330 Prakt. Met. Sonderband 30 (1999) 
The nucleated o ZnAl solid solution of zinc in aluminium phase crystallises in cubic system, with the 
lattice parameter of 0.4045 nm - Table 1. In its crystallographic close packed plane {111} the distan- 
ce between the atoms dui, = 0.286 nm. ALTi phase, introduced with the AlTiS master alloy, shows a 
weighted average spacing in {1 1 2} plane equal to 0.275 nm [6]. Thus the interatomic distance 
difference in this case is: -3.8%. In turn, Ti(AlL Zn); phase, raising both from ALTi and TiZns, cry- 
stallises in cubic system with the mean lattice parameter a = 0.396 nm, with the interatomic distance 
in {111} plane dm» = 0.280 nm - Table 2 and 3. It is well known by practice, that the nucleating pha- 
se is the more effective, the smaller is the difference in the least interatomic distance in relation to the 
phase being nucleated and when it crystallises in the same system as the phase being nucleated. The 
aforementioned crystallographic data indicate that the Ti(Al,Zn); phase shows the difference in dmin 
of -2.1% in relation to the o, and moreover it has the same type of the crystal lattice as o'. This 
justifies high efficiency of the ternary Ti(ALZn); phase. Additionally, the employed ZnTi4 master 
alloy, which requires lower temperature for a quick dissolution in the grain refined Zn-25wt%Al 
melt, could be an effective grain-refiner both for the high-aluminium and low-aluminium zinc alloys. 
3.3. Rietveld analysis 
Results of the investigations of the crystal structure of the Ti(Al,Zn)s trialuminide evolving during 
grain refinement of the Zn-25wt%Al alloy by the AITi5 or ZnTi4 grain refiner master alloys, as well 
as site occupation Occ in the unit cell by Ti, Al and Zn atoms are collected in Tables 4 and 5. 
= A B - De 
xy |e Occ | vy | =z Occ 
Ti 0101] 0037(+/-0005) ; | Ti 0] 0 | 0 | 0.031 (+-0.002) 
Zn [0] 0] Zn [0] 0 | 0 [| -0.010(+/- 0.002) | 
"Al | 0]05]05]0.063(+-0.000)! [Al |[0[| 0.5 | 0.5 | 0.042 (+/- 0.000) | 
| 1 [| ~. 1 Zn [0] 0.5] 0.5 | 0.021 (+/- 0.000) | 
Table 4: Results of the Rietveld analysis of the diffraction pattern of the powder sample extracted 
from the Zn-25wt.%A1-0.05wt.%Ti alloy. Zn site location: A: exclusively in Ti (0,0,0) position; B: in 
position of Ti (0,0,0) and/or (0, 0.5, 0.5). Fitting parameters: A: Rg = 2.04; Rr = 4.15; B: Rg = 1.95; 
Rp = 3.61. Unit cell Pm3m parameter :A and B: a=b=c=0.39531 (+/- 0.00004) nm 
A Co _ _ BB 
x | y | 2 Occ x[ yl z | Oc 
Ti] 0 | 0 | 0 0021-0002) [Ti[O0] 0 | 0 [0.021(-0.000 
Zn | 0 | 05 |05]0020(+:0.000) | [AL] 0] 05 [ 0.5 [0.038 (+/- 0.000) 
"AI 0 [ 05 [05]0042(x0.000) | [Tit] 0 | 0.5 | 0.5 [0.005 (+/- 0.006) 
“Al 0 | 0 [00000-0002] [Zn [| 0 | 0.5 | 0.5 [0.020 (+/- 0.006) 
Table 5: Results of the Rietveld analysis of the diffraction pattern of the powder sample extracted 
from the Zn-25wt.%Al-0.05wt.%Ti alloy. A: Al site location: in Ti (0,0,0) or Al (0, 0.5, 0.5) posi- 
tion; B: Ti site location in position of Ti (0,0,0) and/or Al (0, 0.5, 0.5). Fitting parameters: A: Rg = 
2.13; Rr = 3.11; B: Rg = 3.42; Rp = 3.29. Unit cell Pm3m parametetr:A: a=b=c=0.39535 (+/- 
0.00005) nm: B: a=b=c= 0.39599 (+/- 0.00002) nm 
As it appears from the results collected in Tables 4-5, Ti atoms occupy exclusively (0 0 0) position, 
while Zn and Al atoms occupy (0 0.5 0.5) position. Taking into account the possibility of the Al and 
Zn location on Ti position (0 0 0) and/or (0 0.5 0.5), or Ti location on Al position (0 0.5 0.5) it can 
be observed that the fit-goodness, measured by the Rg (R value for Bragg intensities) and Rg (R va- 
lue for structure F parameter) reliability factors values, decreases and that the number of the site