UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The spectra of indium Archer, Douglas Harley 1948

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-UBC_1948_A8 A7 S6.pdf [ 4.01MB ]
JSON: 831-1.0103785.json
JSON-LD: 831-1.0103785-ld.json
RDF/XML (Pretty): 831-1.0103785-rdf.xml
RDF/JSON: 831-1.0103785-rdf.json
Turtle: 831-1.0103785-turtle.txt
N-Triples: 831-1.0103785-rdf-ntriples.txt
Original Record: 831-1.0103785-source.json
Full Text

Full Text

T H E S P E C T R A O F I N D I U M by DOUGLAS HARLEY ARCHER A Thesis Submitted i n Partial Fulfilment of The Requirements for the Degree of M A S T E R O F A R T S in the Department OF THE UNIVERSITY OF BRITISH COLUMBIA Ap r i l , 1948, P H Y S I C S ACKNOWLEDGMENTS. The writer wishes especially to express his gratitude to Dr. A . M. Crooker, of the University of British Columbia, for his constant help and advice i n conducting the present research. He is also indebted to the National Research Council of Canada, under whose aid the work has been carried out. TABLE OF CONTENTS PA aw I INTRODUCTION . (1) II THEORY (4) III EXPERIMENTAL (25) IV RESULTS (36) V BIBLIOGRAPHY (42) PLATES FACING- PAGE I Arrangement of Apparatus (27) II Circuit Arrangement for Souroe (28) III Spectrum Pictures of In (29) IV Dispersion Curve for Hilger "El (30) V Moseley Diagram (35) THE SPECTRA OF INDIUM I. INTRODUCTION. 1. Object of Research. From the early days of spectrum analysis u n t i l the present time, the importance of data conoerning spectra has been universally recognized as one of the most valuable sources for the understanding of the structure of matter. It was f i r s t observed that the intensities and wavelengths of spectral lines are characteristic of the emitting atoms or molecules. Subsequent investigation of atomic spectra has had a decisive influence on the development of atomic structure, and has shown above a l l that only certain dis-crete energy states are possible for an atom. This i n -vestigation has further given us information about the arrange-ment and motion of electrons i n an atom, and has led to the discovery of electron spin and to a theoretical understanding of the periodic system of the elements. I f , however, the study of spectra were to culminate with these .discoveries concerning the outer part of the atom, the justification for further fundamental research would be somewhat i n doubt. Fortunately, detailed observations with high quality spectral apparatus have brought to light the existence of hyperfine structure i n some spectral lines. The theoretical interpretation of this leads directly to the concept of a (2) nuclear spin and magnetic moment and i n oertain cases to the existence of an electric quadrupole moment i n the nucleus. With our present day knowledge of the nucleus far from complete, additional information of this nature is certainly most welcome and the justification for further spectrum analysis, chiefly with this object i n view, i s immediately evident• As w i l l he seen from the theory below, however, the interpretation of data pertinent to nuclear properties, depends to a large extent on the available information con-cerning the gross structure of the spectrum concerned. It i s therefore the purpose of the present re-search to study the gross structure i n the spectra of Indium and to thereby provide certain necessary foundations for future hyperfine structure measurements on the spectral lines of this element. 2. Review of Previous Work. (a) The Spectra of Indium. The aro spectrum of indium has been f a i r l y ex-tensively studied by numerous investigators since 1920 when de Gramont (11) published a paper on fUltimate lines'. Uhler and Tanch (31), Frayne (9), and more recently Pasohen (21) have given wavelengths and classification of emission lines i n the region 9200A — 2100A. (3) - The spark spectrum in the ultra violet has been especially studied by Saunders (27) and Weinberg (32) who succeeded, by means of a vacuum grating, in reaching wave-lengths as low as 161 angstroms. Later work by Rao (26), Lang and Sawyer (19), Green and Lang (12) and others, and f i n a l l y Paschen and Campbell (22), has extended the wave-length results i n a l l regions. Theoretical investigations into series relations, doublet and triplet separations, screening constants, and other fine structure relations for indium have been carried out chiefly by Lande (17), Sigiura (30), Hartree (14), Badami (1), Lang (18) and Rao (26). (b) Hyperfine Structure of Indium. Considerable work has been done on the hyperfine structure of In I and II with the object of determining the spin and nuclear magnetio moment, and i n investigating the electric quadrupole moment. The nuclear spin of In"5 has been determined by Jackson (16) and by Paschen and Campbell (23). Though no . lines were found where the large spin could be determined directly from the number of components, the aocurate measurements of the separations of 5s 6s aS, which are not expected to show any quadrupole effect indicate quite definitely that for In"* , the spin i s 9/2. (4) The nuclear moment of In by the method of hyper-fine structure has been investigated by Jackson (16), Schuler and Sohmidt (28), MoLennan, A l l i n and Hall (20) and others. The magnetio moment was determined from the' separations observed i n In I, a(6S) • 0.056 cnu and a(6PyJ - 0.76 cm;' , and deduced from In II , a(55) a 0.70 cm"'. These give values of 5.2, 5.9 and 6.0 respectively for the magnetic moment. An electric quadrupole moment was found to exist i n indium, and has been studied by Schuler and Sohmidt (28), Backer and Tomboulin (2), and by Hamilton (13). Mention should also be made here of a very complete work on the hyperfine structure of the lines of In II by Paschen and Campbell (22) and i n which the spectral terms are recorded and discussed. . I I . THEORY 1. Gross Structure of Spectral Lines. (a) Qualitative discussion of Fine Struoture. The f i r s t intensive investigations on line spectra revealed that i n every case a given spectrum consisted of several series of lines converging toward short-wave-length limit s . Success i n finding an analytical expression for (5) such a series was f i r s t met by Balmer i n his investigations on Hydrogen, and later a more general formula applicable to other series of Hydrogen or H-like ions, was put forward by Rydberg. These empirical results, whioh are now completely explained by wave mechanics, showed without a doubt the existence of stationary states i n an atom between which radiative transitions take place giving rise to spectral lines. To a f i r s t approximation i t may be stated that an energy level of an atom i s completely described by the resultant orbital angular momentum of i t s extra nuclear electrons. On a quantum mechanical basis i t can then be shown that this angular momentum i s given by ~t s x/L(L+l) "fc where L is a positive integer or zero, and that transitions take place from a given L lend to another L level i n such a way that AL a 0 or tl with L s 0-»L • 0 forbidden. In general, i f a spectral line arising from suoh a transition i s more closely examined i t is found to consist of two or more oomponents. This s p l i t t i n g , known as fine structure of spectral lines, can be explained by intro-ducing the concept of electron spin. If we assume each electron i n the atom to possess a spin angular momentum as well as an orbital angular (6 ) momentum, then the resultant spin angular momentum given by with S an integer, half integer or zero, coupled with the orbital momentum gives a resultant ~S* + L - " j * a /j(J+l)"h and represents the total angular momentum of the electrons. According to quantum mechanics, ~i l and L can be so combined that the quantum number J taken on the values L+S, L+S-l, , | L-S I that i s 2L + 1 values i f L*-S or ES+1 values i f S^L. Because of the fact that a magnetic moment i s associated with "s* and with~L, a coupling between"!, andlt w i l l now exist. This means that since each state having different J value arises from a different orientation of ~L* and-~S, each w i l l correspond to a different energy level of the atom. Diagramstically, let L, and L x represent the two states between which a transition takes place when spin i s neglected. If spin i s then Included, each level w i l l s p l i t i n some way as indicated by the levels J+l, J, J<-1, eto» (7) The transitions thus actually take place not between the two levels L, and Lx but between the fine structure levels characterized by quantum numbers J. The resulting spectral line s p l i t t i n g i s then explained i f we assume the selection rule AJ" = 0 , ± 1 to hold, and in addition J s 0-*J • 0 to be forbidden. (b) Some Quantitative Results valence electron, the spin quantum number S a £ and therefore j s £. The result i n thi s case i s that a state with 1 > 0 w i l l be associated with doublet levels. Calculations of the energy of interaction due to the spin-orbit effect, which represents this doublet splitting, have been made on a quantum mechanical basis by Pauli (24), Dirao (7), Gordon (10), and others and by Goudsmit, using a semi-classical vector model. The result obtained by both methods for this spin-orbit interaction energy i s where T= / M+Dt J / = JjfiFT)t ; 3 =>/s(S+l)T ; R the Rydberg constant; and «c the fine structure constant shift that a level with given value would experience upon In the case where the atom has only a single given bye)££. The expression for M^j represents the ( 8 ) introducing an electron spin. It i s often written in short as A % s = a SM = aJ.S = aisaKJ.(i,s) where a = R p ^ Z * and i s the i n t e r a c t i o n constant f o r t h i s s i n g l e e l e c t r o n . I t i s to he noted that j = S whenever^ o 0, and the r e s u l t i n g l e v e l i s s i n g l e . T h i s i s i n agreement wi t h o b s e r v a t i o n t h a t a l l S s t a t e s ( i . e . s t a t e s w i t h L = 0) are s i n g l e . For any g i v e n doublet, J > and S have the same val u e s whereas j r £ f o r the upper l e v e l and -\ f o r the lower l e v e l . The s u c c e s s i v e s u b s t i t u t i o n o f these v a l u e s f o r j i n the f i r s t e x p r e s s i o n f o r A W f j 5 g i v e s by s u b t r a c t i o n A* - B«»Z* . . . O i ) n ' JU+l) or f o r s h o r t &2J = a ( i 4 ) The value o f Z appearing i s not n e c e s s a r i l y the charge on the nucleus hut r a t h e r i s the average charge' observed by the valence e l e c t r o n . For atoms w i t h more than one valence e l e c t r o n , t here are s e v e r a l methods of c o u p l i n g the s p i n and o r b i t a l a ngular momenta t o g e t h e r i n order to o b t a i n e x p r e s s i o n s f o r (9) the fine structure level differences. In the Russell-Saunders, or LS type, i t is assumed that the orbital momenta, and similarly the spin momenta, of the individual electrons are respectively coupled strongly together. The resulting L and S are then considered to interact more weakly. In the j-j coupling, i t i s assumed that the spin angular momentum of each individual electron is strongly coupled to give a resultant j for each electron. The re-sultant j's are then coupled more weakly to give a grand resultant J. A large percentage of a l l known spectrum lines and energy levels may be classified with one of these coupling schemes, or with some intermediate but closely associated scheme. Assuming LS coupling to hold, i t i s again found that the shift of each fine structure level from the centre of gravity can be written AWLjS = A («f* - L1 - §M 2 where A i s a constant for given L and S. This expression leads directly to the important Lande interval rule, by taking the difference between one term J and the adjacent term J+-1: W = A / 2 { (J-rl)(J+2) - J(J+l)j = A(J+1) This means that the difference between two adjacent fine (10) structure levels should he proportional to the larger of the two J values. The faot that the interval rule may hold i n a given spectrum identifies the electron coupling as Russell-Saunders, and further proves a valuable aid to the c l a s s i f i -cation of the speotrum lines observed. As an example of th i s interval rule, the theoreti-cal and observed separations for several t r i p l e t states of In II have been obtained and are shown i n the following ohart: Configuration 3 ° , - \-\ Observed Ratio J-KL / J Theoretical Ratio J+l / J 5s 5d 87 130 1.5 1.5 5s 6d 35 52 1.49 1.5 5s 7d 18 29 1.6 1.5 (o) Iso-eleotronio Sequences and Moseley Diagrams. The term iso-electronic sequence refers to a sequence of atoms having the same number of extra nuolear electrons. Because each atom i n such a sequence contains the same nymber of extra-nuclear electrons, the energy levels and spectrum lines arising from each atom w i l l show remarkable similarities from element to element. For atoms with.a single valence electron term values may be represented either i n the form (11) T = R(Z-p)* where n* = (n^u) i s the effective principal quantum, quantum number, and p the number of core electrons, or i n the form T s R(z-<r)*-n where C T i s a screening constant and n. the true principal quantum number of the emission electron. The differences between the term values of a l l atoms as given by the f i r s t formula and those of H-like atoms as given by T - RZ*" n are accounted for by the quantum d e f e c t w h e r e a s i n the second formula by the screening constant CT". Transposing this second formula and talcing square roots: F T r 1 (Z-cr) v R n If now 07 and cr^are screening constants for the two levels having the same principal quantum number n, the difference i s written A nT - CTj - cri_ z ACT = constant. *R T ~ n~ This i s known as the irregular doublet or screening doublet law and can be stated: the difference between the square roots of the term values of the levels having the same principal quantum number n i s independent of the atomic number Z. If Ff i s then plotted against Z, or (Z-p), (12) those levels having the same value for n should l i e parallel to one another. The following experimentally-obtained Moseley diagram of the iso-electronic sequence KI, Ca II, etc. brings out this fact, and also the fact that i n agreement with the above, the slopes of the curves are nearly equal to l/n. Since the doublet fine structure separations are small, they are neglected i n the drawing. J»*D tf5-WTO so* K I Moseley Diagram of Iso-electronio Sequence K I, Ga II, . As a consequence of the parallelism of curves on a Moseley diagram, the irregular doublet law has a more useful form. Sinoe T = R(Z-oj n the difference between two terms having the same n w i l l be (13) A T % R^Z-on2-- (Z-<5)XJ = 2R (C7I- (71) j Z - CT? -<-C?I ( ^ I 2 J = c ( ( z - c j and i t i s thus seen that AT is a linear function of the atomic number Z. It is of importance since, when T (- T t i s known for two members of a series of ions, the value T, - T t oan be calculated for other members of the series. It therefore furnishes a most reliable and accurate means for predicting the frequencies and wavelengths of unidenti-fied spectrum lines i n multiply ionized atoms. It i s the spectra of such atoms, as w i l l be seen below, that prove especially useful i n providing certain information con-cerning the atomic nucleus. 2. Hyperfine Structure of Spectral Lines. (a) Qualitative Discussion It i s well known that the ordinary fine structure due to spin-orbit interaction, does not exhaust the possibilities for fine details i n spectral lines. When individual components of a spectral multiplet are examined with apparatus of high resolving power, i t i s found that i n many atomic spectra, each of these components is s t i l l further s p l i t into a number of components lying extremely close together. This s p l i t t i n g is called hyperfine structure, (14) It was seen i n the f i r s t part of the theory that the assumption of orbital and spin angular momenta of the individual electrons of an atom explain completely the fine structure multiplet structure. As i t is d i f f i c u l t to imagine an additional degree of freedom of the extra nuclear electrons of an atom which would account for the s t i l l further s p l i t t i n g , we are led to assume that this hyper-fine structure is caused by properties of the atomic nucleus. The influence of the nucleus may be due either to i t s mass (isotope effect) or to the spin of the nucleus. We shall considerthis latter effect only. It was found previously, that i n the case of ordinary fine structure, a single spectral line component arises due to a transition of the atom from one energy level characterized by quantum number J, to a lower level characterized by quantum number J . " - F - l In order to qualitatively explain the faot that this spectral line may not be single, i t i s necessary to assume that one or both of the levels J, . J of the atom (15) are s p l i t into a number of components so that transitions may take place from levels of one J level to levels of the other J level and so give rise to the observed hyperfine spectral multiplet. To quantitatively explain such hyperfine structures, i t is neoessary to assume that the nucleus possesses an i n -tr i n s i c angular momentum with which is associated a magnetic moment ja, and the spli t t i n g occurring due to the action of the eleotrons in the f i e l d of this magnetic dipole. With this assumption, -> I and J are now combined to give a resultant -> . . F = v/FlF+l) Ti which represents the total angular momentum of the atom, including nuclear spin. In the same way i n w h i c h a n d " ! combined to give " j , I^ and ~f combine so that ~F* has suoh values that F - I-fJ, I-fJ-l , | I-J | . That i s 2 1+1 values for F i f I<J or 2J-H i f J<il. Because of the magnetic moment of the nucleus, a coupling between I and J results, and gives rise to energy differences between states with different F values. These are just the small energy differences which are needed in order to account for the hyperfine structure of spectral lines. -The selection rule derived from wave mechanics i s that only those transitions take plaoe for which AF : 0 o r t l and further F = 0-/-=>F - 0. (16) It i s now completely certain that the hyperfine structure of spectral lines i s for the greatest part due to the interaction of a nuclear magnetic moment with the electrons not appearing in closed shells. That the inter-action is mainly a magnetic one, follows from the fact that hyperfine structure groups show a striking resemblance to ordinary fine structure multiplets. Since such i s the case, i t can be shown that the energy of interaction between the nuclear moment and 7 may be written and gives the energy which must be added to the J" level i n order to represent the energy of the level with given F quantum number. Substituting f i r s t F+l for F, and then F i t s e l f i n the last equation, and taking the difference, gives for the energy difference between two adjacent hyper-fine levels: W I j 3 = A1 (I.J) - A ' (F*- 14 - ?* ) 2 F+l AW = A (F+l) H f A(F+1) It i s therefore seen that 1 F the same interval rule should apply F- l to the separation of adjacent hyper fine levels as was found to apply for fine structure separations under the assumption of L S coupling. (17) (b) Application of h-fffio Determination of Nuclear Properties Because the hyperfine structure of spectral lines is so closely oonnected with certain properties of the nucleus, i t is to be expected that examination of the hyper-fine structures would reveal, at least indirectly, information concerning such properties. This i s indeed the case, and as w i l l be seen, the interval rule, or deviations from this rule, play a leading role in this connection. Two cases shall be separately considered here; f i r s t , the case where the interval rule holds, and secondly, the case where i t does not hold. In each case, information relating to certain nuclear properties w i l l be discussed. Interval rule obeyed - In this case, the following may be determined: (1) Nuclear Spin n I M It has been pointed out above that the number of hyperfine states into which a given J level i s s p l i t by the presence of a nuclear moment is 21+1 for I*J or 2J+1 for J<E. It i s apparent that the complete analysis of the hyperfine structure for any state which is s p l i t into a number of components less than 2J-M, i s a conclusive determination of I. I t i s frequently the case i n simpler spectra of the one electron type, that no state of ( 1 8 ) sufficiently large J value can be found whioh has appreciable hyperfine structure. In such a case, the interval rule can be used to determine the F values for the hyperfine states and thus the value of the nuclear spin I i f J is known for the states. Such a determination by the use of the interval rule relies ab-solutely on i t s validity and therefore upon the cosine law of interaction. ( 2 ) Nuclear Magnetic Moment This may be written i n the form: where >^(D is the nuclear ^ -factor and M the proton mass. It i s then found that the i n -terval factor A, contains the nuclear magnetic moment ju, and factors which relate to the electrons and the probability of their being near the nucleus. The constant jx, i s related to the absolute size of the hyperfine structure separations. In order to determine jx, i t i s necessary to make a determination of the other factors i n A and then to use the experimental size to find^a. For a purely tiBoretioal determination of JJL, using the experimental size of A, a knowledge of the wave functions near the nucleus (19) i s essential. As such wave functions are i n general not available, i t i s the procedure to use approximate methods and so obtain semi-empirical expressions for JX. • Fermi and Segre (8) and later Breit (4) and Racah (25) have shown that when the state considered arises from an atom with a single S electron 0 Ro^Zi (l+Z)* K(j,Zj) where K(j,Z t) » 4j(j+1) (j+fc) / (4/>*-l)/> i s a r e l a t i v i s t i c correction factor with /> = [(M)* - z V ] S Z-and Z measure the effective nuclear charge i n the inner and outer regions, and n* i s the effective quan-tum number. Further, i f the electron is not of the S variety, then whereAV i s the doublet separation, and i s a correction factor for the doublet separation. Using these expressions, we obtain values forjn when the atom considered has only a single valence electron, I f the atom considered has more than one valence electron, the above relations cannot be applied directly. For many configurations, a l l of the valence electrons w i l l have a considerable (2G) interaction with the nucleus. It i s possible i n suoh cases to find relations which give the hyper-fine structure size i n terms of the separation oonstants of the various electrons involved. For example, for two S electrons, the interval factor is known to be half the sum of the i n -dividual interaction constants More generally, Breit and.Wills ( 5 ) have given formulas for several l electron configurations of the type Sp, Sd, Sf, and the interval factors of terms of suoh configurations which are i n -dependent of coupling are of the non-S electron with j a J+£ and J( -£ respectively. It should be remembered that the c a l -culation of jx. depends upon the experimental determination of the interval factor A, and hence upon the s t r i c t obeyance of the Interval rule. Interval rule not obeyed ( 1 ) More accurate measurements show that the hyperfine structure levels do not always follow A Csy) r £ (a.+ a j where a, and a 4 refer to the interaction constant (21) this rule. These deviations have two different causes. In the f i r s t place, they can he a con-sequence of electric Interaction with the nucleus; i n the second place, they can be due to a second order effect i n the magnetic inter-action. Such a second order effect is to be expected whenever the distance of two fine structure levels i s not large compared with the hyperfine structure separations. Consider f i r s t the oase where two fine, structure levels J, and J x say, have the small separation D. By the magnetic interaction with the nucleus both levels are separated into a number of components; and one F-value can occur twice, once for a level belonging to the f i r s t finestrueture level and once for a level be-longing to the second one. The results of per-turbation calculations i n such a case show that i f J, and Jx arise from the same electron con-figuration, and i f |J, - JJ ss 0 or± 1, then levels with equal F are perturbed i n such a way as to be repelled, and the order of magnitude of the repulsion i s given by (h.f.s. splitting) /(separation of non-perturbed levels ) For s t r i c t LS coupling, there i s the additional restriction that the states must have the same (22) resultant L and S . It i s very d i f f i o u l t to calculate suoh perturbations exactly and since they may have a great influence i n the determination of jxt i t i s desirable to avoid using states i n the determination of jx whioh are subject to great change. It i s therefore desirable i n the de-termination of jx, to use states which have large hyperfine structure and are not subject to violent perturbations. From the expressions shown above for the ^ f a c t o r , i t is seen that the more highly ionized an atom i s , the larger w i l l be the interaction constant a for a given state and hence the larger w i l l be the observed hyperfine structures. Measurements on the higher spark spectra of a given element should then prove most valuable for accurate determinations of jx* (2) Electric Quadrupole Moment nQ/» It has been found that deviations from the interval rule exist which cannot be explained by the above perturbation effect. Schuler and Sohmidt (29) observed that the deviations could be satisfactorily accounted for by the presence of an interaction term which i s pro-portional to the square of the cosine of the angle between i* and J. Such an interaction, (23) they pointed out, would be expected i f the nuoleus has an electric quadrupole moment which may arise i f the protons are not distributed over the nucleus on the average with spherical symmetry. Oasimir (6) has shown that the change i n the energy of a hyperfine state F due to the presence of the electric quadrupole moment can be written where K - F(F+l) - 1(1+1) - J(J+1) ; and where B i s related to the quadrupole moment through the relation over the charge density of the electrons and can be calculated for various electron con-figurations. It i s found that in the oase of an atom with a single valence electron with j « £, l\VF vanishes. There would thus be no quad-rupole effect for S and P^ electrons, however, i t i s found that other states show effects which are roughly proportional to the fine structure doublet separation so that large effects are cm. i s an average (24) to be expected for low and d electrons of the heavy elements. By measuring the deviation from the i n -terval rule for a particular state, and c a l -culating the other factors which enter into the relations, values for Q can be found. Such values are extremely useful for determining the shape of the charge distribution over the nucleus, since i t can readily be shown that the distribution i s prolate, oblate, or spherical i n the direction of spin, according to whether Q is>0,<G, or - 0 respectively. It has now been shown that the interval rule may be violated due to two different causes and that one of these causes proves useful for the description of the shape of the nucleus. To make use of this fact, we must therefore examine states which show a deviation from the interval rule because of the interaction with an electric quadrupole moment only. States which also show deviations due to per-turbations from neighboring states must be avoided i f accurate values for Q are to be obtained. It was further seen that in the determination of the magnetio moment, states which are perturbed i n any manner, must also be avoided. Before any hperfine structure measurements can be (25) made, i t i s therefore quite evident that the gross structure of the speotrum of the element under investigation must he thoroughly studied i n order to pick out those states best suitable for accurate determinations of jx. or Q. III. EXPERIMENTAL 1. Apparatus Used. (a) The Spectrograph The dispersing device used was a Hilger E 1 Quartz Spectrograph. The plate dispersion for this i n -strument was found to vary from about 1.6 A/mm at 2100 A to about 48A/mm at 6600 A. (b) The Light Source The type of source desired for this work was f i r s t l y one i n which the excitation of the spectrum could be controlled to some extent, and secondly, one i n whioh pressure broadening of the lines could be reduced to a minimum. To f u l f i l these two conditions, an electrodeless disoharge was chosen. The action of such a source may be described i n quite general terms as follows: i f a closed vessel con-taining the vapor of an element under low pressure be placed inside a c o i l carrying a high-frequenoy current, the c o i l current sets up i n the closed vessel an oscillating high ( 2 6 ) frequency electric f i e l d , which produces enough electron velocity to excite the gas atoms or molecules. The electron velocities depend on the f i e l d strength, the length of time the f i e l d i s applied, and the distance an electron can travel before colliding with a molecule. These factors can be controlled by adjusting strength and frequency of the ourrent and the gas pressure; accordingly, the type of spectrum to be produced varies within considerable limits. That the excitation may be changed by varying the pressure within the vessel is an immediate consequence of the fact that changes i n pressure of a gas change the mean free path of particles moving in the gas (in this case, electrons and ions) and i n particular, i f the pressure i s lowered, an ion which now traverses a longer path w i l l pick up extra energy and thus cause greater excitation. Any gas, or any metal which can be vaporized by the temperatures to which the tube can be subjected, can be used i n the eleetrodeless discharge. The closed vessel, which was constructed by Mr. W. Pye, consisted of a translucent quartz tube about 14 inches long by 1 inch i n diameter. The ends of the t ube were made by fusing f l a t transparent quartz windows into place. A small bore side tube was provided i n order to fac i l i t a t e the entry of the indium metal into the tube, and also to provide an outlet for evacuation purposes. (27) The c o i l consisted of 12 to 14 turns of #10 copper wire, and was wound so as to extend over nearly 7 inches of the tube. In series with this core was connected a variable iron spark gap; across this series combination, four leyden jars i n parallel were connected and these in turn charged up from the secondary of a transformer giving nearly 100,000 volts. In series with the primary, connected to the 110 volt A.C. supply, was added a duty control resistance of from.3 to 12 ohms. As the f i r s t discharge ionizes the air i n the gap, the resistance becomes so small that oscillations are set up and a high frequency current flows i n the ooi l . Diagramatlcally, the experimental arrangement of the source together with circuit is shown i n Plate I I . 2. Method of Procedure (a) Conditioning the Source The indium metal i n the form of small pellets was admitted to the main quartz tube through the small bore side tube and positioned so as to l i e as nearly i n the centre of the c o i l as possible. The sidetube was then connected to a Mega-Vac producing a pressure of onlyootamm. i n a very short time. Because of the fact that the boiling point of indium i s so high (M-SO'C), i t was found that the emitted lines were extremely weak unless sufficient heat was applied (28) to raise the indium vapor pressure. To accomplish this, i t was convenient to place a Meker Burner directly beneath that part of the tube containing the metal and then to connect the burner ele c t r i c a l l y to the high potential i n order to prevent sparking across the gas flame. With a variable pinch clamp over the rubber hose joining the tube to the pump, the pressure within the tube could be varied within rather large limits. By looking at the emitted light from one end of the tube through a small specroscope, variations i n excitation and Intensity were observed due to such pressure variations. Observations were further made by varying the current in the primary circuit from about 10 to 30 amperes. To start the discharge, i t was found necessary to f i r s t disconnect the pump by means of the pinch clamp so that the Indium vapor pressure could build up. When the Intensity then became sufficiently high, the pinch clamp was slowly released i n order to raise the exoitation of the a toms• Several d i f f i c u l t i e s were met with i n the operation of the source. In the f i r s t place, as mentioned above, a very hot flame i s required to vaporize enough of the indium for useful observations. An ordinary bunsen burner was quite inadequate for this purpose. An oxygen flame, on the other hand, although quite satisfactory for raising the vapor pressure, proved too hot for the walls of the PLATE II. Showing Circuit Arrangement for Light Source. (29) quartz tube, which collapsed due to the low pressure inside. The best heating device was found with a Meker burner. Such a flame, however, easily melts the copper c o i l , which of necessity must surround that region of the tube containing the metal. To overcome this d i f f i c u l t y , i t was found that the central part of the c o i l could be spread apart about 2 inches so as to minimize the effect of heat on the c o i l and at the same time without seriously effecting the electric f i e l d within the tube. Considerable trouble further arose during an actual exposure, where i t was found that the quartz windows of the tube became coated with a thin film of indium. This seriously impaired the intensity of the light trans-mitted by the tube and necessitated constant cleaning of the windows by means of a bunsen flame. (b) Plate Exposures With the source properly conditioned and lined up with the s l i t of the spectrograph, exposures were taken i n the regions 6890 A — 3300 A, 3300A -* 2500 A, and 2500 A. 2080 A. In each case an iron arc comparison spectrum was superimposed upon the main indium spectrum by shining an iron arc beam axially through the quartz tube immediately before or after the indium exposure. This method of procedure introduces no shift between the iron and Indium lines. Region 6890 A — 3300 A Region 3300 A — E500 A PLATE III. Showing Indium Spectrum i n the two wavelength regions together with Iron Arc Comparison Spectrum. (30) The exposure times for the indium varied from region to region. In the visible and f i r s t ultra-violet regions, Eastmann II F plates were used and exposure times of approximately one hour, and l£ hours respectively re-quired. In the lower region from 2500 A to 2080 A, however, special Ilford Q, I plates, sensitive i n this region, were used, and exposure times greater than l i hours required. Prints of plates i n the regions 6890 A — 3300 A, and 3300 A — 2500 A are shown facing page (M ). Plates taken i n the lower wavelength region using the Ilfords, however, showed l i t t l e contrast and were therefore not suitable for printing purposes. 3. Determination of Wavelengths The positions of a l l prospective indium lines found on the plates i n the three wavelength regions, to-gether with suitable iron standard lines, were obtained with a Hilger comparator reading to 0.00 1 mm. For two different regions on each plate, the Hartmann formula * 1 A' 4 a . - d was applied to the indium lines and to the iron standard lines used for determining the inherent deviations of this formula from the actual wavelength values. Because of the long exposure times required for the indium, the resulting spectral lines showed i n general, slight broad-ening due to temperature effects on the prism. As this PLATE IV. Showing D i s p e r s i o n i n A/mm. For H i l g e r E 1 Quartz Spectrograph. (31) introduces a slight relative shift of iron and indium lines, i t was therefore important to use only indium lines i n the determination of the three Hartmann constants. For the identification of such lines, and also in some cases for the identification of the iron standards, a dispersion curve for the Hilger E l Spectrograph through-out the entire region 6800 A — 2080 A. was plotted. This i s shown i n the graph of Plate IV. 4. Preparation of Moseley Diagrams  and Iso-electronio Sequences In view of the fact that the higher spark spectrum of indium w i l l i n general show larger hyperfine structures than ~the arc spectrum, i t i s important to investigate the term structure for In III. As seen froiii the theory, an iso-electronio sequence involving In III is of some value for this purpose, and therefore the following term table for the iso-electronic sequence A^I, C d l l , In III, Sn IV, Sb V was prepared. I. ELEMENT CONFIGURATION SYMBOL J T(cm"' ) II . — / T n * Ag I 4d'°('S) 5s *S 61106.5 247.19 1.3408S od II 4d'° 5s *S 136376.6 369.29 1.7941 In III 5s 226133 475.53 2.0899 Sn IV 5s 328671 573.29 2.3113 Sb V 5s * 449300 670.29 2.4710 (32) ELEMENT CONFIGURATION SYMBOI ' T(om"' ) n * Ag I 5p 31554.5 177.63 1.8649 1* 30631.5 175.02 1.8928 cd II 5p i e 92241.3 303.72 2.1815 89758.1 299.60 2.2114 In III 5p *p° 168948 411.05 2.4178 i£ 164606 405.71 2.4495 Sn IV 5p 8 259112 509.03 2.6031 252594 502.59 2.6365 Sb V 5p 2po 367734 606.41 2.7314 ijr 358746 598.95 2.7654 Ag I 6s ls 18550.4 136.20 2.4322 Cd II 6s l s 53386.4 231.05 2.8674 In III 6s *s • * 99255 315.04 3.1544 Sn IV 6s *s 154540 393.10 3.3707 Sb V 6s z s 224713 474.04 3.4941 Ag I 5d *D 12362.5 111.19 2.9793 2& 12342.3 111.11 2.9818 Cd II 5d lD 1* 46685.3 216.07 3.0663 46531.0 215.71 3.0714 In III 5d XD i£ 97675 312.53 3.1799 2£ 97385 312.07 3.1846 Sn IV 5d XD 1* 163374 404.20 3.2783 2£ 163267 404.06 3.2794 Sb V 5d XD 1* 247393 497.38 3.3301 2* 246564 496.55 3.3357 (33) ELEMENT CONFIGURATION SYMBOL J T(cm"' ) n * Ag I 6p *P° * 12809.3 113.14 2.9271 I t .12605.9 112.28 2.9506 Cd I I 6p * 41665.8 204.12 3.2458 I t 40992.5 202.47 3.2723 I n I I I 6p t 81545 285.56 3.4802 I t 80208 283.21 3.5091 Sn I V 6p ' *P° t 130826 361.70 3.6635 I t 128649 358.68 3.6943 Sb V 6p *P° t I t Ag I 7s *S° t 9219.5 96.02 3.4500 Cd I I 7s *S t 29077.1 170.49 3.8854 I n I I I 7s *s t 56706 238.13 4.1733 Sn I V 7s *s i e 91063 301.76 4.3910 Sb V 7s *s * Ag I 4f •*F* 2t 6901.9 83.08 3.9874 3t 6901.9 83.08 3.9874 Cd I I 4f JF' 2t 27955.1 167.19 3.9626 3t 27942.3 167.17 3.9635. I n I I I 4f XF* 2t 64154 253.28 3.9236 3 t 64154 253.28 3.9236 Sn IV 4f *F° 2t 211243 459.61 2.8830 3t 211231 459.59 2.8830 Sb V 4f *F° 2 t 180673 425.06 3.8967 3t 180673 425.06 3.8967 (34) ELEMENT CONFIGURATION SYMBOL J ' T(cm~' ) n* Ag i : 6d 1* 6903.4 83.086 3.9870 2£ 6892.9 83.024 3.9900 Cd II 6d *D I t 26202.1 161.87 4.0930 2 * 26128.6 161.64 4.0987 In III 6d 2D 1* 55602.0 235.80 4.2146 2£ 55420 235.41 4.2215 Sn IV 6d 2D i t 93885 306.40 4.3245 2* 93553 305.86 4.3322 Sb V 6d 1* sir Ag I 5f a* 4394.8 66.293 4.997 3| 4394.8 66.293 4.997 Cd II 5f '2t 3t In III 5f 2 * 40320 200.80 4.9492 3t 40320 200.80 4.9492 Sn IV 5f' 2 * 115411 339.72 3.9004 3t 115246 339.48 3.9032 Sb V 5f 2t According to the irregular doublet law, the difference between the square roots of the term values of the levels having the same principal quantum number n i s i n -dependent of the atomic number Further i t was seen i n the (35) theory that the term differences between two levels having the same n value should be linear functions of the atomic number Z. Using Table I, the Moseley diagram facing page (35 ) was constructed. The parallelism of lines with the same value of n for the emission electron i l l u s t r a t e s quite clearly the irregular doublet law i n i t s f i r s t form. To il l u s t r a t e the law i n i t s second form, Table I was again applied and the following table constructed to show the linear progression of term differences with Z. ELEMENT 4d'° 5s - 4d 1 0 5p 10 10 4d 5p - 4d 5d t z ° S^ - P<A. (Difference Pi/i. - T>3/2 Difference Ag I 29552.0 • 4 1 18329.0 14583.3 24743.8 Cd II 44135.3 43072.8 13049.7 23858.2 In III 57185 66931 12374.0 22289 Sn IV 69559 89220 12007 22133 Sb V 81566 111353 TABLE 2. Irregular Doublet Law applied to Iso-electronic Sequence Agl - SbV PLATE V. Showing Moseley Diagram f o r I s o - e l e o t r o n i c Sequence Agl - SbV. (36) IV. RESULTS. Over 550 lines covering the region 6890JU-2080A have been measured and their wavelengths calculated. The. results show that the indium spectrum has relatively few lines, for of the total number measured, only some 2£5 can be Identified with those of indium. The remainder, re-presenting impunity lines, could be identified mainly with oxygen, nitrogen, and argon i n the longer wavelength regions, and with tungsten i n the shorter regions. In the following table i s presented the results obtained from the wavelength oalculations. For the most part, these are found to agree with the published l i s t s . Measured wavelengths, however, agreeing with neither the published wavelength l i s t s for indium nor with known im-purity lines, have been marked "new". It i s believed that these have not appeared i n the literature to date. ORIGIN INT. A air REMARKS ORIGIN INT A air REMARKS In III 4 6891.28 In 1 5046.03 new In II 0 6867.71 In II 1 4973.61 In I 4 6847.45 In 2 4965.20 new In II 5 6783.72 In II 1 4906.97 In 4 6684.35 new In III 8 4847.85 In 5 6642.27 new In 9 4737.40 new In 0 6611.90 new In 8 4728.47 new In III 2 6198.06 In II 2 4681.11 (37) ORIGIN INT. A a i r REMARKS ORIGIN INT A air fl<Mff?KS In 2 "6184.32 new In II 3 4638.28 In 2 6116.49 new •In II 2 4620.61 In II 0 6095.72 In II 2 4615.57 In II 0 5918.65 In 4 4612.13 In II 0 5903.42 In 3 4608.20 new In 0 5890.26 In 8 4579.32 In 0 5869.62 new In 1 4564.97 In II 4 5852.83 In 1 4552.82 new In 0 5836.24 new In 1 4530.05 In III 0 5820.08 In I 8 4511.32 In 1 5811.74 new In 1 4479.99 new In I 1 5709.80 In 3 4457.29 new In III 5 5644.86 In 2 4421.30 new In II 0 5535.94 In 2 4376.57 new In 0 5460.38 new In II • 1 4364.77 In III 1 5248.52 In II 0 4358.50 In 1 5142.83 new In II 1 4336.34 In 7 5062.97 new In II 3 4330.02 In 0 4323.52 new In 9 3995.16 In 1 4309.81 new In 2 3992.63 new In 1 4301.50 In 4 .3969.13 In 4 4283.59 new In 4 3957.47 In 0 4277.99 In 3 3944.43 new In III 1 4253.88 In 1 3936.79 In 6 4230.51 new In II 0 3934.12 (38) ORIGIN INT. A a i r REMARKS ORIGIN INT. A a i r ] IEMARKS In II 2 4227.16 In 1 3931.11 In II • 0 4218.72 In II 0 • 3929.53 In 0 4212.32 new In 1 3926.03 new In II 1 4205.08 In II 1 3924.35 In 0 4180.94 In 1 3920.54 In 0 4171.04 new In 6 3884.83 new In 2 4146.49 new In 7 3835.18 In I 9 4101.77 In II 0 3808.11 In 1 4082.90 new In II 2 3801.52 In 10 4073.88 new In II 1 3799.37 In III 3 4071.40 In 0 3739.95 In III 2 4062.86 In II 3 3723.65 In 1 4053.27 new In II 2 3717.89 In 3 4040.38 new In II 3 3716.28 In III 2 4033.05 In II 2 3708.26 In III 1 4024.83 In 1 3640.65 new In II 2 4012.96 In 2 3590.35 In II 3 4013.49 In 2 3583.85 In II 8 4013.93 In 3 3503.62 new In 0 3480.16 In I 8 3039.56 In 1 3467.66 In 2 3034.12 In 2 3466.24 In II 0 3028.57 In 0 3454.73 In 0 3023.39 In II 3 3438.52_ In II 2 3022.37 In II 2 3376.60 In III 9 3009.96 (39) ORIGIN INT. A air REMARKS ORIGIN INT. X air REMARKS In II 3 3346.06 In 4 3002.64 new In 4 3336.26 In II 3 2999.40 In 4 3328.49 new In III 8 2982.93 In III 2 3325.44 In 4 2979.58 In 1 3285.80 In I 3 2957.01 In II 3 3274.02 In II 9 2941.05 In II 2 3264.01 In 0 2938.71 In I 8 3258.56 In I 5 2932.62 In 3 3250.38 In 1 2924.59 new In 2 3247.61 In II 8 2890.19 In II 0 3236.86 In 0 2881.61 In II 0 3198.04 In I 3 2858.74 In II 3 3194.63 In 2 2853.27 new In 4 3187.03 In 3 2843.04 new In II 5 3138.64 In I 7 2836.92 In 4 3078.27 In 1 2823.99 In 0 3067.73 In II 1 2818.93 In 0 3064.72 new In 1 2810.90 In 3 3054.87 new In II 1 2805.40 In I 3 3051.25 In II 3 2798.76 In 1 2785.18 new In II 6 2432.73 In I 3 2775.36 In I 1 2430.99 In I 7 2753.88 In II 4 2425.82 In II 1 2749.71 In I 2 2419.16 In 1 2726.27 In II 1 2418.93 (40) ORIGIN INT. A air REMARKS ORIGIN INT. A a i r REMARKS In I 3 2713.94 In II 8 2415.49 In I 4 2710.27 In II 7 2412.83 In 1 2709.2 In II 8 2410.85 In II 3 2674.48 In I 7 2399.18 In II 2 2654.76 In I 0 2378.14 In 0 2578.9 In o 2360.33 new In I 4 2560.23 In II 4 2350.74 In II 2 2536.67 In I 1 2345.90 In 0 2519.41 In I 1 2332.76 In II 6 2512.34 In II 1 2328.00 In 1 2510.41 In 1 2321.83 In II 8 2508.16 In 2 2318.00 new In II 0 2488.67. In I 1 2315.09 In 2 2475.57 In II 8 2306.12 In I 1 2468.02 In I 1 2302.49 In II 0 245-1*07 In I 1 2298.33 In 0 2455.68 In 8 2297.09 In II 1 2453.86 In 1 2292.8 In II , 1 2453.24 In I 2 2278.3 In II 0 2442.63 In - 0 2254.37 In 8 2438.75 In I 1 2241.66 In I 0 2441.10 In I . 1 2187.40 In III 2 2350.73 In 3 2181.29 new In 2 2229.8 In 5 2177.7 In 0 2227.50 In 3 2171.44 (41) ORIGIN INT. A air REMARKS ORIGIN INT. A air REMARKS . In II 1 2195.67 In 3 2165.87 new In I 0 2190.84 In 0 2154.81 0 2188.26 new In II 0 2079.26 (42) V. BIBLIOGRAPHY (1) Badami, £. Proo. Phys. Soo. 43 : 538 (1931). (2) Bacher, R. F. and Tomboulian, D. H. Phys. Rev. 52 : 836 (1938). (3) Bethe, H. A. Rev. Mod. Phys. 8 : 83 (1936). (4) Breit, G. Phys. Rev. 35 : 1447 (1930). (5) Breit, G. and Wills, L. A. Phys. Rev. 44 : 470 (1933) (6) Casimir, H. B. G., On the Interaction Between Atomic Nuclei and Electrons (Teyler's Tweede Genootschop)(1936) (7) Dirac, P. A. M. Proc. Roy. Soo. A 117 : 610 (1927) (8) Fermi, E. and Segre, E. Zeits f. Physik 82 : 729 (1933) (9) Frayne Phys. Rev. 31 : 152 (1928). (10) Gordon, W. Zeits F. Phys. 48 : 11 (1929). (11) de Gramont. Comte Rendus 171 : 1106 (1920). (12) Green, J. B. and Lang, R. T. Proo. Nat. Acad. Sci. 14 : 706 (1928). (13) Hamilton, D. R. Phys. Rev. 56 : 30 (1939). (14) Hartree, D. R. Proo. Cambridge P h i l . Soo. 23 : 304 (1926). (15) Hertzberg, G., Atomic Spectra and Atomic Structure. (Dover Publications, New York, 1944). (16) Jackson, D. A. Zeits f. Physik 80 : 59 (1933). (17) Lande, A., Zeits f. Physik, 25 : 46 (1924). (18) Lang, R. J. Proo. Nat. Acad. So. 15 : 414 (1929). (19) Lang, R. J. and Sawyer, R. A. Zeits f. Physik. 71 : 453 (1931) (43) (20) McLennan, J. C., A l l i n , E. J., Hall, K. E., Roy. Soc. Proc. 133 : 333 (1931). (21) Paschen, F. Ann, der Physik 32 : 152 (1938). (22) Paschen, F. and Campbell, J. S. Ann, der Physik 31.1 : 29 (1938). (23) Paschen, F. and Campbell, J. S. Naturwiss 22 : 136 (1934) (24) Pauli, W., Zeits f. Physik 43 : 601 (1927). (25) Racah, G. Zeits f. Physik 71 : 431 (1931). (26) Rao, K. R. Proc. Phys. Soo. 39 : 161 (1937). (27) Saunders. Astro Phys. Journ. 43 : 240 (1916). (28) Schuler, H. and Schmidt, Th. Zeits f. Physik 104.5-6 : 468 (1937). (29) Schuler, H. and Sohmidt, Th. Zeits f. Physik 94 : 457 (1935). (30) Sugiura, Jap. Journ. Phys. 3 : 155 (1924). (31) Uhler and Tanch Astro Phys. Jour. 55 : 291 (1922). (32) Weinberg, H. Proc Roy. Soo. 107 : 138 (1925). (33) White, H. E., Introduction to Atomic Spectra (McGraw-Hill, New York, 1934). ABSTRACT The present project has been concerned with the study of the spectra of indium. For this purpose, a Hilger 151 Quartz Spectrograph was used, and plates taken i n the three wavelength regions 6890 1 — 3300 1, 3300 A — 2500 A and 2500 A -- 2080 A. The light source, consisting of an electrodeless discharge i n the vapor of indium, was constructed and was then conditioned so as to permit certain variations i n the ex-citation of the observed spectra. A large part of the ex-perimental work was concerned with the conditioning of the source where many d i f f i c u l t i e s were encountered due to the high boiling point of the indium. The spectral lines were measured using a Hilger comparator and their wavelengths calculated from the Hart-mann dispersion formula. Over 225 known indium lines were identified, and nearly 50 lines, not appearing i n the literature to date, have been found. 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items