th . . . ** 1. M 1 . LATA . WAP . :: . . N MI . 40 CN6 wz. 13: SA UNCLASSIFIED ORNL wo . . . .1 M - * . .. ' 590 ORNI-P-590 CONF-773-1 of OCT 3 0 1964 "EXPLOSION"QE.MULTICHARGED MOLECULAR IONS: CHEMICAL CONSEQUENCES OF INNER SHELL VACANCIES* THOMAS A. CARLSON and R. MILFORD WHITE Oak Ridge National Laboratory Oak Ridge, Tennessee, USA ABSTRACT Molecules containing atoms, which undergo internal conversion or electron capture, are subject to extensive decomposition. This decomposition results fro:d the large number of electrons that an atom loses as it adjusts to a vacancy in one of its inner shells. Electrons are pulled from the rest of the mylecule to the region of high positive charge, and the whole molecule literally" explodes"from Coulombic repulsion. In this paper a short review is first made of the past work on the molecular consequences to inner shell vacancies with particular emphasis on .. T , re ES. a iX - 1. 1 . . . been limited to qualitative observations. A description is then given of a new charge spectrometer, which utilizes X-rays to initiate inner shell vacancies. With this spectrometer it is possible to measure the relative abundances of all the fragment ions formed in the decomposition of the parent molecule without the errors that arose earlier from a dependence of collection efficiency on recoil energy. Furthermore, it has been possible to measure the recoil spectrum for each of the ions. As an example, some recently acquired data are given on the decomposition of CH I following vacancies formed primarily in the L shell of iodine by X-rays. 11 * . T * Research sponsored by the U. S. Atomic Energy Commission under contract with the Union Carbide Corporation. * Summer Participant, present address: Baker University, Baldwin, Kansas. ** -LEGAL NOTICE - The report we prepared uu nocow of anvenus sponsored work, Morthens the tale mane, me the numiestom, por me porno matug on how to counterede A. Make my wurthaty or representation, impreund or tapued, wo mapect to acou- rudy, cumpletua , or w.low of the button outdown Moreport, or that the a way Information, appunto, molt a proones de lowed to the report mey wer bearings petrostaty wodnoty at : 3. korumes any liabilities with respect to the uno af, er for damages resulting from the wameru , mert, wo or more teolowed the reporte . As wed to the home pasta notte a hello the Community melwa m . noch omployee or contractor of the Commission, or employen al much contractor prepares, daw n a, a onda scoutomy formation peront Nasployment of contract wto the Commtunion, or No amployment with much moneter. The decomposition 18 violont, with the molecule decomposing almost entirely into a*, cht, and 14t. The relative abundance of molecular luns 18 very small. The sums of the carbon, iodine, and hydrogen ions are in the approximate ratios of 1:1:3 suggesting that the quantity of neutral species 18 also small. The most abundant carbon ion 18 cat, which po88e88e8 an average recoil energy of about 40 eV. The most abundant lodine species 18 1>, which contrasts with an average charge of eight from an analysis of Xe ions produced with X-rays of the same energy. These and other data on the recoil and charge spectra from CA,I are compared with the calculations using a model of a multiple-ion Coulomb "explosion." 1. Introduction Not long after the discovery of the Szilard-Chalers effect, it was found that an atom undergoing internal conversion was separated from its parent molecule, despite the fact the process usually did not impart sufficient recoil energy to cause severance of the chemical bond.11,4 In fact, Hamill and Young' " demon- strated that essentially all the 5° Berline transitions in methyl bromide were effective in bond rupture. At that time it was correctly realized 14,5d that the decomposition following internal conversion was due to an inner shell vacancy in which an atom readjusts itself by a series of radiative and non-radiative transitions. Each of the non-radiative transitions (Auger processos) results in the loss of an electron, and the atom becomes highly charged. Such a series of Auger processes has been called & vacancy cascade.'°If the atom 18 part of a molecule, the whole molecule violently decomposes. The average charge carried by the lons resulting from internal conversion or electron capture was measured for a number of radioactive atoms and mole-.. cules.lroos Later, mass spectrometric analysis was employed by Kofoed-Hansen and more extensively by Snell and Pleasontonº, Ugdybel in measuring the relative abuncances of ions resulting from internal conversion and electron capture of various rare gases. These studies were extended to molecular systems by Wexler et al.-5,14% and by the authors.1+ The results indicate that the atom undergoing internal conversion or electron capture not only broke away from the parent molecule, but that the parent molecule itself underwent extensive de- composition. The studies on molecules were, however, only qualitative since the fragment ions received considerable recoil energy which strongly influenced their collection efficiencies. A new experimental approach has been made, however, to the problem of molecules undergoing violett decomposition as the result of inner shell vacancies, in which X-rays are used to create the intial vacancy. This new approach will be the main topic of this paper. ensen 197 VIVA RAIS . 8 . X 17. A 22 " 11, : 1 . .. RY . . . . . 14: - 't US ' ... . .... . 4 1 ", Since the chemical consequences of internal conversion are due to the formation of vacancies in the inner shells of an atom, it follows that the same phenomenon can be obtained by producing these holes by other means such as photo- electron emission. With the use of X-rays one has the advantage of (1) not being restricted to certain radioactive 180 topes and (2) being able to select by the proper choice of X-ray energy the shell where most of the vacancies will be produced. Ar. extensive program 18 already underway for measuring the charge spectra of rare gas ions following inner shell vacancies as produced by X-rays.lt1.doy As expected, the data from radioactive studies and X-ray studies are nearly identical in cases where the initial vacancy distributions are the game. In this paper we shall present some recent results on the fragment ions, essentially the same as if the lodine had undergone internal conversion. There is a large improvement, however, over radioactive studies in that the use of X-rays has allowed us to operate our spectrometer in such a way that the un- certainties of the relative abundances due to collection efficiencies are removed. In addition, we have also made measurements on the reccil spectra for most of the fragment ions. With this type of data 1t 18 now possible to have a more quantitative understanding of the violent molecular decomposition that accompanies extensive ionization. In the case of CH I it will be shown that the description of the decomposition most consistent with the data is an "explosion" of ions propelled by Coulombic force. 2. Experimental The charge spectrometer used in our study of methyl iodide has been described previously. -191 Tons born in the source volume (Fig. 1) are extracted, and magnetically analyzed. The times for atomic readjustment are very short, ~10~44 sec, compared to the time it takes an ion to leave the source volume, v10 sec, so the phenomena we are investigating are completed before analysis. The X-ray source is operated at 40 keV and 18 a Machlett AEG-50 tube with a tungsten target. When the irradiated gas 18 CH, I, this source of X-rays will produce initial vacancies mostly in the L shell of iodine. The photoelectron cross sections for carbon and hydrogen are negligible. The experiments reported in this paper were generally carried out at pressures of 2 x 10 torr in the source volume and 4 x 10*° torr in the analyzer. All studies were repeated at higher pressures, and the results were extrapolated to zero pressure. . . LS 17 Experiments on CH I were carried out by two different methods of analysis, which will be described briefly. Analysis No. 1: In this analysis an extraction voltago, which 18 4% of the total voltaga, Va, 18 applied between plates a and $ as shown in F48. 1. When an ion has a negligible recoil energy, as in the case of rare gas atous, the collection efficiency 18 independent of the choice of V.. In the study of the fragment ions from molecules the recoil energy, however, is not negligible, and the collection efficiency 18 a function of V./E, where n 18 the charge of the ion and E, 18 its recoil energy. By measuring the relative counting rates for each of the ions as a function of V, 1t 18 possible to extrapolate these intensities to 100% counting efficiency, and even to evaluate the relative re- coil energies of the different ions. Analysis No. 2: In the second method for analysis, plates a and b are shorted so that the ions are formed in a field-free region and are not extracted, but emerge from the second grid of plates b with only their initial recoil energy. Additional kinetic energy is then obtained as the lons pass through the potential field between plates b and ground. When the spectrum of fragment ions 18 measured, V, 18 set at a sufficiently high voltage to insure that all the ions of a given species are analyzed with nearly the same total energy. By reducing V. so that the recoil energy is a measurable fraction of the accel. erating voltage, we can evaluate the recoil energy spectrum for each species. The two analyses complement each other. The counting rates for analysis No. 2 are an order of magnitude lower than analysis No. 1, but the data are more dependable, particularly for ions of low charge and high recoil energy; and it is possible to make direct measurements of the recoil spectra. The first analysis is useful for those ions possessing high charge and low recoil energy, such as the iodine ions, which were collected with nearly 100% efficiency when the maximum voltage V. was applied. 3. Results and Discussions Table I lists the weighted average of results obtained from analyses No. 1 and No. 2 (cf. Experiment) for the relative abundances of the fragment ions formed following the X-radiation of CH, I. Also listed are the peak values of the recoil spectra for each of the ions examined. We note first the very low yield for molecule ions. Nearly all of the observed fragments are either #*, ch* or Int. This is in sharp contrast to electron impact studies and to decomposition following Bºdecay. The evidence in the present case is one of a violent decomposition. It should also be noted from Table I that the sum of iht, cht, and * lons are in the ratios of 1.0:1.0:3.0, suggesting tha, 911 the atoms originally making up CH I are ionized, and that very few neutral In Fig. 2 ure plotted (1) the spectrum of iodine ions from CH I; (2) the spectrum of I tons from HI;(21 (3) a spectrum of Xe cons; (24) and (4) a spectrum of 15-Xe fons.Cod In the first thres studies the initial atomic vacancies were produced by the same X-ray source. In the fourth study ionization arises from internal conversion. The Xe" lons are compared with the 14+ ions from HI, and the Ints ions from CH I. By plotting the charge spectre in this way, we see that all four spectra have nearly the same shape. From this comparison we may draw the following conclusions: (1) The X-ray source employed in our study gives rise to a charge spectrum for xenon that is very similar to that found with internal conversion. Sad (2) Formation of an inner shell vacancy in HI, which is 180- electronic with Xe, results in a spectrum of iodine ions nearly identical to that obtained from xenon except that one charge is carried away by the proton: (3) Following the vacancy cascade in CH, I, the iodine generally picks up two electrons more from the methyl group than it did from hydrogen. It is also instructive to list the most likely reactions that occur in these four studies, as deduced from the most abundant ions observed, as follows: (1) xe– xe8+ + Be" (2) HI → 17+ + x4 + 8e (3) CH TO → c2+ + 3#* + 15+ + 10e where indicates presence of an inner shell vacancy. The first two equations are restatements of the first two conclusions given above. From equation 3 we see that in the decomposition of CH I the two electrons picked up by the iodine are accompanied by a loss of two additional electrons. This result is consistent with auto-ionization processes, where for each vacancy filled by an electron, one electron goes into the continuum. For an example of a recoil energy spectrum, data taken on c* are plotted in Fig. 3. The results have been corrected for the finite resolution of the spectrometer, and the data have been compared under conditions of equal focusing, 1.e., where the ratio of the acceleration of the ion to its recoil energy is kept constant. In Fig. 4 are plotted the peak recoil energies listed in Table I as vy ..CR a function of the charge of the ion. Since the source of the recoil energy is Coulombic repulsion, the recoil should increase with higher charge. The fact that the recoil increases at a rate faster than the first power of the charge demonstrates that highly charged iodine ions promote greater ionization in the methyl group, resulting in proportionally greater kinetic energy for both the carbon and iodine fragments. 4. Simple Model for the" Explosion" of a Multicharged Molecular Ion Let us examine the chroñology of chemical consequences following an inner shell vacancy in one of the atoms of a molecule. First, a vacancy cascade occurs in about 10°44 sec, proceeding, for the most part, in an orderly fashion from the inner to the outer shell. Most of the ionization in the outer shell occurs in the last steps.Next, there is electron transference from the other atoms of the molecule to the highly charged ion with the possibility that some of the energy released in the electron transfer will create further ioni- zation. Some, if not all, of the atoms of the molecule are now positively charged, and they are repelled from each other with considerably energy. In actuality these three steps may overlap somewhat in time, but we do suggest that such an order of events occurs. (Preliminary studiesland on the decomposition of hydrogen lodide show that the # and I are still close to each other at the end of a vacancy cascade, although there is also evidence for some separation during the cascade.) For a simple model of the decomposition of CH, I we shall assume that at the conclusion of the ionization processes the bond distances and spacial arrangement of the A, C, and I are the same as in the neutral molecule. To calculate values related to the most probable recoil energies, we have assigned to each of the atoms the average charge as obtained from data in Table I. The charges are 1.0, 2.05, and 4.8, respectively for hydrogen, carbon and iodine. In Table II the recoil energies, calculated from the net Coulombic repulsion, are compared with experimental ones. The experimental values for carbon and lodine are interpolated from Fig. 4. Considering the simplicity of the model, the agreement is good. This agreement does not prove that the description given by the model is correct in any of its details. But it dues suggest that loni- zation and the subsequent Coulombic repulsion of the fragment ions, takes place while the ions are still close to each other. 5. Conclusion The chemical consequences of internal conversion or electron capture are due to atomic readjustment to inner shell vacancias. Since these vacancies may also be produced by photo-electron emission, X-rays can be used to study the same phenomena. This has been done for CH, I and a new spectrometer has been described, in which quantitative infni huition is obtained, regarding the relative abundances and recoll energy spectra of 1.3 fragment ions that are formed as the rosult of X-radiation. The data on CH, I give evidence of violent decomposition and have been correlated with a model that explains the results as a Coulombic "explosion" of a multicharged molecular ion. Mw :::. .. ..., 177 PERLU References [1] SEGRE, E., HALFORD, R. 8., and SEABORG, G. T., Phys. Rev. 35 (1939) 321. [2] DE VAULT, D. C. and LIBBY, W. F., Phys. Rev. 55 (1939) 322. 1.3) HAMILL, W. A. and YOUNG, J. A., J. Chem. Phys. 17 (1949) 215. [4] DE VAULT, D. and LIBBY, W. F., J. Am. Chem. Soc. 63 (1941) 3216. [5] COOPER, E. P., Phys. Rev. 61 (1942) 1. [6] PLEASONTON, F. and SNELL, A. H., Proc. Roy. Soc. 241A (1957) 142. PERLMAN, M. L. and MISKEL, J. A., Phys. Rev. 21 (1953) 899. [8] WEXLER, 8., Phys. Rev. 23 (1954) 182. 19 KOFOED-HANSEN, O., Phys. Rev. 96 (1954) 1045. [10] SNELL, A. H. and PLEASONTON, F., Phys. Rev. 100 (1955) 1396. [11] SNELL, A. H. and PLEASONTON, F., Phys. Rev. 107 (1957) 740. 12] SNELL, A. H. and PLEASONTON, F., Phys. Rev. 111 (1958) 1338. (13 WEXLER, S. and ANDERSON, G. R., J. Chem. Phys. 33 (1960) 850. (14 WEXLER, 8., J. Chem. Phys. 36 (1962) 1992. CARLSON, T. A. and WHITE, R. M., J. Chem. Phys. 38 (1963) 2930. Some earlier studies of fragment ions following X-radiation of molecules were carried out with a time-of-flight spectrometer employing coincidence measurements between the charged fragments and ejected electrons. The data because of collection efficiency problems are, however, only qualitative. KRAUSE, M., VESTAL, M., WAHRHAFTIG, A. L., LAMPE, F. W., and JOHNSTON, W.A., Technical Report ASD-TDR-62-10 (1962) (unpublished). [17] KRAUSE, M. O. VESTAL, M. L., JOENSTON, W. 1., and CARLSON, T. A., Phys. Rev. 133 (1964) A385. [18] CARLSON, T. A. and KRAUSE, M. O., Bull. Am. Phys. Soc. 2 (1964) 51; KRAUSE, M. 0. and CARLSON, T. A., Bull. Am. Phys. Soc. 2 (1964) 51. (19)' CARLSON, T. A. and KRAUSE, M. O., (to be published in Phys. Rev.) [20] CARLSON, T. A. and WHITE, R. M., J. Chem. Phys. 36 (1962) 2883; 38 (1963) 2075. [21] CARLSON, T. A. and WHITE, R. M. (unpublished results). This 18 true despite the fact that most of the initial vacancies in our study are in the I shell, while 1945 is mostly converted in the K shell. However, it should be noted that in 87% of the cases a K vacancy in Xe 1s transferred without ionization to a higher shell, usually I, by means of a radiative transition. [23] For the justification of these statements examine the extensive calcu- lations done on Auger transition rates for several atoms: R.A. Rubenstein, Ph.D. Thesis, University of Illinois (1955) (unpublished). 10 Pigure Captions Figure 1. Source volume for studying the charge d18tribution of ions produced by X-rays. Figure 2. Spectra of Xe and I ions resulting from readjustment to inner shell vacancies. The initial vacancies produced by X-rays are primarily in the L shell of I or Xe. The data on 55 Xe 18 taken from F. Pleasonton and A. H. Snell, Proc. Roy. Soc., 241A (1957) 141. Figure 3. Recoil spectrum of c