X-Ray Tube Electron Sources

[0001] The present application is a national stage application of PCT Application Number PCT/GB2010/050125, which was filed on Jan. 27, 2010, and relies on Great Britain Patent Application No. 0901338.4 filed on Jan. 28, 2009.

[0002] The present invention relates to X-ray tubes, to electron sources for X-ray tubes, and to X-ray imaging systems.

[0003] X-ray tubes include an electron source, which can be a thermionic emitter or a cold cathode source, some form of extraction device, such as a grid, which is arranged to control the extraction of electrons from the emitter, and an anode which produces the X-rays when impacted by the electrons. Examples of such systems are disclosed in U.S. Pat. No. 4,274,005 and U.S. Pat. No. 5,259,014.

[0004] With the increasing use of X-ray scanners, for example for medical and security purposes, it is becoming increasingly desirable to produce X-ray tubes which are relatively inexpensive and which have a long lifetime.

[0005] Accordingly the present invention provides an electron source for an X-ray scanner comprising an emitter support block. An electron-emitting region may be formed on the support block and arranged to emit electrons. An electrical connector may be arranged to connect a source of electric current to the electron-emitting region. Heating means may be arranged to heat the support block.

[0006] The present invention further provides a control system for an X-ray scanner. The system may comprise an input arranged to receive an input signal identifying which of a plurality of electron emitters is to be active. The system may be arranged to produce a plurality of outputs each arranged to control operation of one of the emitters. In some embodiments each of the outputs can be in a first state arranged to activate its respective emitter, a second state arranged to de-activate said emitter, or a third state arranged to put said emitter into a floating state.

[0007] The present invention further provides a control system for an X-ray scanner, the system comprising an input arranged to receive an input signal identifying which of a plurality of electron emitters is to be active, and to produce a plurality of outputs each arranged to control operation of one of the emitters. The system may further comprise output monitoring means arranged to monitor each of the outputs, and the monitoring means may be arranged to generate a feedback signal indicating if any of the outputs exceeds a predetermined threshold.

[0008] The present invention further provides a control system for an X-ray scanner, the system comprising an input arranged to receive an input signal identifying which of a plurality of electron emitters is to be active, and to produce a plurality of outputs each arranged to control operation of one of the emitters, wherein each of the outputs can be in a first state arranged to activate its respective emitter, and a second state arranged to de-activate said emitter. The system may further comprise blanking means arranged to fix all of the outputs in the second state irrespective of which state the input signal indicates they should nominally be in.

[0021] Referring to FIG. 1, an X-ray scanner 50 comprises an array of X-ray emitter units 25 arranged in an arc around a central scanner Z axis, and orientated so as to emit X-rays towards the scanner Z axis. A ring of sensors 52 is placed inside the emitters, directed inwards towards the scanner Z axis. The sensors 52 and emitter units 25 are offset from each other along the Z axis so that X-rays emitted from the emitter units pass by the sensors nearest to them, through the Z axis, and are detected by the sensors furthest from them.

[0022] Referring to FIGS. 2a and 2b each of the emitter units 25 includes an electron emitter element 116 which comprises an aluminium nitride (AlN) emitter support block 117 with low work function emitters 118 on its top surface 120 and platinum (Pt) heater element 122, on its bottom surface 121. The emitters 118 are formed from platinum-based ink coated with a highly emitting coating, and the heater element is also formed from Pt-based ink. The emitters cover discrete spaced apart areas on the surface 120 of the block 117, spaced along its length, and a connecting strip 123 of electrically conducting material extends from each of the emitters 118 around the side of the block 117 to its under side 121, where they form connector pads 124. The connecting strips are also spaced apart from each other, so that each emitter 118 is electrically isolated from the other emitters 118. Aluminium nitride (AlN) is a high thermal conductivity, strong, ceramic material and the thermal expansion coefficient of AlN is closely matched to that of platinum (Pt). Alumina (Al₂O₃) can also be used for the substrate as it has similar properties. These properties lead to the design of an integrated heater-electron emitter element suitable for use in X-ray tube applications.

[0023] AlN is a wide bandgap semiconductor material and a semiconductor injecting contact is formed between Pt and AlN. To reduce injected current that can occur at high operating temperatures, it is advantageous to replace the injecting contact with a blocking contact. This may be achieved, for example, by growing an aluminium oxide layer on the surface of the AlN substrate 120 prior to fabrication of the Pt metallisation. The provision of an oxide layer between the AlN and the Pt emitter forms a suitable blocking contact.

[0024] Alternatively, a number of other materials may be used in place of Pt, such as tungsten or nickel. Typically, such metals may be sintered into the ceramic during its firing process to give a robust hybrid device.

[0025] In some cases, it is advantageous to coat the metal on the AlN substrate with a second metal such as Ni. This can help to extend lifetime of the oxide emitter or control the resistance of the heater, for example.

[0026] To form the heater element 122 of this embodiment the Pt metal is formed into a track of 1-3 mm wide with a thickness of 10-200 microns to give a track resistance at room temperature in the range 5 to 200 ohms. It is advantageous to limit the heater voltage to below 100V to avoid electrical cross talk to the emitter pads 118 on the upper surface 120 of the substrate. By passing an electrical current through the track, the track will start to heat up and this thermal energy is dissipated directly into the AlN substrate. Due to the excellent thermal conductivity of AlN, the heating of the AlN is very uniform across the substrate, typically to within 10 to 20 degrees. Depending on the current flow and the ambient environment, stable substrate temperatures in excess of 1100 C can be achieved. Since both AlN and Pt are resistant to attack by oxygen, such temperatures can be achieved with the substrate in air. However, for X-ray tube applications, the substrate is typically heated in vacuum.

[0027] The emitter pads 118, heater element 122, and connecting strips 123, are applied to the surface of the substrate block 117 in the required pattern by printing. The connector pads 124 are formed by applying several layers of ink by means of multiple printing so that they are thicker than the connecting strips 123. The connectors at the ends of the heater element 122 are built up in the same way. The substrate block 117 is then heated to around 1100 C to sinter the ink into the surface of the substrate block 117. The emitter pads 118 are then coated with a Ba:Sr:Ca carbonate material in the form of an emulsion with an organic binder. This coating can be applied using electrophoretic deposition or silk screen printing. When the emitter element 116 is installed, before it is used, the heater element 122 is used to heat the substrate block 117 to over 700 C, which causes the carbonate material firstly to eject the organic binder material, and then to convert from the carbonate to the oxide form. This process is known as activation. The most active material remaining in the emitter pad coating is then barium oxide, and electron emission densities in excess of 1 mA/mm² can be achieved at operating temperatures of around 850-950 C.

[0028] Referring to FIG. 3, each emitter unit 25 comprises an emitter element 116, a circuit board 310 that provides the electrical control of the emitter element 116, a grid 312 arranged to control extraction of electrons from each of the emitter pads 118, and focusing elements 314 arranged to focus the beam of extracted electrons towards a target area on an anode 311. Typically, the underlying circuit board 310 will provide vacuum feedthrus for the control/power signals that are individually controlled on an emitter-by-emitter basis. The circuit board is best made of a material with low outgassing properties such as alumina ceramic.

[0029] The emitter element 116 is connected to the circuit board 310 by means of sprung connection elements 316. These provide physical support of the emitter element over the circuit board 310, and also each connection element 316 provides electrical connection between a respective one of the connector pads 124 on the emitter element 116 and a respective connector on the circuit board 310. Each connection element 316 comprises an upper tube 318 connected at its upper end to the emitter element so that it is in electrical contact with one of the connector pads 124, and a lower tube 320 of smaller diameter, mounted on the circuit board 310 with its lower end in electrical contact with the relevant contact on the circuit board 310. The upper end of the lower tube 320 is slidingly received within the lower end of the upper tube 318, and a coil spring 322 acts between the two tubes to locate them resiliently relative to each other, and therefore to locate the emitter element 116 resiliently relative to the circuit board 310.

[0030] The connector elements 316 provide electrical connection to the connector pads 124, and hence to the emitter pads 118, and mechanical connection to, and support of, the AlN substrate. Preferably the springs 322 will be made of tungsten although molybdenum or other materials may be used. These springs 322 flex according to the thermal expansion of the electron emitter assembly 116, providing a reliable interconnect method. The grid 312 and focussing elements 324 are less affected by thermal expansion and therefore provide a fixed location. The top of the emitter element 116 is kept at a fixed distance from the grid 312 by spacers in the form of sapphire spheres 317. Hence the emitter pads 118 are held stationary by being clamped against the grid 312, via the sapphire spacers 317, during any thermal expansion or contraction of the emitter assembly 116. The potential of each of the emitter pads 118 can therefore be switched between an emitting potential, which is lower than that of the grid 312 such that electrons will be extracted from the emitter 118 towards the grid 312, and a blocking, or non-emitting, potential, which is higher than that of the grid, so that electrons will tend not to leave the surface of the emitter 118, or if they do, will be attracted back towards the emitter.

[0031] Referring also to FIG. 4 the grid 312 is formed from a thin foil of tungsten. It extends over the upper surface 120 of the emitter element 116, and down past the sides of the emitter element 116, through the plane containing the lower surface 121 of the emitter block. It also extends down as far as the circuit board 310, passing through the plane including the front face of the circuit board, and the plane containing the rear face of the circuit board. The grid 312 includes a number of extraction areas 313 each of which extends over a respective one of the emitter pads 118 and has a series of narrow apertures 315 formed in it. The apertures 315 make up at least 50% of the area of the extraction areas. Each extraction area 313 covers an area approximately equal to the area of, and located directly above, the emitter pad 118. The areas of the grid 312 between the extraction areas are solid. The grid 312 therefore helps to focus the extracted electron beams from the individual emitter pads 118. The apertures 315 are formed using chemical etching of the tungsten foil. The grid 312 therefore forms an almost continuous layer over the emitter element 116, apart from the apertures 315. The sapphire spacers 317 maintain a fixed spatial relationship between the top surface 120 of the emitter element and the upper portion of the grid 312, and the side portions of the grid 312 are spaced from the emitter element 116 and the circuit board 310. The grid 312 therefore forms an effective heat shield enclosing the emitter element 116 on both sides and over its top surface. This partial enclosure reduces the radiation of heat from the emitter element. Other materials such as molybdenum can also be used for the grid 312. The grid 312 is connected to an electrical connector on the circuit board 310 so that its electrical potential can be controlled. The grid 312 is supported close to the emitter pads 118, with a gap between them of the order of 1 mm. This enables the extraction voltage (i.e. the difference in voltage between an active emitter pad 118 and the grid) to be kept low, for example below 200V while achieving beam currents in excess of 1 mA/mm².

[0032] The focusing elements 314 extend one along each side of the emitter element 116. Each focusing element 314 is mounted on isolating mountings 323 so as to be electrically isolated from the grid 312 and the emitter element 116. It includes a flat lower portion 324 that extends parallel to, and spaced from, the side portions of the grid 312, and a curved portion 326 that extends upwards from the lower portion 324 beyond the grid upper portion, over in a curved cross section, and back towards the grid 312, with its inner edge 328 extending along the length of the emitter, spaced from the grid 312 and approximately level, in the lateral direction, with the edge of the emitter pads 118. This leaves a gap between the two focusing elements 314 that is approximately equal in width to the emitter pads 118 and the apertured areas of the grid 312. The focusing elements 314 are both held at an electric potential that is negative with respect to the grid 312, and this causes an electric field that focuses in the lateral direction the electrons extracted from the emitters. The focusing elements form a further, outer heat shield, spaced from the grid 312, which further reduces the radiation of heat away from the emitter elements 116.

[0033] Referring to FIG. 3, a heat shield or reflector 330 is located between the emitter element 116 and the circuit board 310. In this embodiment, the heat shield 330 is formed from a mica sheet coated in a thin layer of gold. The addition of a titanium layer between the gold and the mica improves adhesion of the gold. The heat reflector 330 is supported on the sprung connection elements 316, which extend through holes in the reflector 330. Its coated upper surface is located close to, but spaced from, and facing, the lower heated side 121 of the AlN substrate. This reflects heat from the emitter element back towards it, and thereby improves the heater efficiency, reducing the loss of heat through radiative heat transfer. With the grid 312 enclosing the top and sides of the emitter element, and the shield 330 enclosing its under-side, the emitter element is surrounded on all four sides by heat shielding. Heat shielding can also be provided at the ends of the emitter element 116, but this is less important as the emitter elements are placed end-to-end in close proximity to each other. Silica can be used as an alternative substrate for the reflector, and other reflective materials such as Ti or multi-layer IR mirrors can also be used. Further similar reflectors can also be provided between the emitter element 116 and the grid 312.

[0034] Referring back to FIG. 1, the scanner is controlled by a control system which performs a number of functions represented by functional blocks in FIG. 1. A system control block 54 controls, and receives data from, an image display unit 56, an X-ray tube control block 58 and an image reconstruction block 60. The X-ray tube control block 58 controls a focus control block 62 which controls the potentials of the focus elements 314 in each of the emitter units 25, an emitter control block 64 which controls the potential of the individual emitter pads 118 in each emitter unit 25, and a high voltage supply 68 which provides the power to the anode 311 of each of the emitter blocks and the power to the emitter elements 118. The image reconstruction block 60 controls and receives data from a sensor control block 70 which in turn controls and receives data from the sensors 52.

[0035] Referring to FIG. 5, the circuit board 310 includes a high voltage push-pull output stage 500 for each of the emitter pads 118, arranged to provide a high voltage signal to it to control the emission of electrons from it. Each output stage 500 comprises a pair of transistors, in this case FETs, 502, 504 connected in series between the supply 506 and ground 508. The HV output 510 is connected between the two FETs. A drive input 512 is connected directly to the second FET 504 and via an XOR gate 514 and an inverter to the first FET 502. The XOR gate 514 has a second input en (the drive input being the first). This input en is usually low, so that the output of the XOR gate 514 matches the input, but can be used to provide further control as will be described further below with reference to FIG. 6. When the input signal goes low, the first FET 502 is switched on and connects the output 510 to the supply voltage and the second FET 504 is switched off and isolates the output from ground. The output voltage therefore rises quickly to the supply voltage. When the input goes low, the first FET is switched off and isolates the output 510 from the supply voltage and the second FET 504 is switched on and connects it to ground, so that the output voltage falls rapidly to zero. Therefore this output stage 500 allows the emitter to be switched on and off rapidly in a well controlled manner, so that the position of the source of the X-ray beam can be accurately controlled. If the input is at an intermediate level that is not high enough to turn on the second FET 504 or low enough to turn on the first FET 502, then both of the FETs are turned off and the output is in a floating tri-state condition, in which it is disconnected from the fixed potentials of the HV supply and ground and is free to fluctuate. This puts the emitter pad 118 into an electrically isolated state which inhibits electron extraction from the emitter pad 118.

[0036] Referring to FIG. 6, the emitter control block 64 of the control system provides a digital emitter control signal which is input to a number of emitter control devices 600, each of which is arranged to control the operation of 32 electron emitter pads in one of the X-ray emitter units 25. Each control device 600 receives as an input signal a serial digital signal Din which includes data indicating which of the emitters should be turned on and which turned off, and which should be in the floating state. This input signal is fed to a processor 601, and a number of shift registers 602, 604, 608, 614 which control and monitor the output of the output stages 500, of which there is one for each of the 32 controlled emitters.

[0037] The processor 601 is arranged to receive a control signal CTRL as well as the data signal Din and a clock signal SCLK, and to output a number of signals that control operation of the shift registers, and other functions of the device 600.

[0038] One of the registers is a data register 602, in the form of 32 bit serial-in-parallel-out (SIPO) shift register, and is arranged to receive the serial input signal Din, which includes data indicating the required state of each of the emitters 118 for a particular cycle, to load that data under control of a signal ld_dat from the processor 601 and a clock signal SCLK. It is arranged to output the 32 required states to the inputs of a parallel-in-parallel-out data register 604, which loads them under control of a clock signal XCLK. The data register 604 presents the loaded data at its parallel outputs to one of three inputs to respective NAND gates 606. Assuming for now that the other inputs to the NAND gates 606 are all high, the outputs of each NAND gate 606 will be low if its respective emitter 118 is to be active, and high if it is to be inactive. The output from each NAND gate 606 is fed to one input of an exclusive-OR (EOR) gate 609, the other input of which is arranged to receive a polarity signal POL. The output of each EOR gate 609 is input to a respective output stage 500, each of which is as shown in FIG. 5, which therefore provides the controlled HV output HVout to the emitter. The polarity signal POL allows the polarity of the system to be reversed. For example when the grid is fixed at −HV (or ground) potential then a positive voltage is needed on the emitter to turn the beam off. If, however, the emitter potential at −HV (or ground) then a negative potential is needed to turn the beam off. The XOR gate and POL input allows the circuit to be used in either configuration.

[0039] A tri-state register 608, in the form of a second 32 bit SIPO register, is arranged to receive the serial input signal Din which also includes data indicating which of the outputs should be set to the tri-state (or floating state) condition. This data is read from the input signal and loaded into the tri-state register 608 under the control of the signal ld_en and the clock signal SCLK. This data is then output in parallel to the respective output stages 500, with the output en being high if the output stage 500 is to be switched to the tri-state condition, and low if the output stage is to be set to the high or low level as determined by the output from the respective NAND gate 606. Referring back to FIG. 3, when the enable signal en is high, the emitter will be in the floating condition when the input signal to the output stage is low. Data from the serial input signal Din can therefore be used to set any one or more of the emitters 118 to the tri-state condition. This is useful, for example, during initial activation of the emitters 118, when all of the emitters are set to the tri-state condition, or if short circuits occurred affecting one or more emitters, for example connecting it to the grid, in which case setting the affected emitters to the floating state would allow the short circuit to be mitigated.

[0040] Each NAND gate 606 also has one input connected to a blanking signal BLA. Therefore if the blanking signal BLA is high, the output of the NAND gate 606 will be low regardless of the output from the data register 602. The blanking signal can therefore be used to set the outputs of any of the NAND gates to a blanked state, in which they are constant or at least independent of the input data, or an active state, in which they are controlled by the input data. A further chip select input CS is provided to all of the NAND gates and can be used to activate or de-activate the whole control chip 600.

[0041] Each HV output HVout is input to a respective comparator 612 which is arranged to compare it to a threshold signal VREF, and produce a feedback output indicative of whether the output drive signal is above or below the threshold. This feedback data, for all 32 output signals, is input to a parallel-in-serial-out feedback register 614, under control of a signal rd_fb from the processor 601, and the feedback register 614 converts it to a serial feedback output 616. This output 616 therefore indicates if any of the outputs is supplying excessive current, which can be used as an indication of, for example, a short circuit problem. The level of the reference signal VREF is set by the processor 601.

[0042] A serial output 618 is also provided from the data register 602 which is indicative of whether each of the output signals is nominally at the high or low level. These two serial outputs are multiplexed by a multiplexer 620, under the control of a multiplexing control signal mux from the processor 601, to produce a single serial digital output signal Dout. This allows the expected output values to be checked from 618, for example to check the programming of the device, and the actual values to be checked from 616 to check that the correct outputs have actually been achieved.

[0043] The control device 600 is arranged to operate in three different modes: a sequential access mode, a random access mode, and a non-scanning or reset mode. In the sequential access mode, the X-ray beam is scanned around the X-ray sources sequentially. Therefore in each full scan of all emitters, each control device will be active for a single period within the scan, and during that period, will activate each of the emitters it controls in sequence for respective activation periods. In the random access mode, the X-ray source is moved around the X-ray source array in a pseudo-random manner. Therefore, in each scan of all emitters, each control device will activate one of its emitters for one activation period, and then will be inactive for a number of activation periods while emitters controlled by other devices 600 are active, and will then be active again for a further activation period when another of its emitters is active. Some of the control inputs for the sequential access mode are shown in FIG. 7, and for the random access mode in FIG. 8. The signal marked ‘LOAD’ is not a specific signal, but is indicative of the times at which the device 600 is actively providing power to one of the emitters. As can be seen, in the sequential access mode, the blanking signal BLA is held high for a period covering successive activation periods while the emitters are activated in turn. Therefore the outputs track the data loaded in the data register 604. In the random access mode, the blanking signal is high only during the activation periods in which one of the emitters controlled by the device 600 is active. Therefore, for those activation periods, the outputs track the data in the data register 604. For intervening activation periods, the blanking signal is low, so the outputs are all turned off. This ensures that whatever processing is being carried out by the device 600 during those intervening periods does not affect the device's outputs and therefore does not affect the control of the emitters. This mode is suitable for scanning methods such as the random-access method described above, in which the emitter unit is active, in the sense of emitting electrons and hence X-rays, for a number of activation periods during a scan, but inactive for a number of intervening periods during which other emitter units are active.

[0044] Referring to FIG. 9, in the non-scanning mode, which is the default mode, the blanking signal is kept low, so the outputs are all maintained in the off condition and all of the emitters are inactive. This mode is used, for example, to enable calibration of the data acquisition system.

[0045] The format of the data input signal Din is a 5-byte programming pattern having the following format:

	MSB	Control word
		Status/data word 0 (MSB)
		Status/data word 1
		Status/data word 2
	LSB	Status/data word 3 (LSB)

The control word has a bit configuration such as:

MSB	7	1 = load data register	0 = no action
	6	1 = read status register	0 = read data register
	5	1 = set tri-state register	0 = no action
	4	1 = set BLA hi (Mode 1)	0 = normal action (Mode 2)
	3	1 = set BLA lo (Mode 3)	0 = normal action (Mode 2)
	2	Don't care
	1	Don't care
LSB	0	Don't care

[0046] Therefore one 5-byte input signal is required for each emitter activation period, and the signal indicates by means of the four data/status bytes which emitter is to be active, and by means of the control byte which mode the system is in. The emitter control block 64 sends a serial data input signal to a control device 600 for each of the emitter units 25 of the scanner so as to coordinate operation of all of the emitters in the scanner.

[0047] In this embodiment as shown in FIG. 6, the threshold voltage is generated by an on-chip DAC programmed using the lower 3 bits of the control word. The same threshold programming voltage is applied to all 32 output channels. In this case, the control word is assigned the following bit pattern:

MSB	7	1 = load data register	0 = no action
	6	1 = read status register	0 = read data register
	5	1 = set tri-state register	0 = no action
	4	1 = set BLA hi (Mode 1)	0 = normal action (Mode 2)
	3	1 = set BLA lo (Mode 3)	0 = normal action (Mode 2)
	2	1 = set threshold voltage	0 = no action
	1	Threshold voltage DAC bit 1 (MSB)
LSB	0	Threshold voltage DAC bit 0 (LSB)

[0048] In operation, an object to be scanned is passed along the Z axis, and the X-ray beam is generated by controlling the emitter pad potentials so that electrons from each of the emitter pads 118 in turn are directed at respective target positions on the anode 311 in turn, and the X-rays passing through the object from each X-ray source position in each unit detected by the sensors 52. As described above, for some applications the beam is arranged to scan along the emitter in discrete steps, and for some it is arranged to switch between the emitter pads 118 in a pseudo-random manner to spread the thermal load on the emitter. Data from the sensors 52 for each X-ray source point in the scan is recorded as a respective data set. The data set from each scan of the X-ray source position can be analysed to produce an image of a plane through the object. The beam is scanned repeatedly as the object passes along the Z axis so as to build up a three dimensional tomographic image of the entire object.

[0049] In an alternative embodiment the connector elements 316 of FIG. 3 are inverted. However, the advantage of the spring 322 being near to the circuit board and further from the emitter element 118 is that the upper tube 318 runs at high temperature and the spring 322 at low temperature. This affords a greater choice of spring materials since creeping of the spring is lower at lower temperatures.

[0050] As an alternative to the wraparound interconnects 124 of the embodiment of FIGS. 2a and 2b, through-hole Pt interconnects can be used which extend through holes in the AlN substrate 120 to connect the emitter pads 118 to connectors on the underside of the emitter element 116. In a further modification, a clip arrangement may be used to connect the electrical power source to the top surface of the AlN substrate.

[0051] It will be appreciated that alternative assembly methods can be used including welded assemblies, high temperature soldered assemblies and other mechanical connections such as press-studs and loop springs.

[0052] Referring to FIG. 10, in a further embodiment, the output stages of FIG. 5 are each replaced by an alternative output stage 700. In this embodiment, the output 710 is connected to the supply 706 via a resistance 702, and to ground 708 via an FET 704 which is switched on and off by the input signal on the input 712. (The XOR gate is omitted from the drawing for simplicity). While the input signal is high, the FET 704 is switched on and the output 710 is connected to ground. When the input signal goes low, the FET 704 is switched off, and the output is connected via the resistor 702 to the supply 706, switching the source on slowly. When the input signal goes high again, the FET 704 again connects the output 710 to ground, switching the source off more rapidly.

[0053] Referring to FIGS. 11a and 11b, in a further embodiment each emitter element 810 is formed from a ceramic substrate 812, in this case alumina (Al₂O₃) although AlN can again also be used, with individual spaced apart metal emitter pads 814 formed on its upper surface 816 by sputter coating. The emitter pads can be formed of any suitable metal, such as Ni, Pt or W, and they are covered with an active oxide layer to enhance electron emission as in the embodiment of FIGS. 2a and 2b. Patterning of the individual emitter pads 814 is achieved using shadow masks during sputter coating, although photolithographic methods can also be used.

[0054] On the opposite side 818 of the substrate from the emitter pads, a heating element in the form of a continuous conductive film 820 is applied, which in this case covers the whole of the rear side 818 of the emitter element 810. The heating element is also formed by means of sputter coating, and at each end of the emitter element, the conductive film is made thicker, by further sputter coating, to form contact areas 822, 824. Clearly since the substrate is electrically non-conducting, the heating element 820 is electrically isolated from the emitter pads, which in turn are electrically isolated from each other.

[0055] Referring to FIG. 12, each emitter element 810 is supported in a supporting heat shield structure 830, which is formed from two side elements 832, 834 and two cross members 836, 838 which extend between the side elements 832, 834 and hold them parallel to, and spaced apart from, each other. The emitter element 810 is supported between the upper (as shown in FIG. 9) edges of the side elements, parallel to the cross members 836, 838, with the emitter pads 814 facing outwards, i.e. upwards as shown, and a circuit card 840 is supported between the lower edges of the side elements 832, 834.

[0056] The side elements 832, 834 and the cross members 836, 838 are formed from silica plates, which are formed into interlocking shapes by laser cutting. These plates therefore interlock to form a stable mechanical structure. The silica material is coated on one side, the side facing the emitter element 810, with a high reflectance low emissivity material, such as Au or Ti. Alternatively the silica may be coated with a multi-layer infra-red mirror.

[0057] A series of connecting wires 842 each have one end connected to a respective one of the emitter pads 814, and extend around the outside of the heat shield structure 830, having their other ends connected to respective connectors on the circuit card 840. The interconnecting circuit card 840 is used to transfer signals from outside the vacuum envelope of the scanner, either directly through a hermetic seal or indirectly through a metal contact which engages with a hermetic electrical feedthru.

[0058] A grid 844, similar to that of FIG. 3, extends over the top of the emitter element 810 and down the sides of the heat shield structure 830, being spaced from the side elements 832, 834 to leave an insulating gap, through which the connecting wires 842 extend. The top part of the grid 844 is parallel to the upper surface 816 of the emitter element and spaced from it by a small distance of, in this case, about 1 mm. Focusing elements 846 are arranged on either side of the heat shield 830 and emitter 810. Each one extends along parallel to the side elements 832, 834, outside the grid 844 and spaced from it by a further insulating gap, and up over the side of the emitter element 810. Each focusing element has its upper edge extending part way over the emitter element 810, so that a focusing gap is left, between the focusing elements 846, which extends along the emitter over the emitter pads 814.

[0059] As with the embodiment of FIG. 3, the grid 844 and focusing elements 846 are connected to appropriate electrical potentials and serve, as well as their primary functions, as additional heat shields to reduce the radiation of heat away from the emitter element 810.