US3431420A

US3431420A - Crossfield photoelectron multiplier tube having channeled secondary emissive dynodes

Info

Publication number: US3431420A
Application number: US606160A
Authority: US
Inventors: Mahlon B Fisher
Original assignee: Sylvania Electric Products Inc
Current assignee: GTE Sylvania Inc
Priority date: 1966-12-30
Filing date: 1966-12-30
Publication date: 1969-03-04
Anticipated expiration: 1986-03-04

Description

M. B. FISHER 3,431,420 ON MULTIPLIER TUBE HAVING CHANNELED SECONDARY EMISSIVE DYNODES Sheet um om N am .mN om N om INVENTOR.

MAHLON B. FISHER CROSSFIELD PHOTOELECTR ATTORNEY March 4. 1969 March 4. 1969 M. B. msm-:R

CROSS? [ELU PHOTOELECTRON MUL'I'IPLIER TUBE HAVING CHANNELED SECONDARY EMISSIVE DYNODES Filed Dec. 30, Sheet =O.3 FINNED DYNODE ELECTRODE Y (INCH) NVENTOR.

MAHLON B. FISHER VOLTAGE SOURCE ATTORNEY 3,431,420 ANNELED March 4, 1969 M. B. FISHE CTRON MULTIPLIE ONDARY EMISSIVE lDYNODES CROSSFIELD PHOTOELE R TUBE HAVING CH SEC Filed Dec. 30, 1966 Sheet INVENTOR.

MAHLON B. FISHER BY /w M/@K ATTORNEY United States Patent 3,431,420 CROSSFIELD PHOTOELECTRON MULTIPLIER TUBE HAVING CHANNELED SECONDARY EMISSIVE DYNODES Mahlon B. Fisher, Monte Sereno, Calif., assignor to Sylvania Electric Products Inc., a corporation of Delaware Filed Dec. 30, 1966, Ser. N0. 606.160 U.S. Cl. Z50-207 10 Claims Int. Cl. H01j39/12, 3]/50 ABSTRACT F THE DISCLOSURE Plane cathode and dynode electrodes of a crossed-field secondary emission photomultiplier are segmented to prevent transverse spreading of the electron beams and thus minimize or eliminate resultant image distortion. Each electrode has transversely spaced longitudinally extending parallel electrically conductive fins projecting from the electrode surface normal to the plane thereof. Corresponding fins on longitudinally adjacent electrodes are aligned in planes parallel to the longitudinal axis of the device and effect a distortion of the electric field to produce the desired focusing of the electron beams.

Background of invention This invention relates to photodetectors and more particularly to an image intensifier having a frequency response extending to microwave frequencies.

Image intensifiers have the desirable characteristic of preserving an electron beam image of an incident signal during multiplication or intensification thereof. Prior art image intensiers such as disclosed by C. Chilowsky in Patent No. 2,495,697 and by W. L. Roberts et al. in Patent No. 2,821,637 maintain the electron beam image by multiplying or intensifying portions thereof in separate associated passageways such as tubes or ducts. The frequency response of these devices is limited, however, by transit time dispersion in the multiplier sections and by the decay characteristics of phosphor screens which display the intensified image. The cross-field electron multiplier disclosed by R. C. Miller and N. C. Wittwer, IEEE Journal of Quantum Electronics, vol. QE-l, No. 1, April 1965, pages 49-59, has low transit time dispersion. It has been determined emprically, however, that this device produces .significant transverse spreading of the ele-ctron beam image during intensification thereof. By way of example, a beam of light focused to a diameter of approximately 0.010 inch on the photocathode spread to a line approximately 0.875 inch long after five stages of multiplication.

An object of this invention is the provision of a crossedfield electron multiplier wherein the electrom beam image of an incident signal is preserved during intensification thereof.

Summary of invention In accordance with this invention, the plane photocathode of a crossed-field electron multiplier is divided into parallel channels to provide a segmented electron beam image of an incident light beam. The cathode is sectioned by laterally spaced electrically conductive fins which are electrically connected thereto. Each fin is perpendicular to the cathode and parallel to the longitudinal axis of the device. Each pair of laterally adjacent fins together with the intermediate surface of the cathode defines a cathode channeL The fins distort the electric field between the cathode and anode electrodes to provide a transverse component of electric field which focuses the electrons into a path parallel to and between adjacent fins. The dynodes are also divided into such channels by fins which produce a transverse component of electric field that similarly prevents lateral crossover of electrons from one channel to another while passing through the dynode-multiplier section. The electron beam image is coupled from the device by a mosaic collector output circuit comprising a plurality of coaxial collectors with several such a collectors associated with each channel.

DESCRIPTION OF DRAWINGS This invention will be more fully understood from the following detailed description of a preferred embodiment thereof, together with the accompanying drawings in which:

FIGURE l is a schematic diagram of a -crossed-field image intensier embodying this invention;

FIGURE 2 is a longitudinal section taken along line 2-2 of FIGURE l;

FIGURE 3 is a transverse section taken on line 3 3 in FIGURE 1 showing the electric field pattern between a segmented dynode and the anode electrode;

FIGURE 4 is a perspective view of a crossed-field image intensifier electron discharge device embodying this invention which was actually built and tested;

FIGURE 5 is a perspective view of a modified structure for supporting the photocathode and dynode electrodes;

FIGURE 6 is a graph illustrating the transverse electric field distribution as a function of height above the finned electrodes;

FIGURE 7 is a section of the mosaic collector output circuit taken on line 7&7 of FIGURE 4;

FIGURE 8 is an end view, greatly enlarged, of part of the collector output circuit inside the vacuum envelope of the device; and

FIGURES 9 and 10 are section views, similar to FIG- URE 2, of modified forms of this invention.

Techniques for segmenting the electrodes and preventing transverse beam spreading include (l) periodic distortion of the electric field, (2) a periodic distortion of the electrode surface, and (3) a combination of these techniques. In conventional crossed-field photomultipliers, a uniform electric fie-ld exists which is normal to the plane electrodes. Spatially periodic changes in the contour of the electrode surfaces in accordance with this invention produce components of electric field which are parallel to the principal plane of each electrode. Such transverse components of electric field must be (l) symmetrical about the center line of each channel and (2) directed from the sides toward the middle of each channel.

Referring now to FIGURE 1, the image intensifier comprises anode electrode 1, photocathode electrode 2, dynode electrodes 3, and output circuit 7. The electrodes each comprise plane members which are parallel to each other and to the longitudinal axis X-X of the device.

Each dynode 3, see FIGURE 3, comprises a base member 9 which is made of an electrically conductive material capable of secondary electron emission and having a gain greater than one such as beryllium copper. A plurality of electrically conductive ridges or fins 10 connected to one side of plate 9 project therefrom perpendicular to the plane of the plate. The fins preferably have the same height h and thickness t and are spaced apart an equal amount s such that there is a dynode surface 11 of width s between adjacent fins. The surfaces `11 of the several dynodes are parallel to each other and to the anode electrode and extend over the length L of the dynodes, as illustrates in FIGURE 2. The dynode electrodes may, by way of example, be milled from a bar of beryllium copper and subsequently oxidized in an atmosphere containing water vapor at a pressure less than one atmosphere prior to assembly of the device for providing a secondary emissive layer of beryllium oxide on the dynode surfaces 11. Each dynode surface 11 and associated fins comprise a channel which produces an effect on the electric -eld restricting transverse movement of electrons emitted from the associated dynode surfaces 11 as will be described more fully hereinafter.

Cathode 2, which is similar in structure to the dynode, comprises base plate 12, -fins 13, and cathode surface 14 (see FIGURE 2) between adjacent fins 13. A photosensitive cathode such as an S-1 cathode is formed on the surfaces 14 between adjacent fins 13. The base plate 12 of an S-1 photocathode is preferably made of silver.

If it is desirable to directly view the operation of this device, output circuit 7 may comprise a phosphor screen. Such an output circuit comprises a wire screen y17 in an electrically conductive frame 18 (see FIGURES 1 and 2). A transparent glass plate 19A having a P-l phosphor screen formed on the surface 20 thereof is connected to frame 18. The wire mesh 17 and plane surface 20 are each parallel to the plane surfaces of the other electrodes.

Anode 1 also has a plane surface 21 parallel to the plane surfaces of the other electrodes. The anode has an opening 22 in the end thereof adjacent cathode 2 for passing a light beam 23 to the latter electrode. A wire screen 24 is located in opening 22 for providing a continuous electrically conductive electrode across the aperture.

As illustrated in FIGURE 2, cathode fin 10 and iins 13 of the dynodes are longitudinally aligned to provide parallel channels 26-30, inclusive, which extend over the length of the device.

An electric field E, indicated by arrow 34, between electrodes on opposite sides of the longiutdinal axis X-X is produced by connecting voltage source to associated g electrodes.

Adjacent electrodes

2 and 3a-3d are stepped such as the distance AX,L to maintain a uniform electric field over the length of the device. Variable resistors or potentiometers are provided between source 35 and the electrodes for providing adjustment of the voltage applied thereto. The electric field pattern produced by the electric field E is represented in FIGURE 3 by the arrows 37.

A magnetic field B indicated by the feathers of arrow 38 in FIGURE l is produced by magnet circuit 39 (see FIGURE 3). The magnetic field is perpendicular to the electric E-field and the longiutdinal axis X-X and extends into the drawing of FIGURE l as viewed.

In an actual embodiment of the invention, the cathode, dynodes and collector circuit were mounted on glass tubes similar to the tubes -40 and `41 in FIGURE 4 in order to electrically insulate these electrodes from each other. Anode 1 was supported in the glass vacuum envelope 42 by pins similar to the

pins

43 and 44. Alternatively, the cathode and dynode electrodes may be mounted on the steps 45 of the staircase support member `46 illustrated in FIGURE 5. The support member 46 may be made of a ceramic such as aluminum oxide.

In operation, light beam 23 incident on photocathode 2 causes electrons to be emitted therefrom. The electrons are drawn toward the anode by the electric lfield and are caused to traverse an arc 47 and impinge on the adjacent dynode 3a by the action of the magnetic field. In a similar manner, electrons emitted by the dynodes travel in a cycloidal path and are collected by circuit 7. The focusing action of the transverse magnetic field causes electrons emitted at a point P on cathode 2 to be incident on the dynodes 3 at associated points P. The points P and P' are each the same distance d from the left edge of the associated electrode (as viewed in FIGURE l). Thus, compensation for beam spreading along the longitudinal axis is not required.

In a crossed-field multiplier which does not have fins on the electrodes, electrons emitted by the electrodes and having a transverse component of velocity parallel to the magnetic field are caused to move in a transverse direction, i.e., experience beam spreading. This movement of 4 electrons is illustrated in FIGURE 2 by the broken line 48. An electron emitted from the cathode surface 14a associated with cathode channel 27 and having a transverse component of velocity moves in a transverse direction while traveling the cycloidal path 48. As shown in FIG- URE 2, this electron is not incident on the dynode surfaces of all of the axially aligned dynode channels 27, but is incident on the dynode surfaces of adjacent channels 26 of

dynodes

3c and 3d.

In an image intensifier embodying this invention, electrons having a transverse component of velocity parallel to the magnetic field are focused by the transverse component of electric -eld to be perpendicular to the plane electrode surface. This focusing action of this transverse component maintains longitudinal alignment of the electron impact and prevents an electron emitted from the cathode surface y14h of channel 23, for example, from striking the surface of the adjacent dynode within channel 27. Thus, electrons emitted from the cathode surface 14b of channel 28 traverse the cycloidal path indicated by the solid line 49 and remain in the channels 28 over the length of the device.

By way of example, a crossed-field electron multiplier similar to the device illustrated in FIGURE l but omitting fins on the cathode and dynode electrodes was built and tested. The multiplier had an anode voltage of 4 kv., a magnetic field of 200 gauss and an anode-to-cathode spacing of 0.660 inch. The useful areas of the cathode, dynodes, and phosphor screen were eac-h one square inch. When a laser light beam was focused to a spot of approximately 0.010 inch in diameter on the cathode, the image on the phosphor screen was a line approximately 0.875 inches long.

An embodiment of this invention which was the same as the aforementioned photomultiplier except that the dynode electrodes included fins having a height h of 0.093 inch, a thickness t of 0.005 inch and spaced apart a distance s of 0.1938 inch was also built and tested. \Vhen the laser light beam was focused to a small spot on the photosurface of one of the channels of cathode 2, the image displayed on the phosphor screen extended the width of the associated channel regardless of the size of the light beam on the photocathode. As the light beam was moved transversely toward one of the fins of that channel, the image on the phosphor screen remained fixed until the input light beam .was incident on the fin and light struck the photosurface of the adjacent channel. At,fthis time, the image displayed on the phosphor screen filled bot-h channels. When the incident light beam was moved to lie entirely within the adjacent channel, the image on the screen was only in the referenced adjacent channel.

Analysis of the electric field pattern associated with a finned electrode reveals that the transverse component of electric field is zero at a height above the electrode which is several times the Iheight of one of the fins and which corresponds to a particular fin height I1 to channel width s ratio. It is desirable that the transverse momentum of electrons also be zero at this point so the electrons will no longer move in the transverse direction. The fin height-to-channel-width ratio for which the transverse component of electric field and the transverse momentum of electrons are both zero at the same height above the electrode lmay be determined as described below.

In the following description, the letter x refers to the transverse direction parallel to arrow 38 and the magnetic field B in FIGURE l. The letter y refers to the direction parallel to the arrow 34 and the electric field E in FIGURE 1.

The momentum of an electron is the time integral of the force operating on the electron. The x-component pX of momentum of an electron emitted from one of the electrodes is representable as where e is a constant equal to the electric charge, Ex(x,y) is the x-component of the electric field as a function of the variables x and y, dt indicates the time derivative, t is time and pDX is the x-component of initial momentum. This time integral in Equation 1 is changed to a spatial integral by substituting in Equation 1 the identity UEY 1105)- wb2 [1 cos wet] (3) vy(t) =71y sin wat (4) where y(t) is the position of the electron in the y direction as a function of time, 11 is the electron charge to mass ratio, Ey is the y-component of electric eld,

and B0 is the magnetic field strength. The velocity vy(y) of the electron in the y-direction as a function of the position of the electron in the y-direction is obtained by combining

Equations

3 and 4 to eliminate time t and is representable as Operating on Equation l with Equations 2-5, the xcomponent of momentum pX of the electron is representable as a function of spatial variables as Lilly N/l (l 77Ey where y0 is the height above the finned electrode at which the x-component of electric lield is zero. Reference to Equation 6 reveals that the x-component pX of momentum is a function of distance x since the component Ex(x,y) of electric field in the x-direction is a function of the variable x. An approximation of the actual momentum may be obtained by selecting a particular value of the variable x and computing the momentum of an electron at this point in the transverse focusing field. Such an approximation is valid for an electric field Ex(x,y) much less than an electric field Ey, as is the case in this example. A map of the electric field between an anode electrode and a finned electrode reveals that the x-component of electric lield varies approximately linearly from a maximum value at the electrode surface to a minimum value or zero at a height y0 above the dynode surface. The variation of the x-component of electric field of a dynode electrode having a fin height h to channel width s ratio of 0.3, an anode-to-cathode electrode voltage of 4,000 volts, and an anode-to-dynode electrode spacing of 0.770 inch is shown in FIGURE 6. This data was obtained by measuring the x-component of electric field as a function of height y above the dynode surface at a particular value of the variable x which was half the distance between the center line of a channel and an adjacent iin. Since the variation of the x-component of the electric field as a function of the distance y is approximately linear, this variation is representable as EX(y)=b-my where Oygyo, b is the maximum value of the x-component of electric field near the dynode surface and m is the rate of change of the x-component of electric field.

Substitution of Equation 7 in Equation 6, the x-component pX of momentum of an electron is representable in a standard form as In solving Equation 8, values for the constants b, m, and y0 may be obtained from an electric eld map such as FIGURE 6. The x-component pox of initial momentum of the electron is assumed to correspond to several electron volts of initial energy. It is desirable that the x-component pX of momentum of an electron be zero after passing through the transverse electric focusing field. The above mathemeatical analysis may be used to determine this component of electric field and thus lthe optimum design of electrodes. Actual tests of devices incorporating finned dynode electrodes reveal that good image resolution is obtained from image intensiiers comprising finned dynode electrodes having a lin-height-tochannel-width ratio of y0.5.

An actual crossed-field image intensier for operation up through microwave frequencies is illustrated in FIG- URE 4. This device is similar to that illustrated in FIG- URE 1 except that output circuit 7 comprises a mosaic collector circuit 50 of coaxial collectors. A detailed section view of mosaic collector circuit 50 is illustrated in FIGURE 7.

Referring now to lFIGURE. 7, the mosaic collector comprises an array 51 of coaxial transmission lines having rectangular outer conductors, a vacuum Window 52, and a connector circuit 53. Vacuum Window 52 comprises a plurality of coaxial lines 54 secured in a flange 55. The coaxial lines 54 each comprise an inner conductor 56, a coaxial outer conductor 57 and a glass bead 58.

Conductors

56 and 57 are preferably made of Kovar so that the glass bead will bond thereto to provide a quality vacuum window. Flange 55 is brazed to the outer conductors 57 to provide vacuum seals at junctions 59 thereof. The flange 55 is preferably made of monel so that it can be heliarc welded to flange 77 to provide a vacuum seal.

Fabrication of the vacuum window includes brazing monel flange 55 to outer conductors 57 at junctions 5'9. This subassembly and conductor 56 are then heated in an air furnace to oxidize the inner and outer conductor for sealing glass beads 58 thereto. After the parts cool, the subassembly is placed in a stainless steel jig and glass beads 58 are sealed to the center conductors 56. The beaded center conductors are -then placed in outer conductors 57 with a small amount of glass frit at the junction of each glass bead and the associated outer conductor. This assembly is heated approximately 850 C. in a nitrogen atmosphere -to complete fabrication of vacuum window 52.

Connector circuit 53 comprises a plurality of tubular outer conductors 61 which are brazed together. The conductors `61 are also brazed at junctions `62 to a square support Iflange 63. Inner conductors 64 are supported in and insulated from outer conductor 61 by ceramic discs 65. After inner conductors 64 are axially positioned in associated outer conductors 61, the ceramic discs 65 are glazed to the inner and outer conductors to provide a rigid assembly. Sleeves 66 extend over and are welded to the reduced diameter end of inner conductors 64.

Array 51 comprises a grate 70 supported in a square frame 71 (see FIGURE 8). Grate 70 is made up of pluralities of

conductive strips

72 and 73 which form square openings therebetween.

A wire screen 74 extends over the opening in and is welded to the top of frame 71 l(see FIGURE `S) to provide a continuous electric field across the opening in the frame.

Array S1 and connector circuit 53 are assembled by forcing outer conductors 61 into the square holes in array 51 until frames 463 and 71 are flush. Flanges 7S are then welded to

frames

63 and 71 to provide a rigid, structure. As illustrated in FIGURES 7 and 8,

coaxial conductors

61, 64, and 66 are centered in the square openings formed by strips 72 and v73. Thus, the junction of array S1 and circuit 53 provides a transition from square coaxial line to circular coaxial line.

Vacuum window 52 and connector circuit 53 are joined by press-'fitting outer conductors 57 and inner conductors 56 into associated outer conductors 61 and openings 76 in center conductors 64, respectively, of the connector circuit. When assembly of the collector circuit is complete, it is vacuum sealed to the glass envelope 42 enclosing the device by heliarc welding flange S to flange 77 which is sealed to the glass envelope 42.

In practice, focusing of electrons may not be such that all electrons are incident on the ends 78 of center conductors 64. It may be desirable therefore to make center conductors 64 approximately 20 volts more posi- -tive than outer conductors 61 so that all electrons entering the square openings in array 51 are collected by conductors 64. A circuit for accomplishing this is illustrated in FIGURE 7 and comprises a potentiometer 79 connected in parallel with a voltage source 80. One terminal of source 80 is connected to outer conductor 61. The wiper arm of potentiometer 79 is connected to inner conductor 64. Although a bias voltage is applied for illusstrative purposes to only one coaxial line 54 in FIGURE 7, it is understood that in practice the bias voltage is applied to each coaxial line 54 of the mosaic collector circuit 50.

The crossed-field image intensifier illustrated in FIG- 'URE 4 has a spatial resolution of 25 output channels per square inch. This spatial resolution is determined by noting that the collector channel 26 is divided into live equal segments by the coaxial collectors 81 to `85, inclusive. Since there are five channels, the spatial resolution of the device is the product of the number of channels and the number of collectors per channel. An image intensifier like that illustrated in FIGURE 4 was built and tested and had a frequency response of greater than 600 mhz.

Although this invention has been described in relation to specific embodiments thereof, variations and modifications will be apparent to those skilled in the art. For example, the emitting surface of the dynode electrode may be scalloped as illustrated in FIGURE 9. This dynode geometery causes the normal to the emitting surface to vary in a prescribed manner across the channel and thus a variation in the direction of the initial velocities of electrons emitted thereby. A sawtooth-shaped emitting surface would also cause the normal to the emitting surface to vary across the channel. In addition to modifying the direction of initial velocities of electrons emitted by the electrodes, these dynode surface geometries also cause a periodic distortion of the electric field similar to that described above. Also, it may be desirable in certain applications to periodically distort adjacent surfaces of both the dynode and anode electrodes in a similar manner such as is illustrated in FIGURE 10 to provide continuous focusing of electrons. Thus, the scope and breadth of the invention is to be determined from the following claims rather than from the above detailed description.

What is claimed is:

1. In a cross-field electron discharge device having a photosensitive cathode, a plurality of longitudinally aligned and longitudinally spaced dynode electrodes, an anode electrode longitudinally coextensive with and spaced from said cathode and dynodes, means for establishing an electric field from said anode to said cathode and dynode electrodes, and means for producing a unidirectional magnetic field in the space between the anode electrode and said cathode and dynode electrodes transversely of said electric field, the improvement comprising means associated with each dynode electrode for producing a change in the direction of at least part of said electric field with components thereof extending in opposite directions and parallel to said magnetic field whereby to limit transverse deviations of longitudinally moving electrons emitted by said cathode and said dynode electrodes.

2. The device according to claim 1 wherein said last named means comprises a segmented dynode surface proximate said anode, said segmented dynode surface comprising channels parallel to each other, each channel having a center line and opposed sides defining the width thereof.

3. The device according to claim 2 wherein each channel and the transverse components of electric :field supported thereby are symmetrical about the channel center line.

4. The device according to claim 3 wherein transverse components of electric .field supported by each channel are directed toward the center line of the channel from opposite sides of the center line, the magnitudes of said transverse components of electric `field increasing with distance from the center line toward the sides of said channel.

5. The device according to claim 1 wherein said last named means comprises a plurality of laterally spaced electrically conductive fins electrically connected to said dynodes and projecting therefrom toward said anode electrode, said fins being parallel to each other and normal to the plane of said dynodes whereby to define a plurality of longitudinally extending channels across the dynodes.

6. The device according to claim 5 wherein corresponding channels on the dynodes are longitudinally aligned and the widths of longitudinally aligned channels are equal.

7. The device according to claim 6 wherein said cathode electrode is segmented.

`8. The device according to claim 7 wherein said 1fins project equal distances from the associated electrodes and all of said channels have equal widths.

9. The device according to claim 2 wherein the surface of each channel is concavely curved.

10. The device according to claim 2 wherein each channel has two spaced flat surfaces intersecting a third fiat surface at right angles.

References Cited UNITED STATES PATENTS 2,163,700 6/1939 Ploke et al 313-95 2,664,515 =12/l953 Smith 313-105 2,762,928 9/1956 Wiley 313--105 X ROBERT SEGAL, Primary Examiner.

C. R. CAMPBELL, Assistant Examiner.