A significant fraction of the losses in the additive, in the crystal growth (conventional Verneuil process) of ruby, is traced to powder technology. Valence states of the additive in the powder are studied and correlated with their escaping tendencies. A working model is proposed and supported by kinetic studies on the powder. From this model is derived certain predictions which are significant to the Verneuil crystal growth of solid solutions.

1.
The optical reflection spectra of such powders has been studied recently by
C. P.
Poole
, Jr.
and
J. F.
Itzel
, Jr.
,
J. Chem. Phys.
39
,
3445
(
1963
).
2.
See E. Ryshkewitch, Oxide Ceramics (Academic Press Inc., New York, 1960), p. 62.
3.
Higher thermal treatment increases densification and consequently, a decline in σ.
4.
Such powders are obtained from ammonium alum crystals mixed with either chrome alum or ammonium dichromate (atom ratio, CrAl∼10−3) and calcined in air for about 3 h at ∼1000 °C.
5.
The leaching conditions required for the trivalent species render the differentiation between the (s) and (γ) regions quite arbitrary.
6.
The mixture to be calcined was derived from a large stock of Cr‐doped ammonium alum (J. T. Baker) precalcined at 800 °C for 3 h. No Cr VI is formed under these conditions.
7.
See E. B. Sandell, Colorimetric Metal Analyses, Chemical Analysis (Interscience Publishers, Inc., New York, 1959), 3rd ed., Vol. 3, p. 397. A comparison of the Cr VI content of the leaching before and after the oxidation treatment with KMnO4 (excess eliminated by NaN3), gave assurance of negligible leaching of Cr III. The isotypic relation between SO4−2 and CrO4−2 may be significant in the exchange, i.e., leaching. Solutions of alkali metal sulfates are just as effective.
8.
Relevant to this is the tremendous difference in volatility between CrO3 and Cr2O3.
9.
R‐line fluorescence from the powder was also established to be applicable.
10.
Chemical analysis (cf. Footnote 7) is not very effective in this region since the errors are large. Thus, for the 1135 °C sample, the Cr VI amounts to 0.006% and such determination can be off by 30%.
11.
“Percent transformed to a given phase” equals “percent of the phase” when a scale limit, imposed by the physical method of assay, is satisfied.
12.
H. C.
Stumpf
,
A. S.
Russell
,
J. W.
Newsome
, and
C. M.
Tucker
,
Ind. Eng. Chem.
42
,
1398
(
1950
).
See also,
W.
Buessem
and
W. A.
Weyl
,
Glastechn. Ber.
16
,
57
(
1938
).
13.
The blocking effect of low concentration of Cr is reported in the x‐ray work of
S.
Okada
and
K.
Kuwashima
,
Kogyo Kagaku Zasshi
59
,
1301
(
1956
).
More recent work with various additives is reported by
Y.
Wakao
and
T.
Hibino
,
Nagoya Kogyo Gijutsu Shikensho Hokoku
11
,
588
(
1962
).
14.
This is easily seen. Since Ẏmax = 0, then ÿmax = −βymax. However, β>0 and ymax>0.
15.
The temperature is, of course, converted to the Kelvin scale.
16.
E. Ryshkewitch (cf. Footnote 2), p. 118.
A higher value is reported in a more recent work by
F.
Juillet
,
Rev. Hautes Temp. Refractaires
1
, (
2
),
139
(
1964
).
17.
This corresponds to the recommended value of 1130 °C by Stumpf et al. (cf. Footnote 12).
18.
“Two percent chromium” in the present case means 2 wt. % Cr2O3. The “2 at. % Cr” in Eq. (19) is the equivalent of 3 wt. % Cr2O3. This refinement is not taken into account in the inequality limits of (20). However if “weight percent Cr2O3.” is adopted, then Eq. (19) becomes ΔT2% = 36 °C which is now on the same concentration scale as that involved in (20).
19.
The partition constant K becomes independent of x0 as x0→0.
20.
In the limit where the mole fraction of Cr2O3 approaches unity, there is no detectable Cr VI(s) in the powder under the conditions considered.
21.
If estimated at x0 = 0,K = 0.46.
22.
The schemes as well as the kinetics for this initially heterogeneous case appears to be the same as the present case which is initially homogeneous (cf. precalcined stock of Footnote 6). The kinetics leading to a homogeneous distribution of the additive, being much faster than the kinetics of transformation of the host (γ‐Al2O3), has provided a basis for various preparations of the boule powder (cf. Fig. 1).
23.
This is taken up in a forthcoming publication.
24.
In crucible methods of growth from the congruent melt, the redox nature of the environment is imposed by the crucible material as well as its protective atmosphere, if any.
This content is only available via PDF.
You do not currently have access to this content.