PHYSICAL R EVIE W A VOLUME 11, NUMBER 4 APRIL 1975

Nonmetastable Penning ionization in He(3'P)-Ne, Ar, Kr, Xe collisions

S. Kubota, ~C. Davies, and T. A. King

Physics Department, Schuster Laboratory, University of Manchester, Manchester M13 9PL, England(Received 11 June 1974)

Time-dependent spectroscopy techniques have been used to determine the cross sections for the de-excitation of nonmetastable He(3 'P) states by collision with Ne, Ar, Kr, or Xe atoms. The mea-sured cross sections in 10 ' cm' at 600+',00

'K are, for Ne, 27.6+', 8, Ar, 55.6+,'8', Kr, 49.5+4&, and

Xe, 73.0+88 . These cross sections are compared with theoretical Penning ionization cross sectionshaving values of 20 for Ne, 40 for Ar, 39 for Kr, and 35 for Xe, in units of 10 ' cm' (at 600'K).

INTRODUCTION

Interactions between electronically excited atoms(or molecules) and ground-state species are ofinterest both for a basic understanding of energy-transfer mechanisms and for their importancein processes in the upper atmosphere, ' in gaseouslasers, ' in plasmas, ' and in other excited-gasmedia. 4 There have been several experimentalinvestigations' on the deexcitation of metastable-state atoms, but few for nonmetastable excitedatoms, mainly because of the experimental diffi-culties.

We report here on measurements of absolutecross sections for the deexcitation of He(3'P)states in collision with gas atoms X (X =Ne, Ar,Kr, and Xe) at approximately 600'K, observingthe relaxation of He(3'P) states in helium con-taining known concentrations of &. The ionizationpotential of the added gas X is lower than the ex-citation energy of He(3'P) and the process studiedinvolves the ionization of X atoms in a collisionwith an excited state by nonmetastable Penningionization. Until recently, the existence of thenonmetastable Penning ionization process hadbeen confirmed only from the additional ioniza-tion yield after excitation by charged particles inhelium-neon, ' argon-krypton, "and argon-xenonmixtures. '

EXPERIMENTAL TECHNIQUES

The experimental apparatus is similar to thatdescribed by Bridgett, King, and Smith-Savilleand has been further developed by Davies andKing'; it is shown in diagrammatic form in Fig. 1and will be discussed more fully elsewhere. TheHe(3 'P) level is excited by electron impact, usinga pulsed beam of electrons of short duration andrapid time cutoff. The energy of the electrons iscontrolled to be as close to the energy of the 3 'Plevel as possible (threshold excitation). The de-cay function following the pulsed electron beam

was determined from single-photon counting usingthe He 5016-A spectral line (3'P-2'S emission).A time-to-amplitude converter is started by atrigger pulse from the pulse generator at the pulsecutoff time and is stopped by the pulse from anamplifier-trigger unit which occurs when a singlephoton is detected by the photomultiplier. Theoutput from the time-to-amplitude converter isproportional to the photon-emission time intervaland is fed into a multichannel analyzer where thedecay curve is accumulated. Background countsare removed by modulating the electron gun belowthreshold and subtracting at the pulse-heightanalyzer and at threshold, and adding counts.

A 5-cm & 2-cm dispenser cathode is used in theelectron gun. Figure 2 shows the excitation cham-ber filling system. The portion of the system inthe oven is baked at 250'C for 12 h while beingevacuated with an argon stable sputter-ion 50-liter/sec pump. The ultimate vacuum obtained is -5&&10 ' torr. The outgassing rate is less than2&&10 ' torrh ', even when the cathode is opera-ting at 1050'C. Spectroscopically pure helium(obtained from British Oxygen Company and con-taining impurities of less than 10 ppm) is admittedto the excitation chamber via leak valves 1 and 2.In the spectral region from 2500 to 7300 A onlya few emission lines from impurity atoms wereobserved; for a helium pressure of 38 mtorr andelectron excitation energy of 24.3 eV the light in-tensities from the impurity lines were 3 to 4orders of magnitude lower than the intensitiesfrom the helium lines. The excitation chamberwas filled with pure helium and the added gas wasleaked through valves 3 or 4, while measuring thetotal pressure using a capacitance manometer(MKS Baratron, model 77).

He(3 'P) DEEXCITATION CROSS SECTIONS

The lifetime of He(3'P) states in pure heliumis a function of the helium pressure due to thedestruction of He(3 'P) by collision with neutral

1200

NONMETASTABLE PENNING IONIZATION IN He(3 'P )- Ne, . . . 1201

PULSEGENERATOR

DRIVER 8P.S.U.

]IE

P. G.- OUTPUT

STAGE

—DISCRIMINATOR

SYSTEM

CLOCK

threshold belm thre

ELFCTRONGUN 4

EXCITATION

CHAM8ERIII

QUAPtTZ ~~LENS

ahold

STARTTIME-TO-AMPLITUDE

CONvERTERSTOP

JI(

PULSE-HEIGHT

ANALYSER

FIG. jL. Pulsed thresholdelectron excitation andsingle-photon counting de-cay-time apparatus.

MONOCHROMATOR T'-—-- PHOTOMULTPLIER DISCRIMINATOR

helium atoms' "and is also dependent on the ef-fective size of the excitation chamber owing to theslight imprisonment of the He O'P-1'S radiation.For a constant size of excitation chamber, thedecay rate 1/T in pure helium is a function of thehelium pressure, P„„

When added atoms X are introduced we can write(1) as

1/T =F(P„)+n(vv)P», (2)

where o is the added deexcitation cross section in

e W

Excitation

Chamberg / ~d~~.....r~c

50 I/e Sputterlon pump

I

I

I

I

I

I

I

I

I

I

I

I

OvenI

I

I

I

I

I

I

I

I

//

Two-StageRotaryPump

Fore line ~Trap

2 Bakeable, g/ Non-Bakeoble Valves

FIG. 2. Vacuum and gas handling system and excitationchamber. P—Pirani gauges, I—ionization gauge, c-cathode, d and g—control grids, w—exit window, e-condensing system.

N(t) =Ae t/"+B, (3)

where A and 8 are constants and in all cases J3 ismuch smaller than A. As an example to indicatethis type of behavior, A, B, and & for 0.59-torrNe in Fig. 3 were 20 300 +140, 100 +20, and 35.6+0.2 nsec, respectively. The lifetime P obtainedfrom the fit from the straight line up to 110 nsecwas 35.8 +0.2 nsec. The difference in the life-times between t» and p is less than 1.0' over allthe range of added gas concentrations. The meanvalue of & and P is used as a measure of & in thispaper. The origin of the constant term B is notcompletely understood. This may arise from re-ciprocal transfer effects from O'D states. " Asan over-all system check, the lifetimes of theO'S and 4'8 states in the zero pressure limit mea-sured with the above technique are 55.7 +0.5and 87.0 +3.0 nsec, respectively. " These com-pare favorably with the theoretical lifetimes of

question, v the relative velocity, n the number of .

X atoms per torr cm', and P~ the pressure of Xin torr. Provided v does not vary with Jt)x, thetluantity (Irv) can thus be obtained by keeping thehelium-gas pressure constant and varying P~.

Measurements were made at 1-torr helium pres-sure and 22.5-eV excitation energy" and for ex-citation pulse widths of less than 50 nsec whichgave a negligible contribution of cascading photonsto the O'P level from higher excited states, forexample, from the 4'S state. " The He(3'P) stateat 23.09 eV is then only weakly excited and oneloses up to two orders of magnitude of signal in-tensity compared with excitation at 23.0 eV by thistechnique. Figure 3 shows a typical He 5016-Aemission curve with and without added Ne. Asseen from the figure, data points for times greaterthan 110 nsec are not described in terms of asingle-exponential decay. It was found that alldecays studied fit, within statistical error, a func-tional form of the type

KUBOTA) DA VIES9 AND KING

I I

10 ~ He I torr0 ~

0o ~

0 ~ 0.590L 0 ~o ~

LL p Oe0"x LL .L 0 'e X rr

x L O ~A 0X L L 0 ~

I . 570 X& L p 4OLP X L L 00 L 0

0 X4 p ~

x A L 0o

L pX & L p ~X J 0

0 X A L 0 ~X L

X L L 00 ~X & L 00 WeL ~

P x Q L 000 ~ ~0 XP X0 x LI~

OO eP x L L 00 ee

X 0 ~x ~ A 00 o ~ ~

0

Xxx L L op

xx LP XX

AL poQ "x ~ 0

xX L

X

a X@ 0 0

aIjP o

""X A LX

DCQ x

0 X0

+ Ne 033 torr

0.81

EA IO-C

C0o IO—

CL

IO0

I I

IOO I 50

Time (nsec)

I

200X

I

250

FIG. 3. Decay functions for He(3 ~P) atoms observedby the 3 ~P 2~P, 5016-A emission line for varied setpressures of Ne.

55.2 and 89.7 nsec'4 and with 55.2 +2.8 and 84.0+10.6 nsec measured by Bennett et al." The pres-sure dependences of the lifetimes of the He (n =3,n =4) states in pure He have been extensively stud-ied by us" and other workers. Figure 4 shows theplot of the decay rates as a function of added gaspressure, the error bars include the standarddeviations of the data points. The plotted pointslie reasonably mell on straight lines over the mea-sured P~ region. Any systematic errors, whichare expected to be due primarily to possible errorsin calibrating the time-to-amplitude converter,are estimated to be less than 2%.

The measured values of n(vv) for He(3'P) atomsreacting with X are shown in Table I. The velocity-averaged total deexcitation cross section a, p, atan absolute temperature T is obtained by dividingn(vv) by noT '(2kT/M*)' ', and these are alsolisted in Table I. Here, & is the Boltzmann'sconstant, M* the reduced mass of the collidingpartners, and no is the number of atoms per cm'torr at 273'K. The error values quoted in Table Iinclude statistical errors and systematic errorsof +15%/-8% due to the uncertainty in the deter-mination of the atom temperature. The tempera-ture of the atoms in the excitation chamber wasestimated to be 600 +

goo K by measuring the tem-

0.5I

1.0I

1.5I

3.5—

I

OOPV)

o 3.0—

Ne (/

"g.ox IQ

x (~i(

2.00

2.QI

0.5 1.0&dded Gas Pressure ( torr )

I

Q.5

1.5

FIG. 4. Variation of He(3 ~P) decay rate with pressureof added Ne, Ar, Kr, and Xe; initial He pressureis 1 torr.Standard-deviation errors are shown for the data points.

peratures of the parts of the apparatus in andaround the region of the excitation chamber. Theuncertainty in determining the temperature ofthe colliding atoms in the excitation chamber isthe main source of error and is included in thefinal result for the deexcitation cross section.

DISCUSSION

The cross sections reported here include allprocesses that deexcite the He(3'P) atom in acollision with the added noble-gas atom. In allof these collision pairs the excitation energy of theHe(3'P) atom exceeds the ionization potential ofthe target. Thus, the initial state of the collisioncomplex is imbedded in a continuum and is subjectto rapid autoionization. This process (chemi-ionization) is expected to be the predominant modeof deexcitation under these conditions. We thusassociate our measured cross sections primarilywith the chemi-ionization process and, to ourknowledge, they are the first to be measured foran optically allowed excited state.

The ionization process due to the optically-allowed transition A* + X-A + X' + e has beentheoretically studied by Katsuura, ' Watanabe and

suura9MorjandPuj ita and Mori Us ingsemiclassical calculation with rectilinear relativemotion with constant velocity, Watanabe and Kat-suura" have found that the ionization cross sec-tion, averaged over the polarization of the initialexcitation of A*, is given by

] 3 9(~2~2 /gv )2/5 (4)

NONMETASTABLE PENNING IONIZATION IN He(3 'P )- Ne, . . .

TABLE I. Experimental values of s (ov) and comparison of the He(3 P) deexcitation crosssection at 600+1200 'K from experiment and theory.

Addedgas

Xe

s(ov)(10 sec "1torr )

0.77 + 0.04

1.49+ 0.16

1.29+ 0.06

1.88 + 0.16

O'CXP t0-16 cm2)

27.6'4, ,'

55 6+10+6-7e8

49 5+8+ 0

73 0+12,9-8e8

0 calc

(10 16 cm2)

19.8-20.7

39.8-40.5

38.2-40.5

32.8-37.4

0'ph

10 '8 cm2)

6.4-7.135-37

29.5-35.521-26

p@2

(a.u. )

0.93-1.03

5.1-5.44.4-5.23.0-4.1

References 20 and 21.References 20, 22, and 23.

References 20 and 24.References 20, 24, and 25.

where p. is the transition dipole moment for A*-A. , p, ~ is the corresponding moment for X-X'+ e, and v is the velocity of A relative to X.is related to the photoionization cross section,&p)„.by

c» =4m'Ep, s/hc, (5)

where c is the velocity of light and E is here takento be the excitation energy of A*. The reactionrate constant (ov), i.e., the mean value of vvaveraged over the Maxwellian velocity distribu-tion, can be derived from Eq. (3):

He(2'S) and He(2'S) at 320'K to be, in Ar: 0.21and 0.15, in Kr: 0.13 and 0.17, and in Xe: 0.02and 0.11. Considering these measurements we

may conclude that most of the difference betweenthe present experimental results and theoreticalvalues for Ne, Ar, and Kr can be attributed tothe contribution from associative ionization. Thelarge difference for Xe is not explained by thatprocess.

There are two other possible effects that can beconsidered to explain the difference for xenon.One is that the He(3'P) atom is also removed bycollisional deexcitation transfer:

He(3 'P) +X-He*+X + KE. (6)

(6)

with I*being the reduced mass of the collidingpartners. The calculated cross section, o,.„, isobtained as (o'v)(2kT/M*) '~'. The calculated re-sults for He(3'P)+X-He(1'S)+X'+e at 600+',«K are given in Table I. A value, 0.0433 a.u. ,is used for p.' between the ground state and theexcited state 3'P of helium. ' The values of (Tph

have been deduced from experimental photoioniza-tion cross sections. ""

Table I shows that the present measurements ofthe total deexcitation cross sections are largerthan 0'c,i. The difference is significantly large forXe. The associative ionization

He(3'P)+X- HeX'+e

may contribute to the total deexcitation crosssections. Munson et al."mass spectrometricallyobserved all of the heteronuclear ions (HeX') inhelium-X gas mixtures excited by low-energyelectrons and they obtained a rough estimate of theratio of cross sections for associative ionizationto Penning ionization to be of the order of 0.1 to0.5. Hotop et al."obtained the ratio for associa-tive to Penning-ionization cross sections for

It is characteristic of this process" that such ex-changes of internal and kinetic energy (KE) areonly probable for energy differences in those caseswhere the He* state considered is another singletstate in close energy resonance with the originalHe(3'P} The follow. ing reaction is the only im-portant collisional deexcitation mechanism underthe present conditions,

He(3'P}+X-He(3'D)+X+KE.

At present, to our knowledge, there has been nomeasurement of the cross section for process (9).In pure helium the cross section for He(3'P)+He-He(3'D)+ He+KE has been measured to be inthe range (3-35)&&10 "cm' by Bakos and Szigeti, ms

Frish and Ionikh" and Wellenstein and Robertson. 'Experiments have been carried out on He-Xe andHe-Kr mixtures as a function of Xe and Kr pres-

0sure with observation of the 6678-A emission fromthe 3'D-2'S transition in He. Over the pressurerange for which the change in He(3'P) lifetime wassignificant as shown in Fig. 4, no increase in6676-A (3'D O'S) intensity was o-bserved. Sincethe population of the He(3'D) following one excita-tion pulse is small compared to that of He(3'P)reaction (9), if of primary importance, would

KUBOTA, DA VIE S, AND KING

noticeably change the He(3'D} population. Wecon-clude that this reaction under those conditions isa secondary effect.

The second possible process is that the calcula-tion, assuming rectilinear relative motion, isnot a good approximation in the case of xenon.Because of the large polarizability of xenon, thevan der Waal's force between He(3'P} and xenonatoms is large and this attraction may play animportant role such that the relative trajectoryis strongly curved inward. Rectilinear motion isa good approximation for Ne, Ar, and Kr in whichthe polarizabilities are smaller than in Xe. Thedifferences between the observed and calculated

cross sections may be a fruitful field of furthertheoretical analysis together with additional ex-perimental results.

ACKNOWLEDGMENTS

The authors thank the Science Research Councilfor funding for part of the equipment, for a SeniorVisiting Fellowship to S.K. , and a Research Stu-dentship to C.D. S.K. is also grateful to Dr. T.W'atanabe for stimulating discussions and he ismost indebted to Dr. J. B. Birks and the Cottamsfor their encouragement during his stay in Man-chester.

*On leave from St. Paul's {Bikkyo) University, Nisi-Ikebukuro, Tokyo, Japan.

~Present address: Cambridge Consultants Limited,Cambridge, England.

~E. E. Ferguson and H. S, Schluter, Planet. Space Sci. 9,701 (1963).

2W. R. Bennett, Jr. , Appl. Opt. Suppl. 2, 3 (1965).3J. Dutton and J. M. Powell, J. Phys. B 4, 1506 (1971).4W. P. Jesse and J. Sadauskis, Phys. Bev. 100, 1755

(1955).5The present state of knowledge regarding chemi-ioniza-

tion reactions by metastable-state atoms has recentlybeen reviewed by B. D. Bundel and R. F. Stebbings, inCase Studies in Atomic Collision Physics, edited byM. B. C. McDowell and E. W. McDaniel (North-Holland,Amsterdam, 1972), Vol. 2, p. 547. See also Radiat.Res. 59, 337 (1974).

8S. Kubota, J. Phys. Soc. Jpn. 29, 1017 (1970).YC. E. Melton, G. S. Hurst, and T. E. Bortner, Phys.

Rev. 96, 643 (1954).K. A. Bridgett, T. A. King, and B. J. Smith-Saville,J. Phys. E 3, 767 (1970).

9C. Davies and T. A. King, J. Phys. E (to be published).~ K. A. Bridgett and T. A. King, in Proceedings of the

International Conference on Optical Pumping and AtomicLine Shape, edited by T. Skalinski (Panstwowe Wydaw-nictwo Naukowe, Warsaw, 1969), pp. 359-368.

~~S. Kubota, C. Davies, and T. A. King, J. Phys. B (tobe published) .

~2This energy is lower than the energy level of He(3 ~P),i.e., 23.09 eV. Since the energy spread of the electronsfrom the gun is about 0.2 eV, He(3 ~P) states are ex-cited by the electrons which are the high-energy tail ofthe electron energy distribution.

~3W. R. Bennett, Jr. , Ann. Phys. (¹Y.) 18, 367 (1962).~4W. L. Wiese, M. W. Smith, and B. M. Glennon, Atomic

Transition Probabilities, NSRDS-NBS-4 (U. S. GPO,Washington, D. C. , 1966), Vol. 1.W. B. Bennett, Jr., P. J. Kindlmann, and G. N. Mer-cer, Appl. Opt. Suppl. 2, 34 (1965).

~~K. Katsuura, J. Chem. Phys. 42, 3771 (1965).~T. Watanabe and K. Katsuura, J. Chem. Phys. 47, 800(1967).

~8M. Mori and H. Fujita, J. Phys. Soc. Jpn. 20, 432(1965).

~9M. Mori, J. Phys. Soc, Jpn. 21, 979 (1966).OJ. A. B. Samson, in Advances in Atomic and MolecularPhysics, edited by D. B. Bates and I. Estermann (Aca-demic, New York, 1966), Vol. 2, p. 177.

2~A. Pery-Thorne and W. R. S. Garton, Proc. Phys. Soc.Lond. 76, 833 (1960).J. A. R. Samson, J. Opt. Soc. Am. 54, 420 (1964).O. P. Rustgi, J. Opt. Soc. Am. 54, 464 (1964).

24O. P. Rustgi, E. I. Fisher, and C. F. Huller, J. Opt.Soc. Am. 54, 745 (1964).

~H. E. Blackwell, G. S. Bajwa, G. S. Shipp, and G. L.Weissler, J. Quant. Spectrosc. Radiat. Transfer 4, 239(1964).

28M. S. B. Munson, J. L. Franklin, and F. H. Field,J. Chem. Phys. 67, 1942 (1963).

2~H. Hotop, A. Niehaus, and A. L. Schmeltekopf,Z. Phys. 229, 1 (1969).J. Bakos and J. Szigeti, J. Phys. B 1, 1115 (1968).

29S. E. Frish and Yu. E. Ionikh, Opt. Spectrosk. 25, 9(1968) [Opt. Spectrosc. 25, 171 (1968)].H. F. Wellenstein and W. W. Robertson, J. Chem. Phys.56, 1072 (1972).