MASS CRYSTALLIZATION.

II. CRYSTALLIZATION AT MANY LEVELS UNDER CONSTANT FEEDING¹

A. M. KUZMIN

Mass crystallization at many levels under “constant” feeding of the solution had as its aim to follow, over a relatively long period of time, the phenomena of growth and dissolution of crystals. “Constant” feeding of solutions in the present case is understood as an additional feeding which, at definite intervals of time, was introduced into the solution in order to prevent its depletion and thereby to maintain continuous crystal growth.

Experimental investigations of this kind were undertaken with the aim of showing that the concentration flows accompanying the crystallization process are not only capable of mixing a supersaturated solution, but, acting against the force of gravity, are capable of entraining dissolved substance from below upward and, on their way, like vortical jets acting in one direction, drawing in supersaturated solution from the sides and thereby, contrary to expectations, promoting a higher growth rate of crystals of the upper levels in comparison with crystals of the lower ones.

To solve the problem just posed, a series of experiments was carried out, and in the present work we shall dwell on one of them.

In contrast to the experiment of P. A. Zemyatchensky with 3 crystals (1914), in our experiment 33 crystals participated in simultaneous crystallization; they were arranged in three horizons, equally (5 cm) spaced from one another. The upper horizon was 6 cm below the level of the solution, and the lower level of crystals was 5 cm from the bottom of the crystallizer. The crystals of the lower levels were located under the crystals of the upper ones, i.e., three crystals of different levels, being suspended one beneath another, formed one column. The columns, 11 in number, were arranged at equal distances relative to one another. The crystallizer with the 33 alum crystals immersed in the solution was tightly closed, in order to reduce evaporation to a minimum and thereby eliminate this factor of supersaturation of the solution (Fig. 1).

The experiment lasted 2 months. Every 6 days the crystals were removed from the solution, air-dried, weighed, and again lowered into the crystallizer. From February 23 to May 6 the crystallization temperature fluctuated around 13.5°C, only very rarely approach-

Growth and dissolution of crystals, column 4
Table 1

toward the extreme limits of 14° and 13°; the most frequent deviations in either direction did not exceed 0.2–0.3°. Every three six-day periods the solution was fed by the addition of dry salt. To illustrate the behavior of all 33 crystals, we shall present a general diagram

Fig. 1. Diagram of the crystallizer with 11 columns of crystals suspended in it

of the curves of growth and dissolution of the crystals in grams (Fig. 2) and a general diagram of the relative and absolute magnitudes of crystal growth, expressed in percent (Fig. 3). For lack of space we present two tables of growth and dissolution of crystals of two columns, characterizing the behavior of the crystals from one weighing to the next. These tables give figures characterizing the weight of the crystal in grams on the day of its weighing; then figures of relative and absolute growth in grams are given, and the figures of relative and absolute growth in percent are computed.

Studying all the numerical and graphical material on the crystallization of the crystals of 11 columns, and studying all the phenomena accompanying crystallization at this stage, we arrive at the following conclusions.

1. The general diagram of the growth of 33 crystals (Fig. 2) and the general diagram of relative and absolute growth (Fig. 3) clearly show that the growth of crystals at three different levels generally proceeded at different rates. For example, this is indicated by the figures and curves of grow-

…of the crystals that occurred from 1.III to 6.III, indicating a rapid—or, more precisely, hurricane-like—rate of crystal growth. During these 6 days all the crystals showed a very sharp tendency to grow and considerably exceeded their dimensions in comparison with

Fig. 2. General diagram of the growth and dissolution of 33 crystals (in grams), suspended at three levels.
X—days on which the solution was replenished.

their initial size and with the weight of the crystals on March 1. The crystals subsequently showed more gradual growth over a comparatively long period of time. Depletion of the solutions at this stage of crystallization is forestalled by three replenishments, which took place on March 12 and 30 and on April 18.

Fig. 3. General diagram of the relative (1, 2, 3) and absolute (I, II, III) values of increase of the same 33 crystals (in percent).
X—days on which the solution was replenished.

Comparing the curves of absolute and relative increase in the crystals from one weighing to the next, one may say that, as a rule, the crystals of the upper horizon grow considerably faster than the crystals of the middle one, while the rate of growth of the crystals of the middle horizon is noticeably faster than the rate of growth of the crystals of the lower one. This conclusion directly…

directly follows from the general diagrams shown in Fig. 1. From this rule, as follows from the tables, there are comparatively rare exceptions.

1. After relatively vigorous crystal growth, in a number of cases the reverse process sets in—dissolution. Thus, after hurricane-like growth during 1.III–6.III, the crystals extracted from the solution a larger quantity of dissolved substance than was required by the state of equilibrium. Exactly the same picture is found in the interval 6.IV–6.V. As a result of the prolonged and steady growth of the crystals, which lasted from 6.IV to 30.IV, more dissolved substance was removed from the solution than should have been, which led to the subsequent (30.IV–6.V), though very slight, dissolution of the crystals.

2. Throughout the entire course of this experiment on mass crystallization, precipitates of crystals on the bottom of the crystallizer are, as a rule, absent. In the interval 23.II–1.III, 11 columns of crystals were mistakenly lowered into the still-hot solution. This circumstance led, on the one hand, to a comparatively sharp dissolution of the suspended crystals, and on the other—upon cooling of the solution to 13.5°—to the appearance on the bottom of the crystallizer of abundant crystals, which during the first two days continued to increase both in their dimensions and in number; on the fifth and especially on the sixth day it became noticeable that the crystalline precipitate had begun to dissolve, but on March 1 the crystals on the bottom of the crystallizer were still abundant, while weighing showed that all 33 crystals, in comparison with their initial weight, had decreased in size. By March 4 the crystals on the bottom had completely disappeared. Weighing on March 6 showed that, by weight, the crystals considerably exceeded their initial size, and in comparison with the dimensions on March 1 they showed an even sharper increase in their weight.

This observed phenomenon strengthened our intention to apply, during crystallization, dry feeding by adding dry salt to the solution. During the periods 12.III–18.III and 18.IV–24.IV, this latter, on the bottom of the crystallizer, passed completely into solution not later than by the end of the fourth day, thereby maintaining its supersaturation.

During 30.III–6.IV the crystals dissolved somewhat; moreover, the crystals of the lower horizon exhibited a relatively high rate of dissolution; the crystals of the two other horizons showed a noticeably lower rate of dissolution. On the one hand, the dissolution of the crystals was a consequence of the general depletion of the solution; on the other, feeding of the solution may also have played a depressing role, since on March 30, as feeding, twice as much salt was added to the solution as on March 12. At first the feed itself attracted dissolved substance to itself, and its crystals increased in size, and this circumstance somewhat intensified the dissolution of the suspended crystals. However, on April 4 a clear tendency toward dissolution of the crystalline precipitate appeared, but on April 6, as weighing showed, the crystals of the columns had not reached the sizes they had had by March 30.

1. One of the principal factors in the overall growth of crystals, and in the preferential growth of the upper crystals as compared with the lower ones, is concentration currents. To explain the cause responsible for the relatively rapid growth of the crystals of the upper levels in comparison with the lower-lying crystals in the column, it must be said that our ideas about concentration currents, as they were interpreted by G. V. Wulff (1895–1917), were not entirely accurate. It seemed to us that this current, rising above the crystals, consists of solution depleted in its soluble component, and it was assumed that its con-

Growth and dissolution of crystals of the 10th column

Table

case when the bottom of the crystallizer must correspond to a saturated or slightly supersaturated state. It was believed that, as a result of concentration flows, the upper part of the solution is diluted and, in general, becomes depleted in dissolved substance.

However, our experiments on mass crystallization showed that the generally accepted conception of concentration flows does not correspond to reality. The concentration flow (Fig. 4), arising above each crystal and acting in one and the same direction for a relatively long time, is capable not only of mixing the solution, but also, as a vortex flow, of drawing it in—like the action of a centrifugal pump or injector—and carrying it upward; that is, the concentration flow, by its nature, has almost the same supersaturation concentration as the solution. This circumstance is a very important factor in crystallization, and the relatively high velocity of motion of the supersaturated solution in the concentration flow simply explains to us why crystals from the upper horizons, as a rule, showed a higher growth rate in comparison with crystals from the lower horizons.

Fig. 4. Column of crystals; concentration flows washing the crystals of the column. Seed crystals at the bottom of the crystallizer with concentration flows above them. → — direction of crystallization diffusion; the crossed arrow indicates the direction of portions of supersaturated solution drawn into the flow.

Further, when crystallization is being discussed, the leading role in crystal growth was usually ascribed to the concentration flow, and in the case of weakly supersaturated solutions the slow growth of crystals was explained by the action of diffusion. However, it is clear that the growth of a crystal cannot be imagined without diffusion of the medium feeding it (crystallization diffusion) and without the concentration flow. In essence, crystallization diffusion and the concentration flow are two aspects of one and the same phenomenon, which are interrelated and mutually conditioned. Therefore, depending on the state of the solution during crystallization, the rate of the diffusion flow toward the crystal may increase or decrease, and accordingly the strength of the concentration flow above the crystal will increase or decrease.

As indicated above, in the period from February 23 to March 1 there was an abundant crystalline precipitation.

the precipitate, which, closer to March 1, had already begun to dissolve and over the next six days disappeared completely, while the crystals suspended above it vigorously increased in size. It should therefore be explained that above the crystals of the precipitate, concentration plumes rose in the solution; similar concentration jets also rose above the suspended crystals, and in their combined action they set the entire mass of the solution in motion. The suspended crystals in the solution were many times larger than the entire mass of crystals on the bottom of the crystallizer, and therefore they had great capacity to create around themselves feeding regions of relatively large diameter and thereby to produce a high rate of crystallizational diffusion of the dissolved substance to the crystal and, naturally, to create under these conditions powerful concentration currents and thus to bring about the corresponding growth of crystals at different levels.

This first observation of the disappearance of the crystalline precipitate and the simultaneous growth of crystals in the period between March 1 and 6, which confirmed our expectations, gave us, as was already said above, the possibility of periodically (12.III, 30.III, and 18.IV) feeding the solution by adding dry salt to it. And each time, after a certain time, this addition, disappearing completely, passed into the solution and promoted the simultaneous growth of all 33 crystals over a relatively long period.

1. The growth of the octahedral faces of alum proceeds, as a rule, unevenly. Regardless of the position of the faces, in their growth they advance by different amounts. Comparatively often the accretion of a face proceeds in steps, which arise now here, now there, then disappear, then appear again. Alongside the general growth of the crystals, partial dissolution of faces takes place. The unevenness of crystal growth in the case of mass crystallization is explained by the fact that concentration currents are not constant in density and do not wash the growing crystal to the same extent. Further, the general unevenness of crystal growth depends to a certain degree on the part of the crystallizer in which the crystal is located: crystals in the central part of the crystallizer grow relatively faster than crystals at the periphery.

2. The suspended crystals initially represented, in most cases, isometric imperfect crystals, which during the entire crystallization process acquired a tendency to become more or less elongated in the horizontal direction. Such a phenomenon indicates that the law of terrestrial gravitation, despite the complex character of the motion in the solution, manifests itself noticeably in the growth of the crystals of our experiment.

3. The method of crystallization at many levels with the use of feeding provides a basis for the construction of a new type of crystallizer, in which the growth of crystals should proceed from the very beginning under ordinary or any specified conditions, i.e., without preliminary heating of the solution to obtain its saturated state at any relatively high temperature and subsequent slow, gradual cooling.

4. In conclusion it should be said that experimental investigations of mass crystallization and the study of the character and role of concentration currents give mineralogists and petrographers new ways of explaining many processes of mineral formation and the formation of rocks accompanying the evolution of magma, and may be widely used in industry for obtaining technically necessary crystals.

References

1. Wulff, G. V. On the question of the rates of growth and dissolution of crystalline faces. — Warsaw University News. 1895–1896.

2. Wulff, G. V. Crystals, Their Formation, Appearance, and Structure. Moscow, 1917.

3. Zemyatchensky, P. A. Studies in crystallogenesis. IV. Splitting of supersaturated solutions. — Proceedings of the Academy of Sciences, Series VIII, Physico-Mathematical Section, Vol. XXXIII, No. 5, 1914.

1. Report presented at the scientific conference on dielectrics and semiconductors on February 6, 1958, in Tomsk. ↩