University of Southern California Dissertations and Theses

Vacuum Ultraviolet-Radiation Studies For Photoabsorption By Moderate-Temperature Plasmas

(USC Thesis Other)

Vacuum Ultraviolet-Radiation Studies For Photoabsorption By Moderate-Temperature Plasmas

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content .jeeewweiF This dissertation has been microfilmed exactly as received 68-17,015 BLACKWELL, Harvel Eugene, 1937- VACUUM ULTRAVIOLET RADIATION STUDIES FOR PHOTOABSORPTION BY MODERATE-TEMPERATURE PLASMAS. University of Southern California, Ph.D., 1968 Physics, spectroscopy University Microfilms, Inc., Ann Arbor, Michigan VACUUM ULTRAVIOLET RADIATION STUDIES FOR PHOTOABSORPTION BY MODERATE-TEMPERATURE PLASMAS by Harvel Eugene Blackwell A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (Physics) June 1968 UNIVERSITY O F SO U TH ERN CALIFORNIA THE GRADUATE SCHOOL UNIVERSITY PARK LOS ANGELES. CALIFORNIA 9 0 0 0 7 This dissertation, written by under the direction of A±s.„ Dissertation Com mittee, and approved by all its members, has been presented to and accepted by the Graduate School, in partial fulfillment of requirements for the degree of D O C TO R OF P H IL O S O P H Y Harve_lMEu£ene__Bla6^ Dean j)ate... DISSERTATION COMMITTEE Q . I AUTO S' T r h a i tC hairman TABLE OF CONTENTS Page LIST OF TABLES............... . . iv LIST OF FIGURES .............................. v Chapter I. INTRODUCTION............................ 1 Plasma Photoabsorption as a Technique . 3 Absorption Cross Sections ....... 8 Experimental Objectives ............. 10 II. DEFINITIONS AND DISCUSSION OF PLASMA PARAMETERS............................ 12 Equilibrium and Plasma Densities .... 12 Plasma Radiation ................ 18 Cylindircal Sources................. 21 Temperature.......................... 23 Effects of Plasma Perturbations .... 25 III. EXPERIMENTAL PROCEDURES, RESULTS, AND DISCUSSION............................ 31 Apparatus..................... 31 Emission Studies ...................... 34 Reduction of Neutral Argon Absorption. 35 Argon Absorption.................... 37 Hydrogen Emission ,................. 39 Argon Emission.................. 40 Neon Emission...................... 42 Procedures for Absorption Studies ... 43 ii 1X1 Chapter Page Oxygen Study ......................... 50 Molecular.................. 50 Atomic............................. 53 IV. CONCLUSIONS........................... 59 BIBLIOGRAPHY ................................. 62 LIST OF TABLES Table Page I. Observed Emission Lines of C I I ........... 46 II. Oscillator Strengths for the Lyman Series of Hydrogen ......... 47 III. 01 Series........... ................... 54 ! IV. Oil Transitions......................... 58 iv LIST OF FIGURES Figure Page 1. Experimental Arrangement for Measurements of Vacuum Ultraviolet Emission and Absorption............................. 32 2. Details of Differential Pumping Chambers between Light Source, Arc, and Vacuum Spectrograph............■ -r- ; .... 36 3. Absorption and Arc Emission Spectra for Argon near the Series Limit............... 38 4. Photoionization Cross Section of Atomic Hydrogen............................... 49 5. Absorption Cross Section of Atomic Oxygen . 51 v I. INTRODUCTION While considerable progress has been made in the study of photoabsorption cross sections of many molecular and atomic gases, the techniques remain as yet under developed for the measurement of many others. These include atoms which are chemically unstable at room tem perature, found in large concentrations only in the molecular state, and both atomic and molecular ions. The measurement of such atomic photoabsorption cross sections i now is of prime importance for application in the study of upper atmospheric photon-particle interactions for this planet, and for others of our solar system. Ion cross sections have long been a concern of the astrophysicist studying various processes which occur in the tenuous plasmas of hot stars. Theoretical estimates of cross sections have been presented but only meager data have been forthcoming. This work concerns itself with the introduction of a new technique for the investigation of this type of photoabsorption. This laboratory undertook the study of plasma ab sorption using ultraviolet radiation as a probe. In the early work the plasma of a Philips-type discharge was 1 2 employed. A magnetically driven shock tube was used later, as a search was made to find a more convenient and reliable plasma source. The investigation ultimately concentrated upon the measurement of photoionization cross sections, from the early study of atomic nitrogen to a shock tube study of singly ionized xenon. 1 Ehler and Weissler discharged N2 in a Philips ionization gauge type discharge and measured the absorption cross section of the excited gas. It was con- i eluded that ground-state N2 and N were the only strongly j absorbing species below the ionization threshold of N- O atoms in the wavelength range between 700 and 400 A. Although values were obtained for the cross section of N which agreed reasonably well with theory, the assumptions regarding the number densities of the basic constituents were not well founded. In a recent and more careful experiment, absorption cross section of atomic hydrpgen and oxygen was deter- ? mined by Cairns and Sampson using an electrodeless-rf or microwave discharge to produce the above species in their parent molecular gas. Measurements were carried ^A. W. Ehler and G. L. Weissler, J. Opt. Soc. Am. 45, 1035 (1955). 2 R. B. Cairns and J. A, R. Samson, J. Opt. Soc. Am. 56, 769 (1966) 1967. 3 out by passing ultraviolet radiation through the after glow of the discharge which was essentially at room temperature. The products of the gas discharge were molecular and atomic concentrations of about 3 x 10"^ and 15 -3 5 x 10 cm respectively. These products were passed through a 78.5 cm long windowless absorption cell. Since the atomic and molecular absorption cross sections do overlap in the greater part of the spectral range, separa- j tion of the molecular contribution to the total absorption j I becomes necessary in order to obtain the atomic absorption j curve. Such a separation could be done accurately, since j the number densities of all constituents could be measured and cross sections for all but the atomic con stituent were precisely known. Plasma Photoabsorption as a Technique The approach taken here to the measurement of photoabsorption by such atomic species is to avoid the molecular absorption and the resultant separation by the following technique. If the equilibrium temperature of a molecular gas is raised sufficiently, to say 1 eV, the gas becomes nearly completely dissociated, i.e., the number density of molecules becomes negligible in com parison with the number density of the atomic species. j Other products such as molecular ions are also negligible, 4 and furthermore, these densities may be determined ana lytically. In hydrogen, for example, calculation of densities shows that for a temperature of 0.7 eV the + number of gas particles such as H2 , and H is less than 1 in 10^ in comparison with the number of electrons, i | protons, and hydrogen atoms. Therefore, for such a high J temperature gas or plasma it should be possible to attribute the photoabsorption unambiguously to a single particle species, the neutral atom. One aim of this study then will be to give a more detailed analysis for the application of this technique. 3 4 Several investigators ’ in other laboratories were able to demonstrate O2 dissociation in shock heated 0 gases by probing the shock tube with radiation of 1470 A and noting the ultraviolet absorption in the Schumann- Runge dissociation continuum. No cross section measure ments in these experiments were attempted, but the position of the shock front could be marked by this method. 3 M. Camac and A. Vaughn, Bull. Amer. Phys. Soc. Ser. II 4, 291 (1959). ^B. Kivel, J. Quant. Spectry. Radiative Transfer 2, 509 (1962). In this laboratory, Blackwell et al.^ studied the absorption cross section of singly ionized xenon in a shock heated gas. This experiment yielded as a result the product of ion cross section and density. The ion density could not be determined accurately because of the transient nature of the shock tube plasma (shock velocities of the order of 10^ cm/sec). But a new tech nique for analyzing shock tube plasmas was developed. Neutral and ion density shock wave profiles were observed for the first time. This was accomplished by developing a reproducible shock and utilizing vacuum ultraviolet radiation from a high-voltage spark source as a probe. By accurately controlling the time between firing of the shock tube and the short duration spark source, the ultra violet radiation was made to pass, without the use of windows, through either the cold gas ahead of the shock front or through the hot plasma behind. The plasma tem perature could be made sufficiently large, by increasing the shock velocity, such that densities of neutral xenon were reduced to the extent that the photon-ion interaction dominated. Difficulty in using this technique increases for lighter gases because both the shock velocity and the ^H. E. Blackwell, G. S. Bajwa, G. S. Shipp, and G. ^jg^l^ssler, J. Quant. Spectry. Radiative Transfer 4 ^ , 249 6 plasma decay time increases with decreasing atomic number, resulting in inaccuracies in differential probe times. In addition, the time in which the plasma exhibits thermal equilibrium becomes shorter requiring still greater accuracy in time correlation between emission and ab sorption studies. Some form of thermal equilibrium is necessary if studies of plasma radiation are to be related to densities, that is, if a temperature is to be defined and used in the equilibrium relations. In order to further extend the experimental tech nique for determining such cross sections, as outlined for the transient shock heated plasma, methods for producing a time-independent plasma were studied. The d-c cascade arc was selected to provide this type of plasma. The cascade arc was first proposed by Maecker,^ i 7 who with Finkelnburg made an extensive survey of high- current electric arc research before 1954. It has been shown to provide a plasma of moderate temperatures, stably operated for periods of more than one hour. The g convection-stabilized arc investigated by Olsen was ^H, Maecker, Z. Naturforschung 11a, 457 (1956). 7 W. Finkelnburg and H. Maecker, Handbuch der Physik (Berlin: Springer-Verlag, 1956), XXII Gasentladungen. O A. N. Olsen, J. Quant. Spectry. Radiative "Trans fer 3, 303 (1963). 7 eliminated for use in this experiment because its elec trode configuration was not easily adaptable and because it exhibits spatial inhomogeneity in both its longitudinal and radial directions. In contrast, the wall-stabilized arc can be made essentially homogeneous for lengths of 10 cm or greater in its axial direction, but it still has the drawback of being spatially inhomogeneous in the radial direction. Since the probing ultraviolet radiation was to be directed longitudinally, i.e., coincident with the arc axis, this appeared to be not a serious drawback. The rapidly moving plasma of the shock tube was itself, by comparison, spatially inhomogeneous in the direction propagation which also was perpendicular to the radiation probe. The d-c arc as a spectroscopic tool was essentially developed in Germany. Early work utilizing the cascade Q arc includes that of Richter, who measured transition probabilities for atomic transitions producing visible emission lines for carbon, followed by transition proba- 10 11 12 bility measurements in nitrogen, * chlorine, 9J. Richter, Z, Astrophys. 51, 177 (1961). ^ ■ 9F. Mastrup and W. Wiese, Z. Astrophys. 44, 259 (1958). 11J, Richter, Z. Astrophys. 51, 177 (1961). 12P. Hey, Z. Physik 151, 79 (1959). -13 14. 15 . . . . . 16 j silicon, argon, iron, and titanium, and an in vestigation of the recombination and negative ion con- 17 18 19 tinuum in nitrogen, oxygen, , and chlorine. Here, an investigation is to be made of its use as a photo absorption cell. Absorption Cross Sections Absorption of radiation by a gas is governed by the Lambert-Beers relation -No, A Ix = Iox or In (I0x/Ix) = NoxA, (1.1) where Iox is the photon flux of wavelength X incident upon the absorbing gas; A the length of the absorbing path in cm; a x is the photoionization cross section at 2 the wavelength X in cm ; and N represents the number of 3 absorbing particles per cm . If the gas contains more than one type of absorber, then 15Ibid. 14W. E. Gericke, Z. Astrophys. 53^, 68 (1961). 150. Roder, Z. Astrophys. 55, 38 (1962). ■^K, H. Wobig, Z. Astrophys. ' 5 5 _ , - 100 (J962). 17G. Boldt, Z. Physik 154, 330 (1959). 18G. Boldt, Z. Physik 154, 319 (1959). *9H. Henning, Z. Physik 169, 467 (1962). 9 m (1.,/ip = I where N. is the number density o£ the j constituent (1.2) whose cross section is o.. J 20 The total photoionization cross section is de fined as the product of the total absorption cross sec tion, a, and the photoionization efficiency. This efficiency is defined as the ratio of the number of ion pairs, produced by the mechanisms of photoionization, to the larger number of photons, lost by any mechanism what ever from the beam of monochromatic radiation. An effec tive method of distinguishing between various ions produced is the use of the mass spectrometer, but it is not easily adaptable for use with either the shock tube or the arc. In the rare gases, the photoionization efficiency has been shown to be unity in the region of the ionization continuum, including the autoionization region near the threshold. Thus, the absorption cross section and the photoionization cross section are equal for these gases. In the case of other gases, molecular gases in particular, the photoionization cross section is not as clearly delineated, since various products such as neutral atoms and atomic ions all may result from the absorption of photons of a given energy. 2^N. Wainfan, W, C. Walker, and G. L. Weissler J. Appl, Phys. 24, 1318 (1953). 10 If a gas under study is excited or heated in some manner*the problem may be further complicated since addi tional species may contribute to the photoabsorption. The identification and evaluation of the added absorbers contribution is considerably easier if the excitation is such that the gas remains in thermal equilibrium. Experimental Objectives The arc is normally operated at pressures between 200 torr and one atmosphere or greater. This is essential if gas temperatures are to remain high and if thermal equilibrium is to be assured for all plasma constituents. It has been shown that equilibrium exists at pressures down to about 100 torr, below which deviations occur 21 not only between the gas and electron temperatures but 22 also between excited atomic states normally having a Boltzmann distribution. It is the purpose of this experiment to show that such an arc can be constructed and made to operate in conjunction with an ultraviolet optical system for use as a windowless photoabsorption cell. Emphasis will be on 21 I. M. Belousora and D. B. Gurevich, Opt. Spectry. (USSR) 10, 206 (1961). 22 - H. N. Olsen, private communication. 11 the techniques for analysis and no photoabsorption cross section curves will be presented. The cross sections of interest are those of atomic and ionic species which are chemically unstable at normal or room temperature and, as i a result, are found in relatively large concentrations only in the plasma state. II. DEFINITIONS AND DISCUSSION OF PLASMA PARAMETERS The words ionized gas and plasma are sometimes used interchangeably. A plasma is formally defined in terms of what is called the Debye shielding length, i.e., an ionized gas may be called a plasma if its physical extent is greater than its Debye length. Since the ion ized gas as a whole is electrically neutral, each ion may be considered to have surrounding it a spherically symmetric ion cloud whose net charge is equal and opposite to that of the ion itself. The potential which effects such a distribution is of the form* where r is a distance measured from the ion, e and Z refer to the unit electron charge and charge number, and PD is called the Debye-Huckel radius or simply the Debye length, which is given by Equilibrium and Plasma Densities V(r) = (Ze/r) exp (-r/pD), (2.1) ( 2. 2) 1 *J. D. Jackson, Classical Electrodynamics (New York: John Wiley and Sons, Inc., 1^62). 12 13 is the electron density and N. the density of ions of e ' 1 ' type i to which may be ascribed a temperature T^. Since the screened Coulomb potential V(r) becomes quite small at distances much greater than the Debye length, pD may be considered to be a measure of the size of the screening cloud surrounding the ion. It has been described as that distance at which electrostatic forces, which tend to impose charge neutrality, are balanced by the kinetic forces, which tend to produce non-neutrality. The concepts of such screening, which results from j I the large number of charged particles, is important not only in defining a plasma but in understanding certain phenomena such as Stark broadening of spectral lines and, in particular, the lowering of the ionization potential (change in energy required to remove an electron from a particle embedded in a plasma from that required for its removal from the particle in a free state) of atoms and ions. To be easily described analytically and to be generally useful as a spectroscopic tool the plasma should exhibit thermal equilibrium. A plasma may be said to exhibit local thermal equilibrium (LTE) if specific quantum states exhibit densities identical to those of a system in complete thermodynamic equilibrium having the same total density, temperature, and chemical composition as the actual system. In complete thermodynamic equilibrium, the j populations of all states of an atom or ion are com- | pletely specified as are the number of particles in given : molecular states, the number of molecules dissociated to form atoms, the electrons which are in the particular state of an atom, etc. The relative population of the state n is given by the Boltzmann equation V N = (gn/u) exP c - V kT)> C2*3) where N is the total density, Nn, gn, and En the popula tion, statistical weight, and excitation energy of the it. n state, respectively, T the absolute temperature, k the Boltzmann constant, and U = J g exp (-E/kT), the n partition function. Equation (2.3) gives the ratio of densities in the excited state n to the density of atoms in all states or the total density. It is proportional to the ratio of statistical weights. The ratio of such densities of the various states of an atom may be generalized to relate these densities to those of higher ionization stages of the same chemical species and to the density of free electrons. When total ion and neutral densities are desired, this leads to the equation 15 N*N U^CTJH _kT 3/2 . , — -4----- r.i--6- (-- -y ■ ) exp ("E x/kT), (2.4) N1 1 U1 (T) 21** 2 first derived by Saha from considerations of chemical equilibrium and called the Saha Equation, m refers to the mass of the electron, h to Planck's constant, and the superscripts i-1 and i refer to successive stages of ionization for the atom, e.g., may refer to densities of singly ionized constituents and N1 those twice ionized, etc. If the initial particle is a di atomic molecule, say AB, a similar relation describes the dissociative products, and it can be seen that the above equations are just a specialization, to the case of ionization, of the general equilibrium relations, . . tAm] 3 . nAnB m.nukT 3/2 JL7nP = ■ JL3B~ C- — -t -) exp C-BAB_A+B/kT). (2.5) N U 27rmABh Considering ionization of the molecule and of each dis sociative product, the following relations are obtained; NAB N U . t 3/2 “A ' B = ■ JjAB ^ exp ^"'BAB->-AB++e^T^ * (2,6) mA+m — ---e-=FA(T), (2.7) NA A 2 M. Saha, Phil. Mag. 4£, 472 (1920). F(T) is the temperature dependent function appropriate for the reaction, similar to the right side of Eq. (2.6) The densities in a high temperature molecular gas can thus be related. Such relations are basic for an equilibrium gas or plasma which, with the conditions of mass conservation, yield NAB(total) = NAB + NAB + NAB(dissociated) =£nAB And (2.9) & k + + 2NAB (dissociated) = NA + NB + NA + NB * jNk; and with the conditions of quasi-neutrality, N = NAB + NA + NB = JN^ + * 3 relates the densities of the various constituents to the plasma temperature. One other possibility is the forma tion of negative ions, whose densities are governed by relations of the type NANa UAU 3/2 3F-----------------------------C-EA^A+e/kT) (2.10) NA UA 2TThZ ' It should be noted that the maximum concentration of some -3 constituents such as negative ions may be less than 10 that of the major plaisma concentrations and therefore 17 have a negligible effect in the calculation of those numbers. However, they may not be negligible in produc- \ i ing observable effects, such as radition, and if necessary | ! their densities may be evaluated. Again, it should be j noted that for temperatures greateT than about 8000° K, j I molecular concentrations are small enough to be dropped from the calculation of major plasma constituents. Evaluation of plasma composition is usually made ! J on the basis of electron density and temperature measure- j ments with subsequent calculations through the Saha j ! relations. It should be mentioned that the above rela tions are derived assuming only those collisional inter actions necessary to establish the equilibrium. Field interactions such as those discussed in the beginning of this section may cause non-negligible effects, and these will be discussed later. Since a complete measurement of plasma parameters could not be-undertaken here, a detailed discussion of the various techniques for analyzing and measuring these parameters will not be made. But for nearly all types of plasmas, especially those of densities greater than 1014 c m a study of the radiation emitted is essential. 18 Plasma Radiation The two types of radiation that may be observed in studying plasma properties are (1 ) line radiation, resulting from bound-bound transitions between either electronic states of the atom or vibrational- states of a molecule, and (2) continuum radiation, resulting from free-bound or free-free. transitions, which are called recombination and Bremsstrahlung, respectively. Radia tive recombination, the reverse of phdtoionization, occurs when a free electron combines with an ion into a bound state with the excess energy of the electron being carried away by a photon. Bremsstrahlung may occur in any col- lisional process where there is a net loss of kinetic energy by one or both of the constituents to produce a photon. In normal laboratory plasmas the bulk of the Bremsstrahlung is produced in electron-ion encounters. The Bremsstrahlung intensity, while comparable with the recombination intensity of long wavelengths, is negligible in the ultraviolet unless the plasma temperature is ex tremely high, 10 to 100 eV. The intensity per unit wavelength interval per unit solid angle emitted by a homogeneous plasma source may be determined by considering the change in inten sity dl^ across a layer of thickness d a given by 19 dX^ = - k^IxdJl. (2.11) The first term describes the emission from the layer and the second describes the absorption. is the Planck function given by I Bx = (2hc2/X5) [exp (hc/XkT) - l)"1, (2.12) where h is Planck’s constant, c the velocity of light, and kT the product of the Boltzmann constant and tempera ture. k1 is the net absorption coefficient, i.e., the absorption coefficient corrected for induced emission, | i kj^ = k^[l - exp (-hc/XkT)]. . (2.13) j The absorption coefficient kx is identical to the product J of cross section and density (see Eq. 1.1). The solution to Eq. (2.11) gives Ix = Bx [1 - exp (-k^A)]. (2.14) In the case of bound-bound transitions, kj^ may be written H = A7rrofmuNu gu^ exp ("hc/*kT) t1 " exp (-hc/XkT) ] L (X) , (2.15) where r is the classical electron radius, f the o * mu oscillator strength of the transition from the upper state with statistical weight g to the lower state of • °m weight gu, and L(X) is the normalized line shape factor. For bound-free continuum absorption, the coefficient 20 3 derived by Salpeter may be approximated as kX,s - 2*roc4W s /h2x3> ( . 2 . 1 6 ) where Ng is the number density of atoms in the state s, E is the ionization energy for the atom in that state, i s i j and fg is the total continuum oscillator strength for j hydrogen. The plasma is said to be optically thin, i.e., all photons emitted escape from the plasma if « 1. (2.17) In this case, expansion of Eq. (2.14) yields Ix « Bxk^A. (2.18) Except in certain regions of ionization continua and near the centers of some strong emission lines, this is usually valid for most laboratory plasmas. For free-bound continuous emission the intensity may be written in the form 8ca3 ttEtt 3 / 2 g^ E-E -hc/X 1 X s 7 Z ^ IcT-) +”kX,sNeexp( KT (2.19) X gl Here a^ is the radius of the Bohr orbit, Eu is the o n * 4 * hydrogen ionization energy, g and g are statistical S J _ weights of the lower state and the ion ground state, and 3E. E. Salpeter and M. H. Zaidi, Phys. Rev. 125, 248 (1962). 21 E is the ionization energy of the radiating atom. This represents the radiation resulting from transitions to the state s. The total intensity at the wavelength then will be the superposition of this radiation with radiation resulting from transitions to the states of higher energy, if the enrgy states are close enough to provide an overlap. Line radiation may be more simply described. Eq. (2.19) yields h = <2hc2A 5) "VmvAsm/SlP NUL(X)». C2.20) The integrated intensity is then i IA = (4^2hc-2r0 /X3) fmu.(gm/gu) V , £2.21) or, condensing the notation and using the Boltzmann relations, 1 = Ckgmfmu/Ux3:i exp < > V kT:)- (2>22) Cylindrical Sources In most cases, particularly when accuracy in determining plasma parameters is necessary, radiation from a cylindrical plasma such as that of the arc is ob served in a direction perpendicular to its axis. In this case the radiation source is strictly inhomogeneous and must be given special consideration. Most plasma ! i sources are, in fact, never completely hompgeneous. \ 22 Both densities and temperatures of the arc vary along the line of sight with the temperature maxima occurring on the axis or center of the cylindrical column and decreas ing radially to the temperature of the containing walls. The radiation observed perpendicular to the arc axis con sists, therefore, of a mixture of radiation from regions of different temperatures which also has passed through different temperature regions; hence, local emission co efficients must be found from the observed intensities by an integral inversion procedure^ using what is called Abel’s integral equation to express the observed trans verse intensity distribution I(x) in terms of the true radial distribution i(r). Considering the z axis to be coincident with the axis of the arc and x the direction of observation, the equation may be written as Here r is the radial coordinate of the radiating source and y is the transverse linear dimension, such that fR x (r) rdr ( x * 2 - y 2 ) 1 / 2 (2.23) 2 2 2 r = x + y . The inversion is then accomplished by tak ing the Abel transform to obtain 826 (I960). M. P. Freeman and S. Katz, J. Opt. Soc. Am. 50, 23 rR r (y -r ) I'Cvldy 2 2 * .1/2 (2.24) where I' = dl/dy. Once the true intensity distribution is estab lished, the measured radiation may be used for a study of the plasma properties. Several methods exist for the measurement of the plasma temperature. Line intensity ratios are one such method. Line-continuum and continuum-continuum intensity ratios may give considerable error in the case of ad mixed plasma constituents or for those plasmas contain ing a significant impurity concentration, since some constituents will emit more strongly than others and separation and identification of the various transitions becomes quite difficult. Absolute line intensities may be used only when the density of the emitting species is known and may likewise give considerable error for mixed plasmas. Perhaps the simplest method of temperature measure ment utilizes line intensity ratios of atomic or ionic lines having transition probabilities that are reliably known, observing the radiation not in the radial direc tion but along the more homogeneous route of the axis of Temperature 24 the arc. The observed line intensity can be written h = ( k g m f m n / u x 3 ) exP C - V k T : ) l C 2 - 2 5 ) i where £ is the absorption oscillator strength and K is an unknown but constant factor which includes the various instrumental conditions of obtaining data. Condensing the notation, the intensity ratio of two atomic lines, 1^ and I2, can be written as Il/I2 = <SlflX2/g2f2Xl5 exp (E1 " E2)/kT]* C2>26) t Taking the logarithm of Eq. (2.26) yields j ( In (I1 /I2) = In (g1f1 X^/g2f2\^) - (Ex - E2)/kT. (2.27) The usual method of solving for T is to plot In (I I 2) or rather log (I^/I2) versus 1/T, a straight line. The 3 3 intercept of such a graph is log * an^ tke slope is determined by the energy difference E^ - E2> Accuracy here is limited by this energy difference which is usually much smaller than kT if one observes lines I from the same neutral or ionic species. However, if the j 1 intensity of one ionic line and one atomic line is used, the energy difference includes the ionization potential, and the ratio becomes a more sensitive function of T. The ionization potential appears because the ion density must be determined through use,of the Saha equations. It should be noted here that there is an effective shift 25 of the ionization potential of the atom in the plasma which is not precisely known. This shift or lowering of the ionization potential is relatively small compared with the ionization potential itself and must be care fully considered when precision measurements are attempted. It will be discussed in more detail below. Measured values of the intensity ratio from a number of I pairs of lines can be averaged to give accuracy. I Alternatively, the above averaging may be done j in another way. Consider the logarithm of the expression | ] for a single line. Since K does not depend on E, the j I I measured intensities from a number of lines can be used to plot a graph of In (IX /gf) versus E, and I, g, f, and X can be expressed in units which are completely arbitrary for each set of data. The temperature is determined from the slope of the graph. Effects of Plasma Perturbations The ionization potential of an atom is usually calculated by assuming that the individual atom is free, i.e., it is completely independent of its surroundings or it is measured under conditions in which that assumption is valid. j i Briefly,, a change or shift in the ionization | potential can be expected to result from the microfield 26 of a single ion causing a perturbation of the atom or, as in the Debye theory, from a spherically symmetric cloud of shielding charged particles. This modifies the Coulomb potential resulting in a decreased attraction i between the nucleus of the atom under consideration and its electron. Thus the energy required to free the electron from the perturbed atom will be lower than that required for an unperturbed or free atom. ! i This describes a lowering of the ionization poten- j 5 ' tial as first discussed by Elenbaas who gave experimental j evidence from the high pressure mercury discharge. Since ! that time much work has been done in an effort to clarify the calculations of Unsold^ and develop a satisfactory [ theory. Although differing expressions have been derived, j i 7 f i ! as for example by Ecker and Brunner, a representative j relation for the shift in potential may be given by AE = “mAEm + “pAEp = “me2/ro + “pe2/lV (2-28) Here a is some coefficient of order unity with m and p designating the microfield and polarization terms, 5W. Elenbaas, Physica 4, 279 (1937). ^A. Unsold, Z. Astrophys. 2 4 , 355 (1948). ^G. Ecker and W. Weigel, Ann. Physik 17^, 126 (1956). ®J. Brunner, Z. Physik 159, 288 (1960). 27 1/3 respectively. rQ = (3/4'irN^) ' is the mean distance between the perturbed and perturbing particle, and gQ refers as before to the radius of the Debye sphere. It has been generally concluded up to the present time, that ? ! the e /r term should not be included below a critical o 2 electron density N = * 3kT/4ire , but this is not firmly V established. This small correction to the ionization energy, less than 1 eV and usually of the order 0.1 eV, must be applied in all relations in which the ionization | i i energy is involved. The Saha relations and intensity j i equations in given form remain valid, since only the sub- j stitution E - AE for E need be made. ! j Recent attempts to measure the change in ionization} 9 ! potential include the work of Olsen, who determined the j j shift by a measurement of the continuum radiation emitted j by a convection-stabilized argon arc plasma. This J measurement is perhaps the most accurate to date. In ! 10 arc emission studies by Boldt in the visible region of the spectrum, shifts in the photoelectric edges of N and 0 were found from observations of the continuum 9H. N. Olsen, Phys. Rev. 124, 1703 (1961). j i 1 0 G. Boldt, Z. Phys. 154, 319 (1959). 28 emission. These, however, were later explained1* as due to the fading of a system of lines near an absorption edge, a result of broadening, and the shifts were found to be in good agreement with what is called the Inglis- 1 2 Teller limit. This limit is simply the approximate point at which the Stark-broadened line widths become equal to the energy difference between two lines. The series limit is then effectively shifted by an amount given by E = 't(a0Ne)4/1 5EH, (2.29) where the calculation is made in terms of the hydrogen limit, EH. The maximum principle quantum number where merging occurs is found to be NI' 5 = - Qt i°- 1" ■ a « 3 > (2.30) m N o * • e or, rearranging, N = 0.013 N" 7 , 5 a' 3 . (2.31) e m o This can be used as a rough estimate of the electron density in the case of single ionization. 1 1 B. H. Armstrong, J. Quant. Spectry. Radiative Transfer £, 207 (1964). *^D. R. Inglis and E. Teller, Astrophys. J. 90, 439 (1939), 29 Stark-broadening is itself a thoroughly interest ing microfield effect which can be used, through line profile measurements, to obtain electron densities with accuracies from 20 per cent to 5 per cent. Although a detailed discussion cannot be made here, it can be effec tively described by the recen£ theories. In the static ' approximation, the perturbers are assumed to be at rest j and the atoms or ions which emit the radiation are j ! assumed to be in some average electric field. The sharp • I frequency distribution for a given emitter can be cal culated by the ordinary quantum mechanics of the Stark effect. This frequency distribution will be different for other emitters since they will be in slightly differ ent electric fields. As a result, the statistical distribution of these frequencies near the unperturbed line produces a broadened spectral line of specific width and shift. In the adiabatic approximation, the perturbers move in classical paths causing the energy levels of the emitting atom or ion to fluctuate with time. More sophisticated approximations involve quantum consideration of collisions in the generalized impact 13 theory developed by Kolb and Griem. It has recently *^A. C. Kolb and H. R. Griem, Phys. Rev. Ill, 514 (1958). 30 14 been given a thorough treatment by Griem, who has also tabulated Stark-broadening parameters for many of the J visible and ultraviolet lines from the elements H to Ca. As a result of the plasma perturbations discussed, neither the line emission nor the continuum emission and absorption edges of plasma constituents can be expected to be sharp, since there are spreadings, shifts of the upper state energy levels, and an actual shift in the ionization energy. 1 ^H. R. Griem, Plasma Spectroscopy (New York: McGraw-Hill Book Co., Inc., 1964). III. EXPERIMENTAL PROCEDURES, RESULTS, AND DISCUSSION Apparatus A diagram of the apparatus is shown in Figure 1. | Its essential components were a pulsed spark source to act as a radiation probe, the wall-stabilized cascade arc for heating the gas, and a vacuum spectrograph. The spectrograph employed an original, unblazed glass grating lightly ruled with 591 lines/mm with a radius of curvature of 2 m. The grating center was fixed 38 cm from the primary slit, giving an angle of incidence O j of 79°. At 150 A the instrument had a reciprocal dis- O person of approximately 2 A/mm. This value increased O O to about 3.3 A/mm at 1000 A. Dispersed radiation down to O 80 A can be observed with this instrument. The spark source employed consisted of a pulsed high voltage spark confined within a Boron Nitride capillary which was repetitively discharged by using an Ignitron switching circuit. Details of both the spectro graph and the spark sources have been described previously.^ *H. E. Bla£kwell, G. S. Shipp, M. Ogawa, and G. L. Weissler, J. Opt. Soc. Am. 5 6 . , 665 C1966). 31 Photographic Plate Optic Pump Out to Oil Diffusion Pump Rowland Focusing Curve Pump Out to Oil Jet Booster Pump Pump Out to Mechanical Booster r Pumps puinp 0ut ^ ^ Mechanical ^ \ 1 M Pump / C ' ( 2) Figure 1. Capillary Spark Light Source Experimental arrangement for measurements of vacuum ultraviolet emission and absorption. (1 ) Spjark source differential pressure chamber, (2) arc, (3) first spectrograph pressure chamber, (4) second spectrograph differential pressure chamber, and (5) 2-meter grazing incidence vacuum spectrograph. 04 33 The cascade arc consisted of a series of water- cooled centrally bored plates made of brass and copper, electrically isolated from each other, and mounted between two similar water-cooled electrode plates. The copper portion was made from a 6.4 mm thick plate, 37 mm in diameter, which had been centrally drilled to form a bore of 4.4 mm diameter. It was machined to form a spool with 1.4 mm walls. This spool was fitted and soldered into a 60 mm square brass plate containing a water inlet and outlet and designed such that water could be passed through the copper spool. Carbon rods inserted into the i bore of the electrode plates served as electrodes. These electrodes were separated by a distance of 5 cm, which could be varied by removing or inserting one or more plates. The inserted plates were electrically isolated by a 1/16 inch diameter Viton 0-ring which separated them by 1/32 inch and sealed them for vacuum evacuation. Their bore served as the containing channel for the arc plasma, and the stability of the long arc was made * possible by removal of the dissipated energy in the form of heat through the thin copper walls. The flow of water served this purpose. Other apparatus components included a dc power supply, 300 V, 150 A, a stainless steel water cooled j i resistor of approximately 1 ohm, for controlling the arc 34 current, a high voltage dc power supply for the spark source, 25 kV, 100 mA, and the associated electronic circuitry for pulsing the spark source. 5 \ Emission Studies i The first experiments in the vacuum ultraviolet region of the spectrum utilizing a stabilized high- ! 2 current are those of Boldt whose measurements were O restricted to the wavelength range from B100 to 1800 A. The wall-stabilized arc provided a stable, equilibrium, high temperature plasma, with little wear and impurity contamination, if operated at or near atmospheric pressure using argon as a buffer gas in the electrode regions. In one of Boldt's experiments, oscillator strengths of Cl lines were measured by adding controlled amounts of carbon dioxide into the central region of a wall stabilized arc burning in argon. The radiation, ob served end-on through a large bore in the cathode, was required to pass through the cooling argon gas at high pressure. This established an emission cut off in the O neighborhood of 790 A due to neutral argon absorption. Absorption by air was eliminated by providing a series of differential pumping chambers to reduce the ^G. Boldt, Z, Naturforschung 18a, 1170 (1963). 35 atmospheric pressure of the arc to the vacuum pressure of less than 1 micron required because of the long path length of the ultraviolet spectrograph. A typical exposure taken with an open bore elec trode showed strong emission up to the point where ab sorption by the energy states of neutral argon near the series limit and beyond cut off the observed emission entirely. The same type of arrangement as Boldt has been employed here, but an attempt has been made to eliminate the neutral gas absorption observed on the plate by moving the first differential pressure stage nearer to the arc plasma. Reduction of Neutral Argon Absorption Absorption by neutral argon was reduced by utiliz ing the arrangement shown in Figure 2. The carbon electrode with a small central hole 1/32 inch in diameter served to isolate the arc from the first differential chamber. Differential pumping from the ear of the elec trode was accomplished by a Kinney pump system employing a double rotor Roots type pump with a speed of 300 cfm in the range from about 0.01 to 1 torr. A second dif ferential stage, isolated from the first by a 1/32 inch hole located 1/2 inch from the ear of the electrode, utilized an oil diffusion booster pump. Vacuum Spectrograph Optic Axis Primary Slit Cascade Arc To Welch 1398B Mechanical Pump I — f — I To Oil Jet To Kinney-Roots Booster Pump KMBD. 400/KD 30 Mechanical Pump Figure 2. Details of differential pumping chambers between light source, arc, and the vacuum spectrograph. w o* 37 With a typical arc pressure of 500 torr, the first differential chamber could be kept at a pressure of \ 90 microns while the second chamber remained at a pressure below 10 microns and the spectrograph at 0.2 microns. With this arrangement cold gas absorption was not en tirely eliminated but was considerably reduced, that is, decreased to the extent that arc emission below the ionization limit of neutral argon could be clearly ob served. Argon Absorption The neutral argon absorption was quite distinct and, in fact, served as an excellent wavelength calibra tion. Absorption could be observed to the higher series members of argon (3p^ ^S0 - nx). Several autoionizing 2 2 states could be observed between the P3 / 2 an(* Pl/2 o o series limits at 786.71 A and 777.96 A, as shown in Figure 3. The background intensity for the absorption spectra of Figure 3 is the Stark-broadened argon emission lines from the hot arc plasma, which begin to overlap near the series limit. At wavelengths shorter than the ^Pl/2 (actually at some longer wavelength correspond ing to the lowered ionization potential) the background radiation represented the free-bound continuous emission, limit d[l/2] I , ■ ' 11"' . ' 1 "' ■ ■ ■ * • s' [1/2] series (B) 12 11 10 9 8 1/7. l i m i t Figure 3. Absorption and arc emission spectra for argon near the series limit. (A) Open electrode showing strong neutral absorption; (B) neutral absorption reduced by electrode differential pumping. The apparent continuum background is formed by over lapping Stark-broadened line emission1 "(s'§e (i)). w 39 that is, recombination of free electrons with singly ion ized argon to produce netural argon in the ground state, A+ + e" - * ■ A(1So) + hv. (3.1) The clarity of the autoionization region is quite remark- j i able although ideal conditions of temperature and pressure! I for an absorption study were not realized. J I In spectra with the open bore electrode, absorption; i lines near the series limit appeared quite broad indicat- j 1 ing considerable absorption by the hot gas between the j electrode and differential chamber. It can be noted, | from Figure 3, that the emission cut-off in this case appears at a wavelength longer than the ionization limit of neutral argon, since the overlapping energy states exhibit continuous absorption. ! Hydrogen Emission j Hydrogen proved to be an omnipresent impurity in i ! the arc. Since atomic parameters for hydrogen are well | j known, it is often used in the measurement of plasma para-j meters. In certain cases, it was purposely introduced i i into the arc in amounts from 1 to 10 per cent. The j spectra showed the Lyman series from Ly^ (Lya was beyond j the spectral range of the instrument setting) to shorter ! wavelengths, The lines were broadened, as expected, 40 and began to merge at about n = 7. At wavelengths O shorter than the series limit at 910 A the Lyman con tinuum H+ + e" H (2S) + hv (3.2) could be seen. Calculations on Stark-broadening para- meters for Lya and Ly^ are available and line profiles 4 5 for both lines have been experimentally investigated. * i Line widths are essentially proportional to the two-thirds 3/2 power of electron density (N = C (N ,T) AX ) and are 6 6 j often used for the measurement of that parameter. i Argon Emission Argon emission lines begin to overlap in the wings between the 6s [1 1/2] and 6s1 [1/2] states which are paired in this wavelength region with the 4d [1 1/2] and 4d' [1 1/2] states. Towards shorter wavelengths as the series of lines become more closely spaced, the emission lines became indistinguishable from each other. Free- bound transitions to the ground state of argon produce strong continuous emission at wavelengths shorter than ' that corresponding to the series limits. ^H. R. Griem, A. C. Kolb, and K, Y Shen, Astro - phys. J. 135, 272 (1962). 4r. c. Elton and H. R. Griem, Phys. Rev. 135, I A1550 (1964). 5G. Boldt and W. S. Cooper, Z. Naturforschung 19, 968 (1964).__ 41 g Stark-broadening parameters have been computed , for two AI lines in this region of the spectrum, the 3p6 1S0 - 3d [1/2] transition at 876.063 A. The 876 A line appeared to be broadened somewhat more than theory | would predict. In general, the calculated Stark widths and shifts of emission lines in the far ultraviolet appear too small to be accurately measured with the present instrument and are of the same order of magnitude as Doppler widths. No calculations on the profiles of the broader high series members have been made. Line I O radiation of wavelengths shorter than 790 A was due " t * almost entirely to singly ionized argon, A . In fact, a near complete spectrum of transitions from upper states, at least the doublet S,P,D, and F states, to the "two" ground states of the ion could be observed. As in the case of neutral argon, the lines resulting from transi tions involving the higher energy states were broadened and formed a continuum near the unperturbed ionization limit at 27.5 eV (448.79 A). i It was noted that the intensity, of the ion lines decreased with increasing arc current (increasing current implying increased temperature), while there appeared to be no significant decrease in intensity of lines ^H. R. Griem, Plasma Spectroscopy (New York: McGraw-Hill Book Co., Inc., 1964). 42 resulting from neutral argon transitions. A decrease in A+ intensity with increasing temperature would imply a decrease in ion density and hence a temperature in excess of 16,000° K, but there should be a corresponding, if not greater, decrease in the neutral argon line intensity. Observations would seem to imply that much of this ion j i radiation comes from a region of temperature different j from that of the arc column, possibly in the neighborhood of the arc electrode or within its bore where non equilibrium conditions possibly may exist. Further j ! j evidence of this came from neon spectra. I I t i ; j Neon Emission j An arc operating in neon with argon as an impurity | showed neutral neon emission, strong lines representing 6 1 ° the transitions 2p SQ - 3s [1 1/2], 743.70 A, and 2p^ - 3s [1/2], 735.89 A, but neutral argon lines were absent. Surprisingly, the ion lines of argon per- j sisted. These lines grew weaker as the arc current was j increased until only the neon lines could be seen. No j continuum emission was observed. Hydrogen emission, ever present in the argon spectra, was weak but still ob servable. At this point studies were initiated in order i to determine if impurities were present in numbers suf ficient to cause absorption of radiation passing through the plasma. 43 Procedures for Absorption Studies To study the plasma absorption, a high voltage spark light source was constructed and aligned along the optic axis as was shown in Figure 2. The differential pumping chamber, evacuated by a Kinney 1398 mechanical pump, was able to keep the spark source pressure down to i about 700 microns with the arc operating at about 250 torr of neon. The source was tested at pressures between 400 and [ I 2 0 0 0 microns with argon and then with helium as the j carrier gases. With a pressure of 200 torr of argon or helium in the arc channel, the spectral cutoffs of the spark source radiation due to strong absorption in the ionization continuum of either argon or helium were ob served as expected. There was no observable decrease in line inten sities of spark source radiation passing through the arc chamber with a pressure of 2 0 0 torr, when compared , directly with the intensity of lines with a gas pressure of 200 microns in the arc chamber. Therefore, it was concluded that absorption by impurities in the high pressure carrier gas would be negligible. Effects of ; impurities from the walls of the arc channel, if any, | were expected to be minimized within a few minutes after striking the arc. On the other hand, if the impurity 44 concentration were constant it would cause no observable effect during the course of an experimental run. The procedure was then as follows. With the arc operating at a carrier gas pressure of 2 0 0 torr or greater, a photographic exposure of arc emission was made. Next, radiation from the spark source was made to pass through the arc operating under the same conditions, and a second 1 exposure of both spark and arc emission was made on the same photographic plate. The spark lines were then easily distinguished from the arc emission and their in- ! tensities for this exposure established the incident j intensity. For the third exposure, a test gas was leaked j into the arc plasma. Radiation from the spark source was j now selectively absorbed by whatever species remains of j ) the test gas which is now at an elevated temperature. Under these conditions, constant impurities from either the carrier gas or from the walls of arc channel make little difference, since this same background absorption will subtract from the incident as well as from the absorbed spectra. Since the radiation from the argon arc showed very O strong emission below about 860 A, and neon emission was O very weak down to 600 A, the neon arc, with its wider O transparent range, 1 0 0 0 to 600 A, appeared advantageous for observation of radiation from the spark source 45 passing through the length of the arc. A more detailed study of emission from the arc operating in pure neon reveale.d the presence of CII lines. However, these lines, listed in Table I, are easily identified, weak, and spaced such that spark source emission lines in the wavelength region could be distinguished easily. Hydrogen was introduced into the neon arc operat ing at a pressure of 250 torr with a current of 100 A r in order to obtain a rough estimate of the plasma tern- j perature under these conditions. A listing of transition j I probabilities for the Lyman lines is shown in Table II. Using Eq. (2.20) the ratio of intensities for the Lyman lines, n = 4 to n = 7 gave a temperature of 10,000° K. t No error limit is given since the measurement is to serve . only as an indication of temperature and may contain con siderable error due to the addition of radiation coming from the near electrode region. Also, it should be noted that it represented an average over a radius of 0.8 mm of the arc core. This is the size of the elec trode hole used for differential pumping in this case. While the radial temperature variation over this distance is not expected to be severe (of the order 1 0 0 CP K) future measurements will be carried out using the emission in the visible region to obtain a true temperatire 46 TABLE I OBSERVED EMISSION LINES OF ClI* 0 Wavelength, A Intensity O Wavelength, A Intensity 858.561 9 686.480 2 858.094 8 651.342 8 809.770 3 651.262 7 809.682 4 651.216 7 806.846 6 641.875 6 806.684 4 641.772 6 806.555 7 641.591 6 806.384 5 636.247 4 799.947 4 635.988 3 799.664 4 595.032 7 687.355 687.059 11 1 0 594.808 6 R. L. Kelley, 'Vacuum Ultraviolet Emission Lines," UCRL Publication 5612. 47 TABLE II OSCILLATOR STRENGTHS FOR THE LYMAN SERIES OF HYDROGEN n O Wavelength, A Oscillator Energy, eV 2 1215.67 0.4162 10.15 3 1025.722 0.0791 12.04 4 972.537 0.0290 12.69 5 949.743 0.0139 13.00 6 937.804 0.0078 13.16 7 930.748 0.0048 13.26 8 926.226 0.0032 13.33 i i 48 profile in order to compare and substantiate the ultra violet measurements. With hydrogen as an impurity in neon, a first attempt was made to pass the radiation from the spark source through the hot plasma. Plasma absorption was evident from visual observation of the photographic ex posure but the weak source intensity coupled with the strong Lyman emission made it difficult to assess accurately the degree of hydrogen absorption. The H atom has a well known photoionization cross section, which is shown in Figure 4. These first observations pointed to the need of greater source intensity for photographic measurements of this type. Increasing the hydrogen con centration to achieve greater absorption of the spark source line radiation caused increases in the intensity of the Lyman continuum emission by the arc. The maximum intensity of line radiation from the spark.source was less than that of the arc continuum with a Ne - Hg ratio of 50 - 1. This condition necessitated comparison of two line intensities one of which was superimposed upon a continuum of intensity equal to that of the line. While the Lyman continuum decreases with energy or decreasing wavelength, the photoionization cross section of H does likewise, causing the intensity change to be less perceptible for low hydrqgen densities. i i 49 • e o 00 o I— I ' _I P U < U CO ( n t n o u u P i o •H P nl N •H Pi O •H O P O 400 600 800 1000 o Wavelength (A) Figure 4. Photoionization cross section of atomic hydrogen. 50 While hydrogen was thought to be desirable to give indications of absorption because of its simple structure, observations pointed to the use of a gas with a photo ionization cross section which increases with energy in contrast with the hydrogen decrease and, in addition, one which exhibits weaker recombination emission. Oxygen, although its atomic structure is considerably more complex than hydrogen, satisfied these requirements. Its cross section is shown in Figure 5. In addition, it is a gas of primary concern in upper atmosphere studies; there fore, a more, detailed discussion of its structure and of the absorption processes will follow. Oxygen Study For an idealized absorption experiment, particu larly when concerned with gases other than those of the simple rare gas atoms, the gas should contain only those atoms or molecules in the state from which the tran sition is to be observed. This is usually the ground state in which molecular or stable atomic gases are found to be densely populated at room temperature and below. Thus, in a photoionization experiment involving atoms, the cross section is measured which involves an electron transition from the ground state of the atom to one of the continuum states above what is called the ionization * $ } . ? & ! : ■ '-■ ■ i f f':-f:y" : ^ v ' 15 •H •rl 600 800 10 Figure 5. Absorption cross section of atomic oxygen. Theoretical curves A and B represent the dipole velocity and dipole length formula tions, respectively. Points represent . experimental results of Cairns and Samson (G.C.A. Tech. Rept. No. 67-2-N (1967) ). 52 limit. That is to say, the observation is that of the absorption of a photon by a ground state atom producing by the process a singly ionized atom in its ground state and an electron with kinetic energy equal to the differ ence between the photon energy and the ionization potential of the atom, A (ground state) + hv - * ■ A+ (ground state) + e" (KE = hv - E + AE) (3.3) However, if the gas is electrically excited the popula tion of atoms in the excited states increases. The changes in density will follow Boltzmann's law (Eq. 2.3) only when thermal equilibrium is maintained. Starting with a molecular gas and exciting it in order to dissociate the molecules for photoabsorption- studies, the ideal experimental conditions are never fully realized. Nevertheless, the conditions are inter esting and may come near to the ideal for the case of an equilibrium plasma. The discussion of oxygen may serve as an example. Molecular •T The ground state of is Ig* From this state the molecule may dissociate at 5.12 eV to form two 3 oxygen atoms in the ground state P. As the temperature 3 of an 0 , ( l ) gas is raised, the relative total density « s 53 of 0 to O2 increases according to the Saha relations. But there are low lying metastable molecular states, the b state at 1.63 eV and the a ^Ag state which lies closer to the ground state at 0.98 eV. The population i density of these states will increase rapidly as the temperature is raised and then proceed to decrease with the total density of C^, as the gas becomes dissociated. At some temperature, therefore, the gas may be expected 2 — * 1 «|» to contain large concentrations of , O2 ( I ), 1 3 O2 ( Ag), and 0 ( P), with densities near the same order of magnitude. Absorption from a photon beam at some energy then may produce 0 ^ derived from three distinct states as well as 0 + from its ground state, which is the desired transition. Then the temperature should be increased in order to deplete the populations. With the C >2 population reduced to a level negli gible compared with that of 0 , the temperature may be sufficiently high for the plasma to contain large concen trations in the and metastable states of 0 , and these should be given consideration. Atomic 2 4 Atomic oxygen has the configuration 2s 2 p 3 1 1 from which the three states P, D, and S are derived. In addition, the oxygen ion 0 has three low lying energy states, ^S, ^D, and ^P, In Table III is listed the 54 TABLE III 01 SERIES Oil Level Term Value from 01 3P State, cm" 1 Convergence Limit, A 01 Series 2s2 2p3 C4S°3/2) 109837.03 910.440 .(4 S°)ns 3S° (4S°)nd 3d° 2s2 2p3 ^2°5/2^ 136645.4 731.821 (2D°)ns1 3d° 2s2 2p3 <2d3/2> 136666.4 731.709 (2D°)nd» 3S° (ZD°)nd» 3po (2D°)nd’ 3d° 2s2 2p3 (2p3/2) 150303.9 . 665.319 (2P°)nsM 3pO 2s2 2 P 3 (2pi/2^ 150305.4 665.312 (2P°)nd" 3pO ■ (2P°)ndM 3d° 2s 2p4 - 2s 2p5 3P° ss 3 energy of these states above the P ground state of 0 and also the different 0 1 series leading to these states. This table was compiled by Huffman, et al., who made a detailed analysis of the series absorption spectrum they observed. At a temperature of 1 eV, according to Eq. 2.3, the population of the state is about one-twentieth 3 1 that of the P and the S state population smaller by a factor of 10 . Therefore, it may be necessary to con- 3 sider not only the P absorption but also the interaction 0 (1D) + hv + 0+ + e. (3.4) The density appears too small to consider, but the photo interaction cannot be dismissed entirely on the basis of population although the cross section may be too small to cause any appreciable effect. No effect was observed in emission. The oxygen emission spectra observed here showed first continuum radiation resulting from the recombination 0+ (4 S) + e - » ■ 0 (3P) + hv . (3.5) This continuum decreased in intensity toward shorter wave lengths to the point where broadened emission lines of 2 the 01 series leading to the D states of Oil began to 7 R, E. Huffman, V. C. Larrabee, and Y, Tanaka, J. Chem. Phys. 46, 2213 (1967). 56 overlap. Beyond this limit continuous radiation resulting from the recombination 0+ (2D) + e 0 (3P) + hv' (3.6) could be observed. At this threshold, an increase in the continuum level was definite, and the intensity again decreased toward shorter wavelengths. No clear jump in the continuum level at any other wavelength could be ob served. The smooth decrease continued to and beyond the 2 ° P limit at 665.3 A, where a second increase might be expected due to recombination from this state. Thus, on the basis of emission it could be assumed that only ab- 3 4 2 sorption from the P state to form ions in the S and D states was significant. i I o o ; Between the limits at 910 A and 731 A, several 01 i emission lines were observed. These lines were broadened j and more intense than expected. The autoionizing 4s1 j 2 4 3 I and 5s* states from the series 2s 2p ? 2 ^ g ++ I 2s22p3 (2d5/2 3/2^ ns' 3d3 2 1 s^owe<^ characteristics ; normally attributed to such states but had the appear ance more of a broad emission line adjacent to an ab sorption line. Members of the other autoionizing series appeared too weak and diffuse to be clearly observed. It should be noted that these states have never before been observed and identified in emission. Lines result- j i i ing from transitions from the state leading to the . 57 2 ° D limit at 665 A appeared in this region and below the O 731 A limit but also were weak and diffuse. Line radiation from 0+ was observed also. These transitions are given in Table IV. Radiation from the spark source operating in the neon atmosphere resulted from transitions of oxygen and nitrogen in higher stages of ionization. Such oxygen lines were easily distinguished from the 01 and Oil arc emission. 58 . TABLE IV Oil TRANSITIONS o Wavelength, A Transition Comments 834.462 2p4 r4p C 5/2 - > ■ 2p3 f 4 q ° c 3/2' . 833.326 2P4 r4p 3/2 ->* 2p3 (4S° 1 C b3/2J 832.754 2p4 (4pi/2 2p3 r4s° ^ 1 b3/2J 796.661 2p4 (2l>5/2 2p3 r 2 p o 1 F3/2J 718.562 2p4 ( 2 d 3 / 2 2p3 L 3/2 . 718.484 2p. 4 ( 2 d 5 / 2 4. 2p3 r2D° ^ C 5/2' 644.148 2p4 f 2 s C 1/2 -V 2p3 (2 p0 1 C P1/2J 673.768. 3s f 2 p C 1/2 ->■ 2p3 r2p° 1 L P3/2J not observed 672.948 3s r2p 1 3/2 2p3 r2p° 1 1 1/2J not observed 617.051 ’ 3s r 2 p C 1/2 -*■ 2p3 f 2 D ° 1 C D3/2j 616.343 3s r2p 1 3/2 - * ■ 2p3 c 2 d ° ") C D3/2J 616.291 3s (2p3/2 2p3 r2 d 0 1 C U5/2j 600.585 3s' (2d3/2 * 4 2p3 r 2 p o \ C 1/2' not observed IV. CONCLUSIONS This study has been in the nature of a survey to illustrate the feasibility.„ • for observation of plasma absorption by the technique suggested here. Since definite transmission through and attenuation of the radiation probe by the arc plasma was accomplished, the experiment indicated that accurate photoabsorption cross sections can be obtained and it also pointed out direc tions which future work of this type may take. First, while data concerning gases such as atomic oxygen and nitrpgen are of current interest, these gases exhibit a complicated pattern of photon absorption in the ultraviolet, and detailed analysis must be made in ex periments of this type, which will serve both the work concerning stellar atmospheres and studies of the upper atmosphere. Of equal, if not greater importance is the measurement of photoabsorption by ions. For this, the atoms and ions exhibiting the simplest structure, the rare gases, should be studied first. The simplicity of the ion absorption technique was illustrated by Blackwell et al^ in the previously 59 60 mentioned shock tube experiment on Xe+. Here Xe+ could not be studied using neon as a carrier gas since its photo- O ionization threshold lies at 585 A, very near to that of O the high density neutral neon, 574 A. There are other types of measurements which can be performed using the plasma of a stabilized arc in the vacuum ultraviolet region of the spectrum. Ion emission, such as that noted in the case of argon, can be studied in the neighborhood of the arc electrodes to clarify questions relating to inhomogeneity and deviation from i l thermal equilibrium in this arc region. Visible emission can be observed with the hope of establishing a well defined temperature. At the same time, variations in the electrode configuration can be made utilizing techniques developed by Maecker* in order to remove the presumably inhomogeneous region from the optic axis and hence eliminate it from observation, in order to obtain com parison spectra in the vacuum ultraviolet region. Having thus a complete understanding of emission from all regions of the arc, an investigation can be made of ion emission for the purpose of measuring oscillator strengths for these ultraviolet transitions. Some techniques for determining oscillator strengths in the vacuum ultraviolet region have been | *H. Maecker, private communication. 61 2 reported. These range from measurements by Prag, of NI O and 01 multiplets in the neighborhood of 1200 A by after 3 4 glow resonance absorption to those of Bashkin and Kay, who created excited Lithium-like ions by passing an ion ) jbeam through a think foil observing the decaying resonance | radiation as a function of distance from the foil. The techniques of primary interest here are those of Boldt,^ using a stabilized arc, and Lincke,^ who measured the O oscillator strength of the Hel line at 584 A using a shock tube. In total, vacuum ultraviolet measurements of oscillator strength are not only few in number but in most cases are unsatisfactory in terms of accuracy. I f ^A. B Prag, C. E. Fairchild, and K. C. Clark, Phys. Rev. 137, A1358 (1965). 3 S. Bashkin, L. Heroux, and J. Shaw, Phys. Letters 1_3, 229 (1964). 4L. Kay, Proc. Phys. Soc. (London) 85, 163 (1965). 5G. Boldt, Z. Phys. 154, 319 (1959). 6R. Lincke and H. R. Griem, Phys. Rev. 143, 66 (1966). --- BIBLIOGRAPHY BIBLIOGRAPHY Armstrong, B. H. J. Quant. Spectry. Radiative Transfer 4, 207 (1964). | Bashkin, S., Heroux, L., and Shaw, J. Phys. Letters 13, 229 (1964). Belousora, I. M., and Gurevich, D. B. Opt. Spectry. (USSR) 10, 206 (1961). Blackwell, H. E., Bajwa, G. S., Shipp, G. S., and Weissler, G. L. J. Quant. Spectry. Radiative Trans fer 4, 249 (1964). Blackwell, H. E., Shipp, G. S., Ogawa, M., and Weissler, G. L. J. Opt. Soc. Am. 56, 665 (1966). Boldt, G. Z. Naturforschung 18a, 1170 (1963). Boldt, G. Z. Physik 154, 319 (1959). Boldt, G. Z. Physik 154, 330 (1959). Boldt, G., and Cooper, W. S. Z. Naturforschung l j ) , 968 (1964) Brunner, J. Z. Physik 159, 288 (1960). Cairns, R. B., and Samson, J. A. R. GCA Tech. Rept. No. 67-2-N (1967). Cairns, R. B., and Samson, J. A. R. J. Opt. Soc. Am. 56, 769 (1966). Camae, M., and Vaughn, A. Bull. Amer. Phys. Soc. Ser II 4, 291 (1959). Ecker, G., and Weigel, W. Ann. Physik 17, 126 (1956). Ehler, A. W., and Weissler, G. L. J. Opt. Soc. Am. 45, 1035 (1955). Elenbaas, W. Physica £, 279 (1937). 63 64 Elton, R. C., and Grieih, H. R. Phys. Rev. 135, A1550 (1964). ! Finkelnburg, W., and Maecker, H. Handbuch der Physik. j | Berlin, Springer-Verlag, 1956; XXII Gasentladungen. j j Freeman, M. P., and Katz, S. J. Opt. Soc. Am. 5 j 0 , 826 (1960) . | Gericke, W. E. Z. Astrophys. J kS, 68 (1961). j i Griem, H. R. Plasma Spectroscopy. New York: McGraw-Hill | Book Co., Inc., 1964. ! Griem, H. R., Kolb, A. C., and Shen, K. Y. Astrophys. J. ! 135, 272 (1962). j Henning, H. Z. Physik 169, 467 (1962). j Hey, P. Z. Physik 151, 79 (1959). j i 1 I Huffman, R. E., Larrabee, J. C., and Tanaka, Y. J. Chem. ; i Phys. 46, 2213 (1967). i ; i Inglis, D. R., and Teller, E. Astrophys. J. 9 ( ) , 439 | (1939). I Jackson, J. D. Classical Electrodynamics. New York: j John Wiley and Sons, Inc., 1962. i i | Kay, L. Proc. Phys. Soc. (London) 85, 163 (1965). | i ; Kelley, R. L. "Vacuum Ultraviolet Emission Lines," UCRL Publication 5612. Kivel, B. J. Quant. Spectry. Radiative Transfer 509 (1962). Kolb, A. C., and Griem, H. R. Phys. Rev. Ill, 514 (1958).j Lincke, P., and Griem, H. R. Phys. Rev. 143, 66 (1966). | ! Maecker, H. Private communication, 1966. J Maecker, H. Z. Naturforschung, 11a, 457 (1956). j Mastrup, F., and Wiese, W. Z. Astrophys. 44, 259 (1958). j i Moore, C, E. Atomic Energy Levels (N.B.S. Washington, j 1949) Circular 467, Vol. 1. j 65 Olsen, H. N. J. Quant. Spectry. Radiative Transfer 3, 303 (1963). Olsen, H. N. Phys: Rev. 124, 1703 (1961). Olsen, H. N. Private communication, 1967. Prag, A. B., Fairchild, C. E., and Clark, K. C. Phys. j . Rev. 137, A1358 (1965). Richter, J. Z. Astrophys. 51, 177 (1961). Richter, J. Z. Physik 151, 114 (1958). Roder, 0. Z. Astrophys. 55, 38 (1962). i Saha, M. Phil. Mag. 40, 472 (1920). j j Salpeter, E. E., and Zaidi, M. H. Phys. Rev. 125, 248 j (1962). Unsold, A. Z. Astrophys. 24, 355 (1948). Wainfan, N., Walker, W. C., and Weissler, G. L. J. Appl. i Phys. 14, 1318 (1953). Wobig, K. H. Z. Astrophys. 55, 100 (1962).

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Optical And Photoelectric Properties Of Gold And Aluminum In The Extreme Ultraviolet

PDF

The Optical And Photoelectric Properties Of Metals In The Vacuum Ultraviolet Region Of The Spectrum

PDF

Extreme Ultraviolet Photoabsorption Processes In Molecular Oxygen

PDF

Mass Spectroscopic Determination Of Different Ions Produced By The Process Of Photoionization

PDF

Absorption-Spectrum Of Oxygen In The Even Singlet 0A-Delta Metastable Stable In The Region From 1090 To 1700 Angstroms

PDF

Vacuum-Ultraviolet Plasma Spectroscopy On A Double-Arc With Measurements On Line-Strengths And Lineshapes; And On The Lowering Of Ionization Potentials

PDF

A Study Of Low Lying Excited Levels Of Palladium-105

PDF

A Theoretical And Experimental Analysis Of An Eighteen-Stage Radio-Frequency Mass Spectrometer

PDF

Interaction Of 31 Mev Protons With Deuterium

PDF

Quasi-Elastic Proton-Deuteron Collisions In Lithium-6 At 30 Mev

PDF

Energy Loss Of Protons With Energies Up To 30 Mev In Different Materials

PDF

Rotational-Analysis Of High-Resolution Oxygen Molecule Absorption-Bands In The Region From 1085 To 1165 Angstroms

PDF

Absorption, Ionization, And Ion Fragmentation Cross-Sections Of Hydrocarbon Vapors For Vacuum Ultraviolet Radiation

PDF

Fluorescence Spectra Of The Molecular Mono Positive Ions Of Oxygen, Nitrogen And Carbon-Monoxide Excited By Vacuum Ultraviolet Radiation

PDF

Photoionization Induced Ion-Molecule Reactions In A Mass Spectrometer

PDF

Polynomial Operators On Topological And Ordered Topological Vector Spaces

PDF

The Faraday Rotation And Some Magnetic Properties Of Evaporated Thin Films Of Iron, Nickel, Cobalt, And Gadolinium

PDF

Fixed Point Theorems For Set Valued Contractions

PDF

Studies Of Rare-Earth Hydrogen Systems

PDF

Elastic And Inelastic Scattering Of 28 Mev Protons From Iron, Cobalt, Nickel And Zinc

Asset Metadata

Creator Blackwell, Harvel Eugene (author)

Core Title Vacuum Ultraviolet-Radiation Studies For Photoabsorption By Moderate-Temperature Plasmas

Contributor Digitized by ProQuest (provenance)

Degree Doctor of Philosophy

Degree Program physics

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag OAI-PMH Harvest

Format dissertations (aat)

Language English

Advisor Weissler, Gerhard L. (committee chair), Hyers, Donald Holmes (committee member), Ogawa, Masaru (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c18-626717

Unique identifier UC11361010

Identifier 6817015.pdf (filename),usctheses-c18-626717 (legacy record id)

Legacy Identifier 6817015.pdf

Dmrecord 626717

Document Type Dissertation

Format dissertations (aat)

Rights Blackwell, Harvel Eugene

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA