SURFACE-ACTIVE COMPONENTS IN STEEL

N. D. TYUTEVA

(Presented by Professor Doctor A. N. Dobrovidov)

There are many works [1, 2, 3] in which, without sufficient grounds, certain components in steel are incorrectly considered to be surface-active. As is known, surface-active substances are those that lower the surface tension at the boundary between two phases.

Experimental data [4, 5, 6] show that the surface tension of liquid alloys at the liquid–vapor boundary is reduced by the introduction of small additions.

The surface tension of austenitic steel with boron is lower than the surface tension of steel without boron [7]. The surface tension of chromium–nickel melts increases under the influence of boron [8]. On this basis S. M. Vinarov considers that boron, with respect to steel, is a surface-active element; that, by lowering in some grades of steel the surface tension at the liquid–vapor boundary, it must necessarily lower the surface tension at the crystal–liquid phase boundary of steels of any chemical composition.

S. M. Vinarov writes: “The influence of small amounts of boron, as a surface-active element, on the size of the austenite grain and on the kinetics of austenite grain growth should be regarded both as the formation of adsorption layers on austenite grains and, possibly, as the formation of the same layers on the centers of austenite crystallization. The latter, in view of the very small size of the crystallization centers, is practically close to the formation of new crystallization centers in place of the former ones.

In both cases the surface energy is reduced, which in the first case causes a decrease in the rate of growth of austenite grains, while in the second an increase of it may occur” [1].

However, it is entirely incorrect to identify a decrease in surface tension at the liquid–vapor boundary with a decrease in surface tension at the solid phase–liquid phase boundary.

No one has determined the change in the surface tension in steel under the influence of boron at the solid phase–liquid phase boundary; consequently, there are no grounds for considering boron a surface-active element.

On the possibility of formation of a monatomic layer of boron as a result of adsorption along austenite grain boundaries

If a crystal with impurities adsorbed on its faces grows, these impurities will not necessarily be pushed aside by the faces. A new

a layer of molecules of the crystallizing substance cannot be deposited beneath an adsorbed layer of impurity molecules. The adsorbed impurity molecules, becoming covered by molecules of the crystallizing substance, must find themselves included inside the crystal. Such is the situation if, in general, one assumes that chemical adsorption of boron on austenite crystals is possible during the solidification of steel.

In assuming physical adsorption, Vinarov excludes the chemical interaction of boron with the elements of steel. It is known from the literature that boron is capable of forming substitutional solid solutions with iron, according to some literature data [9], and interstitial solid solutions, according to other literature sources [10].

This circumstance alone makes it unlikely that boron is pushed aside, as an impurity, by the growing faces of the crystal, instead of forming a solid solution. The formation of a solid solution of boron in iron is especially probable during the crystallization of structural carbon steels, with which Vinarov worked.

We quote Vinarov’s work [1]: “The calculations we have made show that precisely those amounts of boron which are necessary for the formation of a monatomic layer on the surface of grains prove to be the most effective. Thus, according to calculation, for a grain size No. 3, 0.002% boron is required; for grain No. 10, 0.004% boron is required to form a monatomic layer.

Practically, somewhat larger amounts are required (by 30–50%), since the real grain has tortuous boundaries and, consequently, a larger surface. Grain growth in steel will be slowed in the presence of a monatomic adsorption layer of boron.” Calculations of this kind should be made very cautiously, since they will be calculations with an unknown degree of approximation.

Indeed, “the grain has tortuous boundaries.” Where is the measure of this tortuosity? How can one accurately calculate, determine the surface of a grain? Is it worth using calculations of this kind at all?

Vinarov assumes that boron in steel is present in the form of an atom with an atomic radius of 0.97 Å. We believe that boron in an alloy, in liquid steel, is characterized by high valence.

We have shown experimentally that boron in steel is in an ionic state, in the form of a multiply ionized positive ion. The ionic radius of the triply ionized positive boron ion is 0.2 Å. Chemists concerned with questions of catalysis believe that metals of the transition group, or metals close to them in the periodic system, are the best catalysts. Catalytic activity depends on the ability of the electron orbitals of the catalyst to bind the molecules of the substances participating in the process. Crystals of transition-group metals possess an unfilled $d$ band. The electronic factor in this case is apparently connected with this characteristic [11].

According to Pauling [12], there are three types of $d$ orbitals, namely:

1. Bonding $d$ orbitals, participating in $(d, s, p)$ hybrid bonds.

2. Atomic $d$ orbitals, belonging to individual atoms but not taking part in bond formation.

3. Metallic $d$ orbitals, participating in phenomena of electrical conductivity. In passing from one transition-group metal to another, the electrons first fill the bonding $d$ orbitals, but since after Group VII these orbitals prove to be filled, the electrons begin to fill the atomic $d$ orbitals.

By the end of the period these orbitals also prove to be filled.

A distinguishing feature of transition-group metals, according to the later theory [12], is the presence in them of free atomic \(d\)-orbitals; in the older theory this corresponded to the concept of an unfilled \(d\)-band.

Boron, introduced into steel in the precrystallization period and during the process of crystallization, reacts by means of its electrons with the \(d\)-electron orbitals of the elements of the transition group in the alloy. A profound chemical interaction takes place. Under the influence of boron, the character of individual structural constituents changes; the chemical composition and structural state of the phases change. New carbide phases appear, sometimes not amenable to interpretation on X-ray diffraction patterns. The mechanism by which boron affects the phase state of a multiphase, multicomponent alloy (high-speed steel) may be presented in greater detail as follows.

The properties of crystals are to a certain extent laid down in the liquid phase. Therefore the kinetics of crystallization must be studied in close connection with the phenomena occurring in the liquid and with its structure in the precrystallization temperature range. If boron affects crystallization, then one may assume that this effect is already manifested in the liquid melt.

On the basis of the X-ray diffraction pattern it may be assumed that ordered groupings of scattering centers (atoms, molecules, ions) exist in the liquid, encompassing regions on the order of tens of angstroms. Such groupings may be called quasicrystals.

At any moment in time all particles of the liquid may be divided into two groups.

The first group includes particles that perform oscillatory motions about centers of oscillation coinciding with the statistically disordered structure of the crystal lattice (an arrangement characterized by a single distribution function—the structure of a quasicrystal); the second group includes particles moving according to the laws of random walks and not associated with any equilibrium positions. Therefore the picture of thermal motion in a liquid simultaneously contains features of the thermal motion of a solid body (oscillations about equilibrium positions) and of a gas (random walks).

Any particle may pass from the first group into the second and back, so that the entire picture of the structure of a liquid represents the results of a dynamic equilibrium between particles of the first and second groups. The modifier (boron) introduced into the liquid alloy affects the character of this dynamic equilibrium.

Boron transfers particles from the second group into the first, i.e., into the group of particles that perform oscillations about equilibrium positions; it strengthens interparticle bonds through the participation of its free electrons. In the presence of boron the degree of long-range order increases. What is meant is not long-range order in general, but the formation of quasicrystals with atoms grouped in a definite way.

The change in interparticle bonds in the liquid state (the weakening of one type of bond and the strengthening of another) is effected by means of two \(S\)- and one \(P\)-electron, which boron is capable of donating, since boron, to a first approximation, is trivalent in the alloy. We say “to a first approximation,” because boron in an alloy, introduced as a modifier, may donate three and four electrons, whereas as an alloying addition it may be singly ionized or

does not pass into the ionic state at all. The valence of boron in alloys, unfortunately, is unknown.

Electron distribution in the boron atom

As a result of changes in interparticle bonds under the influence of boron, a new model of liquid iron—chromium—carbon—vanadium—tungsten melts is obtained (grouping into groups richer in iron).

Boron affects the thermodynamic activity of iron; each atom of chromium, tungsten, vanadium, and carbon in the melt has, among its nearest neighbors, a larger number of iron atoms than does the remaining part of the liquid.

On the basis of our experimental data we have come to the conclusion that boron is present in the solid solution in a metallic state; its valence electrons belong to the electron $d$-band of the metallic crystal (to the electronic $d$-orbitals of the metallic crystal).

The hypothesis of the metallic state of boron atoms dissolved in metals of the transition groups makes it possible to explain the extremely high mobility of boron atoms. Recently, works [13] have appeared in which interaction is likewise proposed between the $d$-level of iron atoms and the valence electrons of boron during diffusion of boron into iron and steel. Interatomic bonds in a crystal are one of the forms of chemical bonding. Therefore, in the interatomic bonds of crystals of transition-group metals, not only valence $S$-electrons take part, but also the $d$-electrons of the unfilled shell, which leads to the occurrence of very large cohesive forces in these crystals (the heat of vaporization of tungsten is 200 kilocalories per gram-atom, whereas the heat of vaporization of carbon in the diamond crystal is only 150 kilocalories). By providing new crystallization centers, boron apparently enters wholly into the composition of the forming leading phase of crystallization (carbides).

The solubility of boron in pure iron is small. Nicholson’s measurements [14] gave:

alpha-iron: 0.002% at 910°C;

gamma-iron: 0.001% at 910°C, 0.02% at 1165°C.

M. K. Breid, Spretnak, and Speiser [15] determined, in weight percent, solubility in alpha-iron as 0.0004% at 710°C, 0.0082% at 906°C; in gamma-iron, 0.0021% at 906°C, 0.021% at 1149°C.

In such a highly alloyed steel as high-speed steel, the $\alpha$ and $\gamma$ phases also represent solid solutions of alloying elements in iron. As a result, the solubility of boron in the $\alpha$ and $\gamma$ phases of highly alloyed steel must be still smaller. In practice, boron may be regarded as insoluble in the $\alpha$ and $\gamma$ solid solutions of high-speed steel.

As is known, the distribution of electrons between the $S$ and $d$ bands, when they overlap, depends to a greater degree on the influence of external actions (temperature, pressure, the influence of foreign atoms included in the crystal lattice).

A change in the electron distribution in a solid may cause a change in the crystal lattice [16]. Boron gives a solid

an interstitial solid solution, but since the boron ion is small, 0.2 Å (three times ionized), the carbide lattice in which the boron is dissolved may distort toward a decrease in the parameter. “With a decrease in the interatomic distance, the \(S-d\) energy levels are inevitably blurred” [17].

The total kinetic energy of the electrons in the metal (it is often called the “Fermi energy”) increases with increasing compactness of the lattice, in particular upon compression of the lattice, and correspondingly decreases upon stretching. We cannot rule out such a character of the effect of boron on the carbide phase.

The results of calculating the ratio between the boron atoms entering into carbides and the atoms constituting the carbide phase are as follows: if one proceeds from the assumption that all the boron is in the carbides, and the carbides in the alloy average 15–20%, then one boron atom in the carbide corresponds approximately to 200–250 iron atoms. The participation of boron electrons in the \(S-d\) electron redistribution is also confirmed by the sharp increase in their mobility at high temperatures. This is indicated by the change in the electrical conductivity of boron with temperature.

“Electrical conductivity is negligible at ordinary temperature, but when boron is heated to \(800^\circ\) it increases two million times” [18].

Small impurities affect the structure and, consequently, the mechanical properties of individual phases and of the entire alloy. The effect on mechanical properties is connected with a change in the chemical composition and crystal structure of the phases (indirect action).

The change in the phase state of an alloy is determined by the change in free energy. The change in the density of free energy is connected with a change in the kinetic energy of the electrons, with a change in the behavior of the free electrons. Boron makes the system more nonequilibrium. Small impurities of boron increase the free-energy density of the system.

The increase in the free-energy density is compensated by an increase in the entropy factor. Entropy increases because, under the influence of boron, the actual distribution of iron in the volume of the entire alloy as a whole—in the carbides and in the solid solution—becomes more uniform.

If in the absence of boron the concentration of iron is minimal in the carbides and maximal in the solid solution, then in the presence of boron iron is distributed more uniformly throughout the volume.

Similarly, carbon and alloying elements, under the influence of boron, are also distributed more uniformly throughout the entire volume of the alloy.

Apparently, the essence of the influence of small impurities is that a given element—boron, titanium in steel, silicon, magnesium in cast iron—introduced into an alloy in small quantities behaves differently than the same element introduced in large quantities.

When there is little impurity of an element, its atoms are capable of giving up a greater number of electrons than when there is much of it; that is, the character of interparticle interactions during alloying is entirely different.

The character of interparticle interactions is determined by a change in valence. The valence of boron in the alloy during alloying and during modification is different. It depends not only on concentration, but also on temperature.

LITERATURE

1. Vinarov S. M. Izvestiya Akademii nauk SSSR. OTN, No. 6, 1948.
2. Vinarov S. M. Zhurnal tekhnicheskoi fiziki, Vol. XIX, issue 2, 1957.
3. Vinarov S. M. DAN SSSR, 55, 9, 1947.

1. Pokrovskii M. L. and Galanina N. D. Journal of Physical Chemistry, 23, 3, 324, 1949.

2. Pugachevich P. P. and Altynov I. P. Reports of the Academy of Sciences of the USSR, 86, 1, 117, 1952.

3. Zaletaeva R. P., Kreshchinovskii N. S., Kunin L. L. Foundry Production, 2, 1951.

4. Klyachko Yu. A., Kunin L. L., Kreshchinovskii N. S., and Ginzburg E. S. Reports of the Academy of Sciences of the USSR, 72, 5, 927, 1950.

5. Bobkova O. S., Samarin A. M. Proceedings of the Academy of Sciences of the USSR, 1956.

6. Umanskii Ya. S., Finkelshtein B. N., Blanter M. E. Physical Principles of Metallography, Metallurgizdat, 1955.

7. Hansen M. Structure of Binary Alloys, vol. I, 1941.

8. Trapnell B. M. W. Quart. Rev., 8, No. 4, 404—421, 1954.

9. Pauling, Phys. Rev., 54, 899, 1938; Proc. Roy. Soc., 196 A, 343, 1949.

10. Samsonov G. V., Tseytina N. Ya. Physics of Metals and Metallography, issue 2, 303, 1955.

11. M. E. Nicholson, Trans. A. I. M. E., 200, 185, 1954.

12. C. C. Mc. Bride, I. W. Spretnak and R. Speiser, Trans. A. S. M., 46, 499, 1954.

13. Umanskii Ya. S. Carbides of Hard Alloys, 97—99, 1947.

14. Mott—Jones, The Properties Metals and Alloys, Oxford, 1936.

15. Parfenov V. A. Rare Metals, Metallurgizdat, 1954.