Post on 18-May-2021
W.^sf^j 'f^ ?H
'"fC-^'K.i.y^'tt^ PATL^
Thermo Electron
CORPORATION
/ i
DISTRIBUTION JF THIb 0( LJ ' / t f . ' ! iS UHV. ,*
DISCLAIMER
This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency Thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof.
DISCLAIMER Portions of this document may be illegible in electronic image products. Images are produced from the best available original document.
Thermo
[ ' U i P n R A M O KJ
K Electron
Report No. TE4233/4237-60-78
NASA/DOE
ADVANCED THERMIONIC
TECHNOLOGY PROGRAM
PROGRESS REPORT NO. 28
October 1977
NASA Contract NAS3-20959 ERDA Contract EY-76-G-02-3056
P r e p a r e d By
The rmo Elec t ron Corporat ion 101 F i r s t Avenue
Waltham, Massachuse t t s 02154
t^ UiSTPiBJTiON : T,-SS i v ^^ .
W/M^ Thermo f/c. Electron f ( ) n I •( i n AT I CIN
I . SURFACE STUDIES
A, SURFACE CHARACTERIZATION CHAMBER
Samples from surface charac te r i za t ion and conver ter exper iments
were analyzed by Auger spec t roscopy . A p o s t - m o r t e m Auger analysis
of Converter No. 166 indicated that lanthantim had t r ans fe r r ed from
the lanthanum hexaboride col lector to the tungsten e m i t t e r . In addi
t ion, carbon, oxygen, sodiusxi, cesium^ tungsten, and rhenium were
detected on the e m i t t e r . The e33aitter surface had become a dull g ray
that could be peeled eas i ly with a r a z o r blade. The exposed a r e a s con
tained no rhenitim but did contain a g r e a t e r amount of lanthan\im. The
source of rheniura mus t be the molybdenum-rhenium braze that bonds
the era i t ter to the emi t t e r s l eeve . No molybdentxm peaks were readi ly
apparent ; however , if any weak peaks were p resen t , they could have
been dominated by the tungsten spec t rum. Rhenium and molybdenum
oxides a r e both known to be vola t i le , rhenitim being considerably
m o r e s o .
Suprisingly, no rhenium was evident on the col lector although
carbon, oxygen, ni t rogen, sodium, nickel , ces ium, lanthanum, and
tungsten were detected. The lanthanxim and tungsten signals were of
comparable s t r eng ths . Except for one out of the seven spots analyzed,
carbon concentrat ions were s m a l l . Most l ikely, nickel and sodium
w e r e introduced during handling subsequent to the opening of the
c o n v e r t e r .
The Se len ium-Reservo i r Conver te r , No, 173, was dissected and
i t s contents were analyzed. The coEector had a dark deposit along the
edge of the opening to the se lenium r e s e r v o i r . Auger analysis indicated
1
ZJZ Thermo f/c Electron t I ) t KM I n A n ci N
this deposit to be e lementa l seleniijxn.. The emi t te r surface was shiny
(like e lectropol ished tungsten) except for a number of smal l da rk
specks about 1/2 inm in d i a m e t e r . The highest concentration of these
specks appeared in the "shadow" of the selenium r e s e r v o i r opening.
Auger analys is showed the specks to have a high concentration of
selenixira. A few of the selenium Auger spec t ra obtained from var ious
spots on the emi t t e r surface were d is tor ted - indicative of a selenium
compound.
The emi t t e r of Converter No. 173 was subjected subsequently to
an e lec t ron mic roprobe ana lys i s . An analysis taken over an a r ea
about 1 m m at the center of the emi t t e r surface showed the presence
of se lenium, tungsten, ces ium, and a slight amount of nickel . F igure
I - l shows a port ion of this analyzed a r e a magnified 1000 t i m e s . The
l ighter rec tangular portion of the is land-l ike s t ruc tu re was found to
be ces ium, a s were the nximerous globules; the smoother , open a r e a s
were tungsten. The c rus ty region contained cesitim, xnost likely
ces ium hydroxide , which dr ied and cracked under the vacuum environ
ment of the e lec t ron m i c r o p r o b e . Thus , the emi t te r surface is s e g r e
gated into regions of tungsten, selenitim, and ces ium (hydroxide). The
observed d is tor ted selenitoin Auger spec t ra a r e mos t l ikely caused by
the complicated surface topography of the sample . These ana lyses ,
coupled with the independence of the conver te r cha rac te r i s t i c s as a
function of se lenium r e s e r v o i r t e m p e r a t u r e , suggest that a selenium
cold spot existed in the diode. A second conver ter is being constructed
in o rder to continue these invest igat ions .
The r e su l t s of other Auger ana lyses a r e given in subsequent
sect ions of this r e p o r t .
2
w 7711-15
F igure I - l , Scanning Elec t ron Photomicrograph of Central Region of the E |n i t t e r (Converter No. 173) (Magnified lOOOx)
m/gm Thermo f/c, Electron f ( ) t I P ( ) R A I if TM
I I . PLASMA STUDIES
A. CONVERTER THEORY
Schottky and Goiilorab sca t te r ing effects were added to the analyt ical
raodel of the ignited in te re lec t rode p l a s m a . The computer p rog ram was
improved to accommodate these two effects, with prel inainary resu l t s
plotted in F igure I I - 1 . These calculations indicate that Schottky effects
can produce output cu r r en t s significantly l a rge r than the corresponding
sa tura t ion c u r r e n t s . Coulomb sca t te r ing was found to have liie poten
t ia l of substant ial ly increas ing the a r c d rop .
The requ i red emi t t e r sheath height for the formation of the double
sheath can be calculated f rom a s imple sheath model by matching solu
tions der ived f rom independent analyses of the p lasma and eini t ter
sheath^ as shown in F igure I I -2 . The double sheath is seen to be
formed at a cu r r en t r a t io of approximately 0 . 5 5 . This value is con
s is tent with exper imenta l observations^ assuming such formation is
assoc ia ted with the knee of the I-V cha rac t e r i s t i c s of diodes, and
appears at the c u r r e n t ra t ios between 0 .4 and 0 . 6 ,
B , EXPERIMENTAL PLASMA ANALYSES
Spectroscopic analyses of conver te r p lasmas under operating
conditions were ini t ia ted. Spectra a r e being taken throughout the
4000 to 9000 A region. Line width, and re la t ive line and continuum
intensi ty m e a s u r e m e n t s allow determinat ion of e lec t ron densi t ies and
t e m p e r a t u r e s , as well a s excited state population densi t ies through
out the in te re lec t rode reg ion . Sensitive detection may enable t r ace
consti tuents ( e . g . , oxides) to be moni tored . Changes in p lasma con
ditions due to var ia t ions in T , T j T , e lect rode spacing, and L C R
4
7710^4
1.1
1.0
0.9
(0 - 3
0.8
0.7
0.6
0.5
0.4
1 ! 1 1 ! 1 ! 1 r WITH COULOMB SCATTERING ^
WITH SCHOTTKY EFFECT
/
/
/
/
WITHOUT SCHOTTKY' EFFECT
// ' '^^wiTH couLoye / / / SCATTERING
J I I L J i™__„_L
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 OB 0.9
V^ (VOLTS)
Figure 11-1. Schottky Effect on Converter Per formance
7710-13
>
5.0
4.0
3.0
2.0
LO
0
1 r I 1 1 —
/ -—PLASMA / SOLUTION
/
/
/ SHEATH'S. / SOLUTION >y y^
- ^/L^""""^ 1 ^ 1 1 1 1
-
-
-
0.5 0.6 0.7
J /J ,
0.8 0.9 1.0
Figure I I -2 . Format ion of Double Valve Sheath
migm Thermo i c : Electron f ( ) M n ( i n A r I O N
current were observed. Temporal decay of interelectrode conditions
in piilsed converters are also being investigated. Preliminary mea
surements of a converter with a tungsten oxide collector (T = 1425 K, E
T = 751 K, 0,25 tor r cesium, and l-3xim electrode spacing) have ^ 14 -3
shown maximtim electron densities of 2,8 x 10 cm and electron
temperatures below 3000 K. Relative line intensity measurements
indicate progressively less equilibrated excited state poptxlation den
sities with decreasing energy levels. C. ENHANCED MODE CONVERSION EXPERIMENTS
Refined measurements were performed to determine surface
thermal deformation of a molybdenum lase r -mi r ror electrode during
heating. The apparatus (see Figure 11-3) allowed the splitting of a
l - m m beam from a He-Ne laser into two components and reflecting
one from the center of the electrode and the other from its edge.
These reflected light beams were observed as two spots on a screen
placed approximately 20 m away, Deforxnation of the electrode sur
face is measured by the relative motion between these two spots. To
first order, this relative motion is independent of the thermal distor
tion of the sample support s tructure. In a set of experiments with
electrode temperatures ranging up to 1675 K, maximum deformations -5 never exceeded 8 x 1 0 cm. This value is sufficiently low to allow
efficient operation of close-spaced diodes.
7
7710-15
•SCREEN
20 METERS
MIRRORS BEAM SPLITTER
He~Nt LASER
\«8i-METAL BELL JAR
TO VACUUM SYSTEM
Figure I I - 3 . Exper imenta l Ar rangemen t for Surface Therm.al Deform.ation Measurement
8
wi^ Thermo ^ Electron f I 1M p n f 1A I I f 1 fsj
III. CONVERTER DEVELOPMENT
A . LOW-TEMPERATURE CONVERSION EXPERIMENTS
1. Tungsten E m i t t e r , Sprayed Lanthanum Hexaboride Collector
(Converter No, 180)
This var iab le -spac ing device employs a specially gxiarded col
lec tor s t ruc tu re (see F igure I I I - l ) in o rder to de termine m o r e a c c u r
a te ly the work function of i ts sprayed lanthanum hexaboride surface
a t low cu r r en t l eve l s . Plots of col lector work function v e r s u s T /T^^
const ructed f rom re tard ing inode data show good cor re la t ion with dc
back-emiss ion v a l u e s . Work functions a s low as 1.45 eV were ob
tained at T ^ / T values between 1.4 and 1.5 from re tarding plots
C R
(see F igure 111-2). At tempts to introduce l ow-p re s su re oxygen past
the cesixrm r e s e r v o i r , in o rde r to reduce the col lector work function,
have been unsuccessful . Work is cur ren t ly under way with higher
p r e s s u r e s of oxygen.
B . HIGH-EFFICIENCY CONVERSION EXPERIMENTS
1, Tungsten E m i t t e r , Sintered Lanthanum Hexaboride Collector
("Showerhead" Conver ter No. 166)
Measuremen t s of the effects of oxygen diffusion through the porous
LaB/ col lec tor were continued. Oxygen penetrat ion was observed to
init ial ly deignite the conve r t e r . Pumping on the back of the porous
co l lec tor would subsequently re igni te i t . The saturat ion cu r r en t of
the conver te r was always a factor of 2 to 3 g rea t e r than p r io r to the
oxygen dose . However, this inc reased performance would not l as t
m o r e than 20 to 30 m i n u t e s . Eventual ly the conver ter could not be
9
7510-IOA
COLLECTOR FEEDTHROUGH
COLLECTOR
COLLECTOR INSULATION CERAMIC
SEAL
FLEXIBLE BELLOWS
EMITTER FLANGE
EMITTER SLEEVE
OUTGASSING TUBE
/ EMITTER ^WITH PYROMETER AND
THERMOCOUPLE HOLES TO CESIUM RESERVOIR
Figure I I I - l . Schematic of Guarded Conver ter
2.2
2.0
:0N
(eV
) CD
o 1.6 z
cr il.4 QC O 1 -
i i i . 2 o o
1.0
—
__
+ \ s +
\
'BACK EMISSION M~800K ® - 7 5 0 K X - 7 0 0 K + - 6 5 0 K • - 600K
X X X
1
i
J^ ^
t r CONVERTER NO. 180 SPRAYED La Be COLLECTOR
1
® ®
7710-2
®
M
—
'
1.4 L8 2.0 22
T/Tr
Figure 111-2. Work Function of Sprayed Lanthanum Hexaboride Collector Ve r sus T / T R
r p Thermo f/C Electron ( I 1 w R n R A r I C) N
ignited under any condit ions, and it was assumed that a l l the ces ium
had become oxidized. In fact, when the cesiuin r e s e r v o i r was slowly
heated up to 1000 K, no ignition was possible; the conver te r was taken
off t es t and was Auger analyzed.
A new LaB/ "showerhead" conver te r is being designed with a
window to pe rmi t spec t roscopic ana lyses of the oxygen diffusion
p r o c e s s e s ,
2, Lanthanum Hexaboride E m i t t e r , Guarded Nickel Collector
This conver te r is essen t ia l ly a repea t of No. 168 with a gxiarded
nickel col lector (see F igure i n - 3 ) , The problem with Converter
No, 168 was that the cu r r en t from the tungsten cup could not be
separa ted f rom the cu r r en t frora the LaB , emi t te r s lug. The guarded
col lec tor should al leviate this p rob lem in that the cu r r en t from the
tungsten shoxild flow to the gxiard, whereas the cu r r en t from the LaB . o
slug shoxild flow to the col lec tor button. The construct ion of this
conver t e r is near ly completed.
12
7711-6
GUARD
COLLECTOR
COLLECTOR INSULATION
KOVAR SLEEVE
SAPPHIRE WINDOW
EMITTER HOHLRAUM
SPACING PAD
COLLECTOR FEEDTHROUGH
0.004 SPACING LaBg EMITTER ELECTRODE
RHENIUM WIRE SUPPORT TUNGSTEN EMITTER BODY
Figure 111-3, LaB, Emitter, Guarded Nickel Collector Converter
m» Thermo f/c. Electron ( ( ] n p) n n A I I n f^j
IV, COMPONENT HARDWARE PROGRAM
A, CVD COMPONENT DEVELOPMENT
The f i r s t f ree-s tanding si l icon carbide-graphi te- tungs ten com
posite hot shel l has been fabr ica ted. This shel l , shown in F igure
I V - 1 , m e a s u r e s 9. 53 cm in length by 2. 54 cm in d i ame te r . The
tungsten and graphite l aye r s have uniform th icknesses of 0.02 cm
and 0,1 cm, respec t ive ly . The si l icon carbide outer layer is 0 ,28 cm
thick at the he in i spher ica l end and 0.05 cm throughout the cyl indrical
por t ion . Design studies a r e in p r o g r e s s to determine a method to
grow g r e a t e r th icknesses of s i l icon carbide at the round end, and to
es tab l i sh a hydrogen purge to the inside of the shell during CVD of
the si l icon ca rb ide . The hydrogen purge is n e c e s s a r y to maintain the
in te rna l tungsten deposit free f rom t r a c e amounts of silicon ca rb ide .
The t e m p e r a t u r e and flow r a t e set t ings of the tungsten CVD
appara tus have , for the t ime being, becoine es tabl ished. Tungsten
she l l s , which a r e 0,05 cm thick at the open end and 0 ,04 cm thick
at the round end, a r e now routinely formed after an 85-minute expo
s u r e to a 150-SCCM flow of tungsten hexafluoride.
A capabili ty for producing composite hot shells has been demon
s t r a t ed . The CVD hot shel l effort of the component ha rdware p r o
g r a m wil l now be directed along the l ines of establishing routinely
reproducib le fabrication p r o c e s s e s .
B . ALLOY HOT SHELL DEVELOPMENT
Simulated furnace test ing continued at a furnace gas t empera tu re -7
of 1573 K. The inside of each shel l is evacuated to 10 t o r r . Table
IV-1 p r e s e n t s the status of these t e s t s as of November 2, 1977, A
14
w
• ; • • • » • • • ' .
.-. -ilrWtel
7710-17
F igure I V - 1 . Composite CVD Silicon Carbide and Tungsten Emi t t e r Subassembly
15
•
TABLE IV-1
SIMULATED FURNACE TESTS (November 2, 1977)
TEST P O R T NO,
1
2
3
4
5
6
HOT S H E L L
REACTION BONDED SILICON CARBIDE
KANTHAL A !
KANTHAL A l
INCONEL 671
CHROMIZED STAINLESS S T E E L 446
KANTHAL A l
T E S T HOURS
14 ,213
12.626
12,492
5 , 6 6 3
5, 172
2, 810
COMMENT
LEAKTIGHT THIS HOT S H E L L WAS BRAZED TO A M O L Y B DENUM S L E E V E WITH N I C K E L - C O P P E R
LEAKTIGHT
LEAK OBSERVED REMOVED 3 ! Oct . 77
LEAKTIGHT
LEAK OBSERVED
LEAKTIGHT
DATE OF TEST INITIATION
9 A p r i l 75
25 Aug. 75
9 Sept . 75
17 Nov. 76
17 Nov. 76
7 May 77
»y^ Thermo • ' . - • • f/^ Electron { ( ) H P D F ^ A T I O N
leak was observed on the Kanthal A-1 sample on October 31 , 1977,
This shell has been removed from the furnace for failure ana lys i s .
Two additional hot shel ls were placed on tes t on October 31, 1977.
Both of these shel ls were del ivered to Thermo Elec t ron on October
6, 1977 in connection with Subcontract 7070-4190-411-Mod #3 , The
samples were mounted a s follows;
Tes t P o r t No, 3: 446 CRES P l a s m a a r c sprayed with n ichrome, 0. 01 to 0. 02 cm thick and difftised for 8 hours at 1367 K in a hydrogen a tmosphe re .
Tes t P o r t No. 5; 446 CRES P l a s m a a r c sprayed with Cr O , 0.01 to 0.02 cm thick.
Both samples had 2. 54 c m inside d i a m e t e r s , 0, 38 cm wall th icknesses
and were 20, 3 cm long. The end caps were TIG welded. The asseanbly
was s t r e s s re l ieved for one hour at 1256 K and a i r cooled.
17
m» Thermo W^ Electron r I J M p n ^̂ A ! I c) NI
V. COMBUSTION-HEATED THERMIONIC DEVICE
The objective of this t ask is to design, construct , and tes t a
f lame-heated thermionic device for eventual use in combustion-heated
pane l s . The f i r s t step in this development will be the construct ion and
f larae-heated t e s t of a "miniature device" employing the mos t promis ing
m a t e r i a l s in s izes that a r e readi ly avai lable . The next step wil l be the
fabricat ion of a l a r g e r "workhorse device^ " which will be evaluated in
a t e s t furnace that provides coinbustion p a r a m e t e r s s imi la r to tih.ose in
fossi l - fuel powerplants . The data and experience with the minia ture
and workhorse devices wil l then be factored into a l a rge r "prototype
d e v i c e . "
The design of the min ia ture device (see Figxire V-1) in tegra tes
the r e su l t s of the al loy hot shel l development at TRW with the va r i ab le -
spacing diode t e s t s a t T h e r m o E lec t ron . To date , the mos t promising-
m a t e r i a l for alloy hot shel ls is INCONEL 671. This high chromium
nickel aUoy has shown good stabil i ty in s h o r t - t e r m t e s t s a t TRW and
a l so in a l ong - t e rm t e s t in the Simulated Furnace Faci l i ty a t The rmo
E lec t ron .
The hot shel l d imensions a r e determined by readi ly available
INCONEL 671 tubing (nominally 2. 54 c m in d i ame te r ) . The effective a r ea 2
of the conver te r is 2, 0 cm and is de termined by the maximum d iame te r of the col lector that can be inse r ted into the emi t te r hot shell without shor t - c i r cu i t ing .
The L605 emi t t e r m a t e r i a l was chosen on the basis of the favor
able data f rom the var iable spacing Converter No. 143. These data
were repor ted in the P r o g r e s s Repor t for August 1976.
18
MOUNTING BRACKET
CESIUM RESERVOIR-
7710-4
MOLYBDENUM REINFORCEMENT RING'
NICKEL COLLECTOR
INCONEL WICK / EMITTER
:LLOWS WICK SUPPORT
INCONEL 671 HOT SHELL
Figure V - 1 . Miniature F lame-Hea ted Device
mf^ Thermo ^ Electron
In o rde r for the t empe ra tu r e drop through the collector s t ruc ture
to be min imized , a potassixim filled heat pipe is used in the collector
stejXL. A cooling fin, a n d / o r an axxxiliary hea te r , may be clamped to
the outboard end of the col lec tor for diagnostic tes t ing .
The hea t source for heating the emi t t e r of the device will be a
gas t o r c h . A p r o p a n e / a i r , or propane/oxygen, torch will be used
in the init ial t e s t s .
Construction of the min ia ture device is in p r o g r e s s . Test resxilts
should be available next month .
20