N A S A C O N T R A C T O R

R E P O R T

LOAN COPY: RETURN 7 0

KIRTLAND AFB, N MEX AFWL (wm-2)

MEASUREMENTS OF NEGATIVE ION PRODUCTION I N ATMOSPHERIC GASES

Prepared by Dolores E . Ali, Consultant

Berkeley, Calif. for Awes Research Ceuter

N A T I O N A L A E R O N A U T I C S A N D S P A C E A D M I N I S T R A T I O N W A S H I N G T O N , D . C. O C T O B E R 1 9 6 9

https://ntrs.nasa.gov/search.jsp?R=19690029179 2018-05-21T03:56:41+00:00Z

I TECH LIBRARY KAFB, NM

00b049b NASA CR-1406

MEASUREMENTS OF NEGATIVE ION PRODUCTION

IN ATMOSPHERIC GASES

Distribution of th i s repor t is provided in the interest of information exchange. Responsibility for the contents resides in the author o r organization that prepared it.

Prepared under Contract No. NAS 2-5209 by Dolores E. Ali, Consultant

Berkeley, Calif.

for Ames Research Center

NATIONAL AERONAUTICS AND SPACE ADMINISTRATION

For sale by the Cleoringhouse for Federal Scientific and Tochnical Information Springfield, Virginia 22151 - CFSTl price $3.00

MEASUREMENTS OF NECATNE ION PRODUCTION

I N ATMOSPHERIC GASES

by Dolores E. Ali

SUMMARY

Two electron capture by H? and He+ in coll isions with various gases has been investigated. The method used t o measure effective cross sections for the charge changing reactions al lows direct determination of the energy defects. Preliminary cross sections were obtained for two electron capture by H+ at 0.28 keV and fo r H+ and He+ at 2 keV. In addition, negative ions formed by electron bombardment of gases in the ion source were mass ana- lyzed; t he r e su l t s of this survey are summarized in t he Appendix.

The charge exchange gases used in the cross section measurements were H2, N2, 02, Kr, and Xe fo r H+ at 2 keV; Kr, He, H2, N2, and O2 fo r He+ at 2 keV; and Hz, 02, and Xe fo r H at 0.28 keV. The corresponding cross sections range from lom1’ t o cm2 fo r H+ a t 2 keV; t o cm for He+ at 2 keV; and t o 1O-l’ cm2 for H+ at 0.28 keV. Finally, an attempt was made t o observe two electron capture by a t 2 keV; the resu l t s were inconclusive.

+ 2

INTRODUCTION

Current i n t e re s t i n t he production of negative ions derives from photometric studies of high temperature plasmas i n shock tubes and in arcs . Analysis of the continuum radiation emitted by nitrogen plasmas, fo r example, has led to the suggest ion of a s table N- ion (ref. 1). Hawever, most theoretical calculations by means of isoelectronic extrapolation indi- cate a negative binding energy for the 3P ground state of the N‘ ion (ref. 2 ). Although calculations do show a positive binding energy for the b s t a t e of the N- ion (ref. 3) , a bound exci ted s ta te may be insuff ic ient t o account for the anomalous radiation emitted by nitrogen plasmas.

In support of these spectroscopic studies, an experimental program was i n i t i a t ed a t the Amea Research Center to investigate coll ision processes for negative ion production by using the 4 kV ion accelerator i n the Physics Branch. It vas decided to s tudy i n detail the process of two electron capture by singly charged posit ive ions for several reasons. First , two electron capture may be a more efficient process for the production of

negative ions than capture of free electrons by neutral atoms (ref . 4).

Second, direct observation of the N- ion has been reported by Fogel', Kozlov, and Kalmykov (see ref. 4), who measured a cross section of approx- imately cm for the react ion N+ t o N' i n lmypton at an N+ beam energy of 34 keV. An independent observation of the N- ion at a lower primary beam energy would serve t o confirm i t s existence, and might provide informa- t ion about t he s t a t e of the negative ion.

2

Third, the process of two electron capture i s idea l ly su i t ed t o t he study of the shape of the function a(v)* i n the adiabatic region. The adiabatic region is defined by the condition a1 ~ j / h v >> 1, where a i s the effective range of the interaction and E i s the change in i n t e rna l energy of the par t ic les ( re f . 5 ). In the adiabatic region cross sections should increase with velocity according to the formula (ref. 6):

where k' i s a constant. Deviations of U(V) from the adiabatic formula above have l e d t o proposals that ei ther the effective interaction range (a) is velocity-dependent or the relative velocity (v) i s al tered by the f i e ld of the interaction potential (ref. 7 ).

O f a l l the charge changing processes, ordinary charge exchange has the largest cross section. I n general, it is two orders of magnitude larger than the cross section for two electron capture by singly charged posit ive ions. But, t he s t a t e of excitation of the negative ion product of t h e two electron capture reaction is known and limited; thus, distortion of the a (v) curve is minimal. Since the adiabatic region is attained a t low veloci t ies , measurements of two electron capture cross sections at low impact energies are required.

Because the cross section for conversion of N+ t o was expected t o be small a t low energies, preliminary measurements of the corresponding cross sections for H+ and He+ were made to es tabl ish cal ibrat ion s tandards and experimental procedures. Since helium forms a long-lived negative ion only i n an excited state (ref. 8), and atomic nitrogen may be similar in t h i s respect, a study of the He- ion was considered particularly useful. Results of the cross section measurements fo r and He+ at impact energies of 0.28 keV and 2 keV are found in the main body of this report . The Appendix con- t a ins a br ief summary of posit ive and negative ion formation by electron bombardment of atmospheric gases, n i t r i c oxide, and carbon monoxide.

* a(v) is the effective cross section for an atomic collision process,

and v is the relat ive veloci ty of the col l iding par t ic les .

There has been only one reported measurement of the two electron capture cross section for H+ below 0.5 keV (ref. 9), and none fo r He+ below 3 keV. The main experimental limitation has been detection of the small negative ion currents, a problem overcome i n our experiments by applying multi-channel analyzer techniques.

A P P a T U S AND EXF'EKBE~AL METHOD

A schematic diagram of the apparatus used for the negative ion studies reported here is shown in f igure 1. A detailed description of the ion accelerator may be found in reference 10.

Positive ions are formed by electron bombardment in a Carlston-Magnuson type ion source (ref. 11); the source i s held at a posi t ive potent ia l w i t h respect to ground (V )? and the f i lament a t a negative potential (Ve ) w i t h

respec t to the source chamber. The extracted ions pass through an electro- s t a t i c l ens i n to a 90' magnetic analyzer, where they are separated by the i r charge t o mass ra t ios . A second electrostatic lens then focuses the mass analyzed ion beam into the reaction chamber, C.

P

The reaction by which positive ions are converted into negative ions may be represented by the equation

x+ + Y +x- + Y++ + I s 1 ,

where X i s the incident par t ic le ; Y, the t a rge t par t ic le ; and E, the energy defect .

+

The mixed beam of unscattered particles emerging from the reaction chamber consists mainly of primary ions (e), neutral par t ic les (p) formed by ordinary charge exchange, and negative ions (X-) formed by two electron capture. The content of the beam may be expressed by the d i f fe ren t ia l equations

3

where I , Io, and I- are respectively the posit ive, neutral , and negative currents in the beam; uij is the effective cross section for the change of charge s t a t e froan i t o j; n i s the number density of target gas particles; and x, .the path length.

+

The solution of equation 2c when all quantities except I a r e i n i t i a l l y + zero yields an expression for u1 i n terms of experimentally determined quantit ies. For thin targets the following formula i s obtained:

- 1

I- = nu1 - L , +

where Io is t h e i n i t i a l value of I , and L is the effective path length of the reaction chamber.

+ +

A t the ex i t to the reac t ion chamber is a paral le l p la te capaci tor , D, that may be used t o d e f l e c t charged par t ic les from the beam. Following the deflector plates is a retarding chamber, R, originally intended for use in merging beams type experiments, but used here t o retard the primary beam of positive ions when necessary. Beyond the chamber R is a t r ip l e s t age electrostatic analyzer, A, set to pass negative ions of a specified energy, EA, on t o the single particle detector, B4T. (A collector, S, located near the entrance to the tr iple analyzer is used t o measure positive ion currents.) The s ingle par t ic le detector at the ex i t to the t r ip le ana lyzer is a 14-stage electron multiplier with copper-beryllium dynodes.

Figure 2 shows a block diagram of the electronics used in the detection system. Signal pulses from the electron multiplier pass through a pre- amplifier-discriminator-amplifier system; and the amplified pulses are counted by using a multi-channel analyzer, MCA, operated in the scaler mode.

I n actual operation, the tr iple electrostatic analyzer i s tuned t o t h e primary energy; i . e . , EA = E Therefore, the potential of the reaction

chamber must be adjusted to a value that compensates for the energy l o s t by the incident particle during the reaction X+ t o X-. If V represents the potent ia l appl ied to the react ion chamber when the negative ion signal i s observed, the conservation of energy and momentum equations may be used t o derive an expression for I C I (see eqn. 1) i n terms of Vc. For forward scattering, the expression obtained is

P-

C

4

where M and m are respectively the masses of the t a rge t and the incident par t ic les ; E is the energy f o r which the tr iple analyzer is se t ; Ei is the i n i t i a l energy of the incident par t ic les and it equals E + lsvcI; and 9 is the e lectronic charge.

A

Automatic modulation of the parameter V i s coupled with the m u l t i - channel analyzer t o improve the sensit ivity 8f detection and t o record the var ia t ion of signal with V (see fig. 2 ). The sawtooth output of an oscilloscope may be used t 8 supply a sweep voltage of 140-volt range t o t h e reaction chamber; this sweep must be synchronized t o t h e channel advance sweep of the multi-channel analyzer.

CROSS SECTION MEASUREMENTS FOR H+ AT 0.28 keV

Experimental Procedure

Approximately 10 of H reaches the collector, ear the entrance + to the t r ip le ana lyzer when the source pressure is 5 x t o r r of H2 and the electron bombardment energy is 90 eV. Once the current to S has been measured, posi t ive potent ia ls are appl ied to the plates of t h e t r i p l e ana- lyzer; and these plate voltages are adjusted to optimize the current reach- ing the first dynode of the electron multiplier, Dl. The transmission of the analyzer, or the ratio of I+Dl t o I+o, is measured during the "tuning" procedure. Then the plate voltages are switched to opposite polarity to set the t r iple analyzer for 0.28 keV negative ions.

After the triple analyzer has been tuned, gas is admitted t o t h e reaction chamber until the current to S is reduced t o .half i ts i n i t i a l value. Then the H- signal counts are accumulated for a fixed number of sweeps of the voltage Vc. The counting procedure is repeated twice, once with the ion beam deflected and once wi th the gas off , in order to subtract from the signal any background noise from the neutrals and positive ions i n the beam. To minimize background noise from the positive ions, the retard- ing chamber, R, is held a t a posi t ive potent ia l above ground (+3OO V ) during the counting procedure.

Method of Data Analysis

The two electron capture cross section , cr is calculated by using 1 - 1' the equation 3 and t h e following formulas:

I

n = 3.28 X 10 16 p ,

where q is the electronic charge in coulombs; C-, the number of negative ion counts per second measured a t t h e peak of the s ignal dis t r ibut ion curve; g, the eff ic iency of the detector; T, the transmission of the triple analyzer; and p, the pressure i n t o r r i n s i d e the reaction chamber. In equation 3, L is the length of the reaction chamber i n cm (= 13.5 cm), and al - is the

cross section i n cm . A l l curren ts a re in amperes. 2

The primary current, I . -- The current i s measured a t the entrance + "0

t o t h e t r ip le analyzer, rather than a t t he en t r ance t o the reaction chamber, i n o rde r t o account for cur ren t and s ignal losses from defocusing of the beam between these two points. Since a pos i t ive po ten t ia l i s appl ied to chamber R (see f ig. 1) during the counting procedure, the opposite polarity i s applied when Io+ is measured i n order t o account f o r any focusing effect VR has on the signal. The collector, S, i s held a t +225 V t o suppress secondary electron emission when the current i s measured. Also, the reac- t i on chamber (C) is grounded and there a re no background gases admitted. Residual pressures i n the accelerator are less than torr .

Negative Ion Counts, C- . -- Since the range of energies accepted by the t r iple analyzer i s conziderably w i d e r than the energy dispersion of the beam, the signal counts are taken from the peak of the distribution curve rather than from the area under the curve. Typical distribution curves are shown i n f igure 3. A re tarding potent ia l analysis of the 0.28-keV primary beam indicates a fu l l width a t h a l f maximum of 2 eV.

Pressure, p. -- The pressure i s monitored by a Bayard-Alpert ioniza- t ion gauge located near the reaction chamber. A McLeod gauge must be used t o determine absolute pressure. But, since t h i s pressure calibration would have disrupted operation of the ion accelerator, it was deferred t o a more convenient time. Instead, we have used i n the present analysis pressure calibration data taken by Savage and Witteborn i n the course of t h e i r N* 2 experiment (see ref. 12), which was done on the same apparatus. Ionization gauge pressures (pi) f o r a l l gases have been calibrated against the i r r e su l t f o r N2 a t pi = 2.00 X t o r r , where the corresponding McLeod gauge read- ing is p = 1.30 X t o r r a t T = 22.4OC. This determines experimentally the impedance ( p ) of the reaction chamber; viz., j3 = 65. Absolute pressures (p) then are calculated by multiplying relative pressures (pi) by two fac- tors : j3, the impedance o r d i f f e ren t i a l pumping factor; and y, an adjustment t o the ionization gauge reading for the sensi t ivi ty of the gauge t o t h e par t icular gas used. The "gauging" fac tors for a gauge reading of 1.00 are l i s ted below:

6

TABU I. - GAUGING FACWRS

N2 1.00 O2 1.20 A i r 1.00 He 4.80

A r 0.66 ~r 0.52

H2 2.40 Xe 0.37

These fac tors l i e wi th in 10 percent of those given by Dushman (ref . 13) f o r a standard gauge, with the exception of y f o r argon.

Since the cross section measurements reported here were made a t only one pressure, an investigation of the pressure dependence of the observed signals must be made t o confirm tha t the negative ions result from single coll isions.

Wansmission, - T. -- The transmission of the t r iple analyzer i s the r a t i o of I t o Io , where ID1 is the current to the first dynode of the electron multiplier. A potent ia l of -250 V on the screen dynode (D-0) i n f ront of D l i s sufficient to suppress secondary electron emission during the current measurement. Approximately twenty percent of the beam is trans- m i t t e d through the tr iple analyzer.

+ + +

Efficiencx, 1. -- When negative ions are detected, the screen dynode is connected through a poten t ia l d iv ider c i rcu i t to the anode of the electron multiplier; the applied potentials are +25O V t o D-0 and +3500 V t o t h e anode. To detect positive ions, D-0 is disconnected from the divider c i rcu i t ; and the following potentials are applied: -250 V t o D-0, -450 V t o Dl, and +2600 V t o t h e anode. In both cases, the potential drop across the electron multiplier is +3050 V; and the ions impact Dl a t 730 eV.

The efficiency of the detector was measured for the posit ive ion, and the assumption was made tha t the corresponding negative ion would be de- tected with the same efficiency.

The measurement of E would have been a straightforward comparison of gl with the corresponding counts per second had a more sensit ive ammeter been available at the time. Instead, it was necessary t o obtain curves of

versus a retarding voltage, and then to . re la te these to s imilar curves f o r the counts per second. The best estimate of e from these measurements is 5 = 0.065 k 0.020 when the potent ia l drop across the multiplier i s

+

+3050 V. Extrapolation of the retarding potential curves is responsible f o r the large uncertainty in E.

Results and Discussion

Measurements were made for the react ion H t o H' i n hydrogen, oxygen, and xenon a t an H+ beam energy of 0.28 keV. The cross sect ions are l is ted below, along with the estimated standard deviations:

+

TABU 11. - TWO ELECTRON CAPW CROSS SECTIONS

FOR H+ AT 0.28 keV

Target gas 2 Q1 - 1 J cm

H2 (4.7_+ 1.7) X

O2 (2.45 0.8) x Xe (1.35 0.5) X lom1'

Errors in u a r i se from uncertainties i n the parameters of equations 3 and 5 . The values of these uncertainties were estimated t o be: AL/L = 0.05, Ar/r = 0.10, Ly-/I- = 0.05, a/e = 0.30, &/p = 0.10, and Lyo+/Io = 0.10.

1 - 1

* +

If the transmission is measured j u s t p r i o r t o t h e counting procedure, - 1 may be calculated from the following equation.

obtained by substi tution of &/Io f o r T. This procedure was used t o measure the cross section for H+ i n xenon; the uncertainty in approximately 0.05.

+

G1 was

The cross section l isted above f o r H i n hydrogen is approximately a fac tor of two smaller than that obtained by Kozlov and Bondar a t 0.15 keV (ref. 9 ) . Their r e su l t s fo r H+ in H2 from 0.15 t o 5 keV are graphed i n

+

* The uncertainty i n p quoted here does not include errors i n the

gauging fac tor (r) and the impedance ( 8 ) .

8

figure 4; the original paper should be consulted fo r exact values. Since OUT experimental method does not allow for large angle scattering, th i s discrepancy is understandable. Apparently Kozlov and Bondar measured the total cross sect ion by means of a modified mass spectrometric method s imilar to that descr ibed in reference 14.

The experimental method described i n t h i s report allows simultaneous determination of u1 - f o r forward scat ter ing and €,the energy defect of the charge changing reaction. If E is known, the process by which the primary particle captures two electrons may be specified. This is partic- ular ly interest ing i n the case of molecular particles, for which the reaction may proceed either by double ionization of the target molecule o r by a dissocizrtive ionization process; Le., either by the r e y t i o n X+ + Y2 *X- + Y r + le1 or by the reaction X+ +Ye +X- + Y + Y+ + 1 ~ ~ 1 . A s examples, we calculate e f o r H i n % and O2 by comparison with E f o r H+ in xe.

+

Differentiation of equation 4 gives:

for small m/m+M. From the peak positions of the signal curves for H+ i n xenon a t E = 240 eV and E = 280 e V when EA = 280 eV (see fig. 5), we cal- culate approximately 0.8 V per channel. If le(Xe) I* f o r H+ i n xenon is equal t o 19 eV, then we calculate IE( €$) I = 28 eV and IE(O*) I = 26 e V ( fo r

i n H2 and 02) by using equation 4 and the peak positions shown i n figure 3. These values may be compared to t he i n t e rna l energy defects for the reactions H+ + 5 9 H' + H+ + H+ e(H2) ] and H+ + O2 3 H- + 02" 1 E ( 02) 1;

energy of the two protons (E ) is 18 e V (ref. 15 ) , and 1 E( 02) 1 = I EH I + SH ,- Eo2 1 = 22 eV i f the threshold energy f o r O2 (E ) is 36 e V (ref. 16)

P P

IE(H2) I = IER + SH - Ediss I - 3' - Epot I = 35 eV i f the potent ia l

, I1 Pot ++ I1 02

The energy defect E(%) t ha t we calculate from equation 4 seems t o indicate a smaller value for the quantity E However, uncertainty i n the measurement of Vc may be responsible for thgOhscrepancy between the energy defects calculated from equation 4 and those estimated from the threshold

* The internal energy defect is equal t o 1%' + SH - <; I where

is the threshold for H+; SEI i s the e lec t ron a f f in i ty of H-( = 0.75 eV); and

$: is the threshold for Xe".

9

potent ia ls and SH( and E where applicable). Since the r e su l t s quoted here are preliminary, it i s expected t h a t more accurate values for the energy defects will be obtained when these experiments are repeated.

Pot

CROSS SECTION MEASUREMENTS FOR H+ AND He+ AT 2 keV

Experimental Procedure

The experimental procedure a t 2 keV is s imilar to that previously described f o r H+ a t 0.28 keV. Typical currents were 250 of He+ and 50 ppA of H+. Source pressures were on the order of 10 torr ; and the electron bombardment energies were 187 eV and 137 eV f o r He+ and H+ respectively.

-4

Because the t r iple e lectrostat ic analyzer was des2gned mainly f o r use with low energy ion beams, there were frequent voltage breakdowns when the high potentials needed for discrimination of a 2 keV beam w e r e applied t o the analyzer plates. The result ing loss of resolution and fluctuations i n the transmission are primarily responsible for the large uncertainties i n the data that follow. Also, sh i f t ing of the s ignal peak position made it impossible t o determine a unique Vc f o r each reaction.

Background noise from the positive ions was low enough tha t the retarding chamber i n f ront of the t r iple analyzer was not used; it was kept a t ground potential .

The sweep voltage applied to the reaction chamber ranged from -70 t o +TO V. It was observed that the posit ive voltages defocused the primary beam (see f ig. 6). Although the effect on the negative ion signal i s not certain, positive voltages (Vc) appear t o have had an opposite, or focusing, e f fec t on the negative ions. However, the transmission of the t r iple analyzer d i d not vary greatly with Vc (see fig. 6 ) .

Method of Data Analysis

Again, the two electron capture cross sections for H and He a t 2 keV were calculated from the relationship q C- = 3.28 X 10 16 6 T Io p L, obtained by combining eqwtions 3 and 5 .

+ + +

Efficiency, 5. -- The efficiency of the detector was estimated t o be 0.09 and 0.30 when the applied voltages were respectively +3OOO V and +3500 V t o the anode and -100 V t o D-0.

Pressure, p. -- Pressures in the reaction chamber were calculated by multiplying the-relative pressures, pi, by rfi. Table I lists the values

10

I

used f o r y. llhe impedance, p, has a value of 65 except f o r N2 a t pi =

2.8 X to r r , f! = 66; f o r H2 a t pi = 3.2 X lom5 t o r r , /3 = 68; f o r R2 a t pi = 3.9 X loD5 tor r , f! = 71; and f o r Xe a t pi = 2.0 X 8 = 56.

Results and Discussion

For the reaction H+ t o H- i n hydrogen, the cross section is 5 X 1619 cm 2

with a standard deviation of 3 X cm . The quoted e r ro r r e f l ec t s un- 2

cer ta in t ies y nor those t a i n t i e s in

in I-, I , p, e, I;, and T; it does not include errors in f! and due t o dgfocusing effects. Approximate values for the uncer- the parameters above are: AI-/I- - 0.45, LSL-+/I- - 0.05,

+

+ &/p - 0.10, L $ / E - 0.30, &/L - 0.05, and AT/T ... 0.25. The H- counts varied by about a factor of two under the same experimental conditions; and the value listed above f o r =-/I- r e f l ec t s th i s variation although the cause i s undetermined. The relat ive cross sect ions for H+ in N2, 02, K r and Xe are tabulated below:

v u

I

TABLF: 111. - REZATIVE TWO ELECTRON CAPTURF: CROSS SECTIONS

FOR H+ AT 2 keV

Target gas r r

5 X 10-lg cm 2

Xe K r

O2

N2

24 10

3 0.9

Comparison with the results obtained by Kozlov and Bondar (ref. 9) f o r H+ i n H2, by Fogel ( ref . 7 ) f o r H in Xe, and by Kozlov e t a l . (ref. 17) f o r

H+ i n Kr indicates t ha t our measurements may be too large by about a fac tor of two (see fig. 4).

4-

The cross section for two electron capture by He in helium is 8 X w 2 3 cm2. A standard deviation of 4 X cm2 is calculated by us ing the uncertaint ies in Io , p, e, L, and T l i s t e d above and &-/I- - 0.15. Relative cross sections for He+ i n 5, N2, 02, and Kr a re listed below:

+

+

TABLE N. - RELATIVE TWO EIXCTRON CAPlvRE CROSS SECTIONS

FOR He+ AT 2 keV

Target gas Q r ) 8 X LO'^^ cm 2

O2

N2

I-5 K r

25 10

1.6 0.3

There a re no other measurements of ul - f o r He+ a t 2 keV with which t o

compare our resul ts . Windham e t a l . (ref. 18) measured a cross section of - lom2' cm f o r He i n hydrogen a t an He beam energy of 17.5 keV; the results of Dulreltskii e t a l . ( r e f . 19) f o r He+ i n helium and krypton are graphed in f igure 7.

2 + +

A cursory investigation of the var ia t ion of cross section with elec- t ron accelerat ing potent ia l was made f o r He+ i n He. As shown below, the cross section increases by about 25 percent when Ve is increased from 37 V t o 187 v:

'e 9

vol t s

6.4 6.6 6.9 8.0

Typical negative ion signal curves are shown in figures 8 and 9. The s t ructure vis ible in these curves may be due to instrumental effects , but fur ther invest igat ion is necessary before a satisfactory explanation can be given.

CROSS SECTION MEASUREMENTS FOR "N+" AT 2 keV

Af'ter the capabi l i ty of the technique f o r producing the metastable He' ion had been demonstrated, the next step was t o attempt production of N".

A beam of N ( - 5 mpA) was obtained from N2 a t a source pressure of approximately t o r r and a t an electron bombardment energy of 187 eV. The 2-keV beam was sent through three different charge exchange gases, and i n each case negative ions were observed. The re lat ive s ignals ob= tained (I-/Io+p) and the charge exchange gases used a re l i s ted below:

+

Gas Signal, 10' 4 pe r t o r r

K r 3.6

O2 0.7 N2 0.5

. These signals correspond t o two electron capture cross sections of 8 X cm t o 1 X 1$-22 cm . However, to assure that the observed s ignal r e s u l t e d from N and not from impurit ies in the beam, the analyzer f i e l d was varied while source and focusing conditions were kept constant. As the f i e ld was adjusted toward higher q/m - values, and as the N+ current accordingly decreased, the signal was observed t o increase. Figure 10 shows the curves for N+ current and f o r the I- signal versus B; B i s the magnetic f i e ld in gauss measured ex te rna l t o the analyzer magnet chamber, and thus it is only proportional t o t he ac tua l bending field. Therefore, it was concluded that the negative ion signal probably resulted from O+, OH+, and F+ contaminants i n the N+ beam. Since only high purity (99.95 percent) or research grade (99.999 percent) gases were used, the impurities a r e due mainly t o gases evolved from the filament or the walls of the source chamber. Although a positive ion spectrum of the primary beam showed t h a t it contain- ed less than 6 percent O+ and 3 percent F+ (see the Appendix and fig. 12), the two electron capture cross sections for O+ and F+ are large enough t o account for the observed negative ion signal."

2 2

APPENDIX

SURVEY OF IONS ENITI1ED FROM 'l?IB SOURCE

A more d i rec t method for the s tudy of negative ion production may be mass analysis of ions emitted from the source. Although i n the past the mass spectrometric method has not yielded information about the N ion, the increased sensitivity of our detection system required that it be tried. A number of reactions are responsible for negative ion production in the source, including dissociative ionization, charge exchange, and single electron capture of slow secondary electrons.

Mass analysis was achieved by using the analyzer magnetic f i e l d as the sweep parameter coupled wlth the multi-channel analyzer.* Since there were no charge exchange gases used, residual pressures in the accelerator were about to r r . Chambers C and R in f igure 1 were grounded; and the t r iple e lectrostat ic analyzer was tuned t o t h e primary energy, 280 eV.

Mainly to e s t ab l i sh a mass scale, but also to determine the impurity content of the primary beam, posit ive i o spectra w e r e obtained f o r 02, N2, and Ar ; sources pressures were about lo-' t o r r , and the electron bombard- ment energy was 90 eV. These spectra are shown in figures 11 through 13. The concave shape of the peaks identified as 02+, N+, N2', Ar', and Ar* resulted from saturation of the detector.

Then, negative ion spectra were obtained by reversing the polar i ty of the potentials on the source chamber and focusing electrodes, and by re- versing the direction of the magnetic field. Figures 14 through 19 show the spectra obtained for He, A i r , N2 (research grade) , NO, COY and 02. There are several points of i n t e re s t i n the helium and nitrogen spectra: the mass 16 t o 19 impurities in both; the absence of He' from the helium spectrum ( f ig . 14); and the relatively large peak a t mass 26 i n the nitro- gen spectrum (fig. 16).

It is possible that the impurities in nitrogen (Om, O H 9 F') mask an N signal. But, because He- was not observed, it is doubtful that N- would be observed even with better mass resolution.

* Refer to the discussion under "Apparatus and Experimental Method"

i n the main body of t h i s r epor t fo r de t a i l s of the detection system.

14

The mass 26 peak in the nitrogen spectrum has been identified tenta- t ive ly as CY; the signal intensity varied i n proportion to the ni t rogen pressure i n the source. Utterback (ref. 20) found t h a t from I 2 t o 60 e V N2 reacts with CO t o form predcaninantly NO+ and CN-; and a similar reac- t ion may account f o r the mass 26 peak in the nitrogen spectrum (fig. 16).

A

E:

k’

h

a

X

IO

I+

1-

n

a_

Ioi

SYMBOLS

a difference &erator; e.g., &X = Xi - X2 change i n i n t e rna l energy of a system resu l t ing from col l i s ions

a constant

Planck’s constant

re la t ive veloci ty

cross section as a function of v

effect ive interact ion range

source chamber voltage with respect t o ground

filarzlent bias voltage with respect t o t h e source chamber

the different ia l operator

ef-r”ective path length of ions in the charge exchange gas

neutral current

positive ion current

negative ion current

number of molecules per unit volume in the reac t ion chamber

cross section for the change of ckarge from qi t o q j where i = 1, 0, -1 and j = 1, 0, -1

electronic cherge

the in i t ia l pos i t ive ion cur ren t measured a t the entrance t o t h e

t r i p l e electrostat ic analyzer

effective length of the reaction chamber

mass of the t a rge t molecule

mass of the Grojecti le molecule

16

Ei

EP

vc

EA

i n i t i a l energy of the project i le ion i n the reaction chamber

energy of the projecti le before entering the reaction chamber

reaction chamber voltage

energy f o r which the triple electrostat ic analyzer is tuned

C’ number of negative ion counts per second

t eff ic iency of the electron multiplier detector

7 transmission of the t r iple electrostat ic analyzer

Gl ini t ia l pos i t ive , ion cur ren t measured a t the entrance to the electron multiplier detector

P absolute pressure of the charge exchange gas in the reaction chamber

p i

Y gauging fac tor determined by the sens i t i v l ty of the ionization

re la t ive pressure measured by using an ionization gauge i

gauge t o va.rious gases

B impedance of the reaction chamber

%I, threshold energies for the singly and doubly charged posit ive ion of the molecule X

sx Epot

e l ec t ron a f f in i ty of the neutral molecule X

potential energy of two protons

relat ive cross sect ion ‘r

1. Thomas, G. M.; and Menard, W. A.: Measurements of the Continuum and Atomic Line Radiation from High Temperature A i r . AAIA J., vol. 5, Dec. 1967 , pp. 2214-2223.

Norman, G. E.: The Role of the Negative Ion N in the Production of the Continuous Spectrum of Nitrogen and A i r Plasmas. t ranslat ion: Opt. Spectry. , vol. 17, No. 2, Aug. 1964, pp. 94-96.

Allen, R. A. ; and Textoris, A. : Evidence for the Existence of N- from the Continuum Radiation from Shock Waves. J. Chem. F%ys., vol. 40, no. 11, June 1964, pp. 3445-3446.

Boldt, G.: Recombination and 'Minus" - Continuum of Nitrogen Atoms. Z. Fhysik, V O ~ . 154, 1959, pp. 330-338.

2. Edlen, B.: Isoelectronic Extrapolation of Electron Affinities. J. Chem. pfiys., vol. 33, no. 1, June 1960, pp. 98-100.

Johnson, H. R. ; and Rohrlich, F. : Negative Atomic Ions. J. Chem. phys. , vol. 30, no. 6, June 1959, pp. 1608-1613.

3. Schaefer, Henry F., 111; and Harris, Frank E.: Metastability of the 'D s t a t e of the Nitrogen Negative Ion. phys. Rev. Letters, vol. 21, no. 23, Dec. 1968, pp. 1561-1562.

Bates, D. R . : and Moiseiwitsch, B. L.: Energies of Normal and Excited Negative Ions. Roc. Fhys. SOC. London, se r ies A, vol. 68, June 1955, pp. 540-542.

4. Fogel', Ya. M. ; Kozlov, V. F. ; and Kalmykov, A. A. : On the Existence of the Negative Nitrogen Ion. Zh. Eksperim. i. Teor., vol. 36, May 1959, pp. 134-1356 (translation: Soviet Fhys. - m p , vol. 36(9), no. 5, Nov. 1959, pp. 963-964).

5. Massey, H. S. W. : Collisions Between Atams and Molecules a t Ordinary Temperatures. Rept. Progr. Fhys., vol. 12, 1948, pp. 248-269.

6. Hasted, J. B. : Ine las t ic Ion-Atom Collisions. J. Appl. Fhys., vol. 30, no. 1, Jan. 1959, pp. 25-27.

7. Fogel', Y. M. : The Production of Negative Ions in Atomic Collisions. Usp. Fiz. N a u k . , vol. 71, June 1960, pp. 243-287 (translation: Soviet phys. Uspekhi, vol. 3, No. 3, Nov. - Dec. 1960, pp. 390-416).

0.

9.

10.

11.

12.

13 9

14.

15

16.

17

18.

Holdien, E. ; and Midtdal, J. : On a Metastable Energy State of the Negative Helium Ion. Roc. Fhys. SOC. London, series A, vol. 68, 1955, PP. 815-823.

Koelov, V. F.; and Bondar', S. A.: Double Charge Exchange Between Singly-Charged Positive Ions at Low Ehergies. Zh. Eksperim. i. Teor. Fiz., vol. 50, Feb. 1966, pp. 297-306 (translation: Soviet phys. - JETP, vol. 23, no. 2, Aug. 1966, pp. 195-202).

Nichols, Billy J.: and Witteborn, Fred C.: Measurements of Resonant Charge Exchange Cross Sections in Nitrogen and Argon Between 0.5 and 17 eV. NASA TN D-3265 , Feb. 1966. Carlston, C. E.: and Magnuson, G. D.: High Efficiency Low Pressure Ion Source. Rev. Sci. Inst., vol. 33, no. 9, Sept. 1962, pp. 9@-911. Savage, H. F.: and Witteborn, F. C.: Charge-Exchange Cross Sections of %+ in N2, C02, and Ar and Contamination of N" Beams by <+. J. Chem. phys . , vol. 48, no. 4, Feb. 1968, pp. 1872-1873. Dushman, Saul: Scientific Foundations of Vacuum Technique. Second ed., John Wiley and Sons , Inc. , 1965 , pp. 323-324. Kozlov, V. F. : and Bondar' , S. A. : Method for the Measurement of Double Charge-Exchange Cross Sections of Lar-Eher&y Positive Ions. Zh. Tekh. Fiz., vol. 37, no. 3, March 1967, pp. 542-549 (translation: Soviet phys. - Tech. Phys., vol. 12, no. 3, Sept. 1967, pp. 388-393).

Fogel', Ia. M.; and Krupnik, L. I.: The Formation of Negative Oxygen Ions in the Collisions of Positive Oxygen Ions with Gas Molecules. Zh. Eksperim. i. Teor. Fiz., vol. 29, August 19.55, pp. 209-220 (translation: Soviet phys. - JETP, vol. 2, no. 2, March 1956, pp. 252-261 ).

Dorman, F. H.; and Morrison, J. D.: Ionization Potential of Doubly- Charged Oxygen and Nitrogen. J. Chem. phys., vol. 39, no. 7, Oct. 1963, pp. lgO6-lgO7.

Kozlov, V. F.; Fogel', Ya. M.; and Stratienko, V. A . : Two-Electron Charge Exchange of Low-Energy Protons. Zh. Eksperim. i. Teor. Fiz., vol. 44, June 1963, pp. 1823-1825 (translation: Soviet Phys. - JETP, vol. 17, no. 6, Dec. 1963, pp. 1226-1227).

Windham, P. M. : Joseph, P. J. ; and Weinman, J. A. : Negative Helium Ions. Phys. Rev., vol. 109, no. 4, Feb. 1958, pp. 1193-1195.

19. -elf sk i i , V. M. : Afrosimov, V. V. : and Fedorenko, N. V. : Zh. Eksperim. i. Teor. F iz . , vol. 30, April 1956, pp. 792-793 (translation: Soviet phys. - JETP, vol. 3, 1956, pp. 764-766).

20. Utterback, Nyle G.: Cross Section for Atom Exchange with Chemge Transfer in a Molecular Collision. J. Chem. mys., vol. 44, no. 6, W c h 1966, pp. 2540-2541.

20

..-

LIST OF FIGURES

Figure 1.

Figure 2.

Figure 3.

Figure 4.

Figure 5.

Figure 6.

Figure 7.

Figure 8.

Figure 9.

Figure 10.

Figure 11.

Figure 12.

- Schematic diagram of ion accelerator.

- Block diagram of electronics.

- Typical H- s i g n a l s a t 0.28 keV: 0 II -Hz, a H -02, + +

and 0 H+-Xe.

- Cross sec t ion for two electron capture by H a s a function +

of energy: 0 E+-H2(ref. 9) , A H+-Xe(ref. 7 ) , and +

0 II -Kr(ref. 17).

- m i c a 1 H- signals from two electron capture by H i n xenon + a t H+ beam energies of 0.24 keV (9) and 0.28 keV (0).

- Current (0) and transmission (El) versus reaction chamber voltage.

- Cross section f o r two electron capture by ITe as a function +

of energy: 0 He+-He and A He+-Kr(ref. 19).

- Typical H- s ignal from two electron capture by H i n hydrogen +

a t an H beam energy of 2 keV. +

- Typical He- s ignal from two electron capture by He i n helium + + a t an He beam energy of 2 keV.

- Variation of N current with analyzer magnetic f i e ld (0); and variation of the corresponding negative ion current when krypton is used as the charge exchange gas (A).

+

- Typical mass spectrum of positive ions with oxygen i n the ion source. Some main components are ident i f ied on the graph.

- Positive ion soectrum with nitrogen in the source.

Figure 13. - Positive ion spectrum with argon in the source.

21

Figure 14. - Typical mass spectrum of negative ions with helium i n the

ion source. Some main components are ident i f ied on the

graph.

Figure 15. - Negative ion spectrum with a i r i n t h e source.

Figure 16. - Negative ion spectrum with research grade nitrogen in t he source .

Figure 1.7. - Negative ion spectrum with n i t r i c oxide in t he source.

Figure 18. - Negative ion spectrum with carbon monoxide i n the source.

Figure 19. - Negative ion spectrum with oxygen i n the source.

22

SCHEMATIC DIAGRAM ION ACCELERATOR AND END STATION

ELECTRODES \ 1 .......

....... ...... ...... \ GAS INLET ...... ......

ffl ELECTROSTATIC ANALYZER (A) / COLLECTOR (SI

Iumr

(EMT) ............... PUMP 'RETARDING CHAMBER (R)

Figure 1

OSCILLOSCOPE

SYNCH

INPUT 0"-

Q SAW TOOTH

OUTPUT I

REACTION CHAMBER

SIGNAL PULSE

MULTI-CHANNEL I I I I

r m ANALYZER I I I I

I I I I

x- r "'e"- \I

TRIGGER OUTPUT PULSES

P MULTI-SCALE INPUT

MCA

Figure 2

1 I I I 1 I I I 1 140 145 150 155 160 165 170 175 180 185

CHANNEL NUMBER Figure 3

E, kev

Figure 4

15

IO

5

0 280 0 240

CHANNEL NUMBER Figure 5

tn z tn 2, Qt

-

I- z W a 3 0

a

100

90

80

70

60

50

0 n W u u u

9 0 I + O T -

40 - -100 -80 -60 -40 -20 0 20 40 60 80 100

vc , volts Figure 6

o He A K T

A

Figure 7

w 0

lox lo4

8-

s 6- < I - 4 -

tn 0

I

+

0 - 80 -60 -40 -20 0 20 40 60 80

vc, volt

Figure 8

30x103

25

0 a 20

e u, t 0

c5

'a I

15

5

OL -80 - 40 0

vc, volt Figure 9

40 80

W Iu

P l

MAGNETIC FIELD, gauss Figure 10

w w

IO2 ' O i

IO'

IO0

O+ .t 0;

I

i

8 12 16 28 32 14, 19

mAql

Figure 11

OXYGEN

p = I. 2 X IO-^ torr i,=5ma

44 48

N+

.- t c 3

+ h - IO

IO0

, I

N++

NITROGEN p = I x IO-^ torr ie= 5 ma

Figure 12

c

w w

A++

IO2

IO'

IO0 I 12 16 20

14 28 32 40

ARGON p = 8 X torr Ie = 15 ma

‘ O 3 I IO2

l o l f

IO 0 0 I

i 0-

02-

16 19

m/l Figure 14

HELIUM

p = 1.4 X IO-^ torr i, = IO ma

26 32 40 44

(I‘

0- I AIR

p = 2 x IO-^ torr ie= 3 ma

IO0 I 16 19 26 32 40 44

28 m A q I

m t 3.

t "

IO3!

I

- lo[

..

I oo s I 8

F- 0- I

NITROGEN p = 1.2 x IO

i e= 5 ma

-4

12 16 19 26 32 14 28

40 48 44

mAql

Figure 16

H- c

0- NITRIC OXIDE p = I x IO-^ torr ie =2.5 ma

P

C- I

I,~ A I t I 12 16 19 26 30 42

mAql

Figure 17 W \c)

I

L W

I H a

A

0-

H' t

CARBON MONOXIDE p= 9.3 x 10-5 torr i,=5 ma

A n Ann IO0' I I I I I I 12 16 19 26 30 4042

m/lql

Figure 16

I- Y) 0) W

N

lo3 E

IO

0- OXYGEN p = 1.2 x ~ ~ - 4 torr . 1 ,=5ma

02-

n 1

I 32

03- I

48