Molecular Dynamics in Dissociative Electron Attachment to CO probed by Velocity Slice Imaging
MMolecular Dynamics in Dissociative ElectronAttachment to CO probed by Velocity Slice Imaging
Pamir Nag and Dhananjay Nandi Indian Institute of Science Education and Research Kolkata, Mohanpur 741246, India email: [email protected], [email protected] Abstract
Kinetic energy and angular distributions of O − ions formed by dissociative electron attachment to COmolecule have been studied for 9, 9.5, 10, 10.5, 11, 11.5 eV incident electron energies around the resonanceusing time sliced velocity map imaging spectrometer. Detailed observations clearly show two separateDEA reactions lead to the formation of O − ions in the ground P state along with the neutral C atomsin ground P state and first excited D state, respectively. Within the axial recoil approximation andinvolving four partial waves, our angular distribution results clearly indicate that the two reactions leadingto O − formation proceed through the distinct resonant state(s). For the first process, more than oneintermediate states are involved. Whereas, for the second process, only one state is involved. The observedforward-backward asymmetry is explained due to the interference between the different partial waves thatare involved in the processes. Low energy electron-molecule collision leading to dis-sociative electron attachment (DEA) is an importantprocess from the fundamental as well as the appli-cation point of view. DEA study of molecules arevery important starting from astrophysics to biology.The resonance formation can be used for a single- ordouble-strand break in DNA [1]. Site specific frag-mentation [2] can also lead to selective bond cleavagein DNA [3]. Chandler and Houston [4] first used theimaging technique to study the molecular dynamics.Later, velocity map imaging (VMI) technique [5] andslice imaging [6, 7, 8] helped to study the angulardistribution and kinetic energy distribution simulta-neously and very accurately in the photo-dissociationdynamics. Recently, this method has been modifiedand implemented in the low energy electron moleculecollision studies [9] for the first time. Since then, thevelocity slice imaging (VSI) technique in its variousforms have been successfully employed to study thelow energy electron-molecule collisions by different groups [10, 11, 12, 13] in the recent times. Very re-cently, we developed a modified velocity slice imagingspectrometer to study the low and intermediate en-ergy electron-molecule collision experiments. In thisstudy, the spectrometer has been probed to measurethe kinetic energy and angular distribution of O − ionsproduced from CO by DEA process.The O − ion formation from CO due to electronimpact was first observed by Vaughan [14] back in1931. Rapp and Briglia [15] measured the absolutecross section and reported to observe the dissociativeelectron attachment peak near 9.9 eV. The dominantprocess leading to the O − formation (Process I) is e − + CO( Σ + ) → CO −∗ → O − ( P ) + C( P ) . Through energy analysis of the ions by Chantry [16],proposed a second process for the O − formation (Pro-cess II) as: e − + CO( Σ + ) → CO −∗ → O − ( P ) + C ∗ ( D ) . Hall et al. [17] measured the kinetic energy distribu-tion of the O − ions at three specific angles, and also1 a r X i v : . [ phy s i c s . a t m - c l u s ] O c t I on C oun t s
10 eV9.5 eV9 eV 10.5 eV11 eV11.5 eV
Figure 1:
Ion yield curve of O − produced from DEA toCO. The arrows indicate the energies at which the imagesare taken. the angular distribution of the ions and proposed theintermediate state might be a Π state. Morgan etal. [18] recently computed the potential energy curveof the neutral CO molecule and the resonance statesusing R-matrix formalism. Tian et al. [19] recentlystudied the angular distribution of O − ion from COdue to DEA using velocity slice imaging (VSI) andproposed the presence of coherent interference be-tween the different states that are involved. In thisarticle, we report the kinetic energy distribution ofthe negative ions over a broad incident electron en-ergy range of 9 eV to 11.5 eV around the resonanceand also the angular distribution of the O − ions de-pending on their kinetic energy distributions for theabove mentioned electron energy range. Negative ions are formed due to low energy electroncapture and subsequent dissociation. The measure-ments are performed under high vacuum conditionat the base pressure below ∼ − mbar. A mag-netically well collimated pulsed electron beam of 200ns duration, 10 kHz repetition rate and with con- Figure 2:
Time sliced images at different incident elec-tron energies. The incident electron beam direction isalong the horizontal axis from left to right through thecenter of each image. − ion (eV) N o r m a li z ed i on c oun t s Figure 3:
KER of O − ion at different incident electronenergies. µ s and is applied 100 ns after the electrongun pulse. The delayed extraction provide appropri-ate time spread for better time sliced image. TheVMI spectrometer is like a three field time-of-flightspectrometer [9] which focuses ions starting from afinite volume onto a two-dimensional position sen-sitive detector such that ions with a given velocityare mapped to a point on the detector irrespectiveof their spatial location in the source region. Thetwo-dimensional position sensitive detector consistsof three micro channel plates (MCPs) in Z-stack con-figuration and a three layers delay line hexanode [20].The time-of-flight (ToF) of the detected ions is de-termined from the back MCP signal whereas the xand y positions of each detected ions are calculatedfrom the three anode layer [20] placed behind theMCPs. The x and y position along with ToF ofeach detected particles are acquired and stored in alist-mode format (LMF) using the CoboldPC soft-ware from RoentDek. The central slice through the‘Newton Sphere’ contains the full angular and trans-lational energy information. The central sliced im-age is obtained by selecting appropriate time windowduring the off-line analysis from the stored LMF fileusing the CoboldPC. Such time sliced image corre-sponds to the ions ejected in the plane parallel to thedetector containing the electron beam axis.The typical FWHM of the ToF of the O − ions pro-duced in this energy range is about 250 ns. We havetaken a 50 ns time sliced image from the central partof the entire Newton Sphere. The complete infor- mation about the kinetic energy release and angulardistribution of the negative ions can be obtained fromthis central slice. For incident electron beam energycalibration we have considered the O − /CO resonancepeak (shown in figure 1) to be at 9.9 eV [15]. To mea-sure the kinetic energy release (KER) of the negativeions we have calibrated our system using the energyrelease of O − /O at 6.5 eV [21]. We also have cross-checked the kinetic energy calibration by measuringthe kinetic energy of O − produced by electron at-tachment at 8.2 eV of CO [22].To get the ion yield curve a different set of data ac-quisition system has been used. For this purpose thesignal from MCP only has been taken. The MCP sig-nal is amplified through a Fast Amp and then fed to aConstant Fraction Discriminator (CFD). The outputfrom CFD is fed to STOP of a Nuclear Instrumen-tation Module (NIM) standard Time-to-AmplitudeConverter (TAC) and START is generated from themaster pulse used in the electron gun. The outputof the TAC is connected to a Multichannel Analyser(MCA, Ortec model ASPEC-927) and finally commu-nicated with the data acquisition system installed ina dedicated computer via high-speed USB 2.0 (Uni-versal Serial Bus) interface. A home made LabVIEWbased data acquisition system has been used to getthe ion yield curve. Using this software at first theToF has been obtain, then by selecting only the chan-nel corresponding to a particular mass the electronenergy versus the number of ions produced have beenmeasured. Figure 1 shows the ion yield curve of O − ions pro-duced from CO due to dissociative electron attach-ment (DEA) process. The arrows indicate the en-ergies at which the velocity slice images (VSI) aretaken. The central sliced images at different electronenergies are shown in figure 2. The kinetic energyreleased (KER) in the process is distributed amongthe neutral carbon atom and the O − ion. The kineticenergy distribution of the O − ions for different inci-dent electron energies are displayed in Figure 3. For9, 9.5 and 10 eV incident electron energies ions are3able 1: Fitting parameters for the angular distribution of the O − ions taken at 9, 9.5 and 10 eV incidentelectron energies. The angular distributions are fitted with Σ to Σ and Π transition.
Weighting ratio of different partial waves a : a : a : a : 1: 0.56: 0.25: 0.45: 1: 0.42: 0.13: 0.14: 1: 0.54: 1.03: 0.10: b : b : b : b δ s − p , δ s − d , δ s − f (rad) 3.472, 3.457, 1.486 3.068, 2.387, 0.310 3.968, 2.224, 5.036Phase difference (Π) δ p − d , δ p − f , δ p − g (rad) 2.403, 4.956, 4.876 5.447, 1.48, 0.803 3.885, 0.0, 5.003 created with kinetic energy distribution having a sin-gle peak near 0 eV. The number of counts graduallydecreased to zero near 0.7 eV. But for incident elec-tron energy 10.5 eV onward a second peak appearsin the kinetic energy distribution curve. The secondpeak is located around 0.25 eV for 10.5 eV, around0.40 eV for 11 eV and around 0.58 eV for 11.5 eVelectron energies respectively. All the counts shownin Figure 3 are normalized at the zero eV peak. For9 eV, 9.5 eV and 10 eV incident electron energies theangular distributions of the ions created with kineticenergy between 0 to 0.65 eV are shown on the topof Figure 4. Angular distribution of the ions havingkinetic energy in the range between 0 to 0.1 eV, 0to 0.25 eV and 0 to 0.40 eV for incident electron en-ergies 10.5 eV, 11 eV and 11.5 eV, respectively areshown in the middle of Figure 4. At the bottom ofthe Figure 4 the angular distributions of the ions for10.5, 11 and 11.5 eV incident electron energies andhaving kinetic energy between 0.1 to 0.8 eV, 0.25 to0.65 eV and 0.4 to 1.0 eV respectively are shown. Theangular distributions are fitted using different statesand four partial waves for each state. According toO’Mallay and Taylor [23] the angular distribution ofthe ions have the general form as I ( k, θ, φ ) ∼ (cid:12)(cid:12)(cid:12)(cid:12)(cid:12)(cid:12) ∞ (cid:88) L = | µ | a L, | µ | ( k ) Y L,µ ( θ, φ ) (cid:12)(cid:12)(cid:12)(cid:12)(cid:12)(cid:12) (1)due to involvement of each resonant state. We havefitted the angular distribution using equation I ( θ ) ∼ (cid:88) | µ | (cid:12)(cid:12)(cid:12)(cid:12)(cid:12)(cid:12) (cid:88) j = | µ | a j Y j,µ e iδ j (cid:12)(cid:12)(cid:12)(cid:12)(cid:12)(cid:12) (2)In equation (2), µ = | Λ f − Λ i | , where Λ i and Λ f are the projection of the electronic axial orbital mo-mentum along the molecular axis for the initial andfinal molecular states, respectively. The summationover µ take care of the involvement of the differentstates in the process. The ground state of neutralCO molecule is Σ + (Λ i = 0). So µ =0, 1, 2 and3 represents a transition to Σ , Π , ∆ and Φ state re-spectively, a j ’s are the relative weighting factor ofthe different partial waves, δ j ’s denote the phase dif-ferences of the each partial waves with respect to thelowest order partial wave responsible for that particu-lar transition. The potential energy curve calculatedby Morgan et al. [18] showed in Franck-Condon tran-sition region near the resonance energy Σ , Π , ∆ andΦ are present. So the temporary CO − ion can beformed in any of these states. The angular distri-bution of the ions for 9 eV, 9.5 eV and 10 eV inci-dent electron energies can be fitted with a single statemodel for Σ to Σ transition. Fitting these distribu-tion with a Σ to Π transition only shows the con-tribution of Π state and the contribution increaseswith incident electron energies. The best fit is atwo state model with a Σ to Σ and Π transition.The angular distributions are fitted with the equa-tion, | (cid:80) j =0 a j e iα j Y j, | + | (cid:80) k =1 b k e iβ k Y k, | , shownin top of figure 4. Table 1 shows the parameter used4able 2: Fitting parameter for the angular distribution of ions created with lower kinetic energy for incident electronenergies 10.5, 11 and 11.5 eV. The angular distributions are fitted for a Σ to Σ transition only.
Weighting ratio of different partial waves a : a : a : a
1: 0.56: 0.17: 0.04 1: 0.70: 0.26: 0.09 1: 0.32: 0.15: 0.09Phase difference δ s − p , δ s − d , δ s − f (rad) 2.033, 3.231, 4.49 4.170, 2.922, 1.597 3.881, 2.105, 0.975 for the best fit to the data. The weighting ratio ofthe contribution of different partial waves are shownin the first row of the table. The phase difference(in radian) between different partial waves for eachstates are also shown in the table. Around the 10eV a b Σ + state of CO, as suggested by Sanche andSchulz [24] might be involved. Comer and Read [25]also suggested the presence of this state as a Feshbachresonance. With increasing energy the Π resonancestate near 8 eV shown in figure 2 of [18] also gets in-volved. The angular distribution of the ions for 10.5,11 and 11.5 eV incident electron energies and havingkinetic energy between 0 eV to the first minima valuein kinetic energy distribution curve are shown at themiddle in Figure 4. These near 0 eV O − ions arecreated due to the process II mentioned in the intro-duction, having energy threshold of 10.88 eV. Hall et al. [17] proposed that the intermediate negativeion state might be a Π state. However, our angulardistribution data gives the best fit with a Σ to Σ tran-sition model, using the equation, | (cid:80) j =0 a j e iα j Y j, | .In Table 2, the fitted parameters used for the fittingare shown. With increasing energy the contributionfrom Π state increases. The angular distributionsshow that intensity at 180 ◦ decreases with increasein incident electron energy. According to Dunn’s se-lection rule [26] for heteronuclear diatomic moleculea Σ to Σ parallel transition has non vanishing prob-ability but Σ to Π parallel transition has vanishingprobability. As the contribution of the Π state in-creases with the increase in incident electron energythe intensity at 180 ◦ decreases. Individual fitting forΠ state also shows this increasing contribution. Fit-ting with individual ∆ and Φ states shows that theyare not contributing in the process. Thus we propose that intermediate state to be mostly Σ state withminor contribution from Π state.The angular distributions of the ions with thehigher kinetic energy for 10.5, 11 and 11.5 eV incidentelectron energies are shown at the bottom of Figure 4.They are attributed due to process I [16, 17] as men-tioned in the introduction. The angular distributionshad two peaks near 130 and 230 ◦ , two small lobesaround 30 ◦ and 330 ◦ and almost no ions in 0 ◦ butreasonable number of ions along 180 ◦ . The angulardistribution has been fitted with four different singlestate model for a transition to Σ , Π , ∆ and Φ states,and also with multi-state model having different com-binations of the states. With a single Σ state modelthe angular distribution gives a good fit with R valuegreater than 0.97. But overestimates the intensity at180 ◦ and fails to predict the small lobes around 30 ◦ and 330 ◦ . A single Π state model also depicts theangular distribution reasonably well with R valuegreater than 0.9. The Π state model can successfullypredicts the two small lobes around 30 ◦ and 330 ◦ , butgives vanishing intensity at 180 ◦ as parallel transitionfrom Σ to Π states is not allowed [26]. A Σ to ∆ statetransition model can also fairly describes the angulardistribution but slightly over estimate the intensitiesaround 30 ◦ and 330 ◦ . This model also gives vanish-ing intensity at 180 ◦ as this transition is also forbid-den according to Dunn. A Σ → Φ state model canalso roughly describe the angular distribution, butlargely overestimates the the intensities around 30 ◦ and 330 ◦ and underestimates at 180 ◦ . Fitting withmulti-state models showed that the contribution ofΦ state is vanishingly small and a Σ to Σ , Π and ∆model gives the best fit. A three states model hav-ing contribution of four partial waves for each state,5able 3:
Fitting parameter for the angular distribution of ions created with higher kinetic energy for incident electronenergies of 10.5, 11 and 11.5 eV. The angular distributions are fitted for a Σ to Σ, Π and ∆ transition separately.
Weighting ratio of different partial waves a : a : a : a : 1: 0.99: 0.82: 1.30: 1: 0.65: 1.33: 1.68: 1: 1.19: 0.57: 0.62 b : b : b : b : 0.82: 1.67: 1.27: 0.67: 4.56: 2.91: 1.47: 2.04: 0.29: 0.73: 0.71: 0.80: c : c : c : c δ s − p , δ s − d , δ s − f (rad) 2.850, 4.002, 0.387 2.376, 3.457, 0.471 3.679, 0.225, 1.169Phase difference (Π) δ p − d , δ p − f , δ p − g (rad) 3.879, 0.866, 2.442 3.002, 1.779, 4.649 3.211, 0.361, 2.995Phase difference (∆) δ d − f , δ d − g , δ g − h (rad) 3.323, 4.261, 4.186 4.136, 1.207, 1.786 3.490, 1.389, 1.281 of the form | (cid:80) j =0 a j e iα j Y j, | + | (cid:80) k =1 b k e iβ k Y k, | + | (cid:80) m =2 c m e iγ m Y m, | has been used to fit the angulardistribution data. We have not consider any interfer-ence between different states as proposed by Tian etal. [19]. They have considered the interference effectto minimize the two small forward lobes around 30 ◦ and 330 ◦ predicted by the two state model without in-terference but absent in their experimental data. Butsurprisingly our experimental result shows the pres-ence of the two small forward lobes. Our model canalso predict the forward-backward asymmetry quitewell due to the interference between different par-tial waves involved in the process. In a recent study,Tian et al. [19] reported the angular distributionstaken at only two energies, no kinetic energy distri-butions were reported. The Figure 2 of [19] showsthat the central slice images taken at 9.5 and 10 eVgives completely different behavior. We also have ob-served the similar effect while going from 10 - 10.5eV. We have studied the ion yield curve (Figure 1)and considered the peak energy to be 9.9 eV [15].We have performed the energy calibration checkingbefore and after taking each set of VSI. Also, aboveenergy different could be due to different energy cali-bration used in different experiments. Based on mea-sured angular distribution in a limited angular rangeand considering Dunn’s selection rule [26], Hall et al. [17] concluded the negative ion resonance (NIR) statecould be a Π state. The authors excluded the Σ statebased on the trend of the experimental finding, i.e.zero counts in 0 ◦ and 180 ◦ directions. However, weare capable to measure the angular distribution overthe entire 2 π angle in a very efficient way. We alsoobserved strong forward-backward asymmetry in theangular distribution data, as shown in Figure 4.Our data are fitted with the formalism as men-tioned above and considering Dunn’s selection rule.Here, we conclude the temporary negative ion (TNI)state for the process II to be mainly Σ state withlittle contribution from Π state. In summary, we have measured the kinetic energy dis-tribution of O − ions produced from CO due to DEAand the angular distribution of the ions dependingon their kinetic energies over the resonance electronenergy. Two different dissociation channels are ob-served with distinct kinetic energies and angular dis-tributions. We do not find any evidence to include theinterference effect between different states to describethe angular distribution data. We also conclude thatthe process I is due to the involvement of Σ and Πintermediate states, whereas, unlike [17], we observed6
50 100 150 200 250 300 35000.20.40.60.811.21.41.61.82 Angle (Degree) I () /I ( ) I () /I ( ) I () /I ( ) Figure 4:
The fitted angular distribution for different in-cident electron energies. Angular distribution for entireKER ranges shown in top. Middle one shows the angulardistribution of the ions with KER in between 0 eV to therespective first minima in KER distribution curve(fig 3).Bottom image is the angular distribution for the KER be-tween the respective first minima to the maximum KER. main contribution coming from a Σ state with minorcontribution from Π state for the process II.
References [1] Badia Boudaffa, Pierre Cloutier, Darel Hunting,Michael A. Huels, and Lon Sanche. Resonant for-mation of dna strand breaks by low-energy (3 to 20ev) electrons.
Science , 287(5458):1658–1660, 2000.[2] Vaibhav S. Prabhudesai, Aditya H. Kelkar, Dhanan-jay Nandi, and E. Krishnakumar. Functional groupdependent site specific fragmentation of molecules bylow energy electrons.
Phys. Rev. Lett. , 95:143202,Sep 2005.[3] Diogo Almeida, Filipe Ferreira da Silva, GustavoGarcia, and Paulo Limao-Vieira. Selective bondcleavage in potassium collisions with pyrimidinebases of dna.
Phys. Rev. Lett. , 110:023201, Jan 2013.[4] David W. Chandler and Paul L. Houston. Two-dimensional imaging of state-selected photodissoci-ation products detected by multiphoton ionization.
The Journal of Chemical Physics , 87(2):1445–1447,1987.[5] Andre T.J.B Eppink and David H. Parker. Velocitymap imaging of ions and electrons using electrostaticlenses : Application in photoelectron and photofrag-ment ion imaging of molecular oxygen.
Review ofScientific Instruments , 68(3477), 1997.[6] Christoph R. Gebhardt, T. Peter Rakitzis, Peter C.Samartzis, Vlassis Ladopoulos, and Theofanis N.Kitsopoulos. Slice imaging: A new approach to ionimaging and velocity mapping.
Review of ScientificInstruments , 72(10):3848–3853, 2001.[7] Dave Townsend, Michael P. Minitti, and Arthur G.Suits. Direct current slice imaging.
Review of Scien-tific Instruments , 74(2530), 2003.[8] MNR Ashfold, NH Nahler, AJ Orr-Ewing, OPJVieuxmaire, RL Toomes, TN Kitsopoulos, I Anton-Garcia, D Chestakov, S-M Wu, and DH Parker.Imaging the dynamics of gas phase reactions.
Phys-ical Chemistry Chemical Physics , 8:26 – 53, 2006.Publisher: Royal Society of Chemistry.[9] Dhananjay Nandi, Vaibhav S. Prabhudesai, E. Kr-ishnakumar, and A. Chatterjee. Velocity slice imag-ing for dissociative electron attachment.
Review ofScientific Instruments , 76(5):053107, 2005.
10] H. Adaniya, D. S. Slaughter, T. Osipov, T. Weber,and A. Belkacem. A momentum imaging microscopefor dissociative electron attachment.
Review of Sci-entific Instruments , 83(2):023106, 2012.[11] A. Moradmand, J. B. Williams, A. L. Landers, andM. Fogle. Momentum-imaging apparatus for thestudy of dissociative electron attachment dynam-ics.
Review of Scientific Instruments , 84(3):033104,2013.[12] Bin Wu, Lei Xia, Hong-Kai Li, Xian-Jin Zeng, andShan Xi Tian. Positive/negative ion velocity map-ping apparatus for electron-molecule reactions.
Re-view of Scientific Instruments , 83(1):013108, 2012.[13] EnLiang Wang, Xu Shan, YuFeng Shi, YaGuo Tang,and XiangJun Chen. Momentum imaging spectrom-eter for molecular fragmentation dynamics inducedby pulsed electron beam.
Review of Scientific Instru-ments , 84(12):123110, 2013.[14] Alfred L. Vaughan. Mass spectrograph analyses, andcritical potentials for the production of ions by elec-tron impact, in nitrogen and carbon monoxide.
Phys.Rev. , 38:1687–1695, Nov 1931.[15] Donald Rapp and Donald D. Briglia. Total cross sec-tions for ionization and attachment in gases by elec-tron impact. II. negative-ion formation.
The Journalof Chemical Physics , 43(5):1480–1489, 1965.[16] P. J. Chantry. Dissociative attachment in CO andNO.
Phys. Rev. , 172:125–136, Aug 1968.[17] R. I. Hall, I. ˇCadeˇz, C. Schermann, and M. Tronc.Differential cross sections for dissociative attachmentin CO.
Phys. Rev. A , 15:599–610, Feb 1977.[18] L A Morgan and J Tennyson. Electron impact exci-tation cross sections for CO.
Journal of Physics B:Atomic, Molecular and Optical Physics , 26(15):2429,1993.[19] Shan Xi Tian, Bin Wu, Lei Xia, Yong-Feng Wang,Hong-Kai Li, Xian-Jin Zeng, Yi Luo, and JinlongYang. Coherent interference in the resonant dissocia-tive electron attachment to carbon monoxide.
Phys.Rev. A , 88:012708, Jul 2013.[20] O. Jagutzki, A. Cerezo, A. Czasch, R. Dorner,M. Hattas, Min Huang, V. Mergel, U. Spillmann,K. Ullmann-Pfleger, T. Weber, H. Schmidt-Bocking,and G.D.W. Smith. Multiple hit readout of a mi-crochannel plate detector with a three-layer delay-line anode.
Nuclear Science, IEEE Transactions on ,49(5):2477–2483, Oct 2002. [21] Dhananjay Nandi and E. Krishnakumar. Dissocia-tive electron attachment to polyatomic molecules:Ion kinetic energy measurements.
InternationalJournal of Mass Spectrometry , 289(1):39 – 46, 2010.[22] D S Slaughter, H Adaniya, T N Rescigno, D JHaxton, A E Orel, C W McCurdy, and A Belka-cem. Dissociative electron attachment to carbondioxide via the 8.2 ev feshbach resonance.
Journal ofPhysics B: Atomic, Molecular and Optical Physics ,44(20):205203, 2011.[23] Thomas F. O’Malley and Howard S. Taylor. An-gular dependence of scattering products in electron-molecule resonant excitation and in dissociative at-tachment.
Phys. Rev. , 176:207–221, Dec 1968.[24] L. Sanche and G. J. Schulz. New resonances in thetotal cross section of electrons on CO and O . Phys.Rev. Lett. , 26:943–946, Apr 1971.[25] J Comer and F H Read. Resonance scattering ofelectrons from CO.
Journal of Physics B: Atomicand Molecular Physics , 4(12):1678, 1971.[26] Gordon H. Dunn. Anisotropies in angular distribu-tions of molecular dissociation products.
Phys. Rev.Lett. , 8:62–64, Jan 1962., 8:62–64, Jan 1962.