AO-AO9 9" AIR FORCE INST OF TECH WRIGHT-PATTERSON AFB 0ON ICHOO--ETC F/O 7/4SAS FLOW TLUE FOR SPECTROSCOPIC STUDIES.(U)DEC 80 V R KOYM
UNCLASSIFIED AFIT/GEP/PH/B0-5 (I.
mmmmmmnmmu-EEii/EEll/il-Ehmmhiimnnnu
EEEEEnllEEEE-
AIR UNIVERSITYUNITED STATES AIR FORCE
SCOO, OF ENGINEERING' S t ~liT2
La ~WIOM1PAThSON AMR PORCa SAS. Okio 1
mo 9W i61Appip" i paU ltd
OLEVE-
APPRVU ~R ~ flT -r-~ AFR 190-1"
FREDR~IC C LYNCH. '~r.UADirc7'-r Cf public Adt i'r
NIT force jnsttute Ot (NICVP ~ti aAer0~~P~B~ 454A~33
A\ GAS j;IAI, 1(
FOR SPECTROSCOPIC S1i)B
TIMES IS
Vernon R . I\oyI
AFIT/GEP/Pff/8O-5 Ca (fy USA'
DTICSEL ECTEFEB 1 21981
B
A~pproved for inibic Fc c i s; d i.; ri hit ion rIrc imited.
AI
~~AP IT/P;EP/PHT/ 0- 5
A (;AS FOW TI'II
FOR SPECTROSCOPIC STUDIESI
THLESI S
Pt resented to tihe lFca tilt y oi tlbe School of Fli,iileer ilii
of the Air o ce Iis t itite o I TechoiioIogy
A\l" r niver'sity
in Partizal FulfEilimeit of the
Requirements for the Oei 'ee ol
Master of Science
\.\
by
Vernon R: Koym/ B.S.
Capt UISAF
Graduate Engineering Physics
l)ec I tI80
Al~p ove I' r Itll l i l't lt'l:,'; li:l iihll i~ ll 111 im teI
9Preface
'he results oF this study shoul d he use ful to the
invest igat or interested in uI fi,, a as f low tuhe to (10
excited ,as (1AO or N*) spectroscopy reaction kinetics or
oxidizer/oxidant spectroscopy. Data were compiled on how
to build and use a gas flow tube, and the capability to use
a flow tube to examine gas phase reactions in widely
different systems.
Special thanks are due to l)r. Steve [lavis of the Air
Force Weapons Laboratory, without whose help and generous
loan of equipment this thesi. cotild not have been coinpleted.
I would also like to thank miy thesis advisor, Dr. F.A. r rko
of the AFIT faculty, lfu his pat iilnce aind invaluable advice
in completing my research aind prepa iin' the thesis. T hanI,:;
are also due to Lt Col W. Bailey' and Pr. V'. RIoh of my
committee, and Sharon Gabriel for helping ine prepare this
manuscript. Finally, I would like to thank my wife, Daria,
for her tolerance in this effort, and my C;od for sustaining
me in the rough parts.
Vernon R. Koym
", &n2,o/ortDi~at Specia1
o r
DC
Contenit s
Pa 1,e
Prefac -~le---------------------------------------------
List of ,Symbols- -- -- -- -- -- -- -- --- --- --- -- --- --- - -- - ---
Abstract------------------------------------------------ x i
I. Introduction-------------------------------------1
Bac ki 'ound --------- - - - -
Obj ect ive- -- -- -- -- -- ---- -- -- -- -- -- -- -- -- -I.General Approach -- -- -- -- -- -- -- -- -- -- -- ------ 3
Theory------------------------------------ 5
Introduction--------------------------------- 5Backgr-ound-------------------------------- S12 Chem'i "",i fe scence------------------------- 8PbO Reaction------------------------------- 9Singlet Molecular Oxgn------------------ 11Gas Flow Tubes------------ ---------------- 14
II. Experimental Apparatus ------------- --------- 24
Introduction --------------------------------- 24Flow Tube ------------------------------------ 2 4Vapor Generation ---------------------------- 33Ancillary Equipment ------------------------- 37
IV. Experimental Procedure------------------------ 40
Introduction----------------------------- -40Alignment and Calibration ------------------ 40Data Collection--------------------------- 43Safety -------------------------------------- 45
V. Results and Discussion -------------------------- 46
Gas Flow Tube Performance ------------------ 4612 Experiment ------ ----------------------- 51lPho Exper ililit --------------------------------- 62
9Contnilt s (Cont 'd)
Pa ge
VI. onclusions ind .,, ecoinmendat i o n s ------- 64
Fl1ow Tube Performance ------- 64I:xporimfeflts - - - - - - - - - - - - - -- 6 4Recoimiind"It ions - - - - - - - - - - - -- OSPoss iblIe Iinv esigt i gat----- ons--- 6 9
B ib l i o g r a p hyy - - -- --- - --- - --- - -- --- - - 74
Appendix A: Flow Tube Procedure----------------------- 80
Appendix B: Pictures of Apparatus--------------------- 83
Appendix C: Useful PbO l41ta------------------------- ,
Ap~pendix D : Response Cur-ves for Phot omul t ip1ierTube-- ---- ---- ---- --- 88
Vita----------------------------------------------------- 91
List of Figures
Figure P ag e
1 1 File -, I rgy Level1s ----------------------- 7
20.) Pot ent iaI Cur Ives ----------------------- 12
- ~ ROc i r-Ct I Iat i Oi A+Anl Fxpan s ioil Po int -- 5
il Veloc it y Pr-ofiles Uinde r the To r-ho I cot
Laminar, and Plug Flow Regimes- -- -- -- -- -- ----- 19
s Experimental Apparatus------------------------- 2S
6 Viewing Port Geometries------------------------ 27
7 Combust ion Chamber-------------------------
8 Viewing Por't - - - - - - - - - -
9 Glass Wool Par't iclktc 'I aj--------------31
10 Vacuum Systm --------------------------------- 32
11 Furnace -------------------------------- 34
12 Hlood Designs------------------------------------ 36
13 Calibrat ion Setup for Pho Reict ioni---- 38
14 Alignment Setup-------------------------------- 40a
14a Tube Tip Shapes-------------------------------- 44
1s 12 Potential Curves------------------------- 52
16 Vibrational Populations------------------------ 5S4
17 Plot of Raw 1 2 Data and Pi-ed ictedTransitions------------------------------------- 60
18 Relative Intensity Profile---------------------60a
19 Proposed Hood Shape---------------------------- 65
20 Proposed Furnace-------------------------67
21 Gas Flushing nsr----- --- 68
22 View Port I1e;tr-*.. -- 68
List of Tables
Tahleo Page
I Emission Lines of PbO 2
II Flow R e--s-- 20
I I Flow Tube Performance- 47
IV Franck-Condon Factors for12 (B) 12 (X)v 0 ---------------------- 56
V Calculated 12 (B) 12 (X) Transitions -------- -59
VI PbO Flame (Tbservations ----------------------- 63
I "
' ist of Svihols
A Area of flow tube
Angst rom (1(8 cm
A r A r g o n
C Cond Uct ance
cm Centimeter
C'I 'T l, Chemii cal iIect ron iC z]1l1 it ioa I .a , Ir
f) Flow tube diametCr
F!P Rat io o F i i eld i t en'. it I k) l1illh., I-density in a p lasi!a
ev Electron volt
IlgO Mercuric oxide
I Iodine atom
I Iodine molecule
I 2 (x) A form of spectroscopic notation signifying theground electronic state of the I? molecule
I2 (A) The lowest energy electronic excited state ofthe same multiplicity as the v.rotiid state
I2 (B) The second lowest ie r i' level
VIII2t
Vil
1, i st of F Sym l II s (C(on t 'd)
iota I an2,u I ar momentum (sp in1 and an i'l I ar11(ilcn ct ill) I II at olli i c (, I (Ct ruon
J J A par-t iciia r type of electrion ic coup I i nschemle inl 11o0ecul es; 111und',S case C
.1Total angular momentuia, of an electronic -state
in a molecule
k Boltzmiann's constant
K Knudsen's number-
Rec rlat [O'IiJl LoneIC i~
NI Atomi c miass nuibteir
MIT NMillimeter--
N vPopulat ion number of no] ectiles; in ,ihcrat ional
state v
N 2 Nitrogen molecule
N * Activated nitrogen21
0? Nitrous oxide
NFE Nitrogen trifluoride
0 Atomic oxygen
01, Oxygen mo I ciie
P Pressure
I. ist o C Symbo ls ((oIt d
Pb lead
PO ,eod ox ide
lead I olecule
Franck-Condon Factor,IV V
Q Throughput
" Inte rItc le"a Sep)arIaN t Ol!
, Reyliold'.; lllC 1or
Ro Univer:;sal -s constant
Sp Pumping Speed
T Temperature
Velocity
V Experimental velocityo
V Velocity predicted by p F low asstuption
Atomic diameter
v Energy of an electron in cm
0- Oscil I at loll fre(pueflcy of mIl clii Ie
/e Anlalrmon 1 C I 1 ( ils I ) 111(1 e CtI I
iI
I. is t of Symhols ( (ont d
\Icall t'1ee path
lens itv
V iscees it v
Axial component of electron orbital angularmoment uln of molecule
Axial component of total electronic angularmomentun
[ I Conccnt rat i on of
rt , A = 2, mu t I i iCity of
ct = 2, multiplicity, of
(f(T) Time dependent absorpti oll coe F-fic icut
N I
______ ______
AF IT/C lP/ PI 1/ 80--
Abst ract
A (',,S flow tube Was cons't ructed to allow chemnical
react ion stutdies at pre-ssures From () .I torr to 30) torr.
The tubO was designed to allow introduct ion of oxidants,
diluents, or excited gases. A furnace was constructed to
produce high temperature vapor in the combust ion chamber of
the flow tube.
The flow tuhe was characteri-ed by its piimpinp. speed,
p)ressu~re, throughpu) tt, a nd eva l uar j ni, keyno IdV'S a iid Kni I sen , s
numbers at pressures friom 0. 1- 20 turr . TIwo expertic11t5
were conducted to vie r ify the ft o., tuh [ie erfoia rma ce. Singlet
mnol ecular oxygen was p)roduceud hi ;a min i- oi;ive dischiarge, and
reacted with gaseous 1 2 to y ie Id che2iti i i inesccncu from the
transition 12 (B) - 1,(X). The spect],11 ruinwas r'ecorded and the
bandheads assigned to vibrational trainsitions prtedicted by
theory. PbO was created by react ing vaporized lead with
N2 0. The emissions were compared with liter'ature.
6i
A (,A", T:1. ', , V II
FOR S P IIC'I'IROS(' IC I'lSTUD II
. lit roduct ion
BackgIround
The military community has strong interest in the
chemical electronic transition laser (CETL) (Ref 23:2),
because of a nu h c I" or very des i ril) Ic featLurCs. AI, c
systems generate exci ted state mo lecul]es by cimemiCal
react ion; consequent ly, sca lu-up oi such syst cmis is not
inhibited by the ,.e i 4ht and volt Ime of el ect i-Calt I)olvcr
sources. Some molecules aISO exhib, it Ulect 1onic t ranIsitionls
in the near infrared, vis ible , and it rav i ol ut ; thus , a
laser utilizing these react ions could exhibit t ransmi ssion
properties in the atmosphere superiur to current high
power lasers such as the CO. (Ref 18).
The recent success of the so-called iodine laser
(Ref 34) has demonstrated the viability of reactions
involving the 'A state of the 01 1molecule. This excited
gas species exhibits a number of very desirable properties.
First, the 'AO state has a radiative lifetime of about
forty-five minutes, and is relatively immune to collisional
deactivation with walls (Ref 2:2'21). Second, oxygen in
this state is readily produced in large quantities.
.Tnother candidate Iaser pum iin', react ion is
Ph + N PhW* + N,
This 5 rect ion rI,llCeS the electronically excited species,
PbO* (Ref 12)
Ilertzberg (ReF 29:564) and Shawhan et al . (Ref 43)
list the following transitions of the PhO molecule.
TABI.E I
Lmission Lines of PhO
Transition V0 (ciii-l 1 (A)
E -- x 34 755.
D x 30103.2" 3322
C x 2 4 7 6 2 .0a -11 o
B -+ x 2 2 1 7 3 .4 a 1510
A x 19728.3 a 569
a All transitions shade toward red.
b Emission lines reported.
All transitions are of desirable wavclengths for atmospheric
propagation. Much prieviollS wori, on 'h lihis 1)beaii done
,!
(Refs 0, 7, 8, 10, 3, 35, 36, 42), and the spectral
structure of the system is well described. A fundamental
problem iarises, however, in that c, mission rolm the Pb + N2()
reaction is reported as quite weak (kef 33), although the
reaction 11b + N2( is very efficient :it pirodiicini. iround
state PbO. This raises the possibility that efficient
chemiluminescence could be produced by collisions
ground state PbO with excitel molecular gas species ';ith
active nitrogen (RcF 33). mone ut i\:,t ion Fou tLiil~lin- a
flow tube was to estall ish ai s :Nstem in which tiv inv .-tia-
t ion of the PbO + N"i and -h t , ( could he pei forted
Many designs have been reported i n tL2 I i turat ure (ket s 1,
11, 13, 14, 20, 21 , 2-, 23 , 2-, 2 , 2n , 7 , -1 , -1 , 52) and
extensive work has been done on li. Hi) .yst L em in flow tubes
(Refs 33, 36, 43). Consequently, it i.,as decided to build
a gas flow tube and modify it to allow chemiluninescence
studies with gas phase reaction products, excited gases,
or a combination of the two.
Obj ect ive
This thesis was intended to accomplish tivo goals. First,
a gas flow tube would be built and modified from standard
designs to allow the study of gas phase reactions in several
ways. The flow tube design would be verified by reproducing
the results of two widely studied reactive systems,
O2 ('A) + 12 and Pb + N.,).
2 N,0.
rp
General Approach
A gas flow tube was used to react materials in the
va" l phase with oxidants or excited gases. The resulting
spectra were analyzed and the spectra compared with published
data. The results were used to verify gas flow tubC perfor-
mance. The flow tube was characterized by Reynold's and
Knudsen's numbers and system throughput. Some recommendations
for further work are made.
I
1 1 . heory
Int roduct ion
1 brief account of the theoret ical ac krinpd
to this thesis s presented ill this chapter. I t,
discussion of the genesis of this thesis i,
attention to Chemical Electron ic TIran i I
Second, some general theory of gas low rc;ict,.
List, the two reactionis c hos.in IS i,
aru presented, and the a w c it,, I
Back, round
['his thesis had it:: ()I.I ill i .
Weapons Laboratory. In 1!
reported on the succes. of the I
utilized a gas flow reactor eatI
lated reactions:
Singlet Molecular Iroduct in. ,.cd :
chemical production method. <aw, 1. 1 th roul
a mixture of IH20', and .arll, ;and 1, ,,. ,, 1. react ion
chains:
Cl + if 0 + NaC!! A(1 4 "o, (
221
tr i* )C + !lO + il(l (:l + 1C I I j + H ()
+ ,, + Cil (2)
Pi r t t-t the Imi. partici pates in an enery pooling
.1' ( , 1" ,.(:" + Q . K" :2:" (3)
,. , t v., . lltt[liAIICCi , anii d i K- uc i ated LS
S+ .* + ('I (5)
+ 1 . 15 n m p1oton (6)
PII,,t, re 1 .I , ' t ir, , . levels for this system and for
the I m) I ( Ii I'
I A; ' H, ,nve.i to he very attractive for a CETL.
-i,:lv't , :. . r t ye' I ifdt ly f ot I15 minutes (Ref 2:2529)
It t II i l clad ivat ion (R'ef 2:2529).
•I
- 12 (A ~
~ 1 "03
761)m 1 782c
x2
The ilio] Ces s;ta'tes of, mIolecul ar (:;, ca he p rodtice ei in I a ri e(
(Iptant it ic s2 v i a thle chemT i ca I react i cii liown i ) I.qs (I) and
(2 ),and thle 0?C C F I flent is hen i i,,n . Iodi nie i s recvc I cc].
The oucs of th11is (IT, hs S t imul ated inlte rest in1
a nunhe1) r ofI other chem i calI react ions as I)otent i al lIaser
candid ate s, most of which can h e stu (Iiet def fcct ivelIv inl I
gas flwreactor.
Consequently, this thesis was designed to accomplish
t Ie t ask of bu IlId me ai, ee F Iow I kLILI 1 or ic I c Col 'l g.lc I-:! tC
asi'candidates inl at vapor Stalt(. and] react thenJ~ wilh
ox idizers or exictcd cs such aIs 0('',) to p)IuduLCC
vihronically exci ted specie-s. 11wC sy enchosenU to dlemon-
strate the flow tuhe capabilIitijes csthe 1 .,/0, s\'stcm
which exhibits yellow-igreen chieii lumines'cen~ce. Tlii s react i on
utilizes the same constituents aS the 1,,/(), laser, and there
is a large amount of data in the literature for comparison.
The capability to generate oxidant/high tClempeature vapor
reactions was demonstrated by the Pb + N.,h reaction. This
system was chosen because most of its hand Stl-ltUIcrs lies
in the visible regir~n (thus simplify)inlg the eXperIimental
setup and allowing a simpler measurement system), and
because this system may be a candidate for a CETI.
1 2 Chemiluminescence. The cheiniluminescence is generally
agreed to result from emission From tlie, I (B) state (Refs
15, 16, 52). The reaictioni (ain be dle~cr i bed hN
I, B)- V (y7 + v > 5000A (7)
Since analys is of the data is (ependent on the emitting
state and not on the excitation mechan ;sm , no d iscuss ion
will be presented in this sect ion on the react ion mechanism.
There is, however, a controversy about the I(P,) state
formation and this will be discussed in Section VI.
I I!)U)
The PhO molectI e is k) I' i III cIs hecltlise its ,puct ra
arc e.tensivel)' st tid icd , the Ib)ld molecule is amenal Ie to
excitation b) a iIt r it add it y -c such as d ind tl1 3
(Ref 31:727), and becajuse it is a candidate for a (ITL.
PbO formation from Ph and N2 ) is exothermic, hy 71. 68
kcal/mol (Ref 31).
Reaction Methods . PbO emissions have heen eoenerated
by a wide variety of reactions. Bloomenthal examined PhO
spectra by using a lead arc in air (Ref 9:2,1. lie also
reports earlier efforts by E!der and Va lenta tho used lead
chloride in an oxygen flame, and Grelhe and Konen who used
lead chloride in a carbon arc in air. Shawhan and 'Morgan
(Ref 35:377) measured the A, B, and I) systems and discovered
the F system of PhO, using PbO heated in a furance and
irradiated by an ult raviolet source. M.hore receni IV,
n1denher , Dickerson and Z;or, r .tiild d tI lie( rI,-act i n
(II
11h + - hbo + 0 ( $39
and Li nton and Proida did a comprehensive study ut i 2itin
the react ions
Pb + 07 -1 PhO + R (Reaction Products) (9)
Pb + N - Pb + W) (1 ('2
Pb + 1 Ph. (11
and
Pb + 0* (nicrovavc ,lisch , , Pl, (12
The Pb + N2 0 reaction was chosen Voi t,1w s in'. icitv of
production and the milder foul ing p roblc, '; riloi'ted by
Linton and Broida (Ref 31).
Spectra. Hlerzberg reports the major elet-onic
transitions listed in Table [.
Numerous vibrational bands are reported; Linton and
Broida report 33 bands for the A - X transition and 57 for
the a -+ X transition (Ref 31). For ;i list in1, of PhO
emission lines, see Appendix B. Iiitoil and Pro ida show
that Ph + N20 product -peCti- iiivoIv¢t t ii ; i rin the
a -* X, A -) X, 11 - X, a d . \:, I . ii . I. 1
I Ii .
1r
t rans it ions are reported; this is 1os hi y a limitation of
their detection apparatus. They also report that , as
p ressu rC incre.sed from 3. 1 mm to 3.1 1m, the dominant
emission changes from a pale h1uc flame (A - X , C -, X
systems) to a yellow-red flame (a X, B, X N systems)
Coupling. The electronic states of the Group IV
chalconides can approach jj coupling for the heavy metals.
For jj coupling, or more properly, Ilund's case c, the
coupl ing between 1. and S is stron,er than the coupli og.,
hetv;een L or S and the internuc lear ax is (ReCf 21 :31 I.
Thus, A and Z aic not good qu!anttnm ntiiher s, and ee consider
that L and S couple to, t i \ . J1, ii t Ii ,i a xi a I Co]IipOeiit (
For this case, a numher of general selection rules hold.
The rule J = 0, 7! holds, as does the symretry rule
+ --* -, -I + -+- Since is also a ,ood quantum
number, A = 0, +1; 0 *- 0 ,T---O . :or tle case of both
states ? = 0, AC = 0 is forbidden (i.e., no Q branch is
allowed).
Singlet Molecular Oxygen. Excited oxygen Was generated
by passing the gas stream through a 24St) megahertz micro-
wave discharge in a quartz plasma tube (Ref 48:287).
Energy levels of the 02 molecule are sho'n in Figure 2.
The major production mechanism fr o (',) is direct
electron impact excitation with },round state (2 molecule by
a 0.98 ev electron (R.l 1,1):2-) ) IIi an e ry o 782.39
cm 1 (Ref 29:540).
II
_ _ _ _ _ _ __.. .. _ _
(~-'~ ::t) )
>~. 1
(I
((1
H
/
pc~ -
(Is
16.
(1
- 7.7
(Ilir 2 K - t j a 1 ii
I
The re a rc t i,'() mI .ij product jon mechI1 i ss for the
second netastItle state, (, (':). ' '!e first is excitation
by a 1 ()-2 ev elect ron to the 13120).) cm- I ene ra, leve1
eI 2': 1 . This process ispr o- hlh not competitive with
the excitat ion to 0-( .! ), because the cross sections for
excitat ion to n,,('") electron impict ,ire approxirmatel)' one
order of magnitude lower than (C ( , for the same electron
enery (Ref 50:1482).
A second reclhmi i ;i , is thes p o,1 i.), c ict i(i y
De rwent and 11v r h -I t ate t I t t lie c'I icciil ra t i oi; of
02 (1') and 02('A) are LO - 10 " , rep evct i v ly (Ref 16:721
Although they do not specif ical h .;ataitc lay they ariive at
these numbers , Thrush and Thomas rep rt a met had hy which
the concentrations are calculated from thle emissions of
0,(7) at 1 .91 }m and 0,(1A) at 1.2" pm (Ref 18:290). The
02 ('.) concentration is lower because of the low.;Cc cross
section for excitation in the pla,ma , and huc:ause it is
depopulated by
o2( ) + 02 - 0 A) + 0(112 2 (A
and
i i _n
Ii2 I I _ _ _ _ _ ,. . . . .... .. ..
The second react ion shows a cont ra st to 0,( A which is
h ighly resistant to deact ivat ion by wall col I isions
(Ref 3:1769).
An important by-product of the microwave react ion is
atomic oxygen. In order to suppress (0, a coating of P10
is appl ied to the walls of the quartz tube. This coating
catalyzes surface recombination of atomic oxygen (ReF 4:288).
Gas Flow Tubes
flhe flow lv:ttte ns and velocitie:n of a *u' flow tule are
of fundamental imjoi't ance in dete rainin, the ccsults; of kinietic
experiments. Tivo of" thle Cr1 te ia up 11o,. tube desi gn arose
from an examination of fl ow phcnomue
Recirculation and Ilow iest i ict ons. t first design
criterion was the reqy i rement t njiA iii e fN o1 restrictions
and sudden expansions or contractions of' the tube. Flow
restrictions decrease the conductance (if the system;
conductance being a measure of the efficiency with which
gas can be pumped through the tube. An example of a
restriction would be a series of successively smaller tube
diameters towards the vacuum pumps.
A sudden expansion or contraction of the tub- diameter
can generate a recirculation zone such as the one shown in
Figure 3. Recirculation was eyaimined by IPennicci (Pef 38:11)
as part of the problem of stdde( Iv expailndod gas flows.
Ii
F 7
Rze fer r mo to Fi goure 3 , a schiemrat ic dep ict ion is !'ive n of
the vortex phle nomena known as rec i rCuL at i on . (;a., no leCs
in the r-eci i-culat ion zonie 1, im\ rena in for- somie time be fore
reattach mo,, to the flow in the attachiment zone. Ph i S
plenoie non 11 S S r i"1 i f i canI11t whenl a t temp11)ts a rc made(1 t o C rca t C
aflame , s ince in an improperly) (I es i gnc d chiamber , the
reaction may occur in the recirculation zone rather than
the v iewing area downstream (see Chapter V).
Veloc i tv 1PIo f i I cs an 11, F ow Chaira Cte0r i - alt i on1. Tfhe
seCondl design cr i te i i on was that t lie :;\ ;t em opera;tec inj t he
lam inar f low re, Ilie . A\ more ri- orolls de I- in i t ion i S i\'en
below , hut in generalI t erins ai 'l ow ma)'q hu cha ia ic e r i edCL as
turbulent, laminar, or molecular. Tlie i mpoitamce of l aminarI
f low may be understoodl ais f o 11ows p )ick inlg a s illiple
example, the rate of consumption of reLactant, A, in a
chemical reaction can be expressed as
C11 k[A] n (16)
where [A] is the concentration of A, k is tihe reaction
rate coefficient, and n is the power dependency of the
reaction on the concentration of A. To evaluate a [A] in a
flow tube experiment, we may write
d[A] -dEAl clx (17)(T -J -,- -C
0i
where d [X is the chani,,e in [A] as a function of some
distance scale, usuall) longitudinal in the flow tube. 1%'ed x
can then interpret F as the gas velocit y, and evaluate
-,--- in terms of some observable such as the intensity of
a given emission.d x
This approach requlires that d he 'e, I de fined. A
common assumption is that of plug flow, which assumes the
linear velocity to be everywhere uniform. This assumption
holds true for the turbilent flot.: regime, but breaks down
hzadly I for the laminar case (Ref Si) T. (le tilrlhiluit Case
presents problems, ttlo ,, in that tc .,as', nombeir densities:
flay exhibit lo,.ail dcviations Fro t Iot rujl, (ttulluilit
cel1::), with th le ) [ I:; .ioei ictil: h( il- difficult
to predict.
Io'olf showed that the velocit can he defined for
laminar flow in a flow tube (Ref 5) For the molecular
flow case, however, the assumption that the t ime rate of
change of concentration is proportional to the linear
velocity since there may not be a well defined linear
velocity. Thus the second criterion was to coinfirm that
the system exhibited laminar flow. The following, sections
describe the flow regimes, the equations for evaluating
flow tube performance, and present a limited discussion of
parameters useful in calculatin, flow irates.
I?
:low ei~ Miucs. A g'as Hlow t nbc I))a) be characterized
in part by the conditions under which gas flow takes place.
(;enerallv speaking, flow can be turbulent, laminar, or
11(1lecuiti r. lig'urc 4 depicts velocity l profiles for the
turbulent and laminar cases.
Viscous Flow. A volume at pressure P1 is
connected to a second volume at pressure P, through an
aperture of size A1 with P., < P If P is such that the
mean free path of a gas molecule, ,, is small compared -to
the dimensions of A, the flow is viscous (Ref 0:{h2). The
velocity of the gas can be increasCI l, lobV weringh P. until
the ratio P 2 /P 1 reaches a critical vdlie; the .peed of
sound for that gas (Ref 40:67). At Lhis poiint, further
reductions of P 2 will not increase the fiore. rate. Thus,
we see that merely adding more puiips for a ,iven cavity is
pointless if a flow tube is operating in the viscous regime,
once the critical P2/P1 ratio has been reached.
When operating in the viscous region, a gas flow tube
may exhibit turbulent or laminar flow. The dividin.p line
between turbulent and laminar is determined by the tube's
Reynold's number, Re, where
Re pvD/r (Ref 10:61) (IS)
and I is the gas de.u ;it)', v the 1 icar 1, : ream velocity,
a
Turhul ent Flo(w Prolfil1e
7N
J~cril i o f' Ai ,% 11111 .0 1 0I I'l I-,- \I It i 'i r') F i
Ii V I ' (, I fi I I j l ' drII - l tI l l t ;11iI~
llj (w HH''
t e ViScC Sit', o S 1) the di ameter of the tube.
'ene raIl I , an Re ,rea ter than 2100 indicates completely
t r 1 lnt I low, and Re less than 1 1 0l i indicates c nompletc I y
I a!!1ilnilr flow.
lolecular Plow. The dividing lines between
viscous, intermediate, and molecular flow are determined
by Knud_;en's number, defined as the ratio D/X , where )
i.- tl!e flow tllbe ,IlalIt L i ' I l t h ' ! 1'('C' pttl.
"'abl II summal'i :'(s the lcjimc i oIid,tirjes., alter Rot h
(Rei 40:61).
TABLE II
Flow Regimes
Flow Regime Re,D/
Viscous Turbulent Re > 2100
Laminar Q > 200D (air
)/,>110, Re<Il00
Q < 100D (air)
Transition Intermediate 1 < D/X <110
Rarified Molecular D/X < 1
20!
T'hroui,hput Q. The throuch put is a measure of the
(ulantitv of ,as moving through a pipe per unit time. Roth
defines the throughput Q as PV ('DI"/4 ) idhere the units are
a:; previous ly defined. In genera 1, the Flow in thC viscous
regime will be turbulent if gis reater than 2001D, and
laminar if Q is less than lOO1.
An empirical formula for Q in air is
n = 9.0 x i i- Re), in torr-liter/sec 1I.,air
(Re f 40:61
Knudsen's Number. Xnudsen's nlUml)er (K ) is used
to delineate the bandrii jes bet.ev ii vi,-cous, iii'ermediate,
and molecular flow. K is defined by
nn
K> 1 , (Re f 3: 67)(
where D = diameter of the flow tube, and A is the mean
free path (Ref 42:67). A general expression for the mean
free path of a gas in a flow tube is ,given by' Roth
A = kT//T T c2 P (Ref 40:37) (21
where k = Boltzmann constant, l..)80. x lo) 10 er./ K, T is
the gas temperature in OK , 11 is Ihe , as pvessuve, and C is
,'1
g
the m 1 ecu a r I a et er As an example , an ei, i rical
forl cn a Co of a i r at amihicit t('mu rerature is
5 x 10- 3/P (Ref .1(:37 (22
Conductance. 11,1hen steady state flow is achieved
in a flow tube, the number densities of two adjacent sections
of pipe are related by Eq (23),
N = ( i(1 - 2' (N 1- .F 0 : (23
Here N is the numher 01 mu ] LIS crossiO2 a CFus
sect ion of the pipe I , IeU) I'L SCIit Ic 11c 11 111,c r
densities in the respect i ve ,ect ions, :iiid C is the
conductance (Ref 40 :03). The conductancc foi a pil of
dimension D, length L, and operatin,) at pPres 5 e I' is given
by Eq (24),
C = (P D")/(128 pL) Ref 40:71) (21)
For air at 200 C, Eq (24) reduces to
Cai r = 182 (D 4 /L)P (Ref 40:74) (25)
The pumpini. ;.pleedl, S , il ;I clll mic Cc) nulected by
conductance C to a pump havinc sp(ed p i, ciJveii by lq (26),
!
!IS l/sp + I/C (Ilef 40:66) 26
This can he rewritten as
S/SB = (:~p)/[I + (C/SpK] (2
I'quat ion (27) shows that increasing the pump size Sp does
not increase S if the conductance is the limiting factor
(B t I 0 :66)
Plug IIIow. In li, 1astel s thesis, W o If examined
a trad it ional as :sipt ion empiloyed in igas fN ow %.ork; the
l) In Flow.; assll pt It Ile\ f 50) . eIt t he tIol F tlics i , the
di I:crence )etween lu, Hlow, , I aini i : , and I )lr CI Iint fIo w
can .e visualized as in Figure 3. (Vol f detC. ntilxied deviations
from te plu Flow, assumption on the order oi 1.0 - 1.8
for the ratio Ye/Vpf, where Ve is the experimental
centerline velocity and %Vpf is the predicted or plug flow
velocity (.\fe- 53:55). This phenomenon is very important in
interpreting gas kinetic data, or any experiment in which
raditative lifetimes or downstream intensities are measured.
i - ... .. ' - .... .. .2 3 i
Ill.I Experimental A pparatus
1 n t rod tic t i oni
r\ rieC 1eC r i 1)t io0n1 o f th!e za ri 't u s 5 s i ven i n th)i s
c ha1)t er. Inl 1u i id i nog thle appariatus ,two criter ia were
cons idered paramoiunt. 1:irst , the tube w..as to h~e designed
to al1 low highI flow rates with dynamiiic pressures as low as
(1.1 - 0.2 torr. To achieve this condition, special
at tent ion was 1atil d t th ac IUnn '*.) ;I (.!t iod Lu sy:';l eil :I,'
oilt . Second , rvospzi-e is rejio rIt e Ve te pirob 1 ems' ini
s)ystem foulin(, esjpec iail>I ini t h( ilb N 1 lactionl (Ref :3 .
A 1Ium11ber of sclienies cee va' I triteC folr d1lj il Ing wit tis
probl em and for so v vi iiL thle pi jubdi n I' iji iiit ai inii, zt clear
viewport.
Flow Tube
The flow tube (see Figure S) was assembled from
component parts manufactured by Alloy' Products. Tihe tube
was constructed of three inch inner diarneter stainless
steel. pipe. The ends of each sect ion hiave ani (-ring, groove
to provide a vacuum tight seal . The connect ionis between
sections are made with quick flanges.
Input Section. The input sect ion secrves, T liree purposcl,
in ti s apparatus. f:XC ted t' clss et iU rudnelled t1Irough al
!" , ( Ie n l t lII I,:*I l l ] ol ' t I n 1 1 1 1 ' 0
-iI I I
I I I
___ Jr 1
I C-C- -~ -.II
- ~ ~ 2'I
C- l~ -~C~f. Z
III,! C
C-x
I If.
I j C.)
II
chen I iTne:cence CXpCrimelnt. llIr inn. the PhO exper iment
a1 k i -L i I I P, a i-!l w:s pI 1 ced It t hle 1'1'()It o' tlie i lnitit Sect I ol
aInd i IIet s r,> did Enr flushii ,II 8 (."C. s : i , .I Ire
,: 11l81 V , ([IIl'IlC ll .' 5 S Cal I)c' int I'oll cd I'(I pIQS 1181 of- the
react ion c yatihe r.
(oion11tiSt ion Chaellher. Tilec comhust on ch:imheir in which
the react ion he i n(-, stud i ed took p lace s a comnme rci ] "T"
made by Alloy Products. An additional arm was added as
shown in Figure 7, :Inl for the viCI-i1l prt :in en.l plate
:5e l:igure 8) had a 2.5 inch li)c counte I-- 5 .. in tlie Ilc-,
zn I I centered, 0175 inch h1le as i ej tL)u he,. " 01.25
illch thick glsls i l at .l as .51 t1 ,. ( ,1 r till ' 1 I lil tI the
110 c , anlid the vilcinII I :, I!. , I I a d t I c'I I it. hi "I oS ig .
facil itated rapid clean il o do c1, k its-.
Previous invest i ",tolrs ilhitc uxL, riiciced 5cVCI,.- foul in.
problems in oxidizer/metaI vapor react i ol; (Rief 33:399).
Two approaches were tried to solve this pirohlc l i. 'lhe
viewing port on the combustion chIiaer (a i )ore 7) has a
0.0625 inch inner diameter tube inlet througleh the wa Il in
front of the viewport. Gas could le iiit rducced in front of
the viewport to create a flowing gas Iar'r to lie reaction
product.
The second approach was to 1 lace it 21 i lci , (ection of
p) C llp t 1'.1111 of Or lol11,I1 I Io) he co)IIh II:, ioll 1iihI[I)crI
(Figure 6). The extejis ion Iail as I 1l,t , i teni cl to the
_ ,I
II U,. I
Yi I2r 6'ew m
.jI -l
-/,1
C I.-
/
/7'
~i.
71121 V) I714 //
ii
V lire 8: Vj (Wi tip I'~rI.
U
one on the combust ion chamber, and flushing gases were
introduced ahead of the viewport.
An additional experiment was devised to reduce fouling,
downstream of the comhustion chamber (Figure 1). A 3.()
out er1 diameter stainless steel pipe sect ion had a screen
wire welded to the bottom, and this pipe was inserted into
the downstream leg of the combustion chamber. This pipe was
then filled with glass wool and the effect of the device
on pumping efficiency and tube fouli ng, was evaluated.
Oxidizers were intordticed intt) tle com)tistion chamber
through an injector as shown in 'i i re 7. This shape .'as
ciiosen over a nuiibeir of . ide I used .-liapes (Ref 25:7229
because it gave a hri ,hter :iiid more inil'o rm i C ulle.
Transition Section. A 2.0 foot section was incorporated
downstream of the combustion chamber. (;as kinetic studies
could then be performed by replacing the stainless steel
tube with a glass tube.
Pumping Section. Immediately following the transition
is an adaptor which goes from the 3.() inch stainless steel
to a 1.75 inch inne diameter pipe. The adaptor has a port
for measuring pressures (Figure 10).
Provisions were made to install a cold trap immediately
downstream of the adaptor to remove atomic .iodine before
it reached the vacuum pumps (:i gu 'e I()). The cold trap
7;i
K
I -
-~ /
I,- ~~---- I
'K -
F' I 'ii r*t* ~ I: I i I I I *.lI
'I
I
C d
To Vacumii
Top~ V~iew
LM,
was cooled with a d ichloromethane-dry ice slush (approxi -
mately -60 ° F). Downstream of the adaptor or cold trap,
an 8 .) inch lon;, 1 .75 inch inner diameter pipe connected
to a Consol jdated Vacuum 1 .5 inch ball valve (Fi 10ure 10).
Another 1.75 inch pipe doi,'nst ream of the hal I valve connected
to a "Y" as shown in FIigure 10. The downstream les of the
"Y" connect to 1.375 inch inner diameter pipes which curve
on a 24 inch radius to the vacuum pumps. The connections
to the pumps are made with vacuum hose to minimize tray.s-
mitted vibration. The punps are 17. 0 (t helch flour pumlps.
Vapor Generation
Iodine. Iodine vapor was produced b)) passsiing AR gas
stream over 12 crystals in a ]/I" iinwe, ila ss tuihe. The
vapor pressure of iodine is approximately I mw at room
temperature. This pressure was adequate to give useable
concentrations.
Lead. With a melting point of 327.30 C (Ref 28), Pb
requires a high temperature furnace of some sort. An
electric furnace was constructed, as in Figure 11, along
the lines suggested by Dr. S. Davis (Ref 13).
Power was provided from a 28 volt, 600 amp Rapid
Electric Company Model S-528 D).C. 'nemator. (:mlrre, nt was
fed through No. 4 cahles to electrodes cooled by a water
jacket.
I~~~ rahXL:
k -I
F77
sidu View
r()d
Top~ V ie(v.
I;; 1nl1et
1)ir i (~f 1 1: Ii tl
The current heated a crucible which contained reagent
orade I ead shot. The resulting vapor was entrained in an
Ar carrier which took it to the viewin, region. A numher
of schemes were tried for ha ffling the 10low of the furnace
while shaning the flow of entrained lead for maximum hright-
ness. A few hood designs are shown in Figure 12. Excess
heait was eliminated hy water cooling the electrodes and
wrapping copper iater cooling pipe around the outside of
the Curnace and i e, i clammmlr
Measu rement System
he spectrograph iwas a ,Jarrcl ,AsL .5 ;nscanmin;
monochromator, witli a di [traction [,,t ing, rl' d to 1180
grooves/mm, and blazed with a rccip pocal di spersi-on of0 1
16 A/mm in first order. The resolut iun was 0.2 A in first
order. The light was focused on the entrance slit of the
monochromator with an 8.7 cm focal length lens for iodine,
and a 15.0 cm focal length lens for lead (see Figure 6)
Plots of relative intensity vs. wavelength were recorded.
The signal for the 12/02 experiment was received by
a 1P21 photomultiplier biased at 100)0 volts bN a Keithley
244 high voltage source. The 1TZI has an anode sensitivity
of 1.2 x 10 A/W at 4000 A. For the PhO reaction, a more
sensitive RCA 7265 photomultiplier tube with in anode
sensitivity of 3.0 x 10 6 A/W was substituted. Appendix I)
shows typical responsc, ciriv,-, I Iti l.s, tuibcts. Tiel signal
1
E Ind p 1atc, Ki h
c. Chimnecy
1"': _______lo o d________________________
was amp 1 i Cied by a Ke i thIcy 427 cu rrent ampl if ier and
recorded on a Houston Instruments 0nni.iraphic 2000 X-Y
p I otter.
.icro wave. The microwave source as a Kiva MPG-3
driving an Evenson -Broida cavity cooled with gaseous N .
The gas feod tube was 14 mm diameter quartz, which was
coated inside by blowing 02 over 1 and passing the mixture
through the cavity at 100 watts forward power. This pro-
duced a yello-brown lg() coa ting ihose )urpose ,as to
eliminate atomic oxygeu in the excited products. T] he Las
tube was introduced through a conniect or (lmlaiufat l'ured by
Cajon, Inc.) into the flow tube, wh ich al lured the position
of the tube relative to the 'lame l o , vaied. [h lasma
was ignited with a Tesla coil.
Ancillary Equipment
Calibration Lamps. Calibration was performed with an
Hg lab standard from Ultraviolet Products. The 5460.7,
5769.6, and 5790.7 A lines were selected for calibration
marks. For the 12/02 experiment, a siparate calibration
run was performed using the 11g calibration lamp. During the
PbO reaction, the high sensitivity of the 7265 photomultiplie
tube precluded direct observation of the 11g lamp without
inducing unacceptable noise. There'fore, the setup shown
in Figure 13 was devised.
i~i
icl
13 7 1. n rt t() I t __I__UpI
*.'ml11 i 1-oId Ild heed. A I cas c onn1Ie ct i on 1 e re ma911de
th- Is;ue- oc K- lhokc vi Ives :md -if 10" imer (I iamneter
'-t a inlIcs s stool I p pe The N.,(0 wais routed t hrou(,,;h a ma-,n ifoldI
s~ hat inecessary, it couldl he diluted with inert cIses,
'It55t1rc. A "1'' wais ins~tilIled onI tlhe aidzptor ,)v.eslrc
monitoring porft to allow dynamic andI static prsuesensors
to be attached. Static pressure was measured with a
ihastings 0 - 1 torr ",awle and served as- a vacuum i nte~lritv
chuch. Dynamic pressure was mueaISnrc(I I, th a h aUrn>K
1 pe C p. )r es s u ie g auii,,,e te u~c t h t 11, 1a~ oi 141\e
1) 1 cc se measureffoil )) U ynamic pie s sW I-S 'a" lm.. as 0.0]I
t or r
I V. lxperiment.al Procedurue
Int roduct i on
It is approprIate to treat the experimental procedure
in four sect ions The fir.st, stairting and shut-down for
a flow tube of this type, Is covered as a list in Appendix
A. The second section deals with al i,.nment and cal ibration
of the flow tube and spectrograph, while the third deals
v. ith the requi remient s ot dat a coll _[cl iti in t lIc /(
and Ph/N20 experiments. 'T'lie last si(Ctiu discusses som
particular safety ha.a rds ncouuitercd.
Al ignment and Cal i)rat iOin
Alignment. The tube and scct n iz pli ¢e,. I alt :ned
using a Spectra-Physics 142T) 2 mi', Ilcike l iser . A ini rror
was mounted in the combustion chamhuiir at a 15c ainJe to
the tube axis, and the beam was introdticed tliI'OILjl the
orthogonal viewing port. The exiting heam entered the
entrance slit as shown in Figure 1.1, and the spectrograph
was adjusted for a maximum signal at o328.\. Since the
sensitivity of the IP21 photomultiplier at 6328A is well
below the peak response, it was felt Ilie maxiiiin, the
signal in this fashion would guarantee Pond response between
4000 and 6000 A.
.Ii
Monot'hrornat (w 1
TI r I
I'he ,Jarrel-Ash monochrometer was cal ibrated accordin ,,
to the manufacturer's manual usin - an lltraviolet Products
lIk lab standard lamp. The uncertaiirty in wavelength is0
,,iven as - ).2A by the manufacturier f.ef 30)
The large error induced by the gear lash in the mono-
chrometer required a different calibration technique for
the PbO experiment. The PbO molecule has a very large
number of transitions in the visible, and assignment of
these transitions would be r y Uv di licnIlt ., i Ch an olncertai nt
much ;reater than I A
The solution t,'as to insert a gla:ss slide beticen the
viewport and the lens. The slide was placed at a .15 °
angle to the axis of proplagation, rersulting in a 3'transmission loss of the signal (seI i,,u I?). When the
monochrometcr approached a calibration peal, of the fig
lamp, the shutter would be tripped and a cal mark placed
directly on the spectrum. With this method, the uncertainty
is estimated as IA (from data plots).
During the 12/02 experiment, calibration was performed
by returning the spectrograph to the starting point after
a spectrum had been taken and re-running with the fig lamp
in front of the slit. At pre-selected wavelcngths, the
lamp would be uncovered and the calibration wavelength
recorded. Gear lash on the monocmronoator was estimated0
to add I 3A error to Ihis pr-occliirc.
6-
The w.;idth of the calibration spike at a scanning,o
Sl)peed of, SOfl.,/m in ar1d slit width of 0. 0 11 was aplproximately
L\ F!II, so the total uncerta inty in a transit ion wavelength0
wa .1 2A as ,,iro n by I(q (28)
X calibration + A gear
monochromator lash
+ peak width = 0.2A + 3, + i.oA
- ,1.2x, (28s)
AI iv, nment
An RCA 7265 photomultipl iCI .Zl used for the Pho
experiment. Since the 7265 is very scnsitive to system
noise, the following procedures were used. F:irst, two-
prong adaptors were used to float all electronics, except
for the x-y plotter which was grounded to the building
water system. This prevented ground loops. Next, the
x-y plotter was zeroed with no input , and the 720S biased
at 100 volts. With all light sources off or covered, the
gain on the current amplifier was turned up until it went
to 1/4 full scale, then the dark current bias turned on to
bring it to 0. The voltage was turned up in increments of
100 volts and plotted ajgainst the da i, cu',Ient 11nt il the
tube response went non-I inea r , theni the vol tage t;as
dec rc'i;d by I0)0 voll
II
Data Collection
Niic rowave. Experime ntat ion w ith thle mni crowave was
in two areas. First a variety' of i ',ni tion method-,, power
setti ~ ,and gas flow rates were attempted to opt in li e
thle C I anie . Thec best results; were oht ai ned w ithI ali 01
pressure of 0.5 - I.5 torr with a forward power of 1 00
watts and with reflected power minimized. The Lest ignition
method was found to he tin arc piener. ted by a ih voltage
Tesla coil appli ed j ost UpSt rea1 0of tilie cavity. Pressures
ut le-ss than 0.5 torr (1) g"eneral cd anl ext reicl y hot plasmal,
x i th at tendant coo liii,, diifficul ti cs ,, iii ii e itI p roved very
diff icult to sustaiiin a plasma with ( pruSSures over 1.5
torr. The second area of experiWnt iat ion inivolved optimiz -
ing the posit ion and shape of tilie t iibe t ip Cor the region
where the 09 (1A) left the gl ass tube and entered the
combust ion chamber. Tip shapes evaLnat Cd intded a simple
tubular tip, a fan shaped tip, and one withi a reduced exit
diameter. Tip positions were varied froni 0.5 inches to
6.0 inches from the vapor column. The optii mum H 1-aie was
generated with a simple tubular tip 0.5 inches from the
vapor train. Figure 14 shows some of the tube tips tried.
Iodine Production. 1 2 was produced by enitraining9 1 2
vapor in a pipette with an Ar d fluent. Although other
groups have t r ied methiods stichI as liea Iitig oI I photo0-
exc itat ion (Ref 53), amplle I? w~as not alined if) this experimentl
ninette
'iqure 14a: Tul if
by the vapor pressure at room tciiperatilr-C. The gas was
then introduced to the 0 2 ('L\) streuam %'ith aI pipette.
Flame Tuning. The flame was triim~ed by var~ying the
0 2 ('L) and Ar pressure. An optimum Flame occu-rred at
1.5 mm 0 2and 0 5 mm of Ar. The intenisity iwas relatively
independent of cavity geometries.
Pb + N 20. Ph was heated in an electric furniace (see
Figure 10). The resulting vapor was enitrainied in Ar at
pressures of 0.5 - 20.0 torr. N2 0 was iiitireduced in the
burner ring described earlier at peursof' 0.5 to 20.0
torr. The resiltintr 1H aune wa. le v:ia iight an1gles to
the Flowing gas and fi-ouu anI oin-axil mfeit tlj)-A ica of the
combust i on chamber. Power input to the heating coil to
vaporize the lead varied from 200 - 1100 watts of DC power.
The intensity of the flame was a strong function of the
power input . The brightest flame occurred at the hi ghest
power input.
The Ph atoms were then entrained in Ar. The mixture
was passed through the center of the N.0 ring and N2O was
introduced at various pressures. The N2O was also mixed
w ith an Ar diluent to obscrvc the cl'cct
It was discovered in the course of the I1)( experi-
ment that power levels higher than 600 watts severely
degraded the lifetime of the heating coils. The optimum
value of 540 watts was choscii Ct i fe data uins.
Safety
PbO is listed as a class 3 inhalant hazard (Ref 42),
which means that it is dangerous in milligram quantities,
and so precautions were taken to avoid inhalation. Filter
masks were worn whenever the combustion chamber was opened.
The work area was vacuumed after every chamber opening
and the contaminated areas were sponged. The flow tube
was kept under vacuum between runs.
ifr
V.Results anld, Discussion
Th is ",Oct ioil discuIsses the 11,,S Fl ow tube performance,
the I ,/0' c hemi luminescence experiment , ;1nd( the results Of
the PIb N.-,O experiment. Extensive tabulations of data
from literature arc recorded in Appendix B for the PbO
molecule.
Cas I low Tube Pe rtoriiiince
This sect on d i-;cIlSSCS theIC 110. Of est IIC th ow
tube aind the resf It U XpeCrimelnt S ndrtan to iflpr-ove
flame, intensity ind i-cdiic I~s e11C fn n Ii tevarious
react ions. Table I f(SI I- I i :S t IeC fII oW rube1 perLoIRIICe.
The Reynolds number was evi fla)t (-,I uing ii- LI ( 18) where
the density p is given 1y Eq (2'))
M1 (a1tomic we ight/mole) P (s rzilein mm)
R 'IV K)
(ReF 40:61)
The velocity was determined using the empiricail formula
in EqI (30)
Q P(atmospheric pressure in mmi)%r _ _ _ _ _ _ _ __ _ _ _ _ _---- (30)
60 P (ga)s p re ssure in miiii) A (are a for tube )
(Ref F 3:31)
TAB LE I I I
F. 1 Iiilhe Pe or noi lcC
1:1.0'
RAIL TI BE(c m n ill'R / S1: 11 (To r r) Re (cli) Kn COMMENT
0. 1 SOxi10 152 Laminar Flow
900 2. 0 16.70 2.x] 0 -3 3018 Laminar Flow
1700 3.0 31.55
2,150 4.0 45.57
2950 .0 54.75
3 f)0 , ) 07(.
-I t)O 7 0l 8;'. :%
5250 8.0 97. 5:%750 9.0) 1 ((,. 73
o5o00 1i) 0 1_2'. ,x 1i 1.52-1 Lam i nar I.]o.
61,00 1 l1 12. ?
7200 12 .1 133 .. 1
7600 13 0 141 .00
7950 14.0 i17.5o
8250 15.0 153.13
8500 16.0 157.77
8900 17.0 165.19
9100 18. 0 168.91
9500 19.0 176.33
10000 20.0 185.61 2. 5x 101 30180 Laminar Flow
a All data were taken at 76o torr (30.16 in F1g)
ambient, and 190 C (2920 K)
Meter would not record flow rate.
Page 48 is not missing but is misnum-.17 bered
~MmU*~mZZ
The mean free path, , , was calculated using Eq (22).
Knudsen's numher was calculated usin,t, FIq (2). The results
listed in Table IlI demonstrate that the dCsi ,n goals were
ac h i eved . Pres sures as low as 0.1 torr were recorded and
fI ow twas laminar in ill volocity re i,,nns.
If gas kinetics are done in this system, the velocity
calculated from Eq (30) should be multiplied by a correction
factor ranging from 1.6 to 1.8 to account for deviations
from the plug flow assumption. SolIT's thesis (Ref I7, ")
addresses this issue at length. Section \V[ cOnt ain.
recommendations for an assessment of this correction
factor.
System Features
Various ideas for improving f1ow tube system performance
were evaluated. This section will assess their effectiveness.
Hood Designs. Figure 12 shows the various hood designs
which were inserted above the furnace to shape the vapor
flow. Hood design 12a was effective aerodynamically, in
that no evidence of recirculation (Pbo or Pb. plating) was
seen outside the chimney. Its major defect was that high
Ar pressures were needed to get the Ph vapor up to the
orifice and keep the N20 out. These pressures cooled the
Pb vapor and caused condensation inside the hood unless very
high voltages (16 IS volts) were applied, which reduced
4 1.
the life of the heater w,'ires. Additionall y, the hiIh
currents used created a larg e background glovw which tended
to obscure the s ignal from 60() to 7000A.0
Figure 12b depicts the second hood desig(,n, w'hich
attemp ed to shorten the distance the Ph vapor had to travel
to reach the viewino region. The 1/2" diameter hole was
designed to cut down the background g;low. The poor aero-
dynamics of this design created recirculation zones at
points 1 and 2, as witnessed by, t liu 1 " ]eIc d . .. depo it,. -
I and PbO deposits at 2. It proved imp;Ossihie to uIt the
flame into the vietin.? region with thi, design.
Figure 12c depict-, the :;itccussn 1 i gil. It was
recog.nized that the chimcy t..onld ;ecrate a ruci rculationl
zone, but by placing the tip of thIe chiflfnUy just below the
viewing region, the reaction was forced to proceed in front
of the view port. At low pressures, the flaiie extended
well above the chimney.
Anti-Fouling. Two approaches were taken to solve the
fouling problem. The first was to isolate the viewport by
placing it well upstream of the react ion cliamber. This
approach eventually worked. The second approach was to
use flushing gases to create an air i ndow in front of the
viewing port, as shown in FiLgure 6. Th is appiroach worked
well when the viewing port was well away from the reaction,
but Failed when the port wa; pI ated very close to the
.) I
react ion, as in F iiqure 0 where the view is from 90 to
the tube's axis. An attempt to flow oas directly onto the
v ieOqrt (Fi.gure 7) ent rai ned Ph) and PhO and resulted in
a complete ly opaque coat ing with in 30 seconds. The tactic
of locat ing the monochromator iwiy from the combust ion
chamber did cut down the signal intensity, so further
refinement is required.
I Experiment
It was decided to verify I ch em ilUnino scnCCe in tWo
Iays. First, I)andliead positions for triansitions from the
12 (B) state to grotind twould be c:ilciilated Ironi published
spectroscopic constants and coinpa' d to the data obtained
in this thesis. Second, the relative intensity profile,
corrected for photoinultiplier response ani :ihsorpt ion due
to plating of the view port, would be compared to published
dat a.
Correlations 'ith Predicted Bandheads. Certain
assumptions were made in calculating t lie handhead positions.
Figure IS shows a potential curve for the 12 (B,A,X) states
of the iodine molecule.
Predicting the inter-electronic transitions for these
required a knowledge of which vibrational states were
populated in 12 (B) . Berwent and Thrusi/ showed that 12 ,B)
at room temperature dis k liyed I lie cthitior ;iow'in iii
_ _ _ _ _ _ _ _ _ _ _ _
~I
cm
T(X
V-1'
V=()
F i ,ure I0 (ReF 10 :722) Max imum popii I at ions occurred
bet w e en % = and v 21. TI he relative )opi l 1at ions
were calculated by assumiin,,
N I + I [o-)][N (Reft 1]o:722) (31)V V v V , V V
P~ere R v, is the relative rate of population of the
vibrational level of 1,(B) and the k are the rate
coefficients for de-e,\ ital ion fit,i c to v (K<i 16' 2)
The physical s igni ficance 0C th'c0 relat ive opulaHat oins is
th at, at room telml era;ttlre , tIie ju , o it ' tihe I (I',
molecules will bc in stat es abot v .10, rnd., a substantial
number will be between 8 and 10.
This information helps to pre,ict the bandhead
position. The probability distributions foir a molecule
in the I2 (B) state in the higher vibratioal levels are
highest for molecules on the extremums; i.e., lyin g on the
potential curve. This situation is schematically
represented in Figure 15.
For a transition between the upper and lower electronic
states, the Franck-Condon principle states that the electronic
transition occurs much faster than the vibtational transition
(Ref 5:208). This in effect freezes the k'iirat ional motion
of the molecule while the transit on occurs, so that after
the transition the molecule f irds it.Se I wit Ii the same
internuclear separnt ion r . iI tihe iloer elcct ronic state,
5 0
.in.
50
10
10
.1 I I t
I
this r 00'.. coir responds to a di-t!ercnt p robahility distrilvn -
t ion and i hence a di 'Ferent vibrational state v-"
he p ri~ )il i t t hat a m1 olecule in State (I (B), V ,
will cd zp jn state I,(X), v " in an electronic transition
S . iV k II t Ile 1-.-[ C k - (ond o i fI tC t () 1 . A In l 1• ' ' ' V V
discussiO o o i s hexond the sCoi ' oF thi1 s thesis
!)ut the basic approach is to calculate the transition matrix
element between the upper and lower states for each vV"
pair. A fuller d(scl1' io1nr For thr T, lio ecic presented
hy Tl li nghuisen (l 1 7V).
Table IV show.< s. l cted liajick - :undon factoi. t o" the
transitions v - v ' )u (kc1 47 :153). Ir'oi the values
reported, it can hu :oeli 11at t ransI - Ii ,i:l l(t
which terminate in 12 (X)v' o I , l ely for li. h v'
numbers.
As long as the assumption is made that the upper
vibrational levels are more highly populated, as lherwent
and Thrush have shown, the following correlation scheme
is plausible.
First, the total energy is taken to ho
Etotal Eelectronic + Fvibrational + Frotational (32)
(Ref 5:76)
€ I
"'lAB LUI TV
1:rntnck-Condot 1:Kct ors for
., (P, ,v I )(N ,v = (I
v v l: r,1lic -Coldon v v 1 - ;11 c k - Cnn (101
Fictor ra c t o r
0 0 1.4 x 10 18.0) 1 .2 x 10
1 .0 2.2 x l0 1. 0 1 .4 x 102
2 .0 .17 x 10 2 .1 0 1 .7 x 10 2
.77 -2
7. x 1 7 2 1. 1~i 10n
6( -2)1 3 5 x 10 22. 2 .2 x 10
I.1 x 10 23.0 2.4 x 10
6u 3. 2 x 1 2.1 2 .o x 10
7., 7.8 x 10 15 1 2 .8 x 10
8.o 1.7 x 10 2o.0 3 .0 x 10
9.0 3.4 x 10 27 .0 3. 1 x 102
10. 0 6 .2 x 104 28 . 3. 2 x 102
11.0 1 .1 x 10 290 3. 2 x 10-
12.0 1.7 x 10 30.0 3.2 x 10-
13.0 2.6 x 10 31.0 3.2 x 10 2
14.0 3.8 x 10 32.0 3.2 x 10
15.0 5.3 x 10 33.0 3.1 x 10
16.0 7.1 x 103 34.0 3.0 x l0
17.0 9 .2 x 10 35 . 2 9 x 102
I-
The rotat ional energy will be . iven by
IFro t = B J(J + 1) cm (Ref 5:33) (3)
rot
andi 'riot b)y
A Erot = 2 B(J + 1) cm - (Ref 5:34) (34)
Thbe vibrational enery is iven Iby
Evib i-e (v + 1/2) 2 X ,. K 1/2; (25
(Ref" S'73)
,AI F, 6 (1 - 2X ) (Ref 5:73) (36)
For 12 (B), B = 0.0292 cm e 128.0 cm and
X eWe = 0.834 cm-1 (Ref 29:5-11). For these values,
AErot /AEvib = 0.005. So it is reasonable, to a first
approximation, to neglect the rotational energy for band-
head assignments.
The assumption is made, based on the Franck-Condon
factors, that the transition terminates at v- = 0 in the
I2 (X) state. Then the energies of the states are given
as follows.
- To + AEv- Alv - = 0 (37)
-1
For 1 (B), To = 15641.6 cm 1 X "ae are as given, and for
S(X T X = 0.6127, and 211 .57 (Ref 9:5.1)00
Then v is assumed to be 0, and the energv of the transition
is given by Eq (38).
E (cm-! ) I!.(, + 12 . ((v + I /2}
- 83-I (v 1 2 211 .57(11/2
+ ). b]127(1/2)
S1553.1.6 + 128 (v 1i/2J 0 8.8 34(v'+I/2) (38)
Table V shows the wavelength of the bandheads predicted in this
fashion. Figure 17 shows a data record with calculated band-
heads shown by circles and identified by v. The hands shade
toward red (Ref 13), so the bandheads shonild occur on the
left shoulder or near the peak of the transitions.
Relative Intensity Plots. Fii rlle 18 shows a plot of
the experimental dat a , corrected for piot oiuIliplior response
and absorption, and a plot take n i romh work Iby Thrush (Ref
16:721). ) orrectimin (I oi lv I ii i ( ll ipl rc- I ml :c were made
.8
'IABT V
Calculated I,(B)- I 2 (x) Transitions
0 0
v' (calculated in A) v (calculated in A)
5 01o8 21 5586
6 6123 22 5558
7 6080 23 5530
8 0038 24 5504
9 5997 25 5478
S 9 .957 b 515
11 5918 27 5428
12 5881 28 54 1
13 5841 29 5381
14 58(9 )0 5 358
15 5774 31 533
16 5741 32 5315
17 5708 33 5295
18 5676 34 5275
19 5645 35 5250
20 5615
I.
0~~
0a'
'-fl
In
7.
____________ 0'-
LI
U
~-~.----
C.)
______________________________ _________________ U
I i ptitc 17 I' I 017. 0 1 ~8X. I);i ,.j'~ Ii j it j t. i is
I
// *0<
/£5
/
//
6I
/
A
/
0
cC
NN C
C (C(C C~
N
C.
N Q>)0 ~
oxC0
a)
0 *~
C'3 CCCrc -
0WECCX
4~ CCCE -
a)CC C,
* - ~ .2CC
CCC 0.- "-~ 0a)
I I,
I
from manufacturer's typ ical reSponse Curves (Ref 39.
Initially, it was assumed that the absorption could
be described by assuming that I was proportional to some
I° times an exponential time dependence. After considerable
trial and error, it was decided that the actual behavior
could best be described by Eq (39)
0.4 + [2.3x1- /A] X of A above 5100 Act(0 attenuation) = t
4.5 minutesr39)
o
where t is in minutes required to st.,eej , from '1100A to the
wavelength in question, and the muonJcl'oma tor is sweeping in
the direction of increasing ivave len,,;lb. Ths tile adbsorption
shows a linear time dependence plus a izivelength dependence.
No error estimate is made.
Discussion. The correlation betw.,een predicted and
observed transitions is excellent up to v' 30. Noise
above 5300A made it difficult to assign transitions above
v' = 30.
The relative intensity plots are less convincing. The
lateral symmetry is similar, and the high frequency observa-0 0
tion limit (% S100A ± 50A) is the same. However, there is a
significant difference in the relative intensities across
most of the spectrum, Most of the dilference can be
attributed to differences in the experimental conditions,
since the pressures were diHf'CIrcl (P 0) ; toirr in
this work; 3.1 torr in 1lcrw nit and 'lt-tili (kcl 1h:7211))
(I
I
and teCinpe rat u-re meatsurements of the flow tube a,,ases were
not made in this thesis.
1lowever, based on the correlation between the predicted
.n observed transitions, it ws '*'It that I, chCIi ltminescence
had been demonstrated.
PbO Experiment
Qualitative agreement was obtained with the literature.
Table VI lists some of the observntions made ia a function
oC experimental plartiemkters.
As Linton and I i'uid:t ot ed in tie ii jmpc, 1, the plus
flame at low plressuies was tuitu (ef 1. , I(8 I. It
was determined tha an ii ) oprovent jit iii thl' ,a 1 to noise
ratio of the measurement system was recii'd I o \ i , Id
quantitative results.
A qualitative correlation was obtained between power
dissipated in the coil and the react ion prtuduct-. A t
wattages below 250 - 300 !r, there were no react ion products
and between 300 1"' and 500 W, the major piroducts iwere PhO
'identified by the yellow color) and Pb., plated on the
chamber walls. Above 600 !'V, a black oxide formed; presumably
Pb 2 0 (Ref 23).
62
__ _ __ _ __ _ __ _ _
TABLF. VI
PhO Flame Observations
N 20 Pressure Ar Pressure Voltage Color CommentsFlame
0.5 6.0 13.0 Blue Faint but well
defined
1.0 4.0 13.0 Blue " if
1.5 2.0 13.0 Blue It
2.0 2.0 13.0 Blue to
4.7 1.7 13.5 Blue Very faint
0.1 4.0 14.0 Blue Excellent flame,little fouling,bright, well-formedflame
4 - S 0 - 2 14.7 Blue Well-formed flame,much black oxide
2.0 10.0 14.0 Yellow High following
2.5 4.5 14.5 Yellow Wire burnt in two
10.0 15.0 14.5 Yellow
10.0 20.0 14.4 Yellow
30 Off-Scale 18.0 Yellow
7 .,
V I . on Clus ionls a fld ZccohiiwiI dI t ions
Flow Tube Performance
A working gas flow tube was demonstrated. Experiments
were conducted in the laminar flow reg ,ime from 0.35 to 30.0A
mm pressure. The capability to do oxidation, excited gas,
and high temperature metal vapor reactions was demonstrated.
Exper iiients
12/02 Chemil uminescunce. This e:Xperimc unt demonstrated
the use of a microwave power soillrce to ,;olu rlte s inl lct
molecuilar oxygen. The experimtent rcpr(dncd data acquired
by previous researchers and a-rc eed hi } tlhcIoruvl ica] calcula-
tions of predicted emissions. It was concluded that the
emissions were consistent with emission from the 1,
-. U: ) ) transition in iodine.O+U g
Pb + N10. Severe problems with gas d-nat ics and chamber
fouling were solved, and a qualitative observation of the
flame was made. This observation was in agreement with
descriptions of the flame in the literature. Equipment
problems with the measuring system prevented the acquisition
of quantitative data.
11
I
Recommendat ions
Several improvements to the flow can he made , and
follow-on experiments are suggested in this section.
Gas Flow Tube. A length of 3.0 inch inner dliameter
transparent tube should be purchasedl for installation in
the transition section. The tube should bie of quartz
because of the superior transmission properties of quartz
in the ultraviolet 'pecti ruml. ThiMd tilct loll ti llu.
gas 1,inetic studlies to he performedl in fis tube. lIOCtrod(10
should be made for- the combust ion chaiiihcr to ailloluv' spark
discharge studies to he peCrforme1d.
Furnace. Additional hood-. shapssUold he invcLstigated
such as the one shown in Figure 19. 'Ihle Ioal should he to
Figure 1 9. Proposed Hood Sliiape
obt ain the most eff icient Flow of metail into the combust ion
chamber w~ it h thle lecast depos it ionl in theo furnace. 1- i c-ure 1)
s-hows. proposed method for e 1 in i t i ng rec i rcu I at ion in the
fulrnacechabr The hood should be confi gored with a
heat ing' wire onl the outside of the hiood. Tb lis desiicn would (
allow the hood to be heated and thus prevent plat ing of the
vapor from the furnace on the ins ide of the hood.
Another possibility is a new furnace designed such as
the one in Figureu 20. 1,11 (Uht Cut WO~l d I)C to av o id re-c i rcu -
Lit i on or combust ion aheoi.;i of t hec v ic . iig a t-a , ei 10 s imp I i-
f ) i n - r echa rgoimn'.. 1 r,)c i i-e s atnId a v o i, i 1) foalII in1 g 1 1' rob e111S .
NMeas u remn c ) ii cem i tc. 'HeC mlost ohke Ion MISa for
imp roving the fl1ow tube .ouk 1b e t o omprove- tie, i~wa ii rement
system. Two approaches are recommenided.
Viewing Arranglements . ['lie p resecnt a rrange -
ments require a viewing path of 2 to 3 feet tn atvoid plating.
This cuts down the signal intensity, and so at way to shorteni
the view path is required. Figure 21 shlo%,s, an insert to be
placed in the tube ahead of the vietc.ing iindoi , ei ich would
create a uniform gas blanket over the window to prevent
deposition (Ref 13). This should allow the view port to
be placed adj acent to the combust ion cIhamlheir.
Pl'Ieri studying the ra 1 e ccif Fivient s of a
react ion, it may not be possible to tie, at flush inci, "s.
For these ci rcuimstance5 , theL 11odii teaL b in hi -nc 2 is
o o
0 cdl ~ ~ <1 ute
nea ter N
wire matii C v.a. )Liz(-d
T'i rillro( ): !Prcncp~isd II 111(
- - 0 1
- . . ....
f, L k
proposed . The obh ect would be to prevent deposition of
comtpounds such as I , which have low vapor i zat ion temperatures.
Flectrical Noise Reduction. The electrical no;se
problems encountered on this thesis shoul! be solvahle.
ne possih~ility is to electrically isolate the photo-
multiplier from the monochromator. According to Dr. ;on Rob
of the APIT faculty, this is a major source of noise. A
second possibility is to cool the 7265 to reduce thermionic
cmission. This would requ ire Fabrication ()I ' cool i I..
coll ar and selection of a cooling meII dium.
Experiments
PbO. With a neI% icasuremntI :<yst cm av~aila l, research
should continue on the PhO react ion. A possible approach
would be to enhance the PbO spectra by excitation with
excited nitrogen or singlet molecular 0-,. Linton and Broida
(Ref 33:409) showed that the addition of active nitrogen to
a Pb + N 20 flame strongly enhanced the A - ,, B -, y, C - X,
and D - X systems, with some individual transitions between0
6000 and 4000A being 30 times as intense. Additionally, N1
excited some atomic Pb transitions; probably the 'S0 state.
An obvious approach would be to use the microwa:ve to produce
active N2 or 02 ('A or 1F) and record the spectra. Then
radiative lifetime measurements ca be oak ZOn lhe most
strongly excited states to deterline if' a likely candidate
v-,tem exists for establishing i ilaptiktion iiiversl on.
0m)
An add it ional i arca for invest igat ion was suggested
hy Davis (Ref I1). The spectra enhanced by act ive nitrogen
(Ref 33:408) show evidence of excitation to excited states
of the Ph molecule. Divis suggests that the PhO + N* reaction
be used as an energy transfer mechanism to excite the
3P1 or 1D P - it omic transitions of Ph (Ref 13)
With rate constants of 88/sec and 26 sec, respectively, these
metastable states may be good candidates for a visible laser.
Testing this hypothesis would r'ijti ire u ri ji' -t.
rate constants for the react ions,
P) + N ? ) I'h0 + (li)
PbO + 02('/ or 1T) Ph 0 0, . . (11I
PbO + NT Pb* + N + 0 (-12
and other paths.
[ Chemiluminescence Models. nl Iodels have been2'
proposed to explain the 12 chemiluminescence. There is
general agreement that the bright yellow-green glow resulting
from introducing I molecules into a st ,'aLim of (K ) and
02('F ) molecules is due to the transition (Ref 1,:720)
2 0+u) - 12(I(I) + v SO5 ) A (44)2/ 11
6ri
1lowever, there is no general agreement on the nature of the
exc itat ion mechInism. Arguments put forth hy Arnold,
Finlayson, and Ogrvzlo (Ref 2:2529 represent earlier views.
lowever, later work 1 Perwent and Thrush (ReF 16:721)
represents a radically different view of the riechan i si.
Arnold et al. (Ref 2:2530) hypothesized a recombination
scheme:
19 ( 1 A) +,(1 1 ( +12)
2tl + 12 r +2l-i
I (\ I - J. I (,15)
This reaction, as reported, requ i .:- ,i t h hrcc hody reaction
to proceed, result ing in a second or third ) rJ(.r i itensity
dependence on the concentration of 0,)''.). lhis rielation-
ship does not appear to h1old , based on t ]wii , L done by
Detwent and Thrush.
Derwent and Thrush report a secC:o nd 1,ss i i it)', that
of stepwise excitation via the following processes:
S+2(?g) = 12 (A 3'r +u + (.10)
I2(A3 u) + 2(1,,, ) 1 ) 1)o (17)
2 +u (I
Their argument is based on the vi brat i ona I I eve 1s popul ated
and the spatial distribution of the luminescence in their
flow tube (Ref 16:115). In a later paper, !)erwent et :l.
present evidence that the I., cem ilninc.scen ce e en On
on the first power of the iodine concentrat ion (i . ., :1
plot of relative intensity of the ch emilvluinescence v.
pressure is linear) (Ref 16:724). Based on the l;near
behavior of the intensity, Derwent and Thrush proposed the
fol l , .in, total Imoc iait i si! t i) .CCc)llill I )I, tlic J' hr i It' u iI ', (
12 + ),( l' 3 Ii + , I.) (is)
I2(A) + 0,2 A ) I,(") + ( (51t
S1(B) = " (X) v 51)
I2(B) + 02= quenched prodicts (52)
1 2 (A + 02 = quenched products (53)
Russell and Simons (Ref 41:271) showed that the process
I + I + 02 = 12 + 02 is too slow to compete with the
diffusion controlled wall removal ()I iod inCe, :and so does
not contribute to the reaction r;ates.
li
The various theories of this chemiluminescence need
to be investigated. If the hypotheses of Perwent and Thrush
are correct, the chemiluminescence is the result of
01('11) + 02 (1',,) 0, (1 (54)
+ I1(X) , 1 (A) + 0 2 (:j (55)
O2.,(1..') + I)(.\ ) Iz ~ 1 , ) + iN) ,(7,€)
This contention couild be tested a:; 1l 1 I . :i. i -'st establisli[02) .)]. The [,( c the
inten:;itv of 1 .27 i - iatiol tiom I , 1 j 'ith tl,c 1,tal
gas flow (Ref 48:29,). If lOerwent :iiI hitnw :b a, ie corrl'ect, the
rate of formation of 1 (A) should dcpeiiJ on thti second powei
of the [0.'(11t)], or the first power () I- 1 lI. ihe [0,(]?)]
can he determined by ohserving the intensity o- cmission at
1.91 'P. If the chemiluminescent intensity is dependent on
the first power of 02('A), 02 (W) , and 1.(A, this would
be excellent evidence for the theory '1) Iierwnt and Thrush.
G;,s Dnamics
The flow patterns in various par'ts oF the flow tube
should be observed by means of smoke st l'earn in oider to aid
in the design of a more efficient flow tube system.
I)=
" II a " I I I I II I . ... I I" ..... . . . .. Il .. ... : -- . = --''..- , - -
Bibl iography
1. Anderson, R .A., I,. lanko, and S.J. Davis. "TiMe Resolved.,+
Fluorescence of the A"' State of (;ef." Jot ri.i I of
Chemical Physics, 68: 7 (April 197S8.
2. Arnold, S.J., N. Finlayson, and I.A. Ogryzlo. "Some
Novel Energy-Pooling Processes Involving 0
Journal of Chemical Physics, -14 : 2529 (April 1966).
3. Arnold, S.J., E.A. Ogryzlo, and I!. Witzke. "Some New
Emission Bands of Molecular Oxygen." Journal of Chemical
P lvsics, 40.: 17w) ( i rch l) l .I
4. .Atomic Energyv Levels, Vol. III. 'ietil ai -1(7 - ,1II7C- NtS
l i' ishington L)C : Nat i(.) na I Butr ea:it o " t iid i - , , )
5. BPanwel1, C. N. :li.tm, at Is HI -cUllii ". I I',Copy
(i:econd Edition) lo nd,'il: : v ' i I, I )'72.
6. Barrow, P. . , V.V. Va vt li. '1 ; ssa nh
Spectra."- Procecdin :s of the II, ,: ,' v ,I london,5b : 20,9-211 ('lay 19 14 .
7. Barrow, R.F. and F.1. VaLo. "llt ri -violct \hbsorption
Band-Systems of PhO, PhS, PbSe, and li .'' 'r,,cedi n!s
of the Phys ics Soc iety of London, Ml: 3 I'.' nl 1 %)-) .
8. Barrow, R.F., P... Fry, and '". Le yarg,. "6R t at i ona I
Analys is of the Absorpt ion Spect rlii o PC 1." Proceed -
ings of the Physics .Qoc iety of Ilo ,do ii, S1: I 3 \pr i 1
1963)
9. Barrow, R.F. Introduction to Molecular Spectroscol-'.
New York: cGraw-luill Publishing: 6j., Itlo,.
10. BI oomentha I , S. "Vibrational ( a ntim Analysis and
Isotope Effect for the Lead ()xidlu fijd .'ct r-a."
Physical Review, 35: 34- 15 (.a iairy 1930).
11. Brown, R. L,. "Tuhular Flow Reactors with First-Order
Kinetics." Journal of Research of the National Bureau
of Standards, 83(1): 1-8 (Jan-Feb 1978).
12. Bugrim, H.N., S.N. M4akernko, and I.L. Tsikora.
"Efficiency of the Vibration Deactivation and Quenching
of the Electronically Excited I 9 Molecule." Optical
Spectroscopy, 37(6): 610-612 (December 1974).
13. Davis, S.J, private communication, 1980.
1-1. Davis, S.J. and S.C. !ladley. "Measurement of the
Padiative Tifetime of the A rv'>() "'tate of gi F."
Pl ys ical Rtv i c,. A, 1 1(3 ) I C I( I l &mh I 1 ,17 6 .
15. lerwent, R.(;. mid L. r. TIui'is . oLSI !! t ' l I I , 1
-S:ind 07(17 v+. , ) in M ~schm lrce' Fi1 , :v t (,.111 '1 1 itn ; J i ois
,f the Faraday -Aciety, 07: 2U , - .I.- .itI v 1 1) 71
16. 1lerwent, R. I. aid . . ru, . '"Txcitat ionl of Iod,,ine
by Sin lest Molecuilar Oxy'jtn." i:u .aa, lills (, the
I aradav Soc iety , 721H- '7,' :.
17. Derwent , R. ., ). R. Kearns, anlid ..*\ . lurir;. . "The
Fxcitation of iodine 1y Sin.c_'let ,, l culIar 'Cx\.en
Chemical Physics Letters, 0(2): 11 S-ll1 , (CJulIi 1970
18. 1lectro-Optics llandhcok. RCA Defense IElectronics
Products/;\erospace Systems Division. Mass:iclusetts
1968.
19. Dorko, Ernest A. Lecture materials distributed in
Pll 6.61, Methods of Spectroscopy. School of Engineer-
ing, Air Force Institute of Technology, Wriiht-
Patterson AFB, O1, 1979.
20. Evenson, K.. , J.L. Dunn, and 11. 1. Moida. "Optical
Detection of Microwave Trans it ion:, let ween lxc i I ed
Electronic States of CN and l Ie Idleni i licat ion of the
rransitions Involved. " Phys ic. I icvie_ , , 136(A)
Al 566-1 568 (l)ecm l),' 19'4 .
L 1 , .
21. Fontij n , A., W. Felder, and . J. loughton. "Tubular
Fast-Flow Reactor Studies at High Temperatures.
Kinetics of the A1/0l Reaction.' Chemical Phys ics
Letters, 27(3): 365-368 (August 197,1).
22. Fontiin, A. and S.C. Kurzius. "Tubulnr :nst-FlowReactor Studies at High Temperatures. Kinetics of
the Fe/O 2 Reaction at 16000 K." Chemical Physics
Letters, 13(5): 507-510 (April 1972).
23. Fontijn, Arthur. Kinetic Spectroscopy of Metal Atom/
Oxidizer Chemiluminescent Reactions for Laser Applica-
tions. Interim Report for Period I May 1076
.5i April 1)7/'; . c.io lie 'IX-IT'.;. C'.as ii n>.t ,o IW "
.\ir Force Off ic- f Sciult i l'ic kcsc' irchi , toil iilo AFBIC.
24. I:ontijn, A. and V'. eldei. "Iliie jiil l cc last-Floi.
eactor Study of Si /N, ('lCi i Itii Iinbccicc . C II ica I
Phvsics Letters, i. 1 ): 311- 102 ( I , 972i.
25. Fontijn, A., et al. '"'Tih uir is), I Lw W:ictoi- for flili
Temperature Gas Kinetic Studio,.- ' ',cvi'.. , ".cientifi,
[nstruments, 43(5): 720-724 ( 1i1v 1 721
26. Hlagar, C., et al. "Reactions of Atomic Sil icon and
Germanium with Nitrous Oxide to P'roducC llectronically
Excited Silicon %.onoxide and Germanium Ox ide."
Chemical Phvsics Letters, 27(3): .1)- .111 .\mWiust 1974).
27. Hagar, G., et al. The a . .,. antd h
Band Systems of SiO and the a'7 + X'- Band System
of CeO Observed in Chemiluminescence." Journal of
Chemical Physics, 63(7): 2810-182() (t)ctolicr 1975).
28. Handbook of Chemistry and Physics, (Fiftieth Edition).
Cleveland: The Chemical Ruldbct- (ompany, 1970.
7,
I
29. llerzber C, C. ._pectra of l)i atoic M.olecules, (Second
Edition). New York: Reinhold Co., 1950.
30. Instruction M anual , Model 82-0()0 0.5 Meter bert
Scainnin, Spect rometer. ,laurel-Ash Ilivision/Fisi'er
Sc jenti fic Co., WZ'altham MA, 1971.
31 . Thermo-Chemic. I Tables, Second 1d it ion. t(SDC-NBS, JANAF
I'ashington DC: National Bureau of Standards, 1971.
SN0303-0872.
32. LeRoy, Robert. "Spectroscopic Reassignment and Ground-
State Dissociation Energy of Molecular Iodine."
Journal of Cicm ical Ph) : ics , ("1 !u 7 ; 2o 7" (8larch
197 0).
33. l.inton, C. and II. P. Broida . "Chemi lnuuinc-ce ot Spectra
of PhO from k ea c t i on.; of Ph .A t oi> . ' .1( 1 in l 1 1) f
Molecular 1' 0ecti c t , o 1 (.iepteluer 1 976).
34. MIcDermott , W. 1:., N.P. Pchll kin, 1)..J . ce na vd , and
R.R. Bousele. "An Electronic [mans it ion (Cule ical Laser."
Applied Physics Letters, 32 (8): *hl,- 17 (April 1978).
35. N air, K. P.R., R . B. DingIi , and 1). K. ;ii. 'Pot ential
Energy Curves and Dissociation Ine rins of Ox ides and
Sulfides of (,roup IVA Elements. " Journal oI- (ohemical
Physics, 43: 3570-3574 (November 1Oh5).
36. Oldenborg, R.C., et al. "A New lectronic Rand System
of PbO." Journal of- Molecular , _ctroscopv , 58 : 283-
300 (November 1975).
37. Oriel Spectrophotometer Calibration Set Model C-13-02.
Product Brochure. Stamford C'! : Oricil Opt i- Corp.
38. Pennucci, M.A. "Parametric Evaluat ion of Total PressureLoss and Recirculation Zone length in a Sudden Expansion
Combustor." Unpublished MS; Thesis, School of Engineering,
Department of Aeronautics and Att i'onaiitics, Air Force
Institute of Technology, Wriiht -Patterson AFB Ol,
September 1974.
/I'
.-.
39. RCA Tube Pandboot, !IB-3 . -arrison NJ: 1967.
40. Roth, A. Vacuum Technology. New York: North-Holland
Publishing Company , 1976.
41. Russell , K. 1. and J. S imons . "''tlmdies in .er' v
Transfer. ,- Thc Combustion of I Atoms." Proceedin,,s
of the Royal Society A, 217: 271 (April l953 .j
42. Sax, N.I. Dangerous Properties of Industrial Material,
Third Edition. New York: Reinhold Press, 1968.
43. Shawhan, E.N. and F. Mlorgan. "Absorption Spectrum of
Lead Oxide." Physical Review, 47: 377-37R (March 1935).
4.1. Steinfeld, J.1 . ".loleculcs and Rladiation: :\n Introduct ion
to Molecular Spectroscopy , ___X . (iiil rid. MA: I ]T
Press, 1978.
45. Suchard, S.N. S JLct I2OCO: iC IosA Ic" I' , a 1r Uc rLonuc lear -
D)iatomic Molecule , Vol . I 1. P '1 eni, Pubi ic:t ion:;, 1975.
46. Swearingen, P.M., et al. "Reaction Rate stud ics of
Atomic Germanium ( P0 1 ) and S I ili( ( 1 'I ,i th Various
Oxidizers." Chemical Physics letter'-:, i 2, ) : 274-279
(April 1978).
47. Tellinghuisen, Joel. "Intensity FV!ctrs for the
I2 (B) - 12(X) Band System." Journal of Quantitative
Spectroscopy and Radiative Transicir, 1): 149-161
(June 1977).
48. Thomas, R.G.O. and B.A. Thrush. "Vluergy Transfer in the
Quenching of Singlet iolecular Oxygen, P'arts 1-III."
Proceedings of the Royal Society of London A, 350:
287-314 (September 1977).
49. Thrush, B.A. "Atomic Reactions n ilow Tubies." Science,
154: 470-473 (April 1967).
7:
50. Traimar, S., D.C. Cartwright, and W. Williams.
"Differential and Integral Cross Sections for the
Flectron-Impact lxcitation of the a1 A and 1Z+ Statesg g
of 0) ." Phy'sics Review A, ,1(4): 1482-1192 (October 1971).
51. Vincenti, W.C. and C.11. Kruger. Introduction to
Physical Gas Dynamics. New York: Robert E. Krieger
Publishing Company, 1975.
52. West, J.B., R.S. Bradford, J.D. Eversole, and C.R. Jones.
"Flow System for the Production of Diatomic Metal Oxides
and Halides." Review of Scientific Instruments, 46:
164-168 (1Iibruar; lv)
53. 1l'olf, Paul J. The Validity of- Velocity Cillcul ations
Based on the Plug Floq Assuotiitjn in il 0:1 TUhe
Appl icat ions. MS Thesi s , Wr i ght - Patter,;on AF B 011:
Air Force Inst itute ofI Technology, Deceiihe r 1.979
'/ 9
APPNI)IX A
Flow Tube Procedures
Start Up Procedures
(eneral Procedures
1. Close all vacuum fittings.
2. Shut off ball valve to vacuum pumps.
3. It using furriace, ceplace glass u:ool inpart icul ate trap, chiarge furnace , i thmate:ial to be vaporized.
4 . 'ui i vactiiiii pumps.
S. Slowly open h~all :Ilve.
6. Turn on vacutm pump to rele'encc- sideof Baratron gauge.
7. When pressure in Ht[, ttube (static gauge)is 200 microns, open vaive to Baratrongauge and pump for 1 hour.
8. Start Bake function on Baratron at 5,continu-till red bake light uoes of-f,then set Bake function switch to regular.
9. When using materials which must be platedout before reaching the vacuuim pumps, preparethe cold trap.
a. When using oxygen, use a methyl chloride/dry ice slush to cool the cold trap.DO NOT USE LN2 AS THIS WILL CREATELIQUID 02 IN THE COl.) T !A!
b. With other react ions , use LN2.
8
. icrowave Procedures
1. Turn on flow of oas to he excited (0.5 mm!or (12.
2 . TuIrnl on microwave ,enerator.
3. After 3 minutes, turn on forward power to!)1 ,. If plasm- does not iI'nite, use aTesla coil just upstream of the microwavecavity as an additional electron source.Tune with coarse control until plasma isign i ted.
4. P)) N OT aI I Jov povef to , i;i I,: n o i l ,. I ,i th ; ii nut-es with tl e Sh.k 3, 1,here
I + V 0 lorta Ed p,(ser
S R- re fiected po\'er
1.0oVWa 1'd poJwer
5. Turn on microwave cooling ,,as. Adjust flowuntil exiting gas is slightly, vwarm to touch.
Furnace Procedures
1. Charge with material to he vapor i zed.
2. Turn on cooling water.
3. Ensure tube is grounded
4. Turn on carrier gas
S. Turn on power.
6. Adju. ,i cnt t' ) c d t a:1' p V I)O' e,,lc rature.
Shut Down Procedures
I If us ing furnace, turn off power. Continue carrierClow and water flow. Short tulhe su r.fa:cc with( 1' 0oU d wire.
2.Shut off power, IIV, aInd power- on switch on microwave.
3. Shut off active gas.
4. Shut gauge connecting Paratron to system.
S. Shut off carrier after 5 minutes.
h . Shut o f w a t CI a t cI I S !i [ l_;.
7. Shut haIl val I ve o I , vellt llis ttLroiih puI,vent valve.
S. Ensure a I I 1,.,r off.
I, Leave tub, under Va;ICumm,
DO NOT TURN ON FURNACE W''ITHOUT (O(M1 1 .';At 0,N;
DO NOT TURN OFF COOLING WATER BlIA:0WtRl '1IURNI'I(; 01 [uFF FUIRNACE!
AD-AO9" 94 AIR FORCE INST OF TECH WRIHTPATTERSOtd AFB OH SCHOC-fC F/0 7/4A GAS FLOW TUBE FOR SPECTROSCOPIC STUIES.(U)DEC 80 V R KOYN
UNCLASSIFIED AFIT/EFP/P/80-5 MLBBIIIIIII
I I i 'V
J 1111
I!
AIPEND IX (
Useful PhO Dataa
+ 4
I. A+ == X'T+ System
Bandheads in emission
v,v x Intensity v',v" x Intensity
3.8 6433.63 3 0.4 5910.74 6
0.6 6427.73 3 0.3 5677.78 6
2.7 6342. 1 3 . 01 17-;, ) I( .6 "
1.6 620 .75 5 .2 54, .38 ,
0.5 6160. )52 1 1.2 53-1 11
II. B1 == XzE+ System
Bandheads in emission
v , v x Intensity v',v-" A Intensity
0.5 5353.82 3 0.1 1657.98
0.4 5162.31 6 1.1 4553.71 o
0.3 4983.79 6 1.0 4410.38 5
0.2 4816.90 6 2.0 1317.0O ,1
3.0 4229.01 4
III. CO E System
Bandheads in absorption
v,v_ x Intensity v',v" A InIensity
2.1 4156.20 3 1. ( 3877.852.0 4037.63 3 I 38 8.21 24.1 3987.70 , 5 1) 38 (H .913.0 3955. 04 7 0. 37, .91 15.1 39101.31) 3 7 3o -
. .n
IV. C1j X17 +IV. C'_ = System
In absorption, the bandheads Corm a sing;le intense
progression with v-' = 0. v' numbering is deduced
from isotopic studies.
v v " 6.0 7.0
3612.8 3554.8
V. D1 == XIE + System
Bandheads in emission
v"v' 0.2 (0.1 1.1 1 .0 2.0!
x 3485.t)8 3-401. 9 2 33-11.8,3 3261.36 3 09 2
Intensity 0 S 2 2
VI. EO+ == X2F + System
In emission, bands with v' I are not obserked.
Bandheads in absorption
v_,v__ A Intensity v',v" A Intensity
1.3 3062.67 4 3.2 2925.64 3
2.3 3023.38 2 2.1 2 9(0 0.21 4
1.2 2998.52 4 1.2 2891. 21 1
2.2 2960.73 3 3.1 2860.17 5
1.1 2936.19 2 4.! 2836.57 1
3.0 2808.5 1
a 0
a in8
8 -t
S PLI'TROSCOP I C CONSTANTS
: 1a :I -!1 Ch
State T X 1k' ex 10 1) ex 1r c
34155 451.1 6.95 01.12! 2.6 0. 283 2.165
D1 30194 530.4 2.9 0.2710 2.8 0.283 2.047
C1I 24947 494 3.0 0.2491 1.8 0.25 2.135
CO + _3820 532 3.9 0.2545 2.1 0.25 2.112
BI 22289 48( n.2619 7.6 n.3q 2.071+
AO+ 19862.3 444 .2 0.4o 0.2588 1.1 . . 2.(),)5
X1 + 0 721 .46 3.53 0.30751,} 1.01(7 0.22 1.92181
Dissociation energy -3.87 0.05 cv, 89 kcil/mole, 31211 cm 1
(Ref 44)
a -1cmi
bIev
C0
cA
h7
APPENDIX 1)
Response Curves for Photoultiplier Tubes
Th is section sumnma r izes son1e important chatract cr i stics
of the photomultiplier tubes used in this thesis. The
accompany ing graphs are typical re lat ive sens it ivi ty curves,
and actual tube response may vary
1P21
Typical anode sensitivity 9 4 nn.0 -- . . 1. I ip ,/!,;
Typical cathode sein.itivitv lliA--------. .t 1 ),1
Current amplification- . .-----------.- . x 0( A/b
Anode dark current ------- ........ - - . 1 '
Equivalent anode dark ciirrc ,nt ilipul t . I. 5.( x 10- 13
Equivalent noise input ----------------- .1 x i in
72650
Typical anode sensitivity 9 4200A -------- 3.10 x 10 A/Wo
Typical cathode sensitivity @ 4200A ------- 0.0t1
Current amplification ---------------------- 1.8 x 107
Anode dark current ----------------------- 5.0 x I0-1 3
Equivalent anode dark current input------ -1.2 x 10 W
Equivalent noise input ------------------- 2.1 x 1(is
SFi.,ure; t ke ri 1rom l , r' le o i.
or 8
Relative Sensitivity % P2.P!
i~o
.30r 3,)00 AO(J() (III J'U000
Relative Sensitivitv t
10
0.1
2o n ,n ,(I(-it)-()110 Pn-n 01
V i t'-a
Vernon Raymond Koym was born 17 August 1.951 at
Sinton, Texas. lie graduated. from Callalen H~ig h School in
Corpus Christi , Texas in 1969. lie aittended Texas M
University From which hie graduated with ai Bachelor of
Science degree in Physics in 1973. Upon graduaition, he0
was commissioned in the United States Air Force thirough the
ROTCr~ rogram at Texas .M. Ife entered thc AITT 1 rhon of
Engi nc rngin Junle 1971. I lie is 11a tr1ied toutelutc
Daria Lynn Starlks.
Ic rmitiiat Addiness llox 121 A, Route 6
bieQscot t , Arkanfsas
UNCLASSIF I EDSECURITY CL A,',I IIC A I IN OF T11'I P A(,f th ( II,. t'a ll. er
-REPORT DOCUMENTATION PAGE 1 j. ,()k<l. (A)IP'YIlNG k()IRMI fi NPOH NIJIMI 11 -7 , V I A' I I3 III I t'1, T", I-ATAI I1 1,11011 N
4 ITE U Pi[/R0- __l"11 TYI-1- OF RUPORT 4 A ID OEE
A (,ASF-01 5S) FCT1ROSCO I C I!) I 'S MS TFIesis_____________
0# A ['T U1SAF
9 E F ORMIN 0 ONOANIZ Z , ION N AM IAC I ADDIINI 'A M~ F' ' 4 TI P IT I N F P c T TA K
Ar Force In stitute o f Technolot)y (AU IT-EN) elrt,. l JNI AC j ~.MULj S
Wright-Patterson Air Force Base 011
II CONTROLLING OFFICE NAME AND ADCIRES 1 2 REPORT DATE
December 1980
7_4 -M),JTORING AGENCY N AME d gCl~Iif~ % a ',l T - > ~ a
I INCL SsI FIEI)I., A N 5
16. DISTIIUTION STATEMENT I
Approved for Public Izelease; Dlist ribution Unlimited.
7. DISTRIBUTION STATEMENT r~f th, abstract -rtere~l h, FtcRI .1), it flya, r.R.r
8SUPPLEMENTARY NOTES Approved for Publ ic Release; lA1W. APR 190-17
FREDERICK C. LYNCH, Major, USAF 06JN18
Director of Public Affairs ________
19. K EY WORDS (Continue on reverse srdt, if necessary and idenitti by I.), A numbr,
Flow Tube Iodine
Fast Flow Reactor Lead
Chemiluminescence Lead Oxide
20, ABST RACT nnrr~r~ lr Oee esd. If r,--es,-y) tnd1 idrrnifc I." hl", 1, rr 1-t Ir
A gas f low tube was construc ted t o a I Iow chem ical1 react i onstudies at pressures f rom 0. 1 t o r' to )7) 0t orr. The lube was
(esi gned to allow i utijoduuct ion of1 )xidant s, dient s, or excitedgases. A furnace was, construtct-ed to pi'odiuce hi h 1 Clii)ratu ye vaiporin the combustion chamber of the Iflow tiube.
T[he flow tube Wa1s rcbaract 'r i z'd Dh\ its p)iiiijsul, speed, 1li-essulietliromtbpllt , and ev(I ii,it mi t ' c llill l 1
T. ii111iiL'i ' Ii biiers '111
DD 1473 ' I -ION Of1 tarv,1 I'NC 'IIYu II I I(LI)
SICURITY CLASSIFICATION OF TNIS PAGE(IWh., D.t. Fntetd)
BLOCK 20: Abstract (Cont 'd)
pressures from 0.1 - 20 torr. Two experiments were conducted toverify the flow tube performance. Singlet molecular oxyien wasproduced in a microwave discharge and reacted with gascous 1. toyield chemiluminescence from the transition T,,(B) , ,.
The spectrum was recorded and the hnadheads assigned to vihration-al transitions predicted by theory. PbO was created bv react inl,vaporized lead with N O. The emissions were compared withliterature.
lei
J
II(I,\ ; II III)l
__ _ __ _ _ __ _ _
we
