Government & Nonprofit

Absolute cross sections for dissociative electron attachment to NH 3 and CH 4

Description
Absolute cross sections for dissociative electron attachment to NH 3 and CH 4
Published
of 7
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  Absolute cross sections for dissociative electronattachment to NF 3 D. Nandi, S.A. Rangwala, S.V.K. Kumar, E. Krishnakumar* Tata Institute of Fundamental Research, Homi Bhabha Road, Mumbai 400 005, India Received 20 March 2000; accepted 6 June 2000 Abstract Absolute dissociative electron attachment cross sections for the formation of F  , F 2  , and NF 2  from NF 3  have beenmeasured in a crossed beams geometry by using the relative flow technique. Discrimination against ions with high kineticenergy are eliminated by employing a pulsed electron beam, pulsed ion extraction and a segmented time of flight massspectrometer. The cross sections for the F  formation, which is the dominant channel, is found to be almost a factor of 2 largerthan what is reported earlier. This descrepancy is explained in terms of the discrimination against ions of appreciable kineticenergy in earlier measurements. (Int J Mass Spectrom 205 (2001) 111–117) © 2001 Elsevier Science B.V. Keywords:  Absolute cross sections; Dissociative electron attachment; NF 3 1. Introduction The accuracies in the measurements of cross sec-tions for any dissociative processes in molecules havebeen limited by the kinetic energies of the resultingfragments. This has been the situation particularly forelectron impact processes like dissociative electronattachment (DEA) and dissociative ionization (DI),though there exist a few exceptions in which reliableabsolute cross sections for DEA have been measuredusing the Tate and Lozier apparatus [1] for total ioncross sections [2]. In these exceptional cases, only onetype of negative ion is produced and measurementscould be carried out without mass analysis of theproducts. The use of a mass to charge ratio analysisbecomes imperative for the measurement of partialcross sections when more than one type of ions areproduced. However, in order to obtain absolute oreven relative cross sections, it is necessary that theextraction, mass analysis, and the detection proce-dures for these ions are carried out without discrimi-nating against their initial kinetic energies, angulardistributions or their mass to charge ratios. Theserequirements are almost impossible to be met in thecase of a gas cell which provide an extended source of ions, thus necessitating the use of crossed beamsgeometry employing a molecular beam and anelectron beam, which provide an almost pointlikesource of ions. Even with a point source of ions, it hasbeen found that conventional mass spectrometersare not reliable in cross section measurements whenthe ions produced have appreciable initial kineticenergies [3]. An efficient solution to these problemswas found in the use of a segmented time-of-flight * Corresponding author. E-mail: ekkumar@tifr.res.inDedicated to Professor Aleksandar Stamatovic on the occasionof his 60th birthday.1387-3806/01/$20.00 © 2001 Elsevier Science B.V. All rights reserved PII   S1387-3806(00)00270-0International Journal of Mass Spectrometry 205 (2001) 111–117  (TOF) mass spectrometer along with the pulsed-electron-beam and pulsed-ion-extraction tech-niques and the relative flow technique. Such acombination has been used for both DEA and DImeasurements for a number of molecules until now[4–10].NF 3  is an important molecule from the point of view of various applications as a fluoride source indry etching and in gas lasers. It has been shown thatNF 3  plasma is very efficient in etching silicon andsilicon dioxide. This has been explained as due to therelatively large dissociation of NF 3  as compared toCF 4  in a discharge [11]. A specific advantage of usingNF 3  is the clean surface it provides as compared to thefluorocarbons which tend to give carbonaceous depos-its on the etched surfaces. More recent work on NF 3 plasma has shown the importance of electron impactdissociation of NF 3  to the plasma chemistry as well asthe etching of silicon [12].So far there exist only two measurements on theabsolute cross sections for DEA on this molecule[13,14]. And these differ by as much as a factor of 2.This discrepancy is a manifestation of the difficulties(as mentioned above) in making these measurements.Harland and Franklin [13] employed a linear TOFmass spectrometer where as Chantry [14] used a Tateand Lozier [1] apparatus, under the assumption thatthe ions resulting from the DEA process have rela-tively small kinetic energies. However, recent mea-surements on the ion kinetic energies [15] have shownthat F  ions are produced with appreciable kineticenergies (about 2 eV at electron energies close to theresonant peak). In view of this, the fact that the crosssections obtained by Chantry is about twice thoseobtained by Harland and Franklin is not surprisingsince a linear TOF mass spectrometer is more likely todiscriminate ions of large kinetic energy as comparedto a Tate and Lozier apparatus employed by Chantry.Considering the importance of NF 3  data and the factthat F  ions are produced with relatively large kineticenergies, and that the measurements by Chantry wereunder the assumption of “modest” kinetic energies,we felt it necessary to make DEA measurements onthis molecule with the more accurate technique pres-ently available. 2. Experiment The details of the experimental arrangement withslight variations and the method of measurement havebeen described in earlier publications [3,4,9,10].However for the sake of completeness a summary isprovided below.The measurements were carried out in a crossedbeams geometry employing a pulsed electron beamwhich intersected an effusive molecular beam at rightangles (Fig. 1). The ions formed were extracted by apulsed electric field into a segmented TOF spectrom-eter and detected by a channel electron multiplieroperated in the pulse counting mode.A heated tungsten filament was used to producefree electrons and they were formed into a beam usingthe Pierce geometry of cathode, grid and a groundedaperture. The electron beam was further collimatedusing the magnetic field (50 G) generated by a pair of solenoids immersed in the vacuum chamber. The gun,the interaction region, and the Faraday cup which isused for monitoring the electron current weremounted along the axis of these solenoids. Thepulsing of the electron beam was carried out byinitially cutting off the beam current by raising thenegative bias on the grid with respect to the cathodeand the filament and overriding it with a positive pulseof 1 ns rise time. The width of the pulse used in thisexperiment was about 300 ns. The molecular beamwas produced by flowing the gas through a capillaryarray mounted at right angles to the electron beam.The ions formed by the interaction were extracted Fig. 1. Schematic of the experimental arrangement.112  D. Nandi et al./International Journal of Mass Spectrometry 205 (2001) 111–117   by a pulsed electric field produced between twomolybdenum wire meshes of 90% transmissionplaced symmetrically about the interaction region.The separation between them in the current experi-ment was 10 mm. The direction of the extraction fieldwas at right angles to the electron beam direction andalong the axis of the TOF mass spectrometer. Thefield was produced by applying a negative pulse of 200 V for a duration of 1   s to the grid situated awayfrom the TOF spectrometer. The timing of the pulsewas such that it trailed the electron pulse by a few tensof nanoseconds so that the electron beam did not getaffected by the large extraction field. This largeextraction field ensured that all the ions produced withvarying kinetic energies and angular distributionswere brought into the TOF spectrometer within anarrow, though diverging cone. The segmented flighttube of the TOF spectrometer was designed to act asan electrostatic lens assembly such that the divergingbeam of ions were transported to the detector withoutany loss. Since this experimental setup was used tocarry out DEA to laser excited molecules, the chan-neltron was mounted off axis in order to shield it fromthe scattered UV laser photons. By applying a suitablevoltage between two deflection electrodes mounted atthe end of the flight tube assembly and before thedetector, it was ensured that all the ions reached thedetector.The TOF mass spectra were obtained by using atime to amplitude converter (TAC) and a pulse heightanalyzer. Special care was taken to ensure that theoverall count rate did not exceed one-tenth of therepetition rate of the electron gun pulsing so that pileup problems were negligible. This was particularlyimportant since F  ions were produced with verylarge intensities even at relatively low pressures. Formeasuring the excitation functions for individual ions,the appropriate time windows in the TAC werechosen and the data were stored in a PC using aGeneral Purpose Interface Bus (GPIB) based dataacquisition system which also controlled the electronenergy. The measurement of the excitation functionsfor F 2  and NF 2  needed special attention since thecross section for the formation of F  was a few ordersof magnitude larger than those for these ions. SinceF  is the lightest of the three ions, the pulses due to itcaused heavy pile up in the TAC at the pressures weneeded to run the experiment for obtaining reasonablestatistics for the other ions. This was overcome byapplying appropriate delay for the “start pulse” of theTAC so that the pulses due to F  did not getregistered when the excitation functions for the ionsof lower intensity was measured.The excitation functions obtained for each type of ion was normalized to absolute cross section using therelative flow technique [16]. The basic principlebehind this technique is to compare the relativeintensities of the species of interest to that of astandard species of known cross section. In terms of the various experimental parameters, the equationgoverning this technique could be written as     X    /   AX       Y    /   BY     N    X     N   Y    F    BY   F    AX    M   BY  1/2  M   AX  1/2  I  e   BY    I  e   AX   K   Y    K    X    (1)where  A ,  B ,  X  , and  Y   represent atomic or molecularspecies and  AX   and  BY  , the parent molecules.  N   is thenumber of ions collected for a specific time,  M   is themolecular weight of the parent molecules,  F   is theflow rate,  I  e  is the electron current,     is the crosssection, and  K   is the detection efficiency which is aproduct of the efficiency with which ions are extractedfrom the interaction region, their transmission throughthe mass spectrometer and finally the efficiency withwhich they are detected. The overall efficiency as afunction of the mass to charge ratio could be writtenas K   m  /  e   k  1 k  2 k  3  (2)where  k  1  is the efficiency of extraction from theinteraction region,  k  2  is the efficiency of transmissionthrough the mass analyzer, and  k  3  is the efficiency of detection of the ion by the particle detector. Inpractice, it is difficult to isolate  k  1 ,  k  2 , and  k  3  and onemeasures only  K  .  k  1  is independent of the mass tocharge ratio, but depends on the initial kinetic ener-gies and angular distributions of the fragment ions.  k  1 113  D. Nandi et al./International Journal of Mass Spectrometry 205 (2001) 111–117   could be made independent of these by applying largeenough extraction field in the interaction regions sothat all the ions are extracted into the mass spectrom-eter independent of the kinetic energies and angulardistributions. The effect of such extraction fields onthe electron beam could be eliminated only by pulsingthe electron gun and the extraction field without anytemporal overlap.The transmission efficiency in the mass analyzer, k  2  is independent of   m/e  in a TOF spectrometer unlikein the case of a quadrupole mass spectrometer. How-ever,  k  2  could depend on the initial kinetic energiesand angular distribution of the ions in the followingway. Depending on the kinetic energies and angulardistributions, the ions extracted from the interactionregion will have finite divergence at the entrance of the mass spectrometer. In order to transport these ionsto the detector without loss, one needs to use anelectrostatic focusing lens assembly. This could beachieved by using the mass spectrometer itself as afocusing assembly. Thus in the present case, the flighttube of the TOF spectrometer is made of four separateelements and biased in such a way that the divergentbeam of ions entering it is focussed at its exit wherethe detector is mounted. This lens system was de-signed using the  SIMION  program [17] by taking intoaccount worst case kinetic energies and angular dis-tributions and optimum extraction fields. Thus byappropriate choice of a multi-element TOF massspectrometer we could make  k  2  independent of   m/e and initial kinetic energies and angular distributions.The detection efficiency,  k  3  of the ions by thechannel electron multiplier has been found to dependon the velocity with which the ions strike the detectorsurface. Thus for a given acceleration, ions of smaller m/e  will have larger detection probability [18]. In thecase of measurements on positive ions,  k  ( m/e ) can bedetermined using cross sections for the formation of singly charged ions from their respective noble gasatoms in the mass range from 4 to 132, by the relativeflow technique [19]. But so far it has not been possibleto apply it to the negative ions due to nonavailabilityof accurate cross sections in a wide  m/e  range. Theonly way to take care of the  m/e  dependence of   K  through  k  3  is by increasing the nosecone voltages tosuch levels that there is saturation in the detectionefficiency. It was found that a bias of 1500 V on thenosecone of the channeltron was sufficient to obtainsaturation for the all the ions of relevence in thepresent measurements.The cross section for the formation of O  from O 2 was used in the present case to calibrate the crosssections for the formation of F  , F 2  and NF 2  formedfrom NF 3  as well as to calibrate the energy scale. Ananalysis of the various data on the formation of O  from O 2  [20] have identified the results of Rapp andBriglia [2] as fairly accurate within an error of 10%.The uncertainty in the overall detection efficiency of the ions is estimated to be 5%. The uncertainty in theflow rate measurements is about 5% and the statisticalerrors in counting were 1% each in the case of F  andO  and about 5% in the case of F 2  and NF 2  . Theerror in the electron current measurement may havean upper limit of 5%. Considering all these, thecombined error in the present measurements worksout to be about 15%. 3. Results and discussion It was found that F  is the most dominant ion fromthe DEA process with very small intensities of F 2  andNF 2  . This is consistent with the earlier reports[13,14]. The cross sections for all the three speciesare given in Figs. 2, 3, and 4, and in tabular form inTable 1. Fig. 2. Cross section for the formation of F  .114  D. Nandi et al./International Journal of Mass Spectrometry 205 (2001) 111–117   The positions of the resonant peaks of all the threeions are in good agreement with the earlier reports.For all the ions, there is a finite cross section even atzero energy. This is unlike the results of Ruckhaberleet al. [15], but is similar to those by Harland andFranklin [13]. The high resolution measurements of Ruckhaberle et al. showed finite cross section for F  at zero energy, where as both F 2  and NF 2  appearonly above 1 eV. In the present measurements, as in Table 1Cross section for the formation of various ions from NF 3 Electronenergy (eV)    (F  )(10  16 cm 2 )    (F 2  )(10  19 cm 2 )    (NF 2  )(10  20 cm 2 )0.00 0.53 0.18 0.480.10 0.58 0.23 0.340.20 0.64 0.30 0.380.30 0.72 0.34 0.410.40 0.79 0.35 0.460.50 0.92 0.38 0.500.60 1.04 0.48 0.470.70 1.21 0.51 0.670.80 1.33 0.63 0.710.90 1.52 0.71 0.931.00 1.68 0.82 1.171.10 1.95 1.03 1.781.30 2.03 1.15 2.201.40 2.11 1.25 2.431.50 2.16 1.40 3.341.60 2.18 1.51 3.941.70 2.20 1.60 4.501.80 2.17 1.64 4.461.90 2.14 1.59 4.932.00 2.08 1.65 4.622.10 2.03 1.59 4.012.20 1.93 1.52 3.732.30 1.83 1.40 3.202.40 1.74 1.31 2.682.50 1.63 1.15 2.132.60 1.50 1.07 1.552.70 1.40 0.99 1.412.80 1.28 0.87 1.062.90 1.16 0.78 1.013.00 1.05 0.65 0.633.10 0.94 0.58 0.543.20 0.84 0.50 0.503.30 0.75 0.45 0.333.40 0.66 0.38 0.353.50 0.59 0.31 0.213.60 0.50 0.28 0.293.70 0.44 0.24 0.173.80 0.38 0.20 0.153.90 0.32 0.15 0.204.00 0.28 0.15 0.114.10 0.23 0.2 0.084.20 0.20 0.10 0.114.30 0.16 0.09 0.064.40 0.14 0.07 0.094.50 0.12 0.06 0.164.60 0.10 0.04 0.094.70 0.08 0.04 0.044.80 0.07 0.03 0.034.90 0.05 0.04 0.065.00 0.04 0.02 0.015.10 0.04 0.02 0.035.20 0.03 0.02 0.06 (continued on next page) Fig. 3. Cross section for the formation of F 2  .Fig. 4. Cross section for the formation of NF 2  .115  D. Nandi et al./International Journal of Mass Spectrometry 205 (2001) 111–117 
Search
Similar documents
View more...
Tags
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks