Dipolar dissociation dynamics in electron collisions with oxygen molecules
DDipolar dissociation dynamics in electron collisions with oxygen molecules
Pamir Nag ∗ and Dhananjay Nandi † Indian Institute of Science Education and Research Kolkata, Mohanpur 741246, India (Dated: September 24, 2018)The dipolar dissociation of molecular oxygen due to 21-35 eV energy electron collision has beenstudied using the time sliced velocity map imaging technique. A rough estimation about the thresh-old of the process and the kinetic energy and angular distribution of the fragment negative ions aremeasured. The dipolar dissociation found to be occur due to pre-dissociation of a Rydberg state viaion-pair state for lower incident electron energies as well from also direct excitation to the ion-pairstates for relatively higher primary beam energy. The location and symmetry of the excited stateswere determined from the kinetic energy and angular distribution data respectively.
PACS numbers: 34.80.Gs, 34.80.Ht
I. INTRODUCTION
Dissociative ionization by electron impact plays animportant role in astrochemistry and biology. In dipo-lar dissociation the molecule dissociates into an anionicand cationic fragments. The dipolar dissociation pro-cess starts around above 15 eV incident electron energy.Unlike the well studied dissociative electron attachment(DEA) process [1, 2] in dipolar dissociation the incom-ing electron does not resonantly get captured by themolecule but, might partially transfers its kinetic energyto the molecule and leave it in an excited state. Themolecules populated into an ion-pair state either via pre-dissociation of a Rydberg state or through direct excita-tion dissociate into a cationic and anionic fragments.O + e − → O ∗ + e − → O + + O − + e − (1)The ion-pair dissociation dynamics of O is well stud-ied mainly around the threshold using lasers [3–5]. Pri-marily the pre-dissociation of a Rydberg state into anion-pair state was found to be responsible for the dipo-lar dissociation process around the threshold. But onlya few studies of the dipolar dissociation well above thethreshold were performed. It is difficult to access the ion-pair states using lasers due to the required energy range.But the same ion-pair states can easily be probed usingelectron beam of energy above around 15 eV. Van Bruntand Kieffer [6] studied the kinetic energy and angular dis-tribution of the O − ions produced through dipolar dis-sociation over a limited angular range back in 1974. Tobest of our knowledge no recent study except by Nandi et al. [7] is available in literature. In that article alsothe authors only reported the data but did not performdetailed analysis of the ion-pair dissociation process. Inthe present article the dipolar dissociation dynamics ofmolecular oxygen due to interaction with electrons hav-ing energies in between 21 to 35 eV is studied using well ∗ [email protected] † [email protected] Electron Energy (eV) C oun t s ( a r b . un i t s )
21 eV 25 eV 35 eV30 eV27 eV23 eV
FIG. 1. Ion Yield curve of O − ions produced from O due tointeraction with low-energy electrons. established velocity slice imaging technique [8–10]. Inthe present studied only the O − fragments were probedand it was assumed that a O + ion is always accompa-nied with the anionic fragments as negative ions can onlybe formed due to dipolar dissociation processes [11, 12]around the reported energy region. From the velocityslice images the kinetic energy and angular distributionof the O − ions were obtained for six different incidentelectron energies. The location and symmetry of the ion-pair states involved in the process were determined fromthose measurements. II. EXPERIMENT
To study the dipolar dissociation dynamics in O atime sliced velocity map imaging (VMI) spectrometersimilar with previous report by Nandi et al. [8] with someminor modifications was used. The same apparatus have a r X i v : . [ phy s i c s . a t m - c l u s ] A ug already been employed for the dissociative electron at-tachment (DEA) study to different molecules [13–15] inrecent time. Some details of the current apparatus canbe found elsewhere [16]. For sake of completeness thesetup is described here in brief.The entire experiment was performed under oil-free ul-tra high vacuum condition with a base pressure of the or-der ∼ − mbar. The chamber was heated at 130 ◦ C forseveral days before performing the current experimentsto remove water vapours and other impurities that mightpresent inside the chamber wall. A magnetically well col-limated pulsed electron beam of 200 ns duration and 10kHz repetition and of controllable energy was employedin the experiment. A custom build electron gun consistsof thermally heated filament with typical energy resolu-tion 0.8 eV was used for the purpose. A pair of magneticcoils in Helmholtz configuration, producing around 40 Guniform magnetic field, was mounted outside the cham-ber to collimate the low-energy otherwise diverging elec-tron beam. A Faraday cup was placed opposite to theelectron gun to monitor the time averaged pulsed elec-tron beam. An effusive molecular beam produced froma capillary tube of 1 mm diameter was made to interactperpendicularly with the pulsed electron beam in the in-teraction region of the VMI spectrometer. The molecularbeam was along the spectrometer axis and directed to-wards the detector. The VMI spectrometer is a three fieldtime-of-flight (TOF) spectrometer [8] which focuses ionsstarting from a finite volume onto a two-dimensional po-sition sensitive detector (2D-PSD) such that ions of givenvelocity are mapped onto a single point on the detectorirrespective of their spatial location in the source region.The ions produced in the interaction region due to theelectron-molecule collisions were pulse extracted using amoderate electric field pulse of 4 µ s duration. The ex-traction pulse was applied 100 ns after the electron beampulse had passed. The delayed extraction pulse preventsthe electrons from reaching the detector and also pro-vides an appropriate time spread to the ions to expandfor better time slicing. The 2D-PSD used in the experi-ment consists of three micro channel plates (MCP) in Z-stack configuration and a three layer delay line hexanode[17]. The TOF of the detected ions had been determinedfrom the back MCP signal whereas the x and y positionsof each detected ions were calculated from the three an-ode layer [17]. The x, y position along with TOF of eachdetected particles were acquired and stored in List-ModeFormat (LMF) using CoboldPC software from RoentDek.The entire kinetic energy and angular distribution infor-mation can be extracted from the central slice throughthe ‘Newton Sphere’ in the plane of the detector and con-taining the electron beam axis. The central time slicedimages were obtained by selecting only the ions producedwithin an appropriate time window during off-line anal-ysis of the LMF file using the same CoboldPC software.The typical full width at half maxima (FWHM) of theTOF of the O − ions produced in the current report wasabout 500 ns. A 50-ns thin time sliced images around
10 15 20 25
Electron energy (eV) I on c oun t s ( a r b . un i t s ) Expt. dataFitted curve
FIG. 2. Ion yield curve of the O − ions produced due to dipo-lar dissociation around the threshold value. The fitted curvesuing the couple of equation 3 is also shown by a solid line. the central part of the ‘Newton Sphere’ has been con-sidered to get the kinematically complete information ofthe ions. The electron beam energy calibration was per-formed using the 6.5 eV DEA peak of O − /O [18]. Todetermine the kinetic energy of the ions from the slicedimages the spectrometer was calibrated with the energyrelease of O − /O in DEA range [2] and cross-checkedwith the energy release of O − /CO at 8.2 eV [14, 19].To get the ion yield curve a different set of data acqui-sition system was used. For this purpose only the signalfrom the MCP was taken. The MCP signal was amplifiedthrough a Fast Amp and then fed to a constant fractiondiscriminator (CFD). The output from CFD was con-nected to STOP input of a nuclear instrumentation mod-ule (NIM) standard time to amplitude converter (TAC)and START pulse was generated from a master pulsegenerator controlling the timing sequences of the entireexperiment. The output of TAC was connected to a mul-tichannel analyser (MCA, Ortec model ASPEC-927) andfinally communicated with the data acquisition systeminstalled in a dedicated computer via high-speed USB2.0 (Universal Serial Bus) interface. A LabVIEW baseddata acquisition system was used to obtain the ion yieldcurve and control most of the instruments employed inthe experiment. A complete details of this system can befound elsewhere [16]. III. RESULT AND DISCUSSION
The experiments have been performed using 99.9%pure commercially available oxygen gas. The ion yieldcurve of the O − ions produced due to low-energy electron
21 eV 23 eV 25 eV27 eV 30 eV 35 eV
FIG. 3. Time sliced velocity map images of the O − ions created due to six different incident electron energies as mentioned onthe top of each image. The incident electron beam direction is indicated by small arrows along the x-axis. collision with O molecule is shown in Figure 1. A reso-nant peak due to dissociative electron attachment (DEA)at 6.5 eV and almost constant ion formation due to dipo-lar dissociation after nearly 15.3 eV can be observed inthe ion yield curve. The curve is in good agreement withthe previous report [18]. The main focus of this article isto understand the dipolar dissociation dynamics startingfrom around 15.3 eV. The arrows in Figure 1 indicatesthe energies at which the velocity slice images (VSI) weretaken. From the ion yield curve the appearance energyfor the dipolar dissociation processes has been computedwhereas from the VSI the kinetic energy and angular dis-tribution of the O − ions have been measured.Unlike the DEA in dipolar dissociation process the in-coming electrons do not get attached with the moleculebut might partially transfer its kinetic energy. In thepresent case if V i is the amount of energy transfer tothe molecule, D = 5 .
15 eV [20] is the bond dissociationenergy of O , IP = 13 . A = 1 . E is the kinetic energy of each of the O + and O − ion fragments formed in the process, then fromconservation of energy one can found V e = ( E i + D − A + IP ) + 2 E (2)The threshold energy for the diploar dissociation pro-cess using above mentioned accepted thermochemical val-ues of different parameters and considering the O + frag- ments formed in ground state ( E i = 0) is found to be17.25 eV. From the ion yield curve it is possible to ex-perimentally determine the energy threshold of the DDprocess. The data points in the ion yield curve near thethreshold is fitted using the couple of equation given byFiegele et al. [22] and recently used by Szyma´nska et al. [23] f ( E ) = b for E < E
T h f ( E ) = b + a ( E − E T h ) n for E > E
T h (3)The parameter b is the constant background and E T h is the threshold of the DD process. The parameter a is the scaling factor and set to zero below the thresholdvalue and n is the exponent. The effect of finite energyresolution of the electron beam is not considered in theset of equation 3. Due to the nearly 0.8 eV broad energyspread of the incident electron beam the the ion yieldcurve gets smooth near the threshold which will lead tounderestimation of the true threshold value. To minimizethis effect only the data points slightly above from theapparent threshold upto 23.8 eV are considered in thefitting process and a nearly straight slope is obtained.The straight line is extrapolated until it intersects withthe line obtained due to constant background ( f ( E ) = b ).The intersection point of these two curve is considered asthe threshold value. Due to the effect of contact potentialalso the energy might get shift a little. The best fittedcurve is obtained using the parameter a = 44 . b =42 . n = 1 .
435 and shown using a solid line in Fig.
KE of O − ions (eV) C oun t s ( a r b . un i t s )
21 eV23 eV25 eV27 eV30 eV35 eV
FIG. 4. The kinetic energy distribution of the O − ions pro-duced from O due dipolar dissociation for six different inci-dent electron energies as indicated.
2. The threshold value for dipolar dissociation processis experimentally found to be 15.3 eV. The mismatchbetween the thremochemally obtained value 17.25 andexperimentally obtained 15.3 eV could be due to poorelectron beam energy resolution and the effect of contactpotential.The velocity slice images (VSI) of O − ions produceddue to dipolar dissociation (DD) process for 21, 23,25, 27, 30 and 35 eV energy electron collision with O molecules are shown in Fig. 3. Out of the entire ‘NewtonSphere’ of O − ions with nearly 500 ns time width a 50ns thin flat slice through the central part has been takenfor the kinetic energy and angular distribution measure-ment purpose. The incident electron beam direction isfrom left to right through the center of each images asindicated using small arrows on each of the VSI.The kinetic energy distribution of the ions created dueto dipolar dissociation are shown in Fig. 4. The distri-butions are normalized at the near zero eV peak. Thedistributions show one large peak near zero eV followedby a broad peak in between 1.25 to 1.75 eV. The centreof the broad peak shifts towards right with increasing in-cident electron energy. A flat time slicing method is usedto obtain the central slices containing the kinetic energydistribution information. So, in the kinetic energy dis-tribution data only a fraction of the ions produced withhigher kinetic energy are considered but all the ions withlower kinetic energy are included [10, 19]. Thus the rel-ative contributions of the ions with near zero eV andhigher kinetic energy shown in Fig. 4 are not absolute. Angle (Degree) I ( θ ) /I ( ° )
21 eV23 eV25 eV27 eV30 eV35 eV
FIG. 5. Angular distribution of the O − ions created withkinetic energy in between 0.7 to 1.7 eV energy due to dipolardissociation for six different incident electron energies. In the previous study of kinetic energy distribution byVan Brunt and Kieffer [6] a peak near 2 eV and anotherpeak near 3.3 eV were reported. To best of our knowl-edge this is the only published kinetic energy distributionreport of dipolar dissociation to O molecules due to elec-tron collision. The ions below 0.75 eV were not consid-ered in that reported due to non-reliability. The higherpeak near 3.3 eV has not been observed in the currentstudy. In a recent study Nandi et al. [7] also reportedthe absence of the higher peak from the VSIs. The nearzero eV kinetic energy ions might not be created froma true dipolar states but due to pre-dissociation of theRydberg states. The pre-dissociation is present through-out the entire dipolar dissociation region and ions withall possible kinetic energies in between 0 to around 2eV can be observed. Pre-dissociation can occur if thelife-time of an excited states is of the order of or morethan few vibrational time period and the excited statescross the ion-pair states or interact with the vibrationalcontinuum of ion-pair states. The ions with higher ki-netic energy might be coming from excitation to trueion-pair states. If the ions created due to direct excita-tion to the ion-pair states have kinetic energy in between1.25 to 1.75 eV then from equation 3 the energy trans-fer in the process comes out to be in between 19.75 to20.75 eV. So, in the Franck-Condon transition region theseparation between the ground state of oxygen moleculeand the ion-pair state should be of the order of 19 eV.In ion-pair dissociation study of oxygen molecule usingfemtosecond depletion method Baklanov et al. [3] alsofound the ion-pair state around the same energy range.The symmetry of the ion-pair states involved in thedipolar dissociation process can be determined from theangular distribution of the fragment ions. Due to en-ergy and momentum conservation both the O − and O + Angle (Degree) I ( θ ) /I ( ° ) Σ g to Σ u fit Angle (Degree) I ( θ ) /I ( ° ) Σ g to Σ u fit Angle (Degree) I ( θ ) /I ( ° ) Σ g to Σ u + Π u fit Angle (Degree) I ( θ ) /I ( ° ) Σ g to Σ u + Π g + Π u fit Angle (Degree) I ( θ ) /I ( ° ) Σ g to Σ u + Π u fit Angle (Degree) I ( θ ) /I ( ° ) Σ g to Σ u + Π g + Π u fit FIG. 6. The angular distribuition data for different incident electron energies, as indicated, are shown using the scatteredpoints. In the left most column the solid lines indicate the fitted curve for a Σ g → Σ u transition. A transition from Σ g toΣ u + Π u two states is shown using the solid curves in the middle column. The right most column shows the fitted curve for aΣ u , Π g and Π u three final states transition process. ions formed in dipolar dissociation process will have sim-ilar kinetic energy and angular distribution. The angulardistribution of fragment negative ions for six different in-cident electron energies in between 21 to 35 eV are shownin Fig. 5. The ions created with kinetic energy in between0.7 to 1.7 eV are only considered in angular distributionmeasurements. As already discussed due to flat slicingcontribution from the entire ‘Newton Sphere’ is presentfor the near zero eV ions thus no meaningful angular dis-tribution information can be obtained for these ions fromthe sliced images and are not discussed further. The an-gular distribution data are normalized at 90 ◦ and showtwo strong peaks at 0 ◦ and 180 ◦ and two small peaksaround 65 ◦ and 115 ◦ . For 21, 23 and 25 eV incident elec-tron energies the anisotropy is prominent and increaseswith increasing energy. But, the anisotropic nature ofthe angular distribution starts decreasing after 27 eVincident electron energy and becomes almost isotropicwith only a forward and backward peak. A forward-backward asymmetry observed in the angular distribu-tion data might be caused by the effect of molecular recoildue electron collision [24].The angular distribution of the fragments can be de-scribed using the expression given by Van Brunt [25] andidentical with the expression derived by Zare [26] for theinvolvement of a single final state as I (Θ) (cid:39) K − n (cid:12)(cid:12)(cid:12)(cid:12)(cid:12)(cid:12) ∞ (cid:88) l = | µ | i l (cid:115) (2 l + 1)( l − µ )!( l + µ )! × j l ( κ ) Y l,µ (Θ , Φ) (cid:12)(cid:12)(cid:12)(cid:12)(cid:12)(cid:12) (4)Where, κ is the product of momentum transfer vectorbetween the incident and scattered electron, K , and theclosest approach between the impinging electron and thecenter-of-mass of the molecule. The j l ’s are sphericalBessel function and Y l,µ ’s are spherical harmonics. IfΛ i and Λ f are the projection of electronic axial orbitalangular momentum on the molecular axis for the initialand the final states respectively then, µ = | Λ f − Λ i | .The angular momentum quantum number l ≥ | µ | andrestricted to have only even or odd values depending onwhether the initial and final molecular states are havingsame of opposite parity for homonuclear molecules. Inthe present article the angular distribution is fitted usingexpression 5. I ( θ ) = (cid:88) | µ | (cid:12)(cid:12)(cid:12)(cid:12)(cid:12)(cid:12) (cid:88) l = | µ | a l i l (cid:115) (2 l + 1) ( l − µ )!( l + µ )! j l ( κ ) Y l,µ ( θ, φ ) e iδ l (cid:12)(cid:12)(cid:12)(cid:12)(cid:12)(cid:12) (5)The summation over µ takes care of the involvement ofmore than one final states. The phase differences be-tween different partial waves involved in the transitionwith respect to the lowest one for each final state are ex-pressed by δ l s. The momentum transfer vector K andthe adjustable parameter n remain same for a particularincident electron energy and the term K − n in the ex-pression 4 can be absorbed within the parameter a l ofthe expression 5.The ground state of O molecule is Σ − g , having Λ i = 0.For the present case µ = 0 and 1 represents a transitionfrom Σ g to Σ and Π state and even and odd values of l are responsible for transition to a final state having ga-rade ( g ) and ungarade ( u ) parity. To know about thesymmetry of the ion-pair states involved in the processthe angular distribution data has been fitted using ex-pression 5 and the fitting parameters are shown in Ta-ble I and II. The angular distribution fitted with a Σ g to Σ u transition model is shown in the left most columnof Fig. 6. This transition model can represents the an-gular distribution data upto a limit for 21 to 25 eV butwith increasing incident electron energy the fit becomespoorer. This model also underestimate the value at 90 ◦ .For a Σ u and Π u two final states transition model alsothe goodness of fitting decreases with increasing incidentelectron energy and shown in the middle column of Fig. 6.A transition from Σ g initial state to Σ u , Π u and Π g threestate transition model gives the best fit and shown inthe right column of Fig. 6. So, from the angular dis-tribution data it is evident that a Σ u and Π u ion-pairstate is present through out the entire incident electronenergy range. A Π g state is also getting involved in theion-pair production process and the contribution of thisstate increases with increasing incident electron energyand becomes significance above 27 eV incident electronenergy. Van Brunt and Kieffer [6] also concluded that aΣ u state is responsible for the dipolar dissociation. In thelimited angular distribution data the authors also foundthe presence of a Π u state. The molecules might getexcited to a Rydberg state and then the Rydberg statecan interact with an ion-pair state and dissociate into aO + and O − ions. Both the Rydberg and ion-pair statewill have the same symmetry and in ion-pair dissociationstudy using XUV laser Zhou and Mo [4] found the sym-metry of the excited states to be Π u and Σ u . Dehmer andChupka [5] measured the ion-pair production in 720 ˚A to670 ˚A range (17.2 to 18.505 eV) and found the Rydbergstates might dissociates via pre-dissociation through ionpair-states having symmetry Σ − u and Π u . The authors also mentioned the dissociation is primarily through pre-dissociation of the Rydberg states rather than by directdissociation. But in the present case above 20 eV webelieve both the predissociation and direct dissociationof ion-pair states are present and the higher energy ionsconsidered in the angular distribution measurements areform direct dissociation of the ion-pair states. IV. CONCLUSION
The O − ions produced due to dipolar dissociation of O molecule via interaction with electrons have been studiedusing time sliced velocity map imaging technique. Fromthe kinetic energy distribution curve the location of theion-pair state was found to be around 20 eV above theground state in the Franck-Codon transition region. Pre-dissociation of a Rydberg state through an ion-pair statewas found to be responsible for the dipolar dissociationprocess for relatively lower primary beam energy. Di-rect excitation to the ion-pair states was also found tobe involved in dipolar dissociation process for relativelyhigher incident electron energies. From the angular dis-tribution data a Σ u and a Π u ion-pair state was found tobe present in the entire energy range. The involvement ofa Σ g state is also observed with increasing incident elec-tron energies. To know about the exact location of theion-pair states in the potential energy surface and detaildynamics excitation to a ion-pair state either via pre-dissociation or direct excitation theoretical calculationsare highly demanding. V. ACKNOWLEDGEMENTS
D. N. gratefully acknowledges the partial financial sup-port from “Indian National Science Academy” for the de-velopment of VSI spectrometer under INSA Young Sci-entist project “SP/YSP/80/2013/734”.
REFERENCES [1] R. J. Van Brunt and L. J. Kieffer, Phys. Rev. A , 1899(1970).[2] D. Nandi and E. Krishnakumar, Int. J. Mass Spectrom. , 39 (2010).[3] A. V. Baklanov, L. M. C. Janssen, D. H. Parker,L. Poisson, B. Soep, J.-M. Mestdagh, andO. Gobert, J. Chem. Phys. , 214306 (2008),http://dx.doi.org/10.1063/1.3026613.[4] C. Zhou and Y. Mo, J. Chem. Phys. , 084314 (2013),http://dx.doi.org/10.1063/1.4819079. [5] P. M. Dehmer and W. A. Chupka, J. Chem. Phys. ,4525 (1975).[6] R. J. Van Brunt and L. J. Kieffer, J. Chem. Phys. ,3057 (1974).[7] D. Nandi, V. Prabhudesai, and E. Krishnakumar, Ra-diat. Phys. Chem. , 2151 (2006), new Directions in theStudy of Interaction of Energetic Photons with Matter.[8] D. Nandi, V. S. Prabhudesai, E. Krishnakumar, andA. Chatterjee, Rev. Sci. Instrum. , 053107 (2005). [9] H. Adaniya, D. S. Slaughter, T. Osipov, T. Weber, andA. Belkacem, Rev. Sci. Instrum. , 023106 (2012).[10] A. Moradmand, J. B. Williams, A. L. Landers, andM. Fogle, Rev. Sci. Instrum. , 033104 (2013).[11] W. W. Lozier, Phys. Rev. , 268 (1934).[12] F. H. Dorman, J. D. Morrison, and A. J. C. Nicholson,J. Chem. Phys. , 378 (1960).[13] P. Nag and D. Nandi, Phys. Chem. Chem. Phys. , 7130(2015).[14] P. Nag and D. Nandi, Phys. Rev. A , 052705 (2015).[15] P. Nag and D. Nandi, Phys. Rev. A , 012701 (2016).[16] P. Nag and D. Nandi, Meas. Sci. Technol. , 095007(2015).[17] O. Jagutzki, A. Cerezo, A. Czasch, R. Dorner, M. Hattas,M. Huang, V. Mergel, U. Spillmann, K. Ullmann-Pfleger,T. Weber, H. Schmidt-Bocking, and G. Smith, IEEE Trans. Nucl. Sci , 2477 (2002).[18] D. Rapp and D. D. Briglia, J. Chem. Phys. , 1480(1965).[19] A. Moradmand, D. S. Slaughter, A. L. Landers, andM. Fogle, Phys. Rev. A , 134(1967).[22] T. Fiegele, G. Hanel, I. Torres, M. Lezius, and T. D.M¨ark, J. Phys. B , 4263 (2000).[23] E. Szyma´nska, N. J. Mason, E. Krishnakumar, C. Matias,A. Mauracher, P. Scheier, and S. Denifl, Int. J. MassSpectrom. Ion Phys. , 356 (2014).[24] R. J. Van Brunt and L. J. Kieffer, Phys. Rev. A , 1293(1970).[25] R. J. Van Brunt, J. Chem. Phys. , 3064 (1974).[26] R. N. Zare, J. Chem. Phys. , 204 (1967). TABLE I. Fitting parameters for the angular distribution of the O − ions created due to 21, 23 and 25 eV incident electronenergies for a Σ g to Σ u , Π g and Π u transition. 21 eV 23 eV 25 eVWeighting ratio of different partial waves b : b : 1: 0.181: 1: 0.379: 1: 0.325: c : c : 1.359: 0.733: 0.048: 0.186: 0.076: 0.402 e : e δ Σ u , δ Π g , δ Π u (rad) 0.595, 0.426, 0.530 0.726, 0.553, 0.330 0.901, 0.812, 1.045Parameter κ , κ , κ − ions created due to 27, 30 and 35 eV incident electronenergies for a Σ g to Σ u , Π g and Π u transition. 27 eV 30 eV 35 eVWeighting ratio of different partial waves b : b : 1: 0.832: 1: 0.246: 1: 0.187: c : c : 0.144: 0.032: 0.009: 0.981 0.081: 0.425 e : e δ Σ u , δ Π g , δ Π u (rad) 0.497, 0.847, 0.647 0.606, 1.949, 0.055 0.195, 0.490, 0.326Parameter κ , κ , κ3