Enhanced Negative Ion Formation via Electron Attachment to

Electronically-Excited States

Lal A. Pinnaduwage

Health Sciences Research Division Oak Ridge National Laboratory

P. 0. Box 2008, OakRidge, Tennessee 37831-6122 and

Department of Physics, University of Tennessee, Knoxville Knoxville, Tennessee 37996

Abstract. Recent basic studies on electron attachment to laser-excited molecules show that electron attachment to electronically-excited states can have orders of magnitude larger cross sections compared to the respective ground electronic states. Even though systematic studies have not been conducted, there are indications that electronically-excited states may play a significant role in negative ion formation in gas discharges. The high-lying Rydberg states could be of particular sign5canc.e since, (i) their production efficiencies are high, and (ii) they have comparatively long lifetimes. Such states could be populated in discharge sources via direct electron impact or via excitation trander fiom metastable states of inert gases.

I. INTRODUCTION

Electron attachment to electronically-excited molecules is a field of recent origin in contrast to electron attachment studies on ground electronic states that have been conducted for almost a century. Since 1950’s electron attachment studies on vibrationally-excited states of the ground electronic states of molecules have been conducted; these studies have clearly shown that dissociative electron attachment to molecules is generally enhanced when the rotatiodvibrationd energies are increased. In the well-known example of H, dissociative electron attachment rate constant increased by 4 orders of magnitude when H, was excited fiom the lowest vibrational level ( ~ 0 ) to the u=4 level [ 11.

Dissociative electron attachment to a molecule, AB can be represented as a two- step process,

(Jc AB-* ----- A - + B &?(E) + e - ( € ) -----

where a, is the cross Section for the capture of an electron to form the transient parent anion AB" which subsequently dissociates with probability p to form the fragment anion A'; E and E denote the internal energy of AB and the energy of the attaching electron respectively. The cross section for dissociative electron attachment, a,, can then be written as,

For a molecule AB with large internal energy (E), 0, can be expected to be large

(i) the large polarizabiity of the excited molecule will enhance the initial capture [ 11

(ii) the large internal of the transient species AB" will lead to a high dissociation

Therefore as the internal energy of a molecule is increased (see Fig. 1), one can expect the dissociative electron attachment cross section to increase.

Compared with excitation energies of the order of 0.01 electron volts (ev) for rotationally- excited states and 0.1 eV \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ for vibrationaliy-excited 1.p. \\\\\\\\\ states, the electronically- excited states have energies of a few eV (see Fig. 1); therefore, the electronically- excited states can be expected to display even larger enhancements. However, due to the large excitation energies involved, ultraviolet 0 lasers are needed to populate such states. Even though such lasers (especially the excimer lasers) became available in

since both uc and p.will be large:

of the electron (ac),

probability (p).

MOLECULAR STATE

SUPEREXCITED

'...UIIIII-

ELECTRONICALLY EXCITED - STATES

HIGH-VIBRATIONAL GROUND STATES - ELECTRONIC STATE

LOW-VIBRATIONAL .- STATES

FIGURE 1. A schematic energy level diagram of a molecule.

the late 1970'~~ the first experiments on electron attachment to laser-excited electronically-excited states were reported only in the late 1980's. A brief review of the electron attachment studies that have been conducted up to now will be given in Section II.

It must be noted that the first study on electron attachment to an electronically- excited molecule was conducted on the low-lying (energy - 0.98 ev), long-lived (lifetime - 2700s) a 'Ag state of 4 produced in a microwave discharge [3]; an enhancement in cross section of factor of - 3 was observed for this state compared to the ground state of 0,.

The basic electron attachment studies on electronically-excited states show that negative ion formation via these states can have very large cross sections, which can be > cm2 for highly-excited states, see Section II. Therefore, it is important to address the obvious question of whether this mechanism for negative ion formation play a significant role in gas discharges, including those used for H' formation [4,5] and plasma processing [6]. The possible role of electron attachment to electronically- excited states in a novel pin-hollow cathode discharge source [7] used for H' ion formation was recently pointed out [SI. Mechanisms for production of excited electronic states and their role in negative-ion formation in gas discharges will be discussed in Section III.

11. ENHANCED ELECTRON ATTACHMENT TO LASER- EXCITED MOLECULES

A. Studies on Low-Lying Electronically-Excited States

Threshold energies for electronic excitation of small molecules are typically a eV; thus, ultraviolet 0 lasers such as the excimer laser are ideal for populating these states. The first studies on electronically-excited states were conducted on low-lying (mainly the first-excited) states, where the excited state was populated by the absorption of a single UV laser photon. Only a few electron attachment measurements have been reported on low-lying electronically-excited states up to now: (i) about 5 orders of magnitude enhancement (compared to the ground electronic state) for the first excited triplet state (excitation energy - 4.4 eV, lifetime - 8 ms) [9] (ii) about 2 orders of magnitude enhancement for the A '2+ (H) state of nitric oxide (excitation energy - 5.4 eV, lifetime - 200 ns) [lo] (iii) five to seven orders of magnitude enhancement for the fist excited triplet state of p-benzoquinone [ 111, (iv) about two orders of magnitude enhancement for a mixture of 'A, and 'B, states of sulfur dioxide (excitation energy - 4 eV, lifetimes - 50 and - 80-530 p) [12] and, (v) about 7 orders of magnitude enhancement for a mixture of u=2 ,3 states (excitation energies - 6.4 eV, lifetimes - 200 ns) of the A *E+ state of nitric oxide [13]. These studies have clearly established that electronically-excited states in general

have much larger cross sections for electron attachment compared to the corresponding ground states.

Even though negative-ion formation via electron attachment to low-lying electronidy-excited states is of fundamental significance, it is probably of little value with regards to gas discharges since the relative populations of these states in gas discharges are expected to be small, see Section III. However, the electron attachment studies on highly-excited electronic states (located around the ionization threshold) conducted within the past few years show that the associated cross sections are even higher than those for the low-lying states. Furthermore, these highly-excited states have the largest oscillator strengths and thus are likely to be populated with large number densities in gas discharges; they are likely to have comparatively long liietimes as well, see Section III. Therefore, we will concentrate on electron attachment studies on highly-excited states lying close to the ionization thresholds of molecules; these states are normally Rydberg states.

B. Studies on Highly-Excited Electronic States

In 1991, we developed a new experimental technique [14] for the measurement of electron attachment to highly-excited electronic states; using this technique we have observed efficient negative-ion formation in triethylamine [14a], NO [14b], H2 [4,15], Sa [6(a)], and CH, [6(b)].In the following we will restrict our discussion to the H’ formation in ArF excimer-laser-irradiated H, observed [4,15] using this technique.

In this experiment, a single laser pulse produced the attaching electrons via laser photoionization and also produced the excited molecules responsible for e l e c t r o n attachment. The e x p e r i m e n t a l arrangement is shown in Fig. 2. The laser pulse p r o p a g a t i n g between electrodes 1 and 2 excited the molecules to energies above their i o n i z a t i o n thresholds via m u 1 t i p h o t o n

PULSE

VOLTAGE AMPLIFIER

FIGURE 2. Schematic diagram of the laser/elec&ode arrangement used in the “charge detection” mode [ 14(a)]

absorption (the excitation scheme for H2 irradiated by ArF excimer laser is shown in Fig. 3). This simultaneously produced fiee electrons (by ionization) and also superexcited states of the molecules; superexcited states (SES) are electronically- excited states with total internal energies exceeding the ionization energy of a molecule, see below. The electron attaching species could in principle be the SES themselves; however, the experimental observations are more consistent with highIy- excited bound states that are indirectly populated via the SES [16]. Since the electrons and the excited states were populated concomitantly and in close proximityy electron attachment occurred quickly and efficientiy. The negative ions and the unattached electrons were separated fiom the positive ions (produced by laser photoionization) and were extracted to the detection region located between electrodes 2 and 3 through a grid in electrode 2; this was accomplished by applying suitable electric fields in the two regions.

The laser intensity dependence of the measured total signal, V,, and the negative ion signal, VI, are shown in Fig 4 for 6.67 @a (50 Torr) of H2 at an applied field of 50 V cm”; V, is the.sum of VI and the signal due the unuttached electrons, VE. (The total signal is proportional to the number density of electrons initially produced via photoionization; some of these electrons are converted to negative ions via attachment in the laser-irradiated region). At low laser intensities I, V,=V,, but with increasing I the negative-ion signal grew rapidly at the expense of the electron signal, see Fig. 4; at high& the signa consisted almost entirely of negative ions.

For the data of Fig. 4, the negative-ion density corresponding to VI =lo0 mV is - 3x10’ ions cm3 ; cross section of the laser pulse was - 0.1 cm2 and the “length” of the interaction region was - 10 cm, i.e., the detected negative ions were produced over a volume^' of - 1 cm3. The “width”, d, of the laser pulse in the direction of electron drift was 0.1 cm, and the gap ,ds, between the laser pulse and the grid in electrode 2 was also 0.1 mq i.e., while the electrons produced at the “ t ~ p ’ ~ of the laser pulse traversed the maximum distance of 0.1 cm across the laser-irradiated

251 . I

20

- 15 > al

>. 0 lz w z 10 w

Y

5

0 0 1 2 3 4 5

INTERNUCLEAR DISTANCE (A)

FIGURE 3. A schematic energy level diagram of & relevant to the discussion in the text.

region, those produced at the “bottom” of the laser pulse left the laser-irradiated region immediately (see Fig. 2). Therefore, the maximum time an electron spent in the laser-irradiated region was dL/we, where we is the drift velocity of the electrons. For example, for the data of Fig. 4 at 50 V cm“ applied electric field , we - 4x106 cm s and thus the maximum interaction time for an electron was - 25 ns, Le., electron attachment must have occurred in s 25 ns.

The high efficiency with which H ions are populated in ArF-laser-excited H2 was recently illustrated: Kiekopf [22a] had observed laser emission in aluminum when AI was irradiated by the ArF laser in the presence of H,. Recently it was shown [22b] that efficient H: formation via electron attachment to ArF-laser-excited H2 was responsible for the charge neutralization of A+ that led to the population of the lasing Al state. Lasing in AI was observed within a few ns following the ArF pulse; therefore, H: formation must have been completed within that time.

The new data shown in Fig. 4 have different laser intensity, I, dependences compared to the data shown in Fig 2 of ref 15. The new data set taken at 50

16; the data of Fig. 2 of ref 15 taken at 5 TOIT of H, showed VT =I? andV,=I‘. This is due to the fact that at high H2 pressures, collisional quenching of the E,F state [23] dominates the up- pumping fiom the E,F state. Therefore, the actual three- photon dependence for transitions to energies above the ionization threshold is displayed in Fig. 4 in cotnrast to an effective two- photon dependence observed for the data of Fig. 2 of ref 15; details of a rate equation analysis will be given later; the estimated electron attachment cross sections are > 1012 cm2 (rate constants > IOd em's-'>.

We have recently developed an experimental technique [13] to monitor electron attachment to laser-excited molecules in real

TOIT of H2 show VT I3 and VI a 1000

5

a (3

E A

1

cn

v

1 024 i 025

I (photons cm” s-‘)

FIGURE 4. Laser intensity, I, dependence of the meam& total, V,., measured negative ion, VI, and normalized negative ion, (VJn [14a] signals for the experimental parameters indicated in the figure.

time, which allows us to infer the lifetimes of the electron attaching states. The experimental arrangement is shown in Fig. 5. Only two electrodes are employed and the laser pulse propagated in between the electrodes and parallel to them. As in the previous technique, this laser pulse produced both the excited molecules and the attaching electrons in the laser-irradiated region. A negative voltage was applied to the top electrode as shown in Fig. 5 so that the electrons drifted toward the bottom electrode; the current induced by the motion of the electrons in the drift gap was monitored directly by connecting the bottom electrode to a fast digitizer with 50 Q input impedance. The time constant of the detection circuit was - s and thus the monitored signal voltage was directly proportional to the current carried by the electrons. By Varying the gas pressure in the chamber and/or the applied electric field, the electron drift time through the laser-irradiated region could be varied fiom a few ns to hundreds of ns. If the electrons were lost via attachment during their drift, this was manifested as a reduction in the electron current. It was extremely important to reduce the electrical noise in order to observe the electron attachment in ns time scale.

Signal waveforms obtained using the new experimental technique [13] for 50 TOK ofH2 irradiated by ArF excimer laser are shown in Fig. 6. Figure 6(a) shows a signal waveform at a low laser intensity where no significant electron attachment occurred during the drift through the laser-irradiated region; the electron current reached the maximum value over the duration of the laser pulse [ laser temporal profile is shown in Fig. qd)], and then remained constant until the electrons fiom the closest edge of the laser pulse reached the bottom electrode. The linear decay of the current after - 125 ns [see Fig. 6(a)] was due to the gradual loss of electrons to the electrode. Figure 6(b) shows a signal waveform that resulted when significant attachment of electrons to the excited H2 molecules as the electrons drifted through the laser-irradiated region; the decrease occurred for - 60 ns and then the current remained I Y constantduring the rest of the drift until the electrons reached the lower electrode. The time over which the electron current decreased during their drift is a lower limit to the lifetime of the electron

Q=*:+-1 I ,/

CURRENT AMPLIFIER

J?IGURE 5. Schematic diagram of the lasedelectrode arrangement for the “current detection” mode [13].

attaching species [ 131. Therefore, we can conclude that under the experimental conditions of Fig. 6(b), the lifetime of the electron attaching species must be at least

60 ns. When the laser intensity was hrther increased, electron current decreased rapidly, see Fig. 6(c); this ocmred basically within the duration of the laser pulse [see Fig. k(d)] due to- the rapid attachment of electrons to a high number density of excited states populated at this laser intensity.

The above inferred lifethe of - 60 ns is close to the radiative lifetime (- 100 ns) of the intermediate E,F 'E; state populated by the laser via two-photon absorption, see Fig. 3. However, the E,F state is collisionally-quenched by ground state H2 with a rate constant of - 2xiO9 c d s' [23]; therefore, the effective lifetime of the E,F state at a H, pressure of 50 Torr should be - 0.3 ns. Furthermore, preliminary experiments show that the effective lifetime of the electron attaching species is enhanced at higher pressures. Even though a firm conclusion cannot be made yet, it appears that the observed electron attachment is due to high-lying bound Rydberg (HR) states of H2 populated indirectly via SES; hrther work is in progress in our laboratory.

So far, we have observed efficient dissociative electron attachment in a variety of molecules excited to energies above their ionization

20

15

10

5

0

I = 2.4~10" cnf2 s-'

I = 5.0~10" cm-2 s-'

0.05 I , , I 1 I I i P (4 g 0.04 - LASER PROFILE -

5.g 0.03 - - !z= 0.02 - T~ - 25 ns (FWHM) - US !i O-O' - -

0.00 ry"f 0 100 200 300 400 500

TlME (ns)

FIGURE 6. (a)-(c) Signal waveforms in the "cument detection mode" for the experimental parameters corresponding to those of Fig. 4 and for laser pulse energies indicated (see the text for details). (d) laser temporal profile.

thresholds; we believe that HR states populated indirectly via superexcited states is responsible for those observations as well. We did not observe negative-ion f o d o n in N2via this process; the negative result for N, is reasonable since the N

atom cannot bind an extra electron, i.e., electron afEnity of N is negative. The dissociative attachment process of (1) above is energetically possible when,

E + E > D(AB) - EA(A) (3)

where D(AB) is the dissociation energy of AB, and EA(A) is the electron affinity of A. Dissociation energies of molecules are usually - 5 eV; ionization thresholds for molecules (and hence E for SES and for HR states) are almost always higher than 7 eV. Hence electron attachment to HR states of molecules is energetically possible for any electron energy including thermal energy electrons, where the capture cross sections are quite large. As we mentioned earlier, the initial electron capture cross section is generally large for excited states due to their high polarizabilities. For Rydberg states the polarizability is expected increase as n', where n is the principle quantum number. Even though polarizabilities of molecular Rydberg states have not been reported to our knowledge, enormous polarizabilities of > lo9 A were reported for Rydberg states of Cs [24].

In addition to the enhancement associated with the internal energy of a molecule, dissoCiative electron attachment is also enhanced at lower electron energies; this has been confirmed by our laser studies [ 141 as well as by other studies on vibrationally- excited states [l]. This is due to the large capture cross sections, u, , associated with the slow electrons, see equation (1). Therefore, the optimum conditions for dissociative electron attachment are,

(i) large excitation energy for the capturing molecule, (ii) low energy for the electron being captured.

111. NEGATIVE ION FORMATION IN GAS DISCHARGES

EXCITED STATES VIA ELECTRON ATTACHMENT TO ELECTRONICALLY-

A. Background

In the recent years evidence has accumulated in at least two research areas involving gas discharges, that higher than expected negative-ion densities exist in discharges of certain gasedmixhues : (i) in H' discharge sources that generate H- ions, and (ii) plasma processing discharges that use weakly electronegative gases such as silang (see references in [6]). In this paper we will restrict the discussion to (i) above. Enhanced electron attachment to rotationallyhibrationally-excited states of H2 was reported [l] at about the same time the efficient H' formation in H2 discharges was discovered [26]. Ever since that time, an exhaustive experimental and theoretical effort has been devoted to understand the experimental observations in terms of this

electron attachment process (see the reviews in [25]); even though most model calculations had been able to reasonably reproduce the experimental observations, results of recent studies [27] cast doubt as to whether the experimental measurements can be completely understood within the existing model, i.e., electron attachment to rotationally/vibrationally-excited states. It must be noted that many other possible channels of H' formation have been critically-evaluated [25] including electron attachment to the long-lived (lifetime 2 100 ps) c 314 state of H2 [28].

Possible significance of electron attachment to high-Rydberg states for H' formation in H, discharges was first pointed out in 1994 [4]. A recent preliminary calculation [5] has provided supporting evidence for this possibility. Since electron attachment to highly-excited electronic states ofH, have been shown [4,15] to have rate constants > lo6 cm3se1 compared to a maximum rate constants of - 10' c d s ' for the rotational/viirational states, it is quite possible that such electronic states make a significant contribution to negative-ion formation in H, discharges provided that, (i) they are populated in &cient number densities, and (ii) they have sufficiently long lifetimes under the discharge conditions.

B. Formation Mechanisms of Electronically-Excited States in Gas Discharges

a) Electron Impact Excitation

In discharge sources using unitary gases, excitation of electronically-excited states can expected to occur predoGnantly via electron impact excitation of ground-state molecules. Platanan had shown [ 17a] that , in the optical approximation, most of the oscillator strength for a given molecule is associated with transitions to energies around the ionization threshold, and that the transition across the ionization threshold is smooth, see the excitation spectrum of CH, [17a] in Fig. 7. The HR states lying close to the ionization threshold are especially significant since they have high oscillator strengths and, also, long Wetimes (measured under collision-free conditions) of the order of lo6 s 1291. Even those highly-excited states lying above the ionization threhold may be of importance since our studies (see Section II.B) that they may indirectly populate (via internal conversionl collisional stabilization) HR states that are comparatively long lived, see Section II. B.

However, the optical approximation is reliable only for high-energy electrons where electrons behave like photons in the excitation process and the optical selection rulesare obeyed. At low impact energies (close to threshold energies) electron impact

excitations do not follow optical selection rules. In discharge sources’ used for negative-ion formation, low-energy electron impact is obviously important; it is particularly important for the excitation of metastable states.

Therefore, in order to assess the possible cotributions of electronically-excited states in gas discharges accurate calculations of cross sections for electron impact excitation of HR states are needed.

0.50

0.40

0.30

0.20

0.10

6 8 10 I2 14 16 18 20 22 24 26 28 30 32 0

E. dV

FIGURE 7. Excitation spectrum of CH, [17a].

b) Excitation Transfer

Metastable states of inert gases have fairly high excitation energies and long lifetimes [30]. In contrast to their ground states, they are highly reactive; they can transfer their energies to a wide variety of molecules with cross sections of the order of 10-l6- cm * [3 11 leading to excitationdfiagmentationdionization of these molecules. Therefore, in gas mixtures that use inert gases, excitation transfer (sometimes called the Penning process) may play an important role. The following example illustrates this possibility.

Iiplka et al [A recently reported efficient H’ formation in CH, using a novel pin- hollow cathode discharge [32a]: In these experiments a flowing Ar plasma was * produced using this hollow cathode where Ar plasma was fed into the discharge through a hole in the cathode. This plasma was extracted into a “target region” through a hole in the anode; it was shown [32a] that the energy of the extracted electrons in the target region could be lowered by extending the pins in the cathode. When CH, was fed into the target region efficient H’ production in that region was obse.rved which was seen to optimize at low electron energies, Le., at long pin lengths [7,32a]. It is important to note that CH, did not have to be subjected directly to the discharge . It was shown [7] that of the electrons that were present in the target region in the absence of CH,, 7040% were converted to H’ ions when Cp was added; H’ ion densities of the order of 10 cm-3 were reported in the target region

when the electron temperature was < 0.5 eV [7,32a]. The above observations cannot be explained via electron attachment to the ground

state of CH, since that process is extremely weak [33]. We recently measured electron attachment to CH, at temperatures up to - 700K at electron energies in the range 0- 10 eV, we did not observe [34] enhancement in electron attachment to vibrationally- excited states of CH,. Iizuka et al[7] proposed that the observed H' formation was due to electron attachment to vibrationally-excited H, populated via dissociation of CH,. However, at least in some of the experiments [32a], C q was not subjected directly to the discharge and hence significant dissociation of CH, could not have occurred.

We pointed out [8] that the following sequence of events can explain the observations of the above experiments:

(i) the flowing plasma carried the long-lived metastable Ar atoms, Ai" (energies - 11.7 ev), to the target region, (ii) highly excited CH,' states of CH, (ionization threshold - 12.6 eV) were

produced via efficient excitation transfer [3 13 from Ai" atoms,

Arm + CH, -+ CH,' + Ar (4)

(iii) low-energy electrons in the target region attached dissociatively to CH i,

CH,' + e - 4 H - + CH,

, consistent with our observations [6b] of efficient H - formation via attachment of low-energy electrons to laser-excited states of CH,; our proposed mechanism is also consistent with Iizuka et al. 's observation [32b] of efficient CH, radical formation in the same discharge source, see eq. ( 5 ) above.

C. Lifetimes Of High Rydberg States in Gas Discharges

* Rydberg states are traditionally viewed as fragile species. However, recent [35] as well as early [36] studies show that especially the high Rydberg (HR) states are relatively stable against collisions with neutral, non-polar species. At first this may seem unlikely since HR states have very small binding energies (a HR state of principal quantum number, n, of 50 has a binding energy of only - 5.4 mew. However, the key point is that neutral particles do not interact with a HR state as a whole. This was first pointed out by Fermi [36] in explaining the level shifts of

absorption lines of the Rydberg states of alkali atoms in 1 atmosphere of an inert gas; even though the levels shifted they were not significantly broadened, Le., the lifetimes of the Rydberg states were not significantly affected. The radius of a high-Rydberg atom increases as n3; thus a state with n=50 cotnains about 10' inert gas atoms within the orbit of the Rydberg electron if it is placed in an inert gas at atmospheric pressure. Hence the Rydberg electron and the ion core interact independently with the inert gas atoms (see Chapter 8 of ref. 35 for a detailed discussion).

High-Rydberg states of H2 produced via electron impact have been observed with lifetimes of - 100 ps [29]. Such long-lived HR states are thought to be high orbital angular momentum (high-C) states populated via electron impact near threshold energies [29,37]. Lifetime of the order of ps have been observed for molecules laser excited to energies around [19] and even well above [20] their ionization thresholds. These above mentioned experiments were conducted under collision-fiee conditions. In our electron attachment experiments on molecules laser-excited to energies above their ionization threshold, HR states populated indirectly are seen to have lifetimes of tens and hundreds of ns at ambient pressures of several E a . Preliminary measurements 'using a new experimental technique [13] show that the effective lifetimes may be Zengthened at high ambient pressures; this could be due to the collisional stabilization of vibrationally-excited core of the HR state. Further experiments are in progress.

High Rydberg states are quenched by fairly small electric fields; a HR state of n=50 is quenched by an elect& field of - 160 V cm". In our experiments we observed [4,6,14,15] a rapid decrease in negative-ion formation with increasing electric fields; at the moment it is not clear whether this is due to the field-quenching of the HR states or due to the field detachment of the transient parent negative ion [see equation (l)].

IV. CONCLUSIONS

Recent studies on electron attachment to laser-excited molecules have shown that highly-excited electronic states (high-Rydberg, HR, states)can have electron attachment cross sections that vastly exceed those for the ground electronic states; electron attachment rate constants for the HR states of H2 are estimated to be > lod cm3 s" . There is evidence that electron attachment to HR states may contribute significantly to negative-ion formation in gas discharges, including H' formation in H2 discharges. Further work, both experimental and theoretical, is needed to better understand and quantifjr the contribution fiom electron attachment to HR states in gas discharges.

ACKNOWLEDGMENTS

This research was sponsored by the Office of Health and Environmental Research, U.S. Department of Energy under Contract DE-ACO5-84OR21400 with Lockheed Martin Energy Systems, Inc., and by the National Science Foundation. I thank Dr. Panos Datskos for fhitful discussions.

REFERENCES

1.

2.

3.

4. 5. 6.

7. 8. 9.

10.

11. 12.

13. 14.

15. 16.

(a)Allen, M and Wong, S. F., Phys. Rev. Lett. 41,1791-1794 (1978); @) Wadehra, J. M. and Bardsley, J. N., Phys. Rev. Lett. 41,1791-1798 (1978). (a) Vogt, E. and Wannier, G. H, Phy5. Rev. 95,1190-1 198 (1954); @) mar, D., Ruf, M W., and Hotop, H., Chem. Phys. Lett. 189,448-454 (1992). ( a ) b w , P. D., J. Chem. Phy5.59,4922-493 1 (1 973);@) Belic, D. S., and Hall, R I., J. Phys.

Pinnaduwage, L. A. and Christophorou, L. G., J. Appl. Phys. 76,46-54 (1 994). Garscadden, A. and Nagpal, R, Plasma Sources Sci, Technol. 4,268-280 (1995). (a) L. A Pinnaduwage, M. 2. Martin, and L. G. Christophorou, Appl. Phy5. Lett. 65,257 1-2573 (1994); @)L. A Pinnaduwage, M Z. Martin, and L. G. Christophorou, Conbib. Plasma Phys. (to be published, 1995). Iizuka, S., K o M , T., Takada, T., and Sato, N.,AppZ. Phy5. Lett. 63,1619-1621 (1993): Phaduwage, L. A, Appl. Phys. Lett. 67,1034 (1995). (a) Christophorou, L. G. Bunter, S. R , Pinnaduwage, L. A , Carter, J. G. , Christodoulides, A. A. and Spyrou S. M. J’hys. Rev. Lett. 58, 1316-1319 (1987); (b) Pinnaduwage, L. A. , Christophoroy L. G. and Hunter, S. R , J. Chem. Phys. !IO, 6275- 6289 (1989). (a) Kuo, C. T., 0x10, Y. , Hardwick, J. L., md Moseley, J. T. , J. Phys. Chem. 92 ,5072-5074 (1988); (b)Kuo, C. T., Hardwick, J. L., and Moseley, J. T. ,J . Chem. Phys. 101,11084-11085 (1994). Mock, R S. and Grimsrud, E. P. ,J . Phys. Chem. 94,3550- 3553 (1990). J&, T., Hadmi, R, Christophorou, L. G., Illenberger, E., BaumgMel, R, and Pinnaduwage, L. A. Chem. Phys. Lett. 203,21-27 (1993). Pinnaduwage, L. A., and Datskos, P. G., submitted to J. Chem. Physics (1 995). (a)Phaduwage, L. A, Christophorou, L. G., and Bitouni, A. P. J. Chem. Phys. 95,274-287 (1991); @) Pinnaduwage, L. A. and Christophorou, L. G., Chem. Phys. Lett. 186,4-10 (1991); ibid. 189,486 (1 992). Pinnaduwage, L. A and Christophorou, L. G. Phys. Rev. Lett. 70,754-757 (1 993). When a molecule is excited to an energy above its ionization threshold, it may undergo direct ionization (where the excited electron is ejected within an electron orbital time of - l o ” s) or it may be excited to a superexcited state [17]; a superexcited molecule may decay via one of the following processes: (i) preionization, (2) predissociation, (E) fluorescence, and (iv) internal conversion. While preionizatiOn and predissociation have been generally accepted 8s the dominant decay channels for the SES, evidence has been accumulating slowly over the last decade that other decay channels may be active as wek a) K o i i [ 181 recently reported that preionization of SES of a number of organic molecules is negligible, i.e. the neutral decay channels predominate, b) there have been a number of recent reports [ 1 91 indicating long time stability of molecules with internal energies exceeding the ionization thresholds; lifetimes of several p have been reported for benzene molecules excited to - 5 eV above the ionization threshold PO],

B 14,365-373 (1981)

17.

18. 19.

20.

21. 22.

23.

24. 25.

26. 27.

28. 29.

30.

31.

32.

33.

34. 35.

36.

37.

c) Radiative emission h r n SES of IE, has been reported by Breton et d[21 J which means the lifetimes ofthe SES involved must have been in the ns range. Recent experiments by us indicate that either the SES themselves or other lower-lying states populated via collisional stabilization of SES have lifetimes of the order of 1 Oons, see below. (a)Platanan,R L., Vortex 23,372 -389 (1962); @)Berry, R S., andLeach, S., A h . Electron. Electron Phys. 57,l- 144 (1981).

Reiser, G., Habernicht., W., MMer-Defhlds, IC, Schlag, W., Chem. Phys. Lett. 152,119-123 (1988); Schlag, E.W., Grotemeyer, J., Levine, R D., Chem. Phys. Lett. 190,521-527 (1992); Schlag, W., andlevhe, R D.,J. Pbs. Chem. 96,10608-10616 (1992). Schzer, W. G., Selzle, H L., Schlag, W., 2. Naturfrsch. 48% 1256-1259 (1993); Scherzer, W. G. , Selzle, H. L., Schlag, W., andlevhe, R D., Phys. Rev. Lett. 72,1435-1438 (1994). Breton, J., Guyon,P. M, Glass-Maujean, M., Phys. Rev. A 21,1909-1913 (1980). (a)Kielkopf, J. F. J . Opt. SOC. Am. B 8,212-219 (1991); (b) Kielkopf, J. F., Pinnaduwage, L. A,

(a)I(liger,D. J. andRhcdes, C. IC, Phys. Rev. Lett 40,309- 313 (1978); (b) Pummer, H, Egger, H, Luk, T. S., Srinivasan, T., and Rhodes, C. K, Phys. Rev. A 28,795-801 (1 983). van Raan, A F. J., Baum, G., and Raith, W., J. Phys. B. 9, L349-L352 (1976). Hiskes, J. R, Commentr A t Mol. Phys. 19,59-77 (I 987); Bacal, M and Skinner, D. A, ibid. 23,

Bacal, M. andHamilton, G. W.,Phys. Rev. Lett 42,1538-1540 (1979). Eenshuistra, P. J., Gochitashvilli, M., Becker, R, Kleyn, A W., Hopman, H. J., J. Appl. Phys. 67, 85-96 (1990); (b) Berlemont, P., Skinner, D. A, Bacal, M., Rev. Sci. Instrum. 64,2721-2728 (1993); Young, A T., Clm, P., Leung, K. N., and Stutzin, G. C., " WV Absorption Spectroscopy of H' Ion Source Plasmas", Proceedings of the 6th International Symposium on Laser-Aided PlasmaDiagnostics, 1993, pp 178-183. Bottcher, C. andBuckley, B. D.,J Phys. B. 12, L497-LSOO (1979). (a) Tam, S. M.,Schiavone, 3. A andFreud, R S.,J Chem. Phys. 74,28692878 (1981); (b) Kpiyanov, S. E. Perov, A A,Zayats, 2. Yu and Stepanov, A N., Sov. Phys. JETP. 21, 3 11- 317 (1981). J.-L. Delcroix, C. M. Ferreira, and A. Ricard, in , ed. by G. Bekefi, John Wdey, New York, 1976, ch. 5, pp.159-234. (a) Piper, L. G., Velazco, J. E. and Sester, D. W, J. Gem. Phys. 59, 3323-3340 (1973); (b) Bourene, M. and Le Calve, J., J. Chem Phys. 58, 1452-1458 (1973). (a) Sato, N., lizuka, S., K o W , T. and Takada, T., Appl. Phys. Lett. 62,567-569 (1993); (b) Iizuka, S., Takada, T., and Sato, N., Appl. Phys. Lett. 64,1786-1788 (1994). sharp, T.E. andDoweU, J. T., J. Chem. Phys. 46,1530-1531 (1967); @)Davis, D. K.,Kline, L. E. andBies, W.E., J. AppLPhys. 65,3311-3323 (1989). Datskos, P. G. and Pinnaduwage, L. A, unpublished (1995). R F. Stebbings and F. B. Dunning, (Eds.) Rvdbere States of Atoms and Molecules (Cambridge University Press, Cambridge, 1983). (a) Fermi, E., Nuovo Cimento 11,157 (1934); (b) Amaldi, E., and Segrk, E., ibid. 11,145 (1934). U. Fano, J. Phys. B 17, LAO 1 (1 974).

K O W , R, J. Ch- Phys. 95,5846-5852 (1991).

and Chtistopho~u, L. G., Phy~. Rev. A 49,2675-2680 (1994).

283-299 (1990).

Thii report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsi- bility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Refer- ence herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recom- mendation. or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof. - ~- _-_ __

~