normal practice, though it may run as high as 94 per cent under the best conditions.
Modern recovery furnaces are of several kinds, but all make use of water-cooled walls and recover large amounts of heat. They are similar in principle, in that the liquor is sprayed into the combustion chamber, where it is partly evaporated and burned as it falls to the bottom to form a bed of black ash, into which air is blown to complete the combustion and reduction. The smelt which collects at the bottom flows in a continuous stream into a dissolving tank where it forms what is known as "green liquor" because of its color. This is due to the presence of ferrous hydroxide or possibly ferrosodium sulfide. The smelt contains chiefly sodium carbonate and sodium sulfide with small amounts of sodium sulfate, sulfite, thiosulfate, caustic soda and insoluble matter. The sodium carbonate may run from 65 to 73 per cent, and the sodium sulfide from 20 to 22 per cent of the total smelt respectively.
Conversion of the green liquor into "white liquor," which is the cooking liquor, is by means of the causticizing process with lime, which has already been described. In some mills this is done directly, without attempting to separate iron compounds and other suspended matter. If the calcium carbonate is to be burned for lime recovery, or if it is to be used as a paper filler, the suspended matter must be separated by settling, and only the clear liquor used in causticizing. If this is not done the impurities will build up in the lime to such an extent as to make it unfit for causticizing, and the color of the calcium carbonate will be too poor for use in white papers.
The chemical losses in the sulfate process occur at the same points as in the soda process.. The two processes differ, however, in one respect—the corrosion problem is troublesome with sulfate and practically absent when only soda is used. In the sulfate process the presence of sodium sulfide, organic sulfides and sulfonic acid derivatives, together with the formation of organic acids, which may take place locally in evaporator tubes, accelerates the attack on steel equipment. Digesters are attacked at seams and rivets, but the trouble varies greatly in different mills, and even in different digesters in the same mill. Diffusers also show corrosion as local pitting. These troubles can be overcome by the use of stain-less steel, but at a considerable increase in the cost of equipment.
The yields of fiber from sulfate cooks cannot be stated very definitely because they are so greatly influenced by the kind of wood used, the operating conditions in any given mill, and the type of fiber being produced. Good yields for spruce and fir are about 38 to 42 per cent, for southern pines about 42 to 50 per cent and for gumwood about 48 to 52 per cent, all based on the weights of bone-dry wood and fiber.
The quality and uses of the fiber are also very variable. Since the fiber was originally used for strong papers, mention of the kraft process naturally brings that factor to mind, but excellent fiber for book papers can be made from hardwoods, which are short fibered, and if properly cooked and bleached the long fibers from spruce and pine can also be used to advantage for many of the purposes for which sulfite fiber was formerly considered essential. This versatility of the sulfate process accounts for the fact that, though the last of the three chemical processes to be developed, it has now surpassed the other two in importance and in amount of fiber produced.