US2866914A - Photomultiplier - Google Patents

Description

Dec. 30, 1958 Filed Ju'Iy 2o, '1956 3 Sheets-Sheet 1 HIS ATTORNEY Dec. 30, 1958 A. LALLEMAND 2,866,914

PHOTOMULTIPLIER Filed July 20, 1956 5 Sheets-Sheet 2 FIG.2

INVENTOR. ANDRE LALLEMAND lf2/0%; MW

HIS ATTORNEY Dec. 30, 1958 Filed July 20, 1956 IN V EN TOR.

w ANDRE LALLEMAND BY l HIS ATTORNEY United States had PHOTMULTIPLIER Andr Lallemand, Paris, France, assignor, by mesne assignments, to Schlumberger Well Surveying Corporation, Houston, Tex., a corporation of Texas This invention relates to electron multiplier and, more particularly, to electrostatic secondary-emission photomultipliers.

Photomultipliers are now employed in a wide variety of applications for converting light energy into signals of useful strength. In an electrostatic photomultiplier, electrons ejected by radiation falling on the photocathode are accelerated toward the rst of a seriesr ofv dynodes at successively higher potentials. These dynodes comprising an electron multiplier have the characteristic of yielding large numbers of secondary emission electrons when bombarded with a relatively few energetic electrons. By cascading the path of secondary electrons through the series of dynodes, a stream of electrons fiowing from the last dynode to the photomultiplier anode may be derived of a magnitude proportioned tothe numbers of electrons emitted from the photocathode but ampliiied by a million or more.

In photomultipliers heretofore devised, this performance may be disturbed by a variety of causes, several yof which are particularlysignicant. First, because a perfect Vacuum cannot be maintained in photomultipliers over a period of time, residual gases remain which are subject to ionization by collision with accelerated electrons. A positive gaseous ion formed by `ionization may be drawn to the photocathode with suicient energy to eject electrons, thus creating a spurious signal and leading to unstable operation. This disturbance is frequently termed ion feedback.

A second disturbance arises when the potential gradient at the surface of a dynode becomes so great as to liberate electrons by a process known as cold emission or field emission. This disturbance vmay arise particularly from non-linearities in potential variation along the electron paths extending between successive dynodes.

A third disturbance arises from the luse of an insulating glass envelope to enclose the multiplier structure and of insulators to support parts ol' this structure. Because of the requirement that successive dynodes be at diierent potentials, for example, a relatively open space between successive dynodes is frequently provided through which electrons are guided by focusing arrangements. Such focusing never being perfect in practice, some 'of the electrons traveling between adjacent dynodes escape to the walls of the glass envelope and eject secondary electrons towardthe anode. A suicient positive charge may thus be developed on an insulating surface to result in a cold emission discharge and the consequent generation of spurious noise signals at the photomultiplier output terminals. Noise signals may also be derived from so-called microphonics, that is, vibration of the dynode structure in a manner distorting the electron path.

vIt isian object ofthe presentV invention to providea new and improved photomultiplierin which the above-described disturbances are minimized or eliminated.

yMore particularly, it is an object of this invention to provide Aa photomultiplier having an effective arrangement for confining the travel of secondary emission of arent electrons toa path of substantially uniform potential gradients while shielding against outside perturba-tions.

Another object of this-invention is to provide a` photomultiplier of high sensitivity and stabilityv of operation and with a minimum of residual gases.

Yet another object is to provide a photomultiplier requiring a minimum number of signal carrying leads and hence a minimum number of seals between the photomultiplier `envelope and such leads.

K nStill another object is to provide a photomultiplier of sirrlple,l rugged construction for ease of assembly and reliability of operation. Y I

These and other objectsV of the invention are attained by an improved assembly of photomultiplier Jdynodes arranged not. only to yield secondary .emission electrons but to -focus and confine these electrons while adsorbing positive gaseous ions tending to move toward the photocathode. .The assemblycomprises spacers enclosingthe i'egionbetween each successive pair of dynodes and having low conductivity interior walls both to linearize the potential variation between successive Vdynodes and to discharge any particle deviating from the path between successive dynodes.

Each dynode, except the, last, comprises strips of electron emissive material presenting an emissive surface to impinging electrons with passages between the strips providedfor flow of secondary electrons ejected by such impinging electrons. Tofocus the impinging electrons on the secondary emissive surfaces, each dynode carries a curved wire mesh arched over the dynode strips to form a convergent, electrostatic lens. Appropriate portions of these lenses are constructed of a so-called getter material, that is, a material adapted to adsorb the ions of residual gases contained within the photomultiplier.

TheV invention, together with its objects and advantages will more perfectly be understood from the following detailed description taken in conjunction with the drawings, in which: g

Fig. 1 is an elevation of the improved photomultiplier of this invention with portions including the envelope partially cut away to reveal internal features of construction;

Fig. 2 is an exploded View of the iirst dynode with portions thereof cut away;

"'Fig. 3 is a sectional view of the photomultiplier of Fig. l taken along adiametral plane and illustrating only the lower portion of the photomultiplier, together with a schematic version ofcircuitryfor association therewith; and

Fig. 4 is a schematic view of the photomultiplier to illustrate its circuit equivalentitogether with representative paths of charged particles.

Throughout the figures like reference numerals are employed lto designate similar elements. In accordance with the present invention, a photomultiplier is shown in Fig. vl comprising an evacuated envelope 10 having therein a photocathode 11, an electron multiplier assembly 12 and an anode 13. 1'1' he envelope 10 is conveniently of cylindrical configuration and is composed of a material, such as a shock resistant glass, capable of withstanding externally applied pressures and serving as an electrical insulation. Atone end of the photomultiplier the envelope 10 provides a window 14 transparent to radiation such as light quanta or photons.

lInteriorly of the window 14 are a barrier layer 15 and av photosensitive layer 16 of the photocathode l1. The barrier layer 15 is transparent to the. radiations to which the layer r16 -is sensitive. Preferably, the layer 15 is .deposited o n the inner surface of the window 14 and is composed of an ,electrically conductive substance as vdescribed in copending application Serial Number 599,207,

3 filed July 20, 1956, by J. P. Causse and A. Lallemand for Photosurface and Method of Making Same. In accordance with this application, the layer further serves chemically to isolate the photosensitive material of layer 16 from constituents, such as boric oxide, which may be `the photosensitive layer 16,' eachfhaving the characteristic of emitting electrons known as photoelectrons upon the incidence of radiation passing through the transparent window 14 and layer 15. For example, the photcsensitive layer 16 may be composed of cesium-antimony, sodium-antimony, or any of 4a `variety of intermetallic compounds of alkali or alkaline earth `metals and antimony, sulfur, selenium, tellurium, bismuth, thallium, or lead.

To raise the signal `represented by the photoelectrons to a useful strength, the electron multiplier assembly 12 comprises a series of dynodes 20a, 20b, 20j and 20k aligned with the photocathode 11 and anode 13, the first `dynode 20a being disposed opposite the photocathode 11. As seen in Figs. 1 and 2, the first dynode 20a comprises a centering ring 22a, a plano-convex lens 23a and an axial spacer 24a.

Ring 22a serves to position `the dynode 20a along the 'axis of the envelope 10. To this end, the ring 22a includes a thin, at annulus 26 of insulating material, such `as mica, having a notched circular periphery 27 frictiona1- ly engaging the inner wall of the envelope 10, and a circular inner periphery 28 supported in an inwardly crimped portion 29 of an annular axially extending band 30. The band is composed of a conductive'material such as aluminum. Below the inwardly crimped portion 29 is provided a `downwardly and inwardly directed annular shoulder 31 and above it a at upwardly directed "shoulder 32 for engagement with the lens 23a, as hereafter described. The upper shoulder.32 formed in the band extends radially `outwardly therefrom overlying the insulating annulus 26. The band 30 then provides a conduction path between the outer surface 33 of its axially extending lower portion and'the shoulder 32. At the outer periphery 27 of the insulating annulus, uniformly 'spaced notches may be provided in order that conductor 35, as well as other support rods (not shown) may extend axially along the inner wall of the envelope 10 to the base 19. i

The plano-convex lens 23a is,like the centering ring 22a, of a unitary construction andis, moreover, arranged for detachable clamping engagement with the centering ring 22a. The lens comprises a rigid annular frame 37 of generally circular cross section flattened in the axial direction and composed of a conductive material, for ex-` ample, nickel. Formed snugly"about the ring 37 is a channel shaped member 38 opening inwardly and in interior pressure contact with the flattened upper and lower surfaces of the rigid frame 37.` Member 38 is also conductive. s

Inserted between the upper `,face of the frameV 37 and the channel member 38`are vradially extending end portions 39 of tine, arcuate wires `40 lying in spaced axial planes and circularly contoured to define a relatively open, hemispherical mesh having its convex side facing toward the photocathode 11. ,s

Between the lower face of the rigid ring 37 and the inner surface of the channel member 38 pressing thereagainst, tabs 42 formed at either end of dynode strips 43 are inserted to rigidly hold the strips in spaced parallel relation across the aperture defined by the ring 37. The strips 43 are bent out of the horizontal plane of the tabs '42 at anangle, `for example, of 45 to theaxis ofthe Y22a and lens 23a into firm contact.

-larger radius of curvature.

lens in order to present their inclined secondary-emissivc surfaces 44 to the photoelectrons from the photocathode 11 for impingement at an angle. For good secondary emission characteristics, the upper surface 44 of the strips 43 is, for example, composed of a silver-magnesium alloy. By reason of this inclined strip construction, the dynode 20a may be referred to as a Venetian blind type.

To minimize ion feedback to the photocathode 11, both the mesh wires 40 and the undersurfaces of the dynode strips 43 are composed of a metal having the qualities of a getter, that is, the property of discharging and adsorbing gaseous Amolecules formed when a positively charged gaseous ion impinges. Satisfactory getter ma terials include titanium, molybdenum, zirconium, tantalum, palladium,`and others.

To the underface 46 of the channel member 38 is secured a wire ring 47 sized for snugly fitting within the inwardly crimped portion 29 of the centering ring 22 when the underface 46 is resting on the shoulder 32. Also attached to the underface 46 are a plurality of spring fingers 48 extending downwardly in a generally tangential relation to the annular channel member 38. When the lens 23a is assembled with the centering ring 22a, the spring fingers 48 will extend under the inwardly crimped lportion 29 of the centering ring and bear resiliently against the inclined shoulder 31 to cam the centering ring A good electrical connection is thereby effected, so that the entire lens 23a and the band portion 30 of the centering ring 22a may be at the same electrical potential.

The remaining component of the dynode unit 20a, namely, the axial spacer 24a, has the form of a cylindrical vband with an interior diameter sized to snugly receive the band portion 30 of the centering ring 22a. When thus frictionally interfitted with the centering ring 22a, the

spacer 24a has its upper surface 49 bearing against the insulating annulus 26. As seen in Fig. l, the spacer 24a also snugly receives lens 23h of the second dynode unit 20b. The bottom edge 50 of the spacer is then held against the upwardly facing shoulder 32 of the centering ring for the second dynode unit 2Gb.

In order that the potential gradient between successive dynodes may be rendered more uniform, the spacer 24u has a composition which affords a low conductivity path. Since the electrostatic field lies within the spacer and since the inner surface of the spacer makes pressure con tact both with the centering band 30 and the lens channel member 38, it is only necessary that the inner surface exhibit conductivity. Hence, the spacer may be composed of an insulating material such as glass, a ceramic or the like, preferably with its inner surface having a layer 51 of conductive material deposited on it. The layers may, for example, be composed of gold or silver evaporated upon the inner surface or, if desired, a getter material of appreciable conductivity. Moreover, the bulk of the spacer 24 may itself be somewhat conductive. While the required conductivity of the spacer 24 will depend upon the current drawn from the dynodes and the potential difference existing between adjacent dynodes, its resistance may be on the order of 200 kilohrns with about to 200 volts between successive dynodes. Preferably, the resistance will have a low temperature coefficient, especially where the photomultiplier may be employed in variable high temperature environments such l as are encountered in radioactivity well logging.

While the foregoing description has been particularly directed to the dynode unit 20a, it will apply with equal force to the dynode units 20b through 20. A difference will be observed, however, in the later dynodes of the series as .their spacers are shorter in axial length and the wire mesh of the lenses becomes shallower, that is, of a Moreover, the spacer 24j is provided withan aperture 52.

The anode 13, preferably formed of a wire grid as seen in Fig. l, is disposed within a conductive cylinder `53,

known as a Faraday box, having an aperture l54. A `lead-1n `Wire 55 passing throughY aseal'56 in t'hebase 19 and through the apertures-52am 54-supports the anode 13 firmly. The cylinder -53 may be secured to the dynode 20k conductively.

For efficient production of secondary-emission electrons from the last dynode 20k, this dynode yis formed as a circular. plate closely underlying the grid-like anode'13. The periphery of the dynode-20k Vis sized to 'tit within the bottom end of the cylinder S2 and is in electrical contact` with the inner conductive getter layer 51 of the same. A conductive lead 57 passing through a seal v58 in the base 19 affords electrical contact with the dynode Ztlkand, at the same time, rigidly supports the sameand the cylinder S2 in spaced relation to the dynode 20j.

The rugged mounting of the electron multiplier structure may be observed then to follow from the snug contact between the insulating annuli 26 and the inner wall of the envelope 1t), the snug interfitting of the lens, centering ring', and spacer for the successive dynode units, and the -rigid supports afforded by theV leads 35, 55 and 57 passing, respectively, out of the base 19 through seals 36, 56 and 58. The linear arrangement of the'Venetian blind dynodes, furthermore, facilitates ready manufacture of the dynode units such that snug intertting maybe obtained with economy.

if desired, additional lead-in conductors (notshown) may be' connected to each of vthe dynodes20b, 20j in the same manner that conductor 35 connects with dynode 20a further lto rigidify` the assembly. As will be more fully described hereafter, however, 'these additional conductors are unnecessary to the establishment of suitable potentials 'on the dynodes in view of the employment of theconductive spacers interconnecting the successive dynodes, but may nonetheless Vbe employed for this purpose, if desired.

ln Fig. `3 suitable means for energizing-the photomultiplier is shown comprising a D.C. source 60 connected across a potentiometer 61 having its negative terminal connected to conductor 17 and its positive terminal connected through resistor 59to the anode 13. Connected to a tap 62 in the potentiometer close to its positive terminal is conductor57, whereby the potential on the last dynode 20k is made less positive thanf'the anode 13. Coupling means suchas capacitor 63 connected to the anode 13 serves to couple the signal'developed across the resistor 59 to an indicating circuit orl thel like (not shown). Conductor 35 connects the. r'st V'dynode 20a with a tap 65' toward the negative end` of` the'potentiometer' 61. in this manner,'the potential! between the first dynode 20a and the photocathode 11 may be xed at a suitable value governing acceleration of photoelectrons emanating from the photocathode. If desired, additional conductors, similar to conductor 35, might be arranged for connecting each of the dynodes 20h, 20j appropriately to the potentiometer 61 but such additional connections would be largely superfluous in view of the conductive character of the spacers.

In Fig. 4, the spacers 24a, 241 are schematically represented as resistors connecting between the corresponding pairs of dynodes 2da, Zilb, 20j and 20k. This representation illustrates more clearly the significance of the spacers in establishing successively higher potentials proceeding from the photocathode 11 through the first, thence to the last dynode and nally to the anode' 13. Because the inner walls of the spacers are conductive, moreover, they tend to render substantially uniform the potential gradient in an aXial direction along the walls. The eect of the spacers in conjunction with the planoconve: lenses forming the series of dynodes is exemplied by the representations of cascaded electron paths 67 stemming from a single photoelectron 68 emitted from the cathode 11. Also exemplified are representative paths 69 for positive ions 70 formed as collision products of the accelerated secondary emission-electrons. These posilli higher potential field of the first dynode.

6 tive'ions; indicated in Fig.'4 by a' positivel sign, are shown lmplngingin one iin'stanceupon ahwire 40 of the spherical 'mesh lfor'dynode 20e and, in the other instance, impinging `upon the `undersurface of a dynode strip -43 of the -seconddynode"20b. It willfbe appreciated, of course, that only a fragment of the actual electron paths may be representedfand that the high gain per stage precludes van intelligible representation of paths through the entire 20a. `Whenever photons or other radiant energy impingc' upon the photocathode 11,'photoelectrons are ejected and are ldrawn from the region of thecathode toward the As the photoelectrons approach the convex side of the wire mesh 40,

all of themv convergeA inwardly'toward the axis of the photo-tube and vof the first lens.

This results from the convergence of the elec-tric eld `between lenses produced by the mesh curvature. Hence, the photoelectronsare electrostatically'attractedtowad the `axis, thereby in- `creasing the collection efficiency `of the first dynode and preventingphotoelectrons from straying to the wall.

vlt/Iost" of the ph'itoele'ctronsv willimpinge upon the secondary emissive surfaces of strips 43' to eject secondary electrons therefrom. 'Since' the secondary electrons may number yin the ratio of'4 to l, `on the average, relative to the numbers of incident photoelectro'ns, electrons representing-the diiferencebetweenthe captured and the emitted elcctronswill'flow"from lthe-source 60 through the conductor 3S' to the dyndde` 20a. At'the same time, the secondary electrons will' be drawn convergently toward Ythe second kdynodeZtlb, repeating the-secondary emission process, and 'the process will thusl cascade through the series ofdyno'des. `Finally,some^of the secondary electrons emitted by the dynode 20j will pass directly to the anode 13'while perhaps the greater number will pass through'the anode 13tstriking vthe last dynode 20k to 'produce y'et additional secondary electrons. Substantially 4all of the secondary electrons eventually result in an electron iow tothe anode z13'yielding a current therethrough substantiallyA linearly proportioned to the intensity `of klight incident upon the photocathode 11. The Withdrawalof electrons by the secondary emission processtends to drive'each dynode positively, so that electrons flowv from thesourcev through conductor 35 land the spacers 'to the dynodes to restore'each toits normal potential.

In the event that electrons strike the inner walls of the spacers despite the convergent elds established by the lenses, the conductivity of the spacers will allow such electrons to be discharged through the last dynode 20k to the power supply tap 62. Because the conductivity of the spacers is sufficiently large to permit an appreciable current to pass from the last dynode to the first dynode, however, the very slight current produced by electron collisions with the inner walls of the spacers will have no significant elect upon the proper operation of the photomultiplier.

Within the lenses, the electrons will generally exhibit cathode 11, the surfaces which such positive ions likely might strike are prepared for capturing or adsorbing the ions at the same time neutralizing their charge. Thus, positive ions may impinge upon the undersurfaces 45 of the dynode strips or upon the wires 40 of the spherical meshes, each of which are composed of getter material.

In this manner, regenerative ion feedback and consequent instabilities are avoided. It may be noted that the spherical wre mesh of each dynode,A being at the potential of the corresponding secondary emission surface, enhances both the yield and the collection of secondary emission electrons.

In the event that the photomultiplier is subject to shock or vibrational forces applied to the envelope, these forces will be transmitted through the centering rings 22a, 22; to the firmly connected lenses and spacers so that the entire assemblage rwill have a unitary motion.

As seen particularly in Fig. 4, the lens structure will produce a higher potential gradient along the axis of the phototube, but for a shorter distance, than the potential `gradients in the axial direction `displaced radially from the axis of the tube. Where the expression is used that the spacers tend torender the potential gradient more uniform,1this should `be understood asrneaning that the potentialgradient experienced by electrons traversing an axial path between successive lenses or dynodes will experience substantially a constant potential gradient, although the gradients along adjacent paths may be either greater or smaller. i

While a preferred embodiment of the invention has been shown and described, it will `be evident that various modifications may be made without departure from the principles of the invention. Thus, in lieu of a photocathode formed on the interior face of the window 14,

a photocathode might be disposed in front of the iirst dynode at the focus of a Schmidt type optical system coupled to the source of radiation. The spherical wire mesh defining the convex surface of the electrostatic lenses might be formed of wires woven in a crisscross pattern. In lieu of constructing each of the dynodes with getter material at the undersurface of the wires 40 and at the nndersurfaces 45 `of the `dynode strips, onlythe latter tive or so of the dynodes might be so constructed, these being in the region where ionization is most apt to occur.

which may occur to those skilled in the art, the invention is not to be limited except as delned in the appended claims.

8 I claim: 1. In an electron multiplier, a cathode, an anode, a

4plurality of dynodes, each dynode including a circular which confront said anode composed of an adsorptive gettering material, a plurality of insulating cylindrical spacers stacked in endwise alinement, a plurality of cen tering rings extending between the endwise surfaces of Vadjacent spacers and conductively secured to corresponding ones of said frames to support said dynodes in spaced axial alinement within the stack of said spacers intermediate said cathode and anode, envelope means supported in axial alinement with said stack of spacers by at least two of said centering rings for enclosing said cathode vand anode in sealed relation with said dynodes, a pair of leads extending in sealed relation through said cnvelope means for applying a potential difference between said cathode and anode, and high-valued resistances ex tending along said stack of spacers and in electrical contact with said centering rings to provide a potential divider between said dynodes.

2. A11 electron multiplier, as defined in claim l, where in said centering rings are secured in full annular supporting contact with respect to said dynode frames interiorly of said stack of spacers, each of said centering rings being secured in full annular supporting contact with respect to an adjacent one of said spacers.

References Cited in the file of this patent UNITED STATES PATENTS OTHER REFERENCES Rodda: Photo-Electric Multipliers, MacDonald & Co., Ltd., London, 1953,- pages 53 and 54.

Bruining: Physics and Applications of Secondary Electron Emission, McGraw-Hill Book Co., Inc., New York, 1954, pages 112 and 113.

Zworykin: Photoelectricity and Its Applications, I ohn Wiley & Sons, Inc., New York, 1949, pages 144 and 145.

Images