Study of the Crystallization of Epsomite and Mirabilite

S. A. Stroitelev

(Presented by Professor A. M. Kuzmin)

The significance of dehydration for crystal growth has not yet been definitively clarified. The absence of special methods that would make it possible to study experimentally the dehydration of solution compounds during crystallization creates great difficulties in research.

The present article contains the results of experimental studies of crystallization from solutions by means of a special method developed by the author. The paper briefly sets forth the foundations of this method and discusses the results of the experiments.

Method of Investigation

To solve the tasks posed, it was necessary to apply a method of investigation which, by taking into account changes in the general state of the solution, would help reveal signs of the manifestation of dehydration during crystallization. After lengthy searches we succeeded in establishing that the above requirements are more or less met by a method technically based on the precise measurement of the temperature and volume of the system (solution and crystals). The method itself may therefore be called volumetric-thermal.

Crystallization was carried out in a glass vessel with a capacity of 0.5 l (Fig. 1). Through a rubber hose a glass tube 2 mm in diameter and 1.3 m long was attached to the vessel. It served as a receiver for the solution squeezed out of the vessel during crystallization and for measuring changes in the volume of the system. Changes in the volume of the system were measured by the position of the solution meniscus relative to the graduations of the tube scale, with an accuracy of up to 0.5 mm. To prevent crystallization in the measuring tube, its inner end was brought into the upper, less concentrated layers of the solution, while the outer part was filled with saturated solution up to the required position of the meniscus. The temperature of the system was measured with a thermometer having a scale division of 0.1°C. It was inserted into the vessel through the neck, which was tightly closed with a rubber stopper.

The solution under investigation, heated to a temperature of 48°, was poured into the vessel, filling it completely, and was slowly cooled to the initial crystallization temperature (20°C). Then, through a special tube closed by a rubber hose with a Mohr clamp, 1–2 small salt crystals were introduced into the vessel in order to induce crystallization. During the wo-

systematic observations of crystal growth were conducted throughout the entire experiment. The transparency of the solution and of the vessel walls made it possible to follow the growth of the crystal faces and to draw conclusions about their habit in particular periods of crystallization. In order to determine more precisely the regularity in the change of crystal habit, in a number of cases repeated crystallization was carried out from the same solutions. Each experiment was considered complete when the temperature of the system, having risen from the initial temperature (20°C) to the maximum (22–29°C), again fell to 20°C. The temperature in the laboratory was maintained at about 20° or 19.5°C.

Fig. 1. Diagram of the apparatus: 1—vessel with solution, 2—thermometer, 3—clamp, 4—measuring tube.

The physicochemical method described is based on the fact that the volume of a solution is usually not equal to the sum of the initial volumes of the dissolved substances.

Table 1 presents data on the increase in the volume of the system as a result of crystallization of mirabilite; these data show that the volume effect—the increase in the volume of the system during crystallization of mirabilite—increases with increasing supersaturation of the solution.

From the physicochemical foundations of the theory of solutions it follows that the change in the volume of the system can be used for indirect judgment about the course of dehydration and crystallization (Shakhparonov, 1956).

Table 1

Change in the volume of the system during crystallization of mirabilite

In the case where crystalline hydrates crystallize from an aqueous solution, the dehydration process is limited to the removal only of that part of the water which was bound by the salt upon dissolution. Such salts include those studied here: epsomite—$\mathrm{MgSO_4\cdot 7H_2O}$—and mirabilite—$\mathrm{Na_2SO_4\cdot 10H_2O}$.

The volumetric effect of the system is caused by a change in the interaction of the salt and water during crystallization and dehydration. Crystallization in our case is an exothermic process. When crystals separate from a strongly supersaturated solution, a crystallization stage is observed during which the temperature of the system rises. Dehydration has an endothermic effect and causes a decrease in temperature. Thus, for example, in the thermal analysis of substances on heating curves, the dehydration process is recorded by local changes in the slope of the curves, corresponding to heat absorption and a lowering of temperature. When we constructed curves of the change in the volume and temperature of the system during crystallization (Fig. 2), we likewise found similar local complications in their form, which, as we shall discuss in more detail below, can be associated only with dehydration of the compounds of the solution. At the same time, if dehydration is recorded in thermal analysis only by an endothermic effect, then the volume–thermal method, in studying crystallization from solutions, makes it possible to use for this purpose also the volumetric effect. The degree of hydration of the compounds of the solution may, moreover, be judged from its concentration or supersaturation: the greater the number of water molecules per one molecule of salt, the higher its degree of hydration must be.

The material for the study consisted of supersaturated solutions of the sulfate salts of sodium and magnesium. The latter possess a sufficiently large volume–thermal effect and form large crystals, whose habit changes markedly with changes in supersaturation. This determined the choice of the indicated salts for study.

Crystallization of mirabilite from solutions of different supersaturation

Mirabilite crystals belong to the monoclinic system. Its natural crystals most often have a prismatic habit; they are usually elongated along the $b$ axis and more rarely along the $c$ axis. The crystal forms consist of the faces $[100]$, $[010]$, $[001]$, $[110]$, $[111]$, $[101]$, $[011]$, $[111]$, $[112]$, and $[021]$. Of all those listed, the first seven forms are most strongly developed; the faces $[112]$, $[021]$, and $[111]$ are observed rarely (Fig. 4).

Calculations by known formulas (Ormant, 1950) show that mirabilite possesses a hexahedral lattice. In decreasing order of the relative reticular density of the faces, determined from the tables of I. D. Donnay and the tables of O. M. Anseles, the various simple forms are arranged in the following series: $[100]$, $[001]$, $[101]$, $[110]$, $[111]$, $[111]$, and $[010]$.

The head faces of the crystals $[011]$, $[111]$, $[111]$, and $[010]$ have a lower relative reticular density.

From strongly supersaturated solutions (60–80 g of salt per 100 cm$^3$), crystallization of mirabilite proceeds very rapidly. In two to three minutes, about half the quantity of salt that can separate from the solution in accordance with its initial supersaturation is deposited. The temperature reaches 27–29°C. The meniscus of the solution in the tube moves rapidly, which indicates a sharp increase in the volume of the system.

Mirabilite crystals from a solution at a supersaturation of 80–60 g per 100 cm$^3$, growing from several centers of crystallization, formed radial-rayed aggregates. They had the form of thin, strongly elongated leaves and plates with very broad faces of the first pi-

pinacoid \([100]\) and by narrow faces of the third pinacoid \([001]\) and the prism \([011]\). Other faces, having a lower relative reticular density, were usually not observed.

Having reached a maximum, the temperature of the system decreased smoothly, while the volume continued to increase. During this period the crystals that had precipitated earlier almost ceased growing in length and, chiefly, thickened owing to the predominant growth of the faces \([100]\), \([001]\), \([101]\), which possess the greatest reticular density.

At supersaturations of 35 to 50 g per 100 cm³, crystallization of mirabilite proceeds more slowly. Here three periods of crystallization may be distinguished.

In the first period, over two to five minutes, radial-fibrous aggregates grow from several crystallization centers, consisting of numerous needle-shaped crystals. The transverse section of such crystals is close to rhombic. The temperature and volume of the system increase slowly. The first period of crystallization ends with a certain retardation of growth, or even with a decrease in temperature. However, the volume of the system during the temperature drop continues to increase steadily (Fig. 2, points B₁, B₂, B₃, B₄).

After the drop or cessation of the rise in temperature, in the second period the rate of crystallization increases sharply. The volume increases rapidly, the temperature of the solution rises, and many new crystallization centers appear. The crystals, which previously had a needle-like form, acquire the appearance of elongated plates because of the accelerated growth of the head faces \([011]\), \([111]\), \([1\overline{1}1]\), \([010]\) and of the third pinacoid \([001]\). On the faces of the first and third pinacoids a coarse longitudinal striation appears. In general, crystallization in the second period very closely resembles crystallization from a solution with high supersaturation (60–80 g per 100 cm³). Owing to the rise in temperature and the separation of the excess salt in the form of crystals, the solution for a time becomes saturated. After a brief halt, the temperature of the system decreases and the third period of crystallization begins. The volume of the system continues to increase during this period as well. The crystals almost cease growing in length and mainly thicken. The longitudinal striation of the faces gradually becomes inconspicuous and is replaced by an increasingly pronounced transverse striation. Toward the very end of the experiment, i.e., when the supersaturation of the solution has decreased considerably, faces having a lower relative reticular density become clearly visible on the crystals. These include the faces of the following simple forms: \([101]\), \([110]\), \([011]\), \([111]\), \([1\overline{1}1]\), \([010]\).

From solutions of still lower supersaturation (20–30 g per 100 cm³), crystallization of mirabilite proceeds more slowly. In the first period of crystallization the temperature of the system rises slowly, and the volume changes very little; therefore, on the curves (Fig. 2) it corresponds to almost vertical portions. Along with radial-fibrous aggregates, in this case individual crystals may be observed. They have a needle-like form with a rhombic transverse section.

The first period of crystallization, as a rule, ends with a temperature drop, more rarely with cessation of its rise. This is clearly seen from the curves near points B₁, B₂, B₃ and from fragments of the curves for experiments 8, 13, 24 (Fig. 2)\(^1\).

\(^1\) The temperature drop cannot be explained by fluctuations in the laboratory, since the laboratory thermometer did not show these fluctuations. Nor can it be regarded as a phenomenon of local significance; intense concentration and convection currents, judging from the data of additional experiments, equalize the temperature at different points of the system so rapidly that the differences prove negligible.

In the second period the rate of crystallization increases considerably. The crystals rapidly grow in length. The highest linear growth rate is still shown by the faces of the crystal head \([011]\), \([11\overline{1}]\), \([111]\), \([010]\), which have a lower relative reticular density.

Fig. 2. Graph of the change in volume and temperature of the system during crystallization of mirabilite (supersaturation: I—22 g, II—26 g, III—36 g, IV—40 g per 100 cm³; \(a_{1-4}\)—\(B_{1-4}\)—first, \(B_{1-4}\)—\(C_{1-4}\)—second period of crystallization)

Crystallization of epsomite from solutions of different supersaturation

Epsomite crystallizes in the tetrahedral class of the rhombic system. From supersaturated solutions it separates in acicular, hexagonal, or prismatic crystals. The principal habit faces of epsomite crystals are the faces of the rhombic prism \([110]\) and of the rhombic tetrahedron \([111]\). Of lesser importance are the faces of the pinacoids \([100]\), \((010)\), the tetrahedron \([1\overline{1}1]\), and the prism \([101]\). If the prism faces \([110]\) predominate over the tetrahedron faces \([111]\), epsomite crystals have an acicular, hexagonal, or prismatic form; if the predominant faces are the tetrahedron faces \([111]\), then the form of the crystals is prismatic-tetrahedral.

According to calculations, epsomite has a dodecahedral crystal lattice. In decreasing order of relative reticular density

faces the various simple forms are arranged in the following order: \([110]\), \([111]\), \([11\bar{1}]\), \([010]\), \([100]\), \([101]\). The faces of the figures \([110]\) and \([111]\) have the greatest relative reticular density.

From solutions supersaturated by 55–20 g of salt per 100 cm\(^3\), crystallization at first proceeds with a decrease in the volume of the system, while the temperature changes hardly at all. Then, conversely, the temperature rises, while the volume remains almost constant (Fig. 3). During this period, radial-rayed aggregates arise, consisting of needle-shaped crystals of epsomite. They are readily lifted by concentration currents into the upper horizons of the solution; some of the crystals in the process split off from the aggregate and themselves become centers of formation of new radial-rayed aggregates.

After seven to nine minutes, the second period of crystallization begins. It is distinctive in that, although crystallization of epsomite as a whole causes a decrease in the volume of the system, in the second period, owing to the rise in temperature and the intensification of dehydration, the volume increases considerably. In addition, during crystallization of epsomite from solutions whose supersaturation is close to 25 g per 100 cm\(^3\), a temperature drop accompanied by an increase in the volume of the system can be clearly observed (Fig. 3). This fact, as in the case of crystallization of mirabilite, is of interest from the standpoint of the manifestation of dehydration.

Fig. 3. Graph of the change in volume and temperature of the system during crystallization of epsomite (\(A_1B_1\), \(A_2B_2\), \(A_3C_3\)—first; \(B_1C_1\), \(B_2C_2\), \(B_3C_3\)—second periods of crystallization; supersaturation: 1—53 g; 2—35 g; 3—24 g per 100 cm\(^3\)).

In the second period the rate of crystallization increases considerably. The rate of growth of the faces of the crystal head \([111]\), \([11\bar{1}]\), and \([101]\) increases especially noticeably relative to the rate of growth of the faces of the prismatic zone. Therefore the crystals acquire a typical needle-like habit.

Near the state of equilibrium, the rate of separation of epsomite decreases, and finally a third period of the slowest crystallization begins, with a lowering of the temperature and a decrease in the volume of the system. In this period the rate of growth of the faces of the crystal heads appreciably

but it decreases and probably does not exceed the rate of growth of the faces of the prismatic zone. Therefore the crystals thicken noticeably. The nuclei that appear again in the third period of crystallization grow into individual crystals and usually do not lead to the formation of radial-fibrous aggregates.

Fig. 4. Diagram of changes in mirabilite crystals at different supersaturations: 1) 3–5 g, 2) 5–10 g, 3) 20–25 g, 4) 30–35 g, 5) 40–55 g per 100 cm³ of solution; a — [100], e — [010], c — [001]; n — [111], r — [101], μ — [011].

From solutions supersaturated by 5–15 g of salt per 100 cm³, crystallization proceeds very slowly, while the volume of the system constantly decreases. Epsomite crystals separated from such solutions have a more isometric form and are considerably larger than crystals formed at a supersaturation of 55–20 g per 100 cm³. At the very beginning the length of the crystals relates to the transverse dimension as 2 : 1 or 3 : 1. Then needle-like forms appear, arising from new nuclei. The greatest number of them appears in the second period of crystallization, i.e., after the temperature change.

In the third period, crystallization of epsomite from solutions supersaturated by 5–15 g of salt per 100 cm³ proceeds as the temperature decreases; the volume of the system gradually decreases.

The study of crystallization from solutions of different supersaturation makes it possible to note the following.

1. The rate of separation of epsomite and mirabilite increases with increasing supersaturation of the solution and during the transition from the first period to the second period of crystallization.

2. The volume of the system during crystallization of mirabilite increases, whereas during crystallization of epsomite it decreases.

3. The shape of epsomite and mirabilite crystals proves to be the more isometric, the lower the supersaturation of the solution. The linear growth rate of faces having a lower relative reticular density exceeds the growth rate of less dense faces to a greater extent, the greater the initial supersaturation of the solution (Figs. 4, 5).

4. In the first and third periods of crystallization, epsomite and mirabilite crystals have a more isometric form than in the second period of crystallization. The linear growth rate of faces having a lower relative reticular density, in the second period of crystallization

tion greatly exceeds the growth rate of the denser faces than in the first and third periods of crystallization.

Fig. 5. Diagram of the change in epsomite crystals at different supersaturations (supersaturation: I—20–30 g, II—15–16 g, III—8–10 g, IV—1–2 g per 100 cm³); a—[110], b—[111], c—[100], d—[010], e—[\(\bar{1}11\)].

Analysis of the Curves

The experimental data indicate different changes in the volume and temperature of the system during the separation of epsomite and mirabilite crystals at different supersaturations. Plotting temperature on the ordinate axis, and on the abscissa axis the readings on the tube scale characterizing changes in the volume of the system, yielded the curves shown in Figs. 2 and 3. Let us analyze these curves.

The curves of the change in volume and temperature \((V, T^\circ)\) of the system during crystallization of mirabilite (Fig. 2) form, with the abscissa axis, a figure resembling an isosceles triangle. This indicates that the amount of crystalline substance separated in the first and second periods of crystallization is approximately equal to the amount of substance separated in the third period.

Attention is drawn to the fact that the segments of the curves corresponding to the first period of crystallization are almost parallel to the ordinate axis, while the segments of the curves corresponding to the second period of crystallization are directed toward it at approximately an angle of 45°. Another important fact is the sharp change in the slope of the curves at the boundary between the segments corresponding to the first and second periods of crystallization.

The difference in the direction of the curves can be explained on the basis of the following assumption. In the first period of crystallization, when the temperature of the system is still low, the process of dehydration of the compounds in the solution was extremely weakly expressed. Since the water molecules split off in this process are few in number, the change in the volume of the system proves to be only slightly noticeable. The growth of crystals in this period is accomplished chiefly at the expense of those salt particles in which the number of water mole-

very close to the number of molecules of crystallization water¹. Such particles are, evidently, most numerous in a strongly supersaturated solution. It is precisely for this reason that the segments of the curves indicating small changes in volume become longer the higher the initial supersaturation of the solution. Conversely, the length of the curve segments corresponding to the second period of crystallization and directed at an angle of 45° to the coordinate axes is almost the same².

The boundary between the first and second periods of crystallization is of quite special interest. As already noted, in the transition from the first period to the second, crystallization is accompanied by a decrease in the rate of temperature rise (curve 4 in Fig. 2) or by its drop (curves 1, 2, 8, 13, and 24 in Fig. 2). At the same time the volume of the system increases intensively, although in the case when the temperature decreases one would, as is known, expect a decrease in the volume of the system. This fact, in the author’s opinion, makes it possible to see most clearly the dependence of the separation of crystalline salt on the dehydration of the solution compounds. In fact, the drop in temperature shows that, in addition to crystallization, an endothermic process is occurring in the system which leads to an increase in its volume. In our case this process can, undoubtedly, only be the process of dehydration. It is precisely for the splitting of the solution compounds into salt and water that an expenditure of energy is necessary, while the destruction of hydrates, as denser groupings, must have a positive volume effect. This is also indicated by the fact that, at high initial supersaturation of the solutions, from which during the first period a larger amount of solid phase and a larger amount of heat are liberated, and the particles in strongly supersaturated solutions are hydrated to a lesser degree, dehydration does not lead to a temperature drop, and in the curves \(V, T^\circ\) only a certain slowing of its rise is noticeable (Fig. 2, curves 3 and 4).

The intensification of the dehydration of the solution compounds in the second period of crystallization creates a larger number of salt particles ready to attach themselves to the crystals. This, in turn, causes an increase in the crystallization rate. It is precisely for this reason that in the second period of crystallization the linear rate of growth of the faces and the rate of formation of nuclei increase considerably, although the supersaturation of the solution by this time has already decreased somewhat. The heat evolved is then already sufficient both to raise the temperature and to maintain partial dehydration of the subsequent portions of salt particles. The foregoing reveals the reason for the considerable increase in the increment of the system volume in the second period of crystallization and the reason for the change in the angle of inclination on the curves.

In the third period of crystallization the separation of crystalline salt occurs with a lowering of the temperature of the system; the curves have a smooth character.

The curves of change in the volume and temperature of the system during crystallization of epsomite from solutions supersaturated by 52–24 g per 100 cm³, with the abscissa axis, if certain details are omitted, form a figure resembling an obtuse triangle (Fig. 3). This is explained by the fact that part of the crystallization of epsomite proceeds with a decrease, and part with an increase, in the volume of the system. A decrease in the volume of the system, as already noted, is observed at the very beginning of crystallization, which is depicted—

is expressed by segments of the curves almost parallel to the abscissa axis. Likewise, with decreasing volume, crystallization proceeds in the third period. In the second period of crystallization the volume of the system increases, while in the final stage of the first period of crystallization it remains almost unchanged. All this accounts for the distinctive form of the \(V, T^\circ\) curves for the crystallization of epsomite.

The form of the curves described also depends on the supersaturation of the solution. As can be seen from the graph (Fig. 3), at the highest supersaturation the segment \(B_1 C_1\) of curve 1 forms a less obtuse angle with the abscissa axis than do the segments \(B_2 C_2\) and \(B_3 C_3\) of curves 2 and 3. This indicates an increase in the increment of the system volume in the second period of epsomite crystallization when the supersaturation of the solution is reduced from 53 to 24 g per 100 cm\(^3\) (Table 4).

Table 4

Taking the foregoing into account, it may be considered that the increase in the volume of the solution in the second period of epsomite crystallization is explained by: 1) intensification of the dehydration process and 2) an increase in the temperature of the system. If the second may be regarded as unquestionable, then in favor of the first is the fact that at the highest supersaturation (when each salt particle has a smaller number of water molecules) the volume of the system increased less (by 113.0 mm) than at lower supersaturation (by 219.8 and 163.2 mm), although the rise in the temperature of the system in the first case is greater (6.5°) than in the second (4°) and third (3.7°) cases. These considerations are also confirmed by the temperature drop accompanied by an increase in the volume of the system (curve 3, point \(B_3\), Fig. 3).

The segments of the curves reflecting the change in the volume and temperature of the system in the third period of crystallization are almost parallel to one another. This indicates a great similarity in the course of separation of the crystalline salt in the third period of crystallization at different initial supersaturations of the solutions.

Thus, the experiments showed that, during crystallization from aqueous solutions, dehydration plays an extremely important role. From these experiments we clearly see that the growth of salt crystals occurs through the decomposition of compounds in the solution. Obviously, crystal growth depends on the stability of hydrates in the solution.

Stability of Solutions and Crystal Form

Hydrates of supersaturated solutions separate the crystalline salt and are the more unstable the higher the supersaturation. In strongly supersaturated solutions there is an acute deficiency of solvent, and each salt particle is associated with a minimal number of water molecules; therefore the most unstable hydrates must predominate in them. In weakly supersaturated solutions, each salt particle is associated with a somewhat larger number of water molecules; in them there must predominate more

less stable hydrates. It is precisely for this reason that crystallization at high supersaturation proceeds at a higher rate than crystallization at low supersaturation.

According to the experimental data, the rate of crystal separation increases in the second period of crystallization. As was already stated, this period is distinguished by the most intense dehydration, which leads to an increase in the number of hydrates less stable with respect to crystallization. Consequently, the increase in the crystallization rate in the second period is also explained by the decrease in the stability of the solution hydrates.

A. V. Shubnikov, using alum as an example, was the first to show that the habit of their crystals becomes the more isometric, the lower the supersaturation of the solution from which they separate (1913).

This regularity, as the experiments have shown, is valid for crystals of epsomite and mirabilite. At the same time, the experiments also established that the crystals of the above-mentioned salts in the first period of crystallization are more isometric than their crystals in the second period of crystallization, although the supersaturation in the second period is, of course, less than the initial supersaturation of the solution.

This fact is very interesting in that it compels us to penetrate more deeply into the essence of the influence of supersaturation on crystal habit. The decrease in the degree of isometry of the habit of epsomite and mirabilite crystals in the second period of crystallization can also be explained on the basis of the decrease in the degree of hydration of the solution particles at this time. It is known that the degree of hydration of the compounds of a solution also decreases with an increase in its supersaturation, since in this case the number of H₂O molecules falling to each salt particle decreases. Therefore, it may be assumed that supersaturation affects the form of crystals by changing the degree of hydration and the stability of the compounds of the solution.

Let us now see whether the change in the habit of salt crystals with a change in supersaturation can be explained on the basis of the assumption made above. For this purpose we must agree to regard all other crystallization conditions, except supersaturation, as constant.

A change in the habit of crystals is a consequence of a change in the selective assimilation of salt particles from the solution by the faces of different simple forms. The selectivity with which a crystal assimilates salt particles from a solution is explained by the fact that particles on faces of lower reticular density are less firmly connected with one another and interact more vigorously with the hydrates of the solution than particles on denser faces. Therefore, the less dense faces have a greater linear growth rate, and the crystal, as Bravais’ law asserts, becomes covered with faces of the greatest reticular density (Anseles, 1952).

In the author’s opinion, there is also a second reason for the selective assimilation by a crystal of salt particles from a solution. Its essence is determined by the conditions of crystallization. It must be taken into account that solutions with different supersaturation consist of hydrates differing in composition and stability¹). Such hydrates interact unequally with one another and with the surface particles of the crystal. Probably, during crystallization the strength of the bonding of salt particles on the faces of one or another simple form is “compared” with the stability of their bonding in the hydrates of the solution. This circumstance should determine the ability of particles to attach themselves to the corresponding faces of the crystal or to remain in solution. It is clear that only those salt particles can be extracted from the compounds of the solution, attached, and retained by the crystal which

¹) If some salt particles are combined with n, and others with m, water molecules, then, under the condition m > n, the former are dehydrated more easily than the latter.

are less stable in hydrated form, and these particles can be deposited only on those faces on which they interact more energetically with the particles of the faces than with the hydrates of the solution. Those particles that are more firmly bound into hydrates (strongly hydrated) can transfer only part of their water to the particles of the crystal, facilitating their hydration. Therefore, under these conditions, along with the structure of the faces of the various simple forms, the composition and stability of the dissociating hydrates also determine the selectivity with which the crystal assimilates salt particles from the solution. This explains the essence of the fairly frequent deviations from Bravais’ law observed in real crystals¹.

Meanwhile, a crystal of any salt is capable of growing along the faces of all its simple forms, since a solution of this salt always consists of hydrates that differ in composition and stability. And, depending on which hydrates predominate in the solution in terms of stability, the growth of crystals in some directions predominates over growth in others. Therefore, when the supersaturation changes, i.e., when the composition and stability of the predominant hydrates of the solution change, the habit of the crystals changes accordingly. When less stable hydrates predominate in the solution (high supersaturation), salt particles from them should be extracted chiefly by faces with lower reticular density, provided that the strength of bonding of particles on these faces exceeds the stability of the majority of the hydrates in the solution. Under the conditions indicated above, these faces will have the greatest linear growth rate, since their particles interact more energetically with the hydrates of the solution. Conversely, when more stable hydrates predominate in the solution (weaker supersaturation), salt particles from them can be extracted and retained chiefly by faces with a small reticular density, since only on these faces does the strength of bonding of the particles considerably exceed the stability of the hydrates of the solution. Faces with lower reticular density in this case cannot have a high linear growth rate. Many of the particles attached to them must inevitably return to the solution, because they are held more weakly by their neighbors than particles on denser faces.

On the basis of specific experimental data, we came to the conclusion that at high initial supersaturation, i.e., at a low degree of hydration and low stability of the solution hydrates, the growth rate of the faces with the lowest reticular density predominates over the growth rate of faces of higher density. In exactly the same way, in the second period of crystallization, when the dehydration process intensifies and the stability of the hydrates decreases, the growth rate of the faces of the lowest reticular density increases considerably more than the growth rate of denser faces. With a decrease in the initial supersaturation, i.e., with an increase in the stability of the hydrates, crystals of epsomite and mirabilite become more isometric, since the linear growth rate of the faces of lower density then decreases the more, the lower the supersaturation. All this is in full agreement with the considerations set forth above and is confirmed by examples from the literature.

Thus, for example, alum crystals at high supersaturation, according to A. V. Shubnikov, have an octahedral faceting, while with a decrease in supersaturation cube faces \([100]\) and rhombododecahedron faces \([110]\) appear on them; these, as is known, possess a lower reticular density than the faces \([111]\). It follows from this that, with an increase in the degree of hydration of the salt in the solution and an increase in the stability of the hydrates

the growth of the less dense faces \([100]\) and \([101]\) is slowed, and the growth of the \([111]\) faces is relatively intensified.

From strongly supersaturated solutions, NaCl forms skeletal shapes, in which funnel-shaped depressions are observed in place of the cube faces. It follows from this that the densest faces of the cube, in relation to the others (possible faces), grow more slowly under such conditions. From weakly supersaturated solutions, normal NaCl cubes are formed; the growth of the cube faces is also relatively intensified in this case.

It should also be noted that the degree of hydration and crystallization can be influenced not only by changing the supersaturation of the solution, but also by other means. Thus, the author studied the influence of dehydrating substances on the crystallization and habit of epsomite and mirabilite crystals.

Experiments established that co-dissolved substances change the degree of hydration, composition, and stability of the compounds in the solution and thereby influence crystallization and the habit of the crystals.

As is known, the temperature of the system influences the degree of hydration. Therefore, the influence of temperature on the habit of crystals and crystallization can be explained in part from the standpoint of a change in the degree of hydration of the salt particles in the solution.

At the same time, the author does not consider a change in the degree of hydration and in the stability of the compounds in the solution to be the sole cause of variability in the habit and size of crystals. In his opinion, phenomena of mass transfer are of no less importance for crystal growth (1953). If dehydration is a process of chemical preparation of a substance when its particles are incorporated into crystals, then mass transfer supplies their faces with nutrient mineral.

LITERATURE

1. Anshles O. M. — Principles of Crystallography. L., 1952.
2. Ormont B. F. — Structures of Inorganic Substances. State Publishing House of Technical-Theoretical Literature. M.—L., 1950.
3. Stroitelev S. A. — Proceedings of the Second Scientific Conference of Tomsk State University, 1953.
4. Shkharonov M. I. — Introduction to the Molecular Theory of Solutions. M., 1956.
5. Shubnikov A. V. — The influence of the degree of supersaturation of a solution on the external appearance of alum crystals precipitating from it. Izv. AN, 817, 1918.

1. The deviation from Bravais’ law consists in the fact that, in real crystals, the faces with the greatest reticular density are not always the most developed. ↩↩

2. This remark is also valid for curve 3, if one has in mind the sum of its segments directed at an angle of 45° to the coordinate axes. ↩