US3775065A

US3775065A - Grain control in continuous crystallization

Info

Publication number: US3775065A
Application number: US00787437A
Authority: US
Inventors: Y Aoyama
Original assignee: Daido Namarikakoki Co Ltd
Current assignee: Daido Namarikakoki Co Ltd
Priority date: 1968-12-27
Filing date: 1968-12-27
Publication date: 1973-11-27
Anticipated expiration: 1990-11-27

Abstract

Controlling grain sizes in continuous crystallization by transferring the liquid from the supersaturation side into the unsaturation side across the solubility curve prior to retransferring into the supersaturation side across the solubility curve, and selecting the relative position of liquid to the solubility curve in the unsaturated side and the staying time of liquid at such selected position, so as to dissociate molecular clusters desiredly into free molecules in the liquid prior to supersaturation.

Description

United States Patent 1 Aoyama Nov. 27, 1973 GRAIN CONTROL IN CONTINUOUS CRYSTALLIZATION [75] Inventor: Yoshio Aoyama, Osaka, Japan [22] Filed: Dec. 27, 1968 [21] Appl. No.: 787,437

[52] US. Cl 23/295, 423/551, 423/557,

Ackeren 23/273 3,071,447 1/1963 Bernhardi 23/273 3,218,133 11/1965 Ebner 23/273 3,425,795 2/1969 Howard et a1. r 23/300 3,459,509 8/1969 Aoyama 23/273 3,547,595 12/1970 Olivier et a1. 23/273 Primary Examiner-Norman Yudkoff Assistant Examiner-R. T. Foster Attorney-Hall & l-loughton [57] ABSTRACT Controlling grain sizes in continuous crystallization by transferring the liquid from the supersaturation side into the unsaturation side across the solubility curve 10 Claims, 10 Drawing Figures 423/558 [51] Int. Cl B0ld 9/02 [58] Field of Search 23/301, 302, 295, 23/300, 273 R, 125, 126, 121

[56] References Cited UNITED STATES PATENTS 1,091,721 3/1914 Weil 23/273 2,347,073 4/1944 Beekhuis 23/301 2,479,001 8/1949 Burke et a1. 23/298 2,737,440 3/1956 Roberts et a1. 23/301 3,361,540 l/l968 Peverly et al. 23/301 PAIENTEDNUV27 1973 3,775,065

' SHEET 10F 2 Concentratwn (g-GAS04/|003.H,0J

Temperature (C1 FIGS 6 wmrqa- IN VENTOR Wa y? GRAIN CONTROL IN CONTINUOUS CRYSTALLIZATION The present invention relatesto improved methods of controlling occurence and growth of crystalline grains, and in its more particular aspects has to do with the application of these methods to continuous crystallization a liquid circulating system of the kind in which the mother liquor is initially pumped to a saturator to be supersaturated by cooling, heating and/or evaporating, then sent to a crystallizer to have some part of the solute crystallized on artificially provided seed grains as it goes from bottom to top of the crystallizer, the overflow from the top of the crystallizer being returned to the mother liquor again and re-circulated while the crystallized grains sedimented in the bottom of the crystallizer and are taken out.

The general method of the invention is a saturation controlling one where the mother liquor is initially transferred across the solubility curve from supersaturation side to unsaturation side by heating, cooling and- [or adding liquid prior to its entrance into the saturator, then re-transferred across the solubility curve from unsaturation side to supersaturation side by cooling, heating and/or evaporating liquid in the saturator before it enters into the crystallizer, thereby dissociating the molecular clusters of solute into free molecules in the liquid before it is supersaturated.

In the liquid circulating system for continuous crystallization of the above-described kind the yield of grains is primarily in proportion to the supersaturation of liquid in the saturator; the higher it is supersaturated in the saturator, the more it yields in the crystallizer. However, the natural occurence of fine kernel grains in the crystallizer is also primarily in proportion to the supersaturation of liquid in the saturator; the higher it is supersaturated in the saturator, the more kernels they occur in the crystallizer. When occur in quantity, the fine kernels will substantially restrain the growth of seeds which are artificially provided in the crystallizer, resulting in a yield of grains with small and irregular sizes grown on both kernels and seeds. Since grains of small and irregular sizes are not only inconvenient for separation from the liquid but also inferior in chemical property and commercial value, there will apparently be a restriction on raising the supersaturation of liquid in the saturator. If, on the contrary, the supersaturation is sufficiently lowered in the saturator as is usually done conventionally, the growth and yield of grains will be lowered considerably, or alternatively a much larger size of equipment will be required, resulting in a rise of unit production cost in either case. Besides, it will often be practically required in continuous crystallization that the size of crystalline grains is varied desiredly without disturbing the uniformity of grain size.

In static and theoretical crystallization where grains are crystallized by supersaturation in standing liquid containing negligible impurities such as in a laboratory batch-type operation, two different areas may be identified at the supersaturation side of the solubility curve; one is what is called an unstable area and the other is what is called a semi-stable area intermediate the unstable area and the solubility curve in a manner that the boundary line between both areas may run nearly in parallel with the solubility curve, and this line is the socalled supersaturation curve. In the unstable area new kernels naturally occur as well as grain growth on the existing kernels, while in the semi-stable area new kernels do not occur but grains do grow on the existing kernels.

In dynamic and practical crystallization where grains are crystallized by supersaturation in flowing liquid containing considerable impurities such as in an industrial continuous operation for instance of the liquid cir culating type to which the present invention is related, there is no such semi-stable area adjacent the solubility curve but fine kernels always occur when the liquid is supersaturated; this fact was confirmed in experiments carried out by the inventor.

But it was discovered by the inventor in his experiments that there is a time lag from supersaturating of the flowing liquid before the formation of fine kernels in it, provided that the liquid is supersaturated across the solubility curve after it is unsaturated. The time lag was primarily a decreasing function of supersaturation in the inventors experments, given the solubility and concentration of liquid; the higher the liquid was supersaturated, the shorter the time lag was from supersaturating of liquid to occuring of kernels. The relations of supersaturation to time lag obtained in the inventors experiments for aqueous solutions of CuSO 'SH O, Na SO and FeSOfll-I O are given in the following table.

According to the above table, for instance, an aqueous solution of CuSO -5H O supersaturated at 6C in terms of temperature does not begin to crystallize at all within 2 minutes after it is supersaturated in the saturator following being unsaturated; therefore it does not begin to produce fine kernels within the 2 minutes of time lag. Similarly, an aqueous solution of Na SO supersaturated at 18C in terms of temperature does not begin to have any fine kernel formation within 2 minutes of time lag.

It follows from this discovery that the natural occurence of fine kernels will be avoided if the liquid is brought into the crystallizer from the saturator before the time lag elapses, provided that the liquid is unsaturated prior to its entrance into the saturator; when the time lag elapses in the crystallizer in which seeds are provided artificially, the liquid will get the solute crystallized not into natural fine kernels but on the seeds which are sufficiently larger than such natural kernels.

It further follows that the natural occurence of kernels will be controlled in a desired manner if the time lag itself is varied by any means, given the solubility, concentration, and supersaturation of the liquid.

An important object of the invention is to obtain crystalline grains of substantially large and uniform size continuously in a liquid circulating system without lowering the yield of grains and increasing the size of equipment.

Another important object of the invention is to vary the size of crystalline grains desiredly without disturbing the uniformity of grain size substantially in a continuous crystallization system of the liquid circulating A more specific object of the invention is to avoid the natural occurence of fine kernels substantially in a continuous crystallization system of the liquid circulating type by bringing the liquid into the crystallizer from the saturator within a given length of time after the liquid is supersaturated in the saturator following its having been unsaturated.

Another more specific object of the invention is to control the natural occurence of fine kernels desiredly in a continuous crystallization system of the liquid circulating type by varying the time lag from supersaturating to crystallizing of the liquid which is supersaturated following its having been unsaturated.

Other objects and various features of the invention will be more apparent from the following description when read in connection with the accompanying drawings, in which:

FIG. 1 is a diagramatic view, in vertical elevation, of a liquid circulating system for continous crystallization in which the method of the invention may be carried out;

FIG. 2 is a diagramatic view, in vertical elevation, of a modification of the liquid circulating system in FIG. 1 for continuous crystallization in which the imethod of the invention may be carried out;

FIG. 3 shows process diagrams of crystallization for copper sulfate in the liquid circulating systems of FIGS. 1 and 2;

FIG. 4 is a diagramatic view, in vertical elevation, of another modification of the liquid circulating system in FIG. 1 for continuous crystallization in which the method of the invention may be carried out;

FIG. 5 is a diagramatic view, in vertical elevation, of a modification of the liquid circulating system in FIG. 4 for continuous crystallization in which the imethod of the invention may be carried out;

FIG. 6 is a diagramatic view, in vertical elevation, of another modification of the liquid circulating system in FIG. 4 in which the method of the invention may be carried out;

FIG. 7 shows process diagrams of crystallization for copper sulfate in the liquid circulating system of FIG.

FIG. 8 shows process diagrams of crystallization for copper sulfate in the liquid circulating system of FIG.

FIG. 9 is a diagramatic view, in vertical elevation, of still another modification in which the method of the invention may be carried out; and,

FIG. 10 shows a process diagram of crystallization for Glaubers salt in the liquid circulating system of FIG. 9.

The liquid circulating system in FIG. I primarily comprises a crystallizer (l a pump (3) and a saturator (5). The saturator (5) is preferably an evaporator which is combined with a condenser (6). The mother liquor is fed into the circulating system through a pipe (2), and sent to the evaporator (5) by the action of pump (3) by way of a pipe (4). The mother liquor is supersaturated in the evaporator (5) and sent into the crystallizer (1) through a pipe (7) which opens nearly at the bottom of crystallizer (l). The supersaturated liquid then floats upward having some part of its solute crystallized on seeds which are artificially provided in the crystallizer (I). The liquid finally overflows the crystallizer (l) and joins the continuous new feed of mother liquor through the pipe (2). the joined flow being recirculated further. Meanwhile crystalline grains grown on the seeds in the crystallizer (I) sink down by their own gravity relative to the liquid against the upward velocity of liquid. The sedimented grains are finally taken out of the crystallizer (I) by the action ofa slurry pump (8).

conventionally, continuous crystallization in such a circulating system as shown in FIG. 1 is generally subjected to those disadvantages mentioned already. For instance, in crystallization of Glaubers salt for grain size of about 0.5 mm in diameter (about 32 mesh) the maximum allowable supersaturation in the saturator is practically as low as about 0.56 g Na sOJ 100 g H O so as to avoid excessive occurence of fine kernels; in crystallization of copper sulfate for grain size of about 0.8 mm in diameter the maximum allowable supersaturation in the saturator is practically as low as about 0.35 g CuSOJ l 00 g H 0 for the same reason. And such low supersaturation inevitably results in a corresponding low rate of yield per running hour or a corresponding necessity of enlarging the equipment.

The method of the invention may practically be carried out in such a liquid circulating system as shown in FIG. 1. The inventor made a series of experiments with aqueous solution of CuSO -5I-I,O in a liquid circulating system shown in FIG. 1, where the mother liquor with a comparatively low concentration, say, about 35 g Cu- SOJ 100 g H 0 heated at about C was continuously fed through the pipe (2) and the liquid was kept at about 25C in the evaporator (5).

Considering the crystallizing process of the above experiments in reference to FIGS. 1 8c 3, the circulating liquid is positioned at a point (0,) in the supersaturation side when it overflows the crystallizer (l), and then the liquid position is shifted across the solubility curve to a point (b,) in the unsaturation sidebut relatively near to the solubility curve for the purpose of minimizing the required heating calories in the circulating system-as the heated new liquid is continuously fed through the pipe (2). The liquid position is further shifted from the point (b,) in the unsaturation side to a point (c in the supersaturation side again across the solubility curve as the liquid passes the evaporator (5). Here the liquid circulation rate is pre-determined so that the liquid may enter the crystallizer (1) within a given length of time after it leaves the evaporator (5), the given length of time being roughly equal to the time lag to exist from supersaturating to crystallizing of the liquid. Therefore it may be expected that the liquid does not begin crystallization substantially before it enters the crystallizer (l). The liquid position is shifted from the point (0,) to (0,) as the liquid has some part of its solute crystallized in the crystallizer (1).

Actually in the above-described experiments relating to FIG. 1 the natural occurence of fine kernels was restrained considerably though the liquid was supersaturated as high as the point (c resulting in a continuous yield of sodium sulfate grains of quite large and uniform sizes, say, 2.0 mm in diameter with the supersaturation being 0.55 g CuSO /l00 g ",0 at the point (c and 1.5 mm in diameter with the supersaturation being 0.6 g CuSOJlOO g H 0 at the point (c and 1.5 mm in diameter with the supersaturation being 0.6 g Cu- 80 g 11,0 at the point (0,); the higher supersaturation resulted in the smaller grain size because more fine kernels occured before the liquid entered the crystallizer. These yields obtained by the method of the invention are definitely better in size and uniformity than those yields by conventional methods.

The method of the invention may also practically be carried out in such a liquid circulating system as shown in FIG. 2 with aqueous solution of a substance the solubility of which is generally an increasing function of temperature, for instance, CuSO,,-5H O. The circulating system in FIG. 2 comprises the same members as those in FIG. 1 indicated with the same numbers, except an indirect heater (9) is provided intermediate the crystallizer (l) and the feed pipe (2). Accordingly, the liquid is circulated in the system shown in FIG. 2 quite similarly to the system shown in FIG. 1, except that the overflow from the crystallizer (1) is heated before it joins the new feed of mother liquor through the pipe (2).

The inventor made a series of experiments with aqueous solution of CuSO -5l-I O in a liquid circulating system shown in FIG. 2, where the mother liquor with a comparatively high concentration, say, about 55 g Cu- SO 100 g H O kept at room temperature was continuously fed through the pipe (2) and the liquid was kept at about 25C in the evaporator (5).

Considering the crystallizing process of the above experiments in reference to FIGS. 2 & 3, the circulating liquid is positioned at a point (a,) in the supersaturation side when it overflows the crystallizer (l), and then the liquid position is shifted across the solubility curve to a point (a in the unsaturation side as the liquid is heated through the heater (9). The liquid position is further shifted from the point (a,) to a point (b,) in the unsaturation sidebut relatively near to the solubility curve for the purpose of minimizing the required heating calories in the circulating systemas the new liquid is continuously fed through the pipe (2). From (b back again to (a via (0 the liquid position is shifted quite similarly to the circulating system in FIG. 1. And actually in the above-described experiments relating to FIG. 2 quite same results were obtained as in the experiments relating to FIG. 1.

Moreover in another series of experiments relating to FIGS. 1 & 2 carried out by the inventor in accordance with the invention phosphorous acid which is practically impossible to crystallize continuously by conventional methods could even be crystallized continuously into substantially large and uniform sizes.

However the above-mentioned experiments relating to FIGS. 1 & 2 failed to yield larger sizes of copper sulfate grains than 1.5 to 2.0 mm in diameter without disturbing the uniformity of grain size and decreasing the supersaturation in the saturator considerably, though they were quite successful in yielding grains sized not larger than 1.5 to 2.0 mm in diameter. Since grains of large and uniform sizes are not only convenient for separation from the liquid but also superior in chemical property and commercial value, there is often practical need to obtain larger grains than 1.5 to 2.0 mm in diameter for instance in case of copper sulfate.

Then the inventor introduced a new theory of crystallization, re-analyzed the above-mentioned experiments relating to FIGS. 1 8L 2, and established the method of the invention not only for obtaining much larger sizes of grains but also for varying the grain size desiredly without disturbing the uniformity of grain size, which will be fully disclosed in the following part of description.

According to conventional theories crystallization and dissolution of grains are explained on different bases. On one hand according to those theories crystallization occurs by what is called monomolecular transferring summarized in the following. As the mother liquor is supersaturated, the molecular movement of so lute gets activated so much that free molecules collide with each other and are associated into integral clusters, some of which are stabilized as solid kernels when they get larger than a critical size given relatively to the outstanding supersaturation. And then each kernel grows larger as free molecules are attracted and transferred to the kernel in a manner to form the crystal lattice.

On the other hand dissolution occurs by what is called polymolecular transferring summarized in the following. As the mother liquor including crystalline grains is unsaturated, each grain is disintegrated into molecular clusters and then each cluster is dissociated into free molecules.

If crystallization were of monomolecular transferring, it should follow that kernels would grow in a substantially supersaturated liquid condition where most of the free molecules have already been associated into clusters and kernels occur rapidly. In fact, however, kernels grow considerably even in a substantially supersaturated liquid condition where most of the free molecules have already been associated into clusters and kernels occur rapidly, for instance, in the unstable area of standing liquid as alreadly referred to. Therefore the monomolecular transferring theories of crystallization contradict the fact in this respect, thus necessitating new theories instead.

One of the new theories introduced here is that of polymolecular transferring crystallization. According to this new theory, each kernel grows larger as molecular clusters are attracted and transferred to it, while each molecular cluster is formed as free molecules become associated with each other in' the supersaturated liquid. This new theory correctly takes account of the energy balance and molecular collision frequency in crystallization.

Supposing crystallization to occur by polymolecular transferring, it will follow that kernels occur and grow concurrently even in a supersaturated liquid condition where most of the free molecules have already been associated into clusters, since such clusters not only form new kernels but also enlarge the existing kernels, thus explaining the fact satisfactorily in this respect.

Moreover, the new crystallization theory is completely compatible with the established dissolution theories already referred to as they are all of polymolecular transferring. From the standpoint of polymolecular transferring, crystallization and dissolution of grains occur by the same reversible phenomenon carried out symmetrically across the solubility curve. In other words, crystallization is carried out with a positive potential energy in relation to the solubility curve while dissolution is carried out with a negative potential energy in relation to the solubility curve, such positive energy being proportional to the supersaturation of the liquid and such negative energy being proportional to the unsaturation of the liquid, given the solubility and concentration of the liquid.

Since the above-introduced new theory of polymolecular transferring not only explains the actual crystallization satisfactorily but also stands completely compatible with the established dissolution theories; it may be quite right to re-analyze the inventors experiments based on this new theory.

As already referred to, it was discovered by the inventor in his experiments that there is a time lag from supersaturating of flowing liquid to occuring of fine kernels in it, provided that the liquid is supersaturated across the solubility curve after it is unsaturated. According to the polymolecular transferring theory, the time lag will be explained as a length of time in which free molecules are crystallized into kernels via the state of clusters provided that prior to being supersaturated the liquid is unsaturated with the solute completely dissociated into individual free molecules, in other words, a length of time in which free molecules are initially associated into clusters and then such clusters are secondarily associated into kernels. It can further be said that the length of time is shortened if the liquid is unsaturated but with the solute partially dissociated into individual free molecules'and partially remaining as clusters; the more clusters remain in the unsaturated state of the liquid, the shorter the length of time between su-' persaturating and crystallizing of the liquid. In conclusion, the time lag from supersaturating to crystallizing of flowing liquid is determined by the dissociation of molecules in the unsaturated state of the liquid, provided that the liquid is supersaturated'across the solubility curve after it is unsaturated, given the solubility, concentration and supersaturation of the liquid.

Similarly it can be said that there is a time lag from disappearance of kernels in flowing liquid to dissolving of solute in it, provided that the liquid is unsaturated across the solubility curve after it is supersaturated. The time lag will be explained as a length of time in which kernels are dissolved into free molecules via the state of clusters provided that prior to being unsaturated the liquid is supersaturated to a substantial extent with the solute completely crystallized into kernels, in other words, a length of time in which kernels are initially dissociated into clusters and then such clusters are secondarily dissociated into free molecules. And the length of time is shortened if the liquid is not supersaturated sufficiently but with the solute partially crystallized into kernels and partially remaining as clusters. It can similarly be concluded that the time lag from disappearance of crystals to dissolving of solute in flowing liquid is determined by the association of molecules in the supersaturated state of the liquid, provided that the liquid is unsaturated across the solubility curve after it is supersaturated, given the solubility, concentration and unsaturation of the liquid.

Now referring back again to FIGS. 1, 2 & 3, the liquid is positioned at the point (a,) when it overflows the crystallizer (1). In the position (a,) which is still in the supersaturation side of the solubility curve the liquid contains a considerable amount of fine grains and molecular clusters as well as free molecules. The liquid is then transferred from (a to (b,) either directly or via (a,') in order to dissociate such fine grains and molecular clusters into free molecules at the position (b which is in the unsaturation side. But it is taken note that the point (1),) is relatively near to the solubility curve for the purpose of minimizing the required heating calories in the circulating system, therefore providing relatively small potential energy for the dissociation of molecular clusters. It will then require a relatively long time to dissociate the clusters completely into free molecules with such relatively small potential energy, in other words, the liquid will require to remain in the position (b,) for a relatively long time for all the clusters to become dissociated into free molecules.

If however the liquid is sent into the evaporator (5) too early, the liquid position will be shifted from (b,) to (0 before the clusters are completely dissociated into free molecules at (b,), thereby shortening the time lag from supersaturating to crystallizing and thus permitting those remaining clusters to associate with each other into fine kernels even before the liquid enters into the crystallizer (1) from the evaporator (5), and therefore there will be a considerable amount of fine kernels floating together with artificially provided seeds in the crystallizer (l). The more the fine kernels occur, the more the growth and uniformity of grains are restrained in the crystallizer. This might actually have happened in the inventors experiments relating to FIGS. 1 & 2, which failed to yield any larger size of copper sulfate grains than 1.5 to 2.0 mm in diameter without disturbing the uniformity of grain size and decreasing the supersaturation in the saturator substantially.

Therefore it is practically indispensable to obtaining crystalline grains with substantially large and uniform size in a continuous liquid circulating system of the kind to which the present invention relates, that the molecular clusters are almost completely dissociated into free molecules before the liquid is sent into the saturator. And there are generally two ways to dissociate the molecular clusters almost completely into free molecules before the liquid is sent into the saturator, that is, before the liquid position is shifted from the unsatu ration side to the supersaturation side across the solubility curve following a prior reverse crossing over the same curve, or in other words, at a liquid position corresponding for instance to the point (b,) in FIG. 3; one is to have the liquid stand or age for a relatively short time at an unsaturated position sufficiently far from the solubility curve prior to supersaturating across the same curve, and the other is to have the liquid stand relatively age for a sufficiently long time at an unsaturated position relativly near to the solubility curve prior to supersaturating across the same curve.

By either of the above-mentioned two ways the molecular clusters will be almost completely dissociated into free molecules in the unsaturation side, thereby fully extending the time lag from supersaturating to crystallizing thus preventing the free molecules from associating into fine kernels via the state of molecular clusters too early before the liquid enters into the crystallizer from the saturator. In this case, the free molecules may mostly be associated into clusters before the liquid enters into the crystallizer from the saturator, and the given time lag will be over after the liquid enters into the crystallizer.

When the given time lag is over in the crystallizer, the clusters will begin either to be crystallized into fine natural kernels or to be crystallized on the artifically provided seeds. The choice of kernel or seed will be determined under the law of universal gravitation. Since a molecular cluster which is so fine as to be invisible has much less mass than an artificially provided seed which is large and visible, the molecular cluster will be much more frequently attracted to a seed than to another cluster or a free molecule which is still smaller than a cluster, in other words, the growth of artificially provided seeds will be much more likely to happen than the occurence of natural fine kernels in the crystallier, thus making it possible to restrain the occurence of natural fine kernels almost completely in the liquid of substantially high supersaturation, and therefore permitting the seeds to grow much larger and faster than otherwise.

Based on the above-described re-analysis the inventor carried out further experiments and finally has established the method of the invention not only for obtaining substantially large sizes of crystalline grains but also for varying the grain size desiredly without disturbing the uniformity of grain size in the liquid circulating system of the continuous crystallization.

The method of the invention may practically be carried out in such a liquid circulating system as shown in FIG. 4 with an aqueous solution of a substance the solubility of which is generally an increasing function of temperature, for instance, CuSo '5H O. The circulating system in FIG. 4 comprises the same members as those in FIG. 1 indicated with the same numbers, but with an indirect heater (9) provided intermediate the crystallizer (l) and the feed pipe (2). Accordingly, the liquid is circulated in the system shown in FIG. 4 quite similarly to the system shown in FIG. 1, except that the overflow from the crystallizer (1) is heated before it joins the new feed of mother liquor through the pipe (2).

The inventor made a series of experiments with aqueous solution of CuSO -5H O in a liquid circulating system shown in FIG. 4, where the liquid was kept at about 25C in the evaporator (5).

Considering the crystallizing process of the above experiments in reference to FIGS. 4 & 7, the circulating liquid is positioned at a point (a in the supersaturation side when it overflows the crystallizer (l), and then the liquid position is shifted across the solubility curve to a point in the unsaturation side as the liquid is heated to a comparatively high temperature through the heater (9), the point (a being sufficiently far from the solubility curve. The liquid position is further shifted from the point (a to a point (b in the unsaturation side as the new liquid is continuously fed through the pipe (2), the point (b being sufficiently far from the solubility curve so that the molecular clusters may be dissociated completely into free molecules. Then the liquid position is still further shifted from the point (b-,.) in the unsaturation side to a point in the supersaturation side again across the solubility curve as the liquid passes the evaporator Here the liquid circulation rate is pre-determined so that the liquid may enter the crystallizer (1) within a given length of time after it leaves the evaporator (5), the given length of time being roughly equal to the time lag existing from supersaturating to crystallizing of the liquid. Therefore it may be expected that the liquid does not begin to crystallize fine kernels substantially before it enters'the crystallizer, though it permits the free molecules to be associated mostly into molecular clusters. The liquid position is shifted from the point (c,) to (a,) as the liquid has some part of its molecular clusters crystallized on the seeds in the crystallizer (1) after the time lag elapses.

Actually in the above'described experiments relating to FIG. 4 the natural occurence of fine kernels was restrained almost completely though the liquid was supersaturated as high as the point (0 resulting in continuous rapidyield of copper sulfate grains with a substantially large and uniform size, say, 4.0 mm in diameter and over percent in uniformity with the supersaturation being 0.6 g CuSOJ g I-I O at the point (c Moreover, the substantially same size and uniformity of grains were obtained even over 20 hours of continuous operation.

It may be as good to provide a feed pipe (2') before the heater (9) in place of the pipe (2) provided after the heater (9) as illustrated in FIG. 4. In case the feed pipe (2') is provided before the heater (9) or in other words intermediate the crystallizer (l) and the heater (9), the liquid position is shifted from (a to (b via (a in FIG. 7; that is, initially from (a to (a as the new liquid is continuously fed through the pipe (2') and then from (a to (b as the liquid is heated to a comparatively high temperature through the heater (9), the point (a being in the supersaturation side and relatively near to the solubility curve. From (b back again to (a via (0 the liquid position is shifted quite similarly to the case with the feed pipe (2) provided after the heater (9) in FIG. 4. And actually in the inventors experiments he obtained quite same satisfactory results with the feed pipe (2') as with the feed pipe (2).

The method of the invention may also practically be carried out in such a liquid circulating system as shown in FIG. 5 with aqueous solution of a substance the solubility of which is generally an increasing function of temperature, for instance, CuSO -5I-I O. The circulating system in FIG. 5 comprises the same members as those in FIG. 4 indicated with the same numbers, except but with a dissociator (10) provided intermediate the feed pipe (2) and the pump (3). Accordingly, the liquid is circulated in the system shown in FIG. 5 quite similarly to the system shown in FIG. 4, except that the liquid is kept staying for some while in the dissociator (10) after the new mother liquor is joined to the circulation through the pipe (2).

The inventor made a series of experiments with aqueous solution of CuSO '5I-I O in a liquid circulating system shown in FIG. 5, where the liquid was kept at about 25C in the evaporator (5) and designed to stay for about 3 minutes in the dissociator (10).

Considering the crystallizing process of the above experiments in reference to FIGS. 5 & 8, the circulating liquid is positioned at a point (a in the supersaturation side when it overflows the crystallizer (l), and then the liquid position is shifted across the solubility curve to a point (a;,') in the unsaturation side as the liquid is heated through the heater (9), the point (a being relatively near to the solubility curve. The liquid position is further shifted from the point (a;,) to a point (h in the unsaturation side as the new liquid is continuously fed through the pipe (2), the point (b being relatively near to the solubility curve. The liquid position is kept at (b for about 3 minutes as the liquid is kept staying for the same minutes in the dissociator (10), so that the molecular clusters may be completely dissociated into free molecules. Then the liquid position is still further shifted from the point (h in the unsaturation side to a point (0;, in the supersaturation side again across the solubility curve as the liquid passes the evaporator (5). Here the liquid circulation rate is pre-determined so that the liquid may enter the crystallizer (1) within a given length of time after it leaves the evaporator (5), the given length of time being roughly equal to the time lag existing from supersaturating to crystallizing of the liquid. Therefore it may be expected that the liquid does not begin to crystallize fine kernels substantially before it enters into the crystallizer, though it permits the free molecules to be associated mostly into molecular clusters. The liquid position is shifted from the point to (11 as the liquid has some part of its molecular clusters crystallized on the seeds in the crystallizer (1) after the time lag elapses.

Actually in the above-described experiments relating to FIG. the natural occurence of fine kernels was restrained even more completely than the experiments relating to FIG. 4 though the liquid was supersaturated as high as the point (0 resulting in continuous rapid yield of copper sulfate grains with a substantially largge and uniform size, say, 5.0 mm in diameter and over 95 percent in uniformity with the supersaturation being 0.6 g CuSO /100 g H O at the point (c Now the restraint of natural kernel occurence in the above-described experiments relating to FIG. 5 according to the invention may be evaluated in comparison with operations by conventional methods on the basis of copper sulfate crystallization. As already mentioned, copper sulfate grains sized 0.8 mm in diameter were obtained by conventional methods with the maximum allowable supersaturation of about 0.35 g CuSO 100 g H O, while grains sized 5.0 mm in diameter were obtained by the method of the invention relating to FIG. 5 with an allowable supersaturation of 0.6 g CuSOJ 100 g H O.

The total number of grains N contained in the total crystalline yield W kg will be given in the following relation:

N W/dkD where;

d Density of grain k Volumetric factor D Diameter of grain Then the total number of grains N contained per unit kilogram of the total crystalline yield by conventional methods will be as follows:

And the total number of grains N, contained per unit kilogram of the total crystalline yield by the method of the invention will be as follows:

Thus the restraint of natural kernel occurence by the method of the invention will be numerically expressed in terms of a ratio N,/N as below:

Neil filly/2&3? 0:5.12l12i5l99fl In other words, the occurence of natural kernels is restrained by the method of the invention roughly one two-hundred-fiftieth of that of the conventional methods, even with the supersaturation increased about 1.7 times that of the conventional methods.

It may be as good to provide a feed pipe (2) before the heater (9) in place of the pipe (2) provided after the heater (9) as illustrated in FIG. 4. In case the feed pipe (2) is provided before the heater (9) or in other words intermediate the crystallizer (1) and the heater (9), the liquid position is shifted from (a,) to (b,) via (a,") in FIG. 8; that is, initially from (a,) to (a;") as the new liquid is continuously fed through the pipe (2), arid then from (a,,") to (b;,) as the liquid is heated through the heater (9), the point (11 being in the supersaturation side and relatively near to the solubility curve. From (12 back again to (u via (0 the liquid position is shifted quite similarly to the case with the feed pipe (2) provided after the heater (9) in FIG. 5. And actually in the inventors experiments he obtained quite same satisfactory results with the feed pipe (2') as with the feed pipe (2).

It may also be good to provide a bypass pipe (10') in parallel with the pump (3) as illustrated in FIG. 6, in place of the dissociator (10) shown in FIG. 5, so that the liquid may stay for some time sub-circulating by way of the bypass pipe (10) to dissociate the molecular clusters completely into free molecules before the liquid enters into the evaporator (5).

It will be possible in accordance with the invention that the liquid position is shifted far from the solubility curve in the unsaturation side in such a liquid circulating system as shown in FIG. 1. This can be done by adding through the feed pipe (2) new mother liquor of comparatively low concentration and high temperature for a solubility curve which rises slowly in proportion to temperature, and of comparatively high concentration for a solubility curve which rises acutely in proportion to temperature.

So far consideration has been given to situations in which the solubility is generally an increasing function of temperature. The method of the invention may practically be carried out in such a liquid circulating system as shown in FIG. 9 with aqueous solution of a substance the solubility of which is generally a decreasing function of temperature, for instance, Na,SO

The liquid circulating system shown in FIG. 9 generally comprises similar members, but somewhat in reverse arrangement, to the one shown in FIG. 1. The overflow from a crystallizer (11) is initially sent into an evaporator (15) which is in combination with a condenser (16). Then the liquid is unsaturated through an indirect cooler (19). The unsaturated liquid is joined with new mother liquor fed by way of a pipe (12). The joined liquid is then kept staying for some time in a dissociator (20). The dissociated liquid is sent into a saturator (15') by the action of a pump (3). The saturator (15') is preferably an indirect heater. The liquid supersaturated through the heater (15) is sent into the crystallizer (l 1) through a vertical pipe which opens nearly at the bottom of crystallizer (11). The supersaturated liquid then floats upward having some part of its solute crystallized on seeds which are artificially provided in the crystallizer (11). The liquid finally overflows the crystallizer (11) for recirculation, while crystalline grains grown on the seeds in the crystallizer (l1) sink down by their own gravity relative to the liquid against the upward velocity of liquid, the sedimented grains finally being taken out by the action of a slurry pump. The inventor made a series of experiments with aqueous solution of Na,SO in a liquid circulating system shown in FIG. 9.

Considering the crystallizing process of the above experiments in reference to FIGS. 9 & 10, the circulating liquid is positioned at a point (0 in the supersaturation side when it overflows the crystallizer (ll and then the liquid position is shifted to a point (a,') as the liquid passes through the evaporator (15), the point (0 being in the supersaturation side and relatively near to the solubility curve. The liquid position is further shifted from (a,') to (a,") across the solubility curve as the liquid goes through the cooler (19), the point (a,") being in the unsaturation side and relatively near to the solubility curve. The liquid position is still further shifted to a point (b in the unsaturation side as the new liquid is continuously fed through the pipe (12), the point (b being relatively near to the solubility curve. The liquid position is kept at (b.,) while the liquid is kept staying in the dissociator (20) until the molecular clusters may be completely dissociated into free molecules. Then the liquid position is shifted from the point (b.,) in the unsaturation side to a point in the supersaturation side again across the solubility curve as the liquid passes the heater Here the liquid circulation rate is pre-determined so that the liquid may enter the crystallizer (11) within a given length of time after it leaves the cooler (15'), the given length of time being roughly equal to the time lag existing from supersaturating to crystallizing of the liquid. Therefore it may be expected that the liquid does not begin to crystallize fine kernels substantially before it enters into the crystallizer, though it permits the free molecules to be associated mostly into molecular clusters. The liquid position is shifted from the point (c.,) to (a.,) as the liquid has some part of its molecular clusters crystallized on the seeds in the crystallizer (11) after the time lag elapses.

Actually in the above-described experiments relating to FIG. 9 the natural occurence of fine kernels was restrained quite satisfactorily though the liquid was supersaturated as high as the point (0 resulting in continuous rapid yield of Glaubers salt grains with substantiall ylarge and uniform sizes, say, 2.0 mm in diameter with the supersaturation being 1.0 g Nasonioo g H O at the point (0 and 0.5 mm in diameter (about 32 mesh which is the size of Glaubers salt most demanded commercially) with the supersaturation being 1.67 g Na SO 100 g H2O at the point (6 which means a yield about three times the conventional methods by which the maximum allowable supersatyxatiqaifl-sEaSQ4/ Q9; H O- H, a

In accordance with the invention the size of crystalline grains is varied desiredly without disturbing the uniformity of grain size substantially in a continuous crystallization system of liquid circulating type. As already revealed, the time lag from supersaturating to crystallizing of flowing liquid is determined by the dissociation of molecules in the unsaturated position of liquid, provided that the liquid is supersaturated across the solubility curve after it is unsaturated, given the solubility, concentration and supersaturation of the liquid.

It follows that the time lag is controlled desiredly by the molecular dissociation in the unsaturated position of liquid. The natural occurence of fine kernels is controlled desiredly by the time lag. And the grain size is determined by the natural occurence of fine kernels; the less the fine kernels occur, the larger the grains grow, as already fully revealed. Therefore it can be said that the grain size is determined by the dissociation of molecules in the unsaturated position of liquid, provided that the liquid is supersaturated across the solubility curve after it is unsaturated, given the solubility, concentration and supersaturation of liquid.

Practically in such liquid circulating systems as shown in the accompanying drawings, the molecular dissociation is controlled at the points (b (b (b and (b in FIGS. 3, 7, 8 and 10, respectively. If such 6 a point is positioned sufficiently far from the solubility curve, the molecular clusters are dissociated fully into free molecules in a relatively short time because of a comparatively high potential energy for dissociation there; and the nearer the point is positioned to the solubility curve, the less quickly are the molecular clusters dissociated into free molecules. If such a point is positioned relatively near to the solubility curve, a sufficiently long time is required to dissociate the molecular clusters fully into free molecules at such liquid position because of a comparatively low potential energy for dissociation there; and the shorter the given time is at the liquid position near to the solubility curve, the less the molecular clusters are dissociated into free molecules.

It will thus be seen that the crystallization in accordance with the invention continuously produces crystalline grains with substantially large and uniform size without lowering the yield of grains and increasing the size of equipment as well as varies the size of crystalline grains desiredly while still obtaining substantial uniformity of grain size in a liquid circulating system.

Since certain changes and modification may be made in the invention, some of which have been herein suggested, it is intended that the foregoing shall be construed in a descriptive rather than in a limiting sense.

What I claim is:

1. An improved crystallization method of the type that comprises, in combination, the steps of:

a. withdrawing a supersaturated solution of a crystallizable solute from a top region of a liquid upflow crystallizer, said withdrawn solution constituting essentially the entire liquid flow through said crystallizer,

b. augmenting the withdrawn solution by mixing added feed solution therewith and altering the temperature thereof to render the augmented solution unsaturated, the concentration of solution in the added feed solution being substantially higher than that of the solution withdrawn from the top region of the crystallizer,

c. subsequently rendering the augmented solution supersaturated by treatment in a saturator, such augmented solution constituting essentially the entire flow through said saturator,

d. delivering the solution supersaturated in step (c) through a central conduit into a bottom region of said crystallizer,

e. causing the solution delivered in step (d) to flow upwardly in said crystallizer in surrounding relation to the central conduit while a part of its solute crystallizes and sediments countercurrent to said flow of liquid therein, and communication of fluid in said conduit with fluid in surrounding relation thereto being only at the bottom of the crystallizer, and

f. discharging the sedimented crystals from a bottom region of the crystallizer, wherein the improvement comprises:

g. subjecting the solution following its unsaturation in step (b), to aging for a predetermined time sufficient to effect essentially complete dissociation of the solute into free molecules in the solution, before subjecting it to the treatment of step (c), for assuring that essentially no solute in crystal or visible kernel form is supplied to step (c), and

h. in step (d) delivering the solution to the crystallizer following its supersaturation in step (c) within a predetermined time which is sufficiently short to not allow molecular clusters of solute forming therein to attain the state of visible kernels before such delivery for assuring that essentially no solute in crystal or visible kernel form is contained in the solution supplied to said crystallizer.

2. An improved method as claimed in claim 1, wherein the solute is selected from the class consisting of copper sulfate, sodium sulfate and iron sulfate.

3. An improved crystallization method as claimed in claim 1, wherein the solubility of the solute increases with increase of temperature, and in step (b) the temperature is altered by raising the same.

4. An improved method as claimed in claim 3, wherein the solute consists essentially of copper sulfate.

5. An improved crystallization method as claimed in claim 1, wherein the solubility of the solute increases with decrease of temperature, and .in step (b) the temperature is altered by reducing the same.

6. An improved method as claimed in claim 5,

wherein the solute consists essentially of sodium sulfate.

7. An improved crystallization method as claimed in claim 5, wherein in step (c) the augmented solution is heated in the saturator in efi'ecting the supersaturation thereof.

8. An improved crystallization method as claimed in claim 5, wherein in step (c) the augmented solution is concentrated by evaporation in the saturator in effecting the supersaturation thereof.

9. An improved crystallization method as claimed in claim 3, wherein in step (c) the augmented solution is cooled in the saturator in efi'ecting the supersaturation thereof.

10. An improved crystallization method as claimed in claim 3, wherein in step (c) the augmented solution is concentrated by evaporation in the saturator in effecting the supersaturation thereof.

Claims

2. An improved method as claimed in claim 1, wherein the solute is selected from the class consisting of copper sulfate, sodium sulfate and iron sulfate.
3. An improved crystallization method as claimed in claim 1, wherein the solubility of the solute increases with increase of temperature, and in step (b) the temperature is altered by raising the same.
4. An improved method as claimed in claim 3, wherein the solute consists essentially of copper sulfate.
5. An improved crystallization method as claimed in claim 1, wherein the solubility of the solute increases with decrease of temperature, and in step (b) the temperature is altered by reducing the same.
6. An improved method as claimed in claim 5, wherein the solute consists essentially of sodium sulfate.
7. An improved crystallization method as claimed in claim 5, wherein in step (c) the augmented solution is heated in the saturator in effecting the supersaturation thereof.
8. An improved crystallization method as claimed in claim 5, wherein in step (c) the augmented solution is concentrated by evaporation in the saturator in effecting the supersaturation thereof.
9. An improved crystallization method as claimed in claim 3, wherein in step (c) the augmented solution is cooled in the saturator in effecting the supersaturation thereof.
10. An improved crystallization method as claimed in claim 3, wherein in step (c) the augmented solution is concentrated by evaporation in the saturator in effecting the supersaturation thereof.