Mat. Res. Bull. Vol. 2, pp. 865-869, 1967. Pergamon P r e s s , Inc.

Printed
in the United States.

TRANSPARENT YTTRIUM OXIDE CERAMICS*

R. A. Lefever and John Matsko

Sandia Laboratory, Albuquerque, New Mexico

(Received in final form June 29, 1967; Communicated by R. A. Huggins)

ABSTRACT
Transparent discs of yttrium oxide, one centimeter in
diameter by several millimeters thick, have been pro-
duced by hot forging under vacuum at 950°C and I0,000-
12,000 psi for 48 hours. Lithium fluoride, required
to produce full transparency, is eliminated during the
sintering process. Optical absorption spectra of the
hot-forged samples are nearly identical to those
reported for single crystals.

Introduction

Yttrium oxide is an excellent host for rare earth ions and

as such has been the subject of recent research involving
electron spin resonance, luminescence and laser action. In
addition, yttrium oxide is a promising window material with a
transmission range extending from 0.3 ~ in the violet region of
the visible spectrum to approximately 8 ~ in the infrared.
Single crystals of yttrium oxide have been grown but, as a
result of the high melting point (2410°C), only in r e l a t i v e l y
small sizes (I). However, since the material is optically iso ~
tropic, a transparent ceramic form consisting of randomly
oriented grains would serve for many studies and applications.
The preparation of transparent discs of yttrium oxide

*This work was supported by the U. S. Atomic Energy Commission.

865
866 YTTRIUM OXIDE Vol. 2, No. 9

has been reported (2). However, experimental parameters and

optical properties were not fully specified. The fabrication of
similar material has been reported by General Electric (3). The
product (Yttralux) is sintered at 2200°C and contains I0 per cent
thorium. This paper describes the preparation of transparent
discs of yttrium oxide by vacuum hot forging at 950°C and I0,000-
12,000 psi.

Experimental

Lindsay Code 1116 yttrium oxide was used for the majority
of this work. The average particle size (determined with an
MSA Particle Size Analyzer) was about 3 microns, with 90 per
cent of the sample in the 1-6 micron range. The Y203 powder
was milled with -325 mesh lithium fluoride (3-5 wt per cent)
for 12 hours in a plastic container, using several alumina
milling cylinders to break up agglomerates and acetone as the
milling vehicle. The acetone was removed by evaporation and the
powder was dried at 60°C in vacuum for 12-24 hours. The result-
ing powder was cold pressed at 3000 psi in a ?/8 inch I.D.
conventional steel die.
The arrangement shown schematically in Fig. i was used

ROD

ACUUM PORT
THERMOCOUPLE
;ISTANCE
IG ELEMENT

MULLITE TUBE PUSH ROD

IESSED ~O=
DIE SUPPORT" PUSH ROD

VACUUM JACk

Fig. I
Vacuum Hot Forging Assembly.
Vol. 2, No. 9 YTTRIUM OXIDE 867

for vacuum press forging. The cold-pressed compact was loaded

into a i inch I.D., thin-walled mullite tube die. The compact
was embedded in a medium grain zirconia setting sand to pre-
vent the pressed sample from adhering to the push rods. After
the die assembly was placed in the furnace, the system was
evacuated to about 0.05 Torr and a pressure of 500 psi was
placed on the compact. Power was then applied to the furnace
and at 5400C pressure on the compact was increased to 1200 psi.
When the furnace temperature traversed the temperature range
of 850-900°C, the pressure was increased to 2000 psi, which
resulted in bursting of the die. ~t this point the forging
action began. The furnace temperature was allowed to increase
to 9500C while maintaining a pressure of 2000 psi. When the
operational temperature of 9500C was reached, the pressure was
increased at the rate of about 500 psi per minute until a pres-
sure of 10,000-12,000 psi was reached. These conditions were
then held for about 48 hours. The furnace was allowed to cool
over a 6 hour period, after which the sample was removed.

Results and Discussion

The procedure described above produced discs of yttrium

oxide approximately one centimeter in diameter by several
millimeters in thickness. After grinding and polishing,
optical transmission spectra were obtained in the visible and
infrared regions of the spectrum using Cary Model 14 and
Beckman IR-4 spectrophotometers, respectively. Transmission in
the visible region is shown in Fig. 2 (Sample 34). The results
compare favorably with previously-published data on single-
crystal Y203 . The infrared transmission, shown in Fig. 3, also
approximates the single crystal data. Samples have also been
doped with europium oxide. Sample 41 of Fig. 2 shows the base
line of the visible spectrum of a Y203 sample containing 0.5
,mole per cent Eu203. The europium absorption lines, which are
not shown, are nearly identical in wavelengths, intensities,
and widths to those previously published for single crystal
material (4).
Surface microstructures were developed by etching polished
samples for several minutes in hot concentrated phosphoric acid.
 868 YTTRIUM OXIDE Yol. 2, No. 9

*'//3*
/,'"/' /

.J ,///SINGLE CRYSTAL

¥
,q • /
m
er
-.c

*OO0 ~.e',).. 4"~'~o so'oh 2000

WA~£LF.N~TH~N ANGSTROMS
FIG. 2
Transmission in the visible. Curves normalized to 1.0 mm
sample thickness. Single crystal data from Ref. 4.

2z ~ .... ~IGL[ CRYSTAL YzO•

Z
ql
,I,- PRESS FORGED Y,Os \
l-
0_. °
Z
~J
e,,
, , , , , ~\'!
,'~'~,
o
3 4 S Q 7 e 9 I0
WAVELENGTH IN MICRONS

FIG. 3
Infrared transmission. Curves normalized to 1.0 mm sample
thickness. Single crystal data from Ref. 5.

The average grain size, determined by the linear intercept

method was about 3 microns, indicating that grain growth had
not occurred. Examination with a Siemens texture diffractometer
failed to reveal any evidence of preferred orientation. Lithium
was not detected in analyses by flame photometry, indicating
that the majority of the lithium was removed during processing.
Vol. 2, No. 9 YTTRIUM OXIDE 869

The density of a transparent Y203 sample, determined at

23.5°C by carbon tetrachloride displacement, was 5.030 g/cc.
This is within experimental error of the theoretical density
of 5.031 g/cc, determined by x-ray methods. Knoop microhardness
values using a lO0-g load ranged from ?20 to 800 kg/mm 2. Values
of ?00-?50 were obtained on unoriented surfaces of single
crystals grown by the flame fusion process.
We are indebted to Drs. G. H. Haertling and C. A. Hall for
the loan of equipment and assistance with the hot-forging exper-
iments. We also wish to thank Dr. E. K. Beauchamp for valuable
discussions and Mr. J. E. Hesse for assistance in determination
of optical spectra.
References
i. R. A. Lefever, Rev. Sci. Instr. 33, ?69 (1962); A. C. Pastor
and R. C. Pastor, Mat. Res. Bull]--2, 555 (1967).
. L. A. Brissette, P. L. Burnett, R. M. Spriggs and T.
Vasilos, J. Am. Ceram. Soc. 49, 165 (1966).
. Chem. & Eng. News, Oct. I?, 1966, p. 38; New Scientist,
Dec. 29, 1966, p. ?2?.
. K. A. Wickersheim and R. A. Lefever, J. Electrochem. Soc.
iii, 4? (1964).
. K. A. Wickersheim and R. A. Lefever, J. Opt. Soc. Am. 51,
1147 (1961).