Diffusion correlation factors

Units employed in diffusivity correlations commonly followed the cgs system. Similarly, correlations for mass transfer correlations used the cgs or Enghsh system. In both cases, only the most recent correlations employ SI units. Since most correlations involve other properties and physical parameters, often with mixed units, they are repeated here as originally stated. Common conversion factors are listed in Table 1-4. 

Diffusion has often been measured in metals by the use of radioactive tracers. The resulting parameter, DT, is related to the self-diffusion coefficient by a correlation factor/that is dependent upon the details of the crystal structure and jump geometry. The relation between DT and the self-diffusion coefficient Dsclf is thus simply... 

There is not sufficient experimental evidence to continue this discussion quantitatively at the present time, but the sparse experimental data suggests that for a given compound, the D0 value is significantly lower than is the case in simple metals. This decrease may be attributed to a low value in the correlation factor which measures the probability that an atom may either move forward or return to its original site in its next diffusive jump. In simple metals this coefficient has a value around 0.8. 

The vacancy will follow a random-walk diffusion route, while the diffusion of the tracer by a vacancy diffusion mechanism will be constrained. When these processes are considered over many jumps, the mean square displacement of the tracer will be less than that of the vacancy, even though both have taken the same number of jumps. Therefore, it is expected that the observed diffusion coefficient of the tracer will be less than that of the vacancy. In these circumstances, the random-walk diffusion equations need to be modified for the tracer. This is done by ascribing a different probability to each of the various jumps that the tracer may make. The result is that the random-walk diffusion expression must be multiplied by a correlation factor, / which takes the diffusion mechanism into account. 

In the case of interstitial diffusion in which we have only a few diffusing interstitial atoms and many available empty interstitial sites, random-walk equations would be accurate, and a correlation factor of 1.0 would be expected. This will be so whether the interstitial is a native atom or a tracer atom. When tracer diffusion by a colinear intersticialcy mechanism is considered, this will not be true and the situation is analogous to that of vacancy diffusion. Consider a tracer atom in an interstitial position (Fig. 5.18a). An initial jump can be in any random direction in the structure. Suppose that the jump shown in Figure 5.18b occurs, leading to the situation in Figure 5.18c. The most likely next jump of the tracer, which must be back to an interstitial site, will be a return jump (Fig. 5.18c/). Once again the diffusion of the interstitial is different from that of a completely random walk, and once again a correlation factor, / is needed to compare the two situations. 

Table 5.1 lists some values of the correlation factor for a variety of diffusion mechanisms in some common crystal structure types. 

The correlation factor,/, is defined by the ratio of the tracer diffusion coefficient to the random-walk diffusion coefficient (Section 5.6) ... 

When ionic conductivity is by way of interstitials, both conductivity and diffusion can occur by random motion, so that the correlation factor and HR are both equal to 1. In general, the correlation factor for a diffusion mechanism will differ from 1, and in such a case D can be described by the following relationship ... 

Atomic jumps in random walk diffusion of closely bound atomic clusters on the W (110) surface cannot be seen. A diatomic cluster always lines up in either one of the two (111) surface channel directions. But even in such cases, theoretical models of the atomic jumps can be proposed and can be compared with experimental results. For diffusion of diatomic clusters on the W (110) surface, a two-jump mechanism has been proposed by Bassett151 and by Cowan.152 Experimental studies are reported by Bassett and by Tsong Casanova.153 Bassett measured the probability of cluster orientation changes as a function of the mean square displacement, and compared the data with those derived with a Monte Carlo simulation based on the two-jump mechanism. The two results agree well only for very small displacements. Tsong Casanova, on the other hand, measured two-dimensional displacement distributions. They also introduced a correlation factor for these two atomic jumps, which resulted in an excellent agreement between their experimental and simulated results. We now discuss briefly this latter study. 

There are two possible directions for both the atoms in both the first and second atomic jumps. If the jumping direction is completely random and the two atoms have the same probability of performing a jump, then these atomic jumps are said to be uncorrelated. A correlation factor, /, has been introduced for the two atomic jumps, which is defined as the extra probability that the atom making the first jump will also make the second jump in the forward direction. The rest of the probability, (1 — /), is then shared equally for either of the two atoms jumping in either of the two directions. Two experimental displacement distributions measured at 299 K and 309 K fit best with a Monte Carlo simulation with / = 0.1 and /=0.36, respectively. The correlation factor increases with diffusion temperature as can be expected. It is interesting to note that when/= 1, only a and steps can occur. 

For very dilute solid solutions of B in A, the basic physics of diffusional mixing is the same as for (A, A ). An encounter between VA and BA is necessary to render the B atoms mobile. But B will alter the jump frequencies of V in its surroundings and therefore numerical values of the correlation factor and cross coefficient are different from those of tracer A diffusion. Since the jump frequency changes also involve solvent A atoms, in addition to fB, the numerical value of fA must be reconsidered (see next section). 

Solutions for this type of kinetics can only be achieved numerically. Model calculations with constant kinetic parameters have been made [H. Wiedersich, et al. (1979)], however, the modeling of realistic transport (diffusion) coefficients which enter into the flux equations is most difficult since the jump rate vA vB. Also, the individual point defects have limited lifetimes which determine the magnitude of correlation factors (see Section 5.2.2). Explicit modeling for dilute or non-dilute alloys can be found in [A.R. Allnatt, A.B. Lidiard (1993)]. 

Since the fraction of empty sites in a zeolite channel determines the correlation factor (Section 5.2.2), as is well known from single-file diffusion in the pores of a membrane, the strong dependence of the diffusion coefficients on concentration can be understood. This is why a simple Nernst-Planck type coupling of the diffusive fluxes (see, for example, [H, Schmalzried (1981)]) is also not adequate. Therefore, we should not expect that sorption and desorption are symmetric processes having identical kinetics. Surveys on zeolite kinetics can be found in [A. Dyer (1988) J. Karger, D.M. Ruthven (1992)]. 

For a random walk, f = 1 because the double sum in Eq. 7.49 is zero and Eq. 7.50 reduces to the form of Eq. 7.47. In principle, f can have a wide range of values corresponding to physical processes relating to specific diffusion mechanisms. This is readily apparent in extreme cases of perfectly correlated one-dimensional diffusion on a lattice via nearest-neighbor jumps. When each jump is identical to its predecessor, Eq. 7.49 shows that the correlation factor f equals NT.6 Another extreme is the case of f = 0, which occurs if each individual jump is exactly opposite the previous jump. However, there are many real diffusion processes that are nearly ideal random walks and have values of f 1, which are described in more detail in Chapter 8. 

Equation 8.19 contains the correlation factor, f, which in this case is not unity since the self-diffusion of tracer atoms by the vacancy mechanism involves correlation. Correlation is present because the jumping sequence of each tracer atom produced by atom-vacancy exchanges is not a random walk. This may be seen by... 

K. Compaan and Y. Haven. Correlation factors for diffusion in solids. Trans. Faraday Soc., 52 786-801, 1956. 

Calculate the correlation factor for tracer self-diffusion by the vacancy mechanism in the two-dimensional close-packed lattice illustrated in Fig. 8.22. The tracer atom at site 7 has just exchanged with the vacancy, which is now at site 6. Following Shewmon [4], let p. be the probability that the tracer will make its next jump to its kth nearest-neighbor (i.e., a 7 — k jump). 6k is the angle between the initial 6 —> 7 jump and the 7 —> k jump. The average of the cosines of the angles between successive tracer jumps is then... 

Powerful methods for the determination of diffusion coefficients relate to the use of tracers, typically radioactive isotopes. A diffusion profile and/or time dependence of the isotope concentration near a gas/solid, liq-uid/solid, or solid/solid interface, can be analyzed using an appropriate solution of - Fick s laws for given boundary conditions [i-iii]. These methods require, however, complex analytic equipment. Also, the calculation of self-diffusion coefficients from the tracer diffusion coefficients makes it necessary to postulate the so-called correlation factors, accounting for nonrandom migration of isotope particles. The correlation factors are known for a limited number of lattices, whilst their calculation requires exact knowledge on the microscopic diffusion mechanisms. 

Here capacitance effects are absent and modifications are due to the different weight of the different charges. If we ignore correlation factors, the general result for the tracer diffusion coefficient is... 

The pore cantrol kinetics are given for the situation where pore separation is related to grain size. Changes in distribution during growth would change the kinetics. fis the correlation factor for diffusion. 

It should be pointed out that the Bardeen-Herring correlation factor has been left out of the present discussion. In rigorous treatment the factor a, as defined by Equation 9, should in Equation 10 be replaced by flw = fl/f, where f is the correlation factor (see References 5, 16, and 17). The continued use of the term a instead of in this paper will be justified when bearing in mind that the symbols D in our equations denote uncorrelated diffusion coefficients, as measurable—e.g., by NMR—and not tracer diffusion coefficients, as measured in most actual studies.)... 

The Haven ratio may deviate from unity when correlation effects and possibly different jump distances and jump frequencies can not be neglected [51]. For a vacancy diffusion mechanism Hr equals the well-known tracer correlation factor /. 

---