si
.
..
en
w
.
*
I
fim
.
a
*
.
*
-
1
.
TOFT
ORNL P
27 45
.
..
.
.
it
.
17.
in
.
is,
:
.
.
.
.
EEEFEFFE
.
i
MICROCOPY RESOLUTION TEST CHART
NATIONAL BUREAU OF STANDARDS -1963

13
:
2.
net
T
.
VE
COnt'6641
mam
.
HC.
THERMOCFMICAL DEPOSITION AND EVALUATION OF RHENIUM
AND TUNGSTEN-RHENIUM ALLOYS*
HE
.
TI PUCE
J. I. Fecerer
A. C. Schaffhauser C. F. Leitten, Jr
Metals and Ceramics Division
Oak Ridge National Laboratory
Oak Ridge, Tennessee
RELEASED FOR ANNOUNCEMENT
.
S
IN NUCLEAF. SCIENCE AESTRICIS
10)
ABSTRACT
Cena: 535. boy.WINT1
uki
h
Rhenium and tungsten-rhenium alloys liave been deposited by hydrogen
reduction of WF6 and Relo ai low pressures, Tungsten-rich alloys having
ps: compositional unifornity ir: both the axial and radial direction of tubular
deposits were obtained at 600 to 900°C. An embrittlirig B-tungsten phase in
tungsten-rich alloys was dissolved by annealing at 2000°C. As-deposited
tungstenich alloys have a strong orientation of (100) parallel to the sub-
strate whether single phase solid solution, single phase B-tungsten, or &
mixture of these phases. Smooth, dense rhenium coatings deposited at 600°C
have a strong orientation of basal planes parallel to the substrate,
luti fyt. K
?
sar
#,
.
..:
:
INTRODUCTION
-
.
Rhenium and tungsten-rhenium alloys of high purity and density have many potential high-
temperature applications including themnionic converters and nuclear reactor components, such
as, fuel element cladding and tubing. Although the high-temperature strength of tungsten is
decree.sed by alloying with rheniwn, the alloys, particularly W-25% Re, have ductility at low
temperatures. The purpose of this study was to develop the thermochemical deposition process
for these materials so that deposits of consistent quality can be prepared in a variety..OL.
shapes and sizes. The process studied was the hydrogen reduction of tungsten and rhenium
hexafluorides (WF. and ReFo) at low pressures. Materials prepared by this process may be used
in the as-deposited condition or wrought by conventional metal working techniques.
Tungsten-Rhenium Alloys
**
.
L
.
it.
H
i
.
.
.
."
1.'
.
.
.
.
.
Alloy deposition studies have been mainly concerned with compositional uniformity in
3/4 in. OD tubular deposits having surface areas or about 25 in... Studies were initially
conducted in an apparatus similar to that shown in Fig. 1 except that a WF.-Ref. injector was
not used and the deposition mandrel was stationary. The gas mixture which was metered into
the mandrel encountered a temperature gradient Iram room temperature to the deposition tempera-
ture, typically 600°C. Under this condition différences in thermodynami stability between WF-
and ReFo caused a disproportionate depletion of fluorides, and resulted in an axial variation in
composition. The deposits were always richer in rheniva near the inlet end of the deposition
mandrel than farther downstream; the composition gradient was typically 10% Re/in. When the
deposition mandrel was moved at a slow rate through the furnace, without the WF6-ReFinjector
shown in Fig. 1, in order to increase the length of the deposit, the deposit exhibited axial
Uniformity of composition. Metallographic examination and electron microprobe analysis,
however, revealed compositional variations through the thickness, the result of super position
of layers of different composition which formed as the mandrel passed through the temperature
igradient of the furnace.
.
.
une
.
.
.
.
.
-
1
.
.1
.
.'
.
rik.J
i
;
.
-
.
K
-
.
",
:18
,-
Subsequently, a thermodynamic analysis of the tungsten-rhenium codeposition process
indicated that homogeneous reduction, deposition of alloys having the same metal content as the
metered fluorides, can be achieved by proper selection of temperature, pressure, and gas com-
position. In general, increasing the rhenium content of the alloy requires a higher tempera- :
ture at a given pressure to achieve homogeneous deposition. The purpose of the WF6-ReF6 :
injector is to deliver the gases directly into the uniform temperature zone of the deposition
mandrel, thereby minimizing deposition in a tenperature gradient and resulting compositional
variations. The microstructure of nominal 22% Re alloys deposited in a moving hot zone both
with and without an injector are shown in Fig. 2. Except for the injector the deposition
-
.
-
.
-
-.-.
*Research sponsored by the U.S. Atomic Energy Commission under contract with the Union
Carbide Corporation.
'
.
.
A
.

'
U
"Do
AB
*
..'.
"
..
1
7.
"
.
.
BLANK PAGE
1
1
..
.
"
.::
:.
th: 19.
.
.
.
...
. *
FTW !!' '7.' ..HII
HOMM
E
conditions were the same for these deposits: 900°C, 3.0 torr, He: (WE's t Ref) = 15:1, and rate
of mandrel travel of 1.8 in./hr. The deposit shown in Fig. 2a which was prepared without ari
injector etrihed nonuniformly, an indication of different ca positions in the layers. The
deposit prepared with ar. injector, snowa in Fig. 2b, appears to be more uni crm in compos. tion.
The metallographic evidence of degree of uniformity in these deposits was verified by electron
probe microanalysis, shown in Figs, 3 and 4. The darkor. etch ag layer of the deposit shown in
118. 2a, which was prepared without an injector, coatained !! biglier rhenim content than the
remainder (Fig. 3), whereas the nicz'oprobe analysis revealed a substantially uniform composi-
tion through the thickness of the deposit prepared with an injector (F2.8. 4). The microar.&-
lyzer traces are not, quantitative since neither trace gives the composition of the binary alloy
at any position. In addition, different scale factors caused a compositional variation to be
more obvicus in the rhenium trace. The traces do indicate quantitatively the variation of
either element across the thicknes6. ... . .. . ....::
The microstructures of nominally 7 and 90% Re alloys 'are shown in Fig. 5. The deposition
conditions for the 7% Re alloy were 800°C, 5 torr, and H2: (WF e + ReFo) : 20:1; the conditions
for the 90% Re alloy were 900°C, 3 torr, ari H2: (WF6 + ReFo) = 1.2:1. In both cases, the nari-
drel travel rate was 1.8 in./hr. There is not much mi
variations in either case, but the 90% Re alloy (12g. 5b) did not deposit as a dense coherent
coating. The composition along the length of typical tubular deposits about 3/4 in. dian pre-
pared in a moving lioti zone with an injector was determined by standard analytical techniques
and the results are presented in Table 3.. For the lengths shown the axial uniformity of con-
position was about 0.5% RE for nominal 7% Re alloys, about +1% Re Por nanina i 22% Re alloys,
and somewhat less uniform for nominal 90% Re alloys. Deposition rates of about 0.6 g/cm2. nr.*
and metal recoveries approaching 100% were obtained using an injector'.
RE
Alloy deposition has also been conducted in the external coating apparatus shown in Fig. 6.
The reacting gases expand into the 4-in. dicm x 12-in. long chamber, thus proviùing a more uni-
forma gas mixture for reaction at the 3/4-in. o mandrel surface than occurred during deposition
on the inner surface of relatively long, narrow tubes in initial experiments. In other words,
a more rapid depletion of Ref. is minimized in this apparatus. Typical deposition conditions
in this apparatus were 600°C, 10 torr, and H2: (WE6 to ReFo) = 25:1. Under these conditions the
aeposition rate was about 0.3 g/cram.hr, and the metal recovery was about 75%. Axial uniformity
Ol compositon was comparable to that obtained in a moving hot zone with an injector Micro-
structures of deposits prepared in the external coating apparatus are shown in Fig. 1. The
deposit of highest rhenium content (26.9% Re) etched nonuniformly, indicating a compositional
variation through the thickness. Electron probe results are not yet available for the higher
rhenium alloys, but microanalyzer traces in Fig. 8 for a deposit having an average 2.4% Re
content shows that the rhenium is substantially uniforin through the thickness,
19
The tungsten-rhenium phase diagram. in Fig. 9, shows, that the equilibrium phases in tungsteri-
rich alloys at 3.500°C are the body-centered cubic sclid solution (B) and tetragonal sigma (0).?
Alloys deposited at lower temperatures were found by X-ray diffraction to contain two phases
which coexist in certain composition ranges, the solid solution (ac = 3.155 A) and a cubic
B tungsten phase (a = 25.010 A).1,3 The S-tungsten phase, which has a hardness V8
2000 DPH compared to about 400 DPH for the solid solution, erabrittles the deposits and
certainly will have a substantial effect on mechanical properties. The as-deposited 1-25% Re
alloys are, in fact, quite hard and brittle. The temperature and compositional limits for
formation of the B-tungsten phase are currently being investigated. The phase has been found
in alloys deposited at temperatures up to 1200°C. Preliminary results indicate that higher
rhenium contents are required with increasing deposition temperature for the phase to be stable,
approxiinately 15% Re at 800°C and approximately 22% Re at 1100°C. Sigma phase has not been
detected in alloys containing up to 42.4% Re deposited at 1100°C nor in alloys containing up to
25,8% Re deposited at 1200°C. Annealing treatments are being conducted at 1500°C art above to
investigate the thermal stability of the as-deposited phases. After annealing at 2000°C ior
20 hr in vacuum alloys containing 15, 20, and 25% Re were single phase solid solution, a 29% RO
alloy contained the solid solution and a small amount of sigma, and alloys contuining about
50% Re were single phase sigma. The 25% ho alloy was quite ductile following the anncal.
*For tungsten-rhenium 2l.loys multiplying depoeition rute in units of' gm/cm2. or by a factor
of 20 approximately converts to rate expressed in mils/hr.
PR

BLANK PAGE
.
G
X-ray diffractoleter ECCS on surt'aces parallel to the substratt indicated that the kino
orientation was datinant over the runge 0 to 25% Re whether single phase solid solution, single
phase B-tungsten, or a mixture of these phases. The 90% Re alloys were found to be single
phase hexagonal. Lattice parameters were identical to those of unalloyed thermochenically
deposited rhenium, a. 2.76 A and c = 4.45 A. The hardness of the alloy was 600 DPH, however,
compared to 1.00 DPH Por unalloyed rhenium.
Rhenium
Rhenium metal has been reposited in the external coating apparatus at temperatures of 500
to 1150°C. In the higher region of this temperature range (1000 to 1150°C) the coatings were
metallic in appearance, but metallographic examination revealed a noncoherent grain structure
similar to the Re-10% W deposit shown in Fig. 4b. The deposition rate was about 0.4 cm.hr.
At lower temperatures the coatings were dark and porcus. The metal was also deposited on the
inner wall of tubuiar substrates at temperatures of 600 to 900°C. At the higher temperatures
in this range the coatings consisted of loosely joined columnar grains, although metal recover-
ies approached 100%. Smooth dense coatings were obtained in a 600°C static hot zore at metal.
recoveries of 25 to 50% and a maximum deposition rate of 0.05 g/cm2. hr. X-ray diffractameter
traces on sections parallel to the substrate in the latter coatings revealed a strong (0002)
texture.
Summary
Tungster-rhenium alloys containing up to 90% Re have been deposited in the temperature
range 600 to 900°C. Tungsten-rich alloye (7 and 22% Re, having uniformity of composition in
both the axial and radial directions of tubular deposits have been deposited in a moving hot
zone with a WF.-Ref. injector to minimize deposition in a temperature gradient. Deposition on
the outer surface of mandrels in a statis. hot zone also shows promise for obtaining unifornity
of composition. Rhenium-rich deposits deposited noncoherently in the temperature range 600 to.
900°C. Dense coatings of pure rhenium were ootained at 600°C, but rhenium deposited noncoher-
ently at higher temperatures. The B-tungsten structure WuS found in tungsten-rich alloys
deposited at, temperatures up to 1200°C. Tungsten-rich alloys huve (100) of both the bočky-
centered cubic solid solution and the B-cungsten structure parallel to the substrate, whereas
pure ihenium has (0002) parallel to the substrate.
Acknowledgments
. The authors acknowledge contributions to this study by other persons: J. B. Flyan who
performed the experimental work, C. F. Haltom who prepared metallographic specimens, .
R. M. Steele who performed x-ray studies, H. W. Dann who performed clectron probe microang. lyses,
and the Graphic Arts Department which prepared the line drawings. The Metals and Ceramics
Reports Office prepared the manuscript which was reviewed by W. C. obinsori, W. R. Vartin, and
G. M. Adamson, Jr.
Table 1. Composition of Tungsten-Rhenium Alloy Deposits Prepared
in a Moving Hot Zone with a WF6-ReFo Injector
Distance Along
Deposit Length
(in. ja
Rhenium Content, gb
23.3
22.8
22.6
UOCOU IWUM
22.1
23.2
23.4
23.2
22.7
22.5
22.5
94.9
93.2
90.7
89.
91.6
7.85
6.72
7.5).
6.82
6.96
6.62
7.02 6.16
6.94
6.66
7.26
7.10
7.23
7.33
7.28
7.26+0.596.68+0.42
0.32
0.52
89.8
90.2
92.2
91.6
91.1
22.0
21.5
21.7
21.6
21.0
22.1
21,2
22.5+0.9
2.0
Average Re, O
22.2-2.
91.02:8 91.113:3
97 19
+2.2
-1.3
.3
put
Measured îran the injector at experiment startup.
bRemainder tungston except for trace impurities,
WAA249,

i
BLANK PAGE
2
.1
i..
,.."
., -T
i
ti'
.
... :-;
7.11.
:
,
:...
leſexences
13. I. Federer and C. F. Leitten, Jr., Muci. Appl. 1, 575-580 (December 1965).
2J. J. Duglioh, BMI - DMC 152, 92 (April 1961).
J. I. Federer and R. M. Steele, Nature 205(4971), 587 (1965).

-
.
!
+
-
1
+
WATER-COOLED
WFG-Ref
INJECTOR
(STATIONARY)
Ų
x MFG
- WF6
DEPOSITION
FURNACE
(STATIONARY)
XT REFE
DEPOSITION
MANDREL
MANOMETER
B!IRNOFF
SCAVENGER
FURNACE
COLD TRAP
COLD TRAP Q10
VACUUM PUMP
: HF SCRUBBER
DRAIN
Fig. 1.
Tungsten-rhenium deposition apparatus.

M
2
.
.
..
-.
.
...
"
!
.
more
.
1
-
it.
.
I
.
2
.
e
.
is
Fig. 2. W—22% Re alloys deposited in a moving hot zone. Conditions: 900°C,
5 torr, H2: (WF6 + ReFo) - 15:1, mandrel travel rate 1.8 in./hr.
Etchant: 1 NH4OH(conca)-i H202 (30%). (a) Deposited
without injector. (b) Deposited with injector.
i
SUBSTRATE
SURFACE
.
.

DARK ETCHING ZONE
CO!!POSITION (%)
.
Re
:
5.0 : 7.5 10.0 12.5 15.0 17.5
THICKNESS (in. ä 103,
Fig. 3. Election probe microanalyzer traces for a tungsten-rhenium alloy
deposited in a moving hot zone at 900°C. Average
composition 22% Re. Traces are not quantitative.
.
.
...
...
..
.
90
.
cig/SURFACE
SUBSTRATE
80
11
ini?
70

60
COMPOSITION (%)
30
n
. - rad
.
L
0 2.5 5.0 7.5 10.0 12.5 16.0 47.5 20.0 22.5 25.0 27.5 30.0
THICKNESS (in. X 100)
Fig. 4. Electron probe microanalyzer traces for a tungsten-rherium alloy
deposited in a moving hot zone at 900°C with a WF.-Ref. injector.
Average composition 22% Re. Iraces are not quantitative.

ting
4
.
.
1
-
1
Fig. 5, W-Re alloy's deposited in a moving hot come with a WFG-ReFG injector.
(a) W7% Re alloy deposited at 800°C, 5 torr, H2: (WFG + ReFo) = 20:1,
mandrel travel rate 1.8 in./hr. Etchant: 1 NII OH(conca)-1 H202 (30%).
(b) W-90% Re alloy deposited at 900°C, 3 torr, H2: (WF6 + ReFo) = 12:1,
mandre.l travel rate 1.8 in./hr. As polished.
CHAMBER
HEATER

ma
GAS OUTLET
DEPOSITION MANCREL
Wwwww
GAS INLET
Fig. 6. Apraratus for deposition on the external surface of mandreis.

..
HA
Tot
..
i
.
w
new
mento en el more than
.
AL
20
-
Ny TA-
.
*
Nie
+
*
*
*
-
-
w
"
,
Y
T
:
W.
9% Ne
W-8.2%. Ro
W-26.7Kc
Fig. 7. Tungsten-rhenium alloys deposited in an external coating apparatus,
LEGAL NOTICE
Ouvi.
This report was prepared as an account of Government aponsored work. Neither the United
Statos, por the Com'aiosion, nor any person acung on behalf of the Commission:
A. Makes any warranty or representadon, exprossed or implied, with respect to the accu-
racy, completeness, or usefuluess of the information contained in this report, or that the use
o any informadon, apparatus, method, or process disclosed in this report any not infringe
privately owned rights; or
B. Assumou any liabilities with respect to the use of, or for damages rosulung from the
ur of any information, apparata, method, or proceso disclosed in this report.
As used in the above, "person acting on bobali of the Commi8810n" Includes acy em-
ployoe or contractor of the commission, or employee of such contractor, to the extent that
such oinployee or contractor of the Compassion, or employee of such contractor prepares,
disseminates, or provides ACC886 to, any information pursuant to his employment or contract
with the Commission, or big employment with such contractor.
L
SURFACE
Temperature, F
6000
5000
4000
3000
.
.
.
otaal
htx
. Liquid
.
For
c
..
2
1
o
-- -
-
25.0
Traces are not quantitative,
22.5
20.0
17.5
15.0
THICKNESS (IN *103)
Electror probe microanalyzer traces for a tungsten-rhenium alloy deposited in an
12.5
10.0
7.5
external coating apparatus. Average composition 2.4% Re.
,0
.
5
2.
5
:
Fig. 8.
*
:
...
Re
80
60
Weight Per Cent Rhenium
Fig. 9. Tungsten-rhentum phase diagran.
2040
W
:
.:
tri
*
..
-..-.......
:
B. Sto
a
.
.
..
B+L
....
.
oose.
3000
2000-
.
-
!SUBSTRATE

Temperature, C
it
-7.-
1. NOLLISOOP:03
.)

'
11ULL
END

DATE FILMED
3 / 28 / 67
ST
...
D
NI
AR
.
..
:
SUNS
14.
*
H
.
.the
1
AL