Time of flight energy measurement apparatus for an ion beam implanter

[0001] [Technical field to which the invention belongs] The present invention relates to an ion beam injection device, and in particular, a flight time processing technique for measuring the average kinetic energy of ions in an ion beam used for ion doping a semiconductor work piece in an ion beam injection device. With respect to the device using. [0002] [Conventional technology] The ion beam injection device is widely used in the injection process of a semiconductor wafer having a desired ion species. The ion beam injection device generates an ion beam containing positively charged ions of the desired nuclide. The ion beam collides with the exposed surface of the work piece of the semiconductor wafer, thereby "doping" or injecting the desired ions into the work piece surface. In some ion implanters, continuous implantation is performed, in which the work piece of a single semiconductor wafer is placed on a support in the implantation chamber. The support is directed so that the work piece is within the ion beamline, and the ion beam is repeatedly scanned over the work piece to inject the desired ion implantation amount. When the injection is complete, the work piece is removed from the support and another work piece is placed on the support. [0003] Other types of ion beam injectors use disc-type supports that are fitted with workpieces that rotate and move. The plurality of semiconductor workpieces are mounted on a disk-type support. The support is supported at the end of the ion beam injection device or in the injection chamber of the injection station. The rotation and movement of the support allows each of the plurality of workpieces to be exposed to an ion beam during the manufacturing process. [0004] It is extremely important to maintain the accuracy of both (a) the amount of ions implanted into the semiconductor wafer during the injection process and (b) the implantation depth of ion implantation on the machined surface when manufacturing acceptable products. Become. This tolerance between the ion implantation depth and the ion implantation amount, that is, the dose amount when manufacturing many semiconductor devices, is currently in the range of 1% in many uses. [0005] The ion implantation depth of the workpiece in the ion implanter is directly related to the energy of the ion beam. Therefore, precise control, measurement, and monitoring of the energy of the ion beam is required to achieve the desired injection depth. [0006] Conventional high-energy ion implanters control ion beam energy using a finite energy magnet (FEM). Ion beam energy using FEM<u style="single">To</u>By controlling, the energy of the ion beam needs to be selected and bent, i.e. the desired ion species constituting the ion beam is directed towards a workpiece supported by a support in the injection chamber. You can choose the strength of the magnetic field to move in the arched passage. [0007] This ion beam is directed via the FEM. This magnetic field moves the ions that make up the ion beam in the arched passage. The strength of the magnetic field is adjusted so that the desired ionic species with a particular moment is directed from the curved passage to the injection chamber where the workpiece is located. [0008] Unfortunately, however, the accuracy of the FEM method for calculating ion beam energy has significant drawbacks. That is, the bending strength of the ion beam is a function of the angle of incidence of the ions in the strength of the magnetic field of the FEM. The difference in the angle of incidence (2 to 3 °) of the ions is small, and when the magnetic field of FEM is applied, it has a significant effect on the bending strength of the ion beam. FEM energy read back Has been shown by numerical simulations and empirical tests to be correctable within ± 10% of the desired energy of the ion beam under certain conditions. [0009] What is needed is a more accurate ion beam energy measuring device for the ion implanter. Also, in such an energy measuring device, what is needed is that it is relatively inexpensive, durable, and that the ion beam can be updated rapidly and in real time. Moreover, what is needed is that the improvements used for the generation of ion beam currents in the injection device can be made without the need for extensive modifications of the injection device. [0010] [Problems to be Solved by the Invention] In view of such circumstances of the prior art, it is an object of the present invention to provide a flight time energy measuring device for measuring and controlling the energy of an ion beam and an ion beam injection device including the same. Another object of the present invention is to provide an ion beam energy measuring device and a method for measuring the average kinetic energy of ions. [0011] [Means for solving problems] In order to achieve the above object, the present invention has the configuration described in each claim. According to the present invention, an ion beam injection device for directing an ion beam to a plurality of workpieces of a semiconductor wafer is disclosed. This ion implanter includes an implantation station that forms an implantation chamber. The work piece is supported on a disc-shaped rotating and translating support located in the injection chamber. The ion implantation device further includes an ion source that generates an ion beam and a beam forming and directing device that forms an internal region that directs the ion beam from the ion source to the implantation station. A pump device for increasing and decreasing the pressure in this internal region is provided. A radio frequency (rf) ion accelerator is placed between the ion source and the injection station to accelerate the ions with the ion beam to achieve the desired high beam energy. [0012] This ion implanter includes a novel ion beam energy meter that utilizes time-of-flight processing techniques to determine the average kinetic energy of ions in the ion pulse of an ion beam. Energy measuring devices use radio frequency ion accelerators for high energy injection applications and take advantage of the fact that ion beams are a series of beam packets or pulse trains. The flight time processing technique is effectively applied by the energy measuring device to accurately determine the average ion energy in the ion beam. The timing circuit of the energy measuring device determines the time t for a pulse (including a plurality of ions) traveling a predetermined distance between two sensors. Importantly, the energy measuring device is used to roughly estimate the energy of the ions in the ion beam generated by the FEM, ensuring that the pulse elapsed time determined by the timing circuit is accurate. [0013] The average velocity of a bundle of ions or an ion pulse is determined by a conversion circuit using the following equation. v (pulse) = d / t Here, d is the distance between the first and second sensors, and t is the elapsed time for the pulse to move from the first sensor to the second sensor. [0014] Once the pulse velocity is determined, the ion energy E is E (ion) = (1/2) mv²Where v is the velocity of the pulse and m is the mass of the ions in the pulse. The mass of ions of the desired injected ion species is known exactly. [0015] The energy measuring device includes the first and second sensors arranged adjacent to the ion beam and separated by a predetermined distance, and the second sensor is downstream of the first sensor. The first sensor generates a signal as a pulse of an ion beam passing through the first sensor, and the second sensor generates a signal as a pulse of an ion beam passing through the second sensor. [0016] In the first embodiment of the energy measuring device, the device utilizes the fact that the radio frequency ion accelerator generates a series of beam pulses at a constant frequency, and as a result, the interval between adjacent pulses in the beam line is known. It becomes. The timing circuit of the energy measuring device is between the first pulse passing through the first sensor and the second pulse passing through the second sensor.<u style="single">delay</u>Determine the time Δt. Approximate pulse rate from FEM<u style="single">value</u>Once the interval between continuous pulses is known, the first pulse that moves the distance between the first and second sensors<u style="single">Progress</u>The time t is calculated by the timing circuit as shown in the following equation. [0017] t = [N × T] + Δt here, N = number of pulses between 1st and 2nd sensors T = time interval between pulses As described above, the time interval T between pulses is known because the frequency of the pulse is known from the radio frequency ion accelerator. The number N of pulses between the first and second sensors is known based on the estimated energy value of FEM. [0018] The timing circuit determines Δt using the correlation of the digital waveforms received from the first and second sensors. When the time t of the pulse traveling the distance d between the first and second sensors is known accurately, the conversion circuit calculates the velocity of the ion pulse as follows. v (pulse) = d / t Here, d is the distance between the first and second sensors. [0019] In the second embodiment, the energy measurement device, Ion'ene the FEM in a slightly different way includes timing circuits using Energy estimate. In this form, energy estimation<u style="single">value</u>Is used to ensure that the flight time of a single pulse travels between the first and second sensors. The timing circuit includes a first timing differencer, a second timing differencer, and a delay circuit. [0020] The first timing diffifier is electrically connected to the first sensor and generates the first signal when it receives a signal from the first sensor. The second timing diffifier is electrically connected to the second sensor and generates a second signal when it receives a signal from the first sensor. [0021] [0021] The delay circuit is electrically connected to the first and second timing diffs, and inputs the first and second signals generated by the two diffs. The delay circuit also accepts data values for approximate measurements of ion beam energy. This estimated energy measurement data is accepted from the first sensor on the upstream side of the FEM. The delay circuit transforms the approximate energy measurement data and uses the accelerated mass of the ion species to generate the delay time for the ion beam pulse. [0022] For the selected ion pulse, the timing circuit uses the first and second signals corresponding to the selected ion pulse, and the course of the selected pulse that moves a predetermined distance between the first and second sensors. Calculate the time. [0023] The energy measuring device further includes a conversion circuit for converting the elapsed time for the selected ion pulse into a measured value of the energy of the ion beam. The energy measuring device of the present invention is useful as an ion beam dose monitor by providing an alternative means for measuring current, regardless of the Faraday cage located behind the support of the injection chamber. [0024] The internal region formed by the beam forming and directing device on the upstream side of the injection chamber is in a state of being constantly depressurized during injection. The amplitude of the signal from the energy measuring device depends on the level of the ion beam current. Since the energy measuring device is located in the internal region of the ion implanter having a relatively constant pressure, the ion beam current is downstream of the energy measuring device regardless of the collision that neutralizes the charge generated in the region of the beamline. Accurate measurements can be made using a sensor located in a location where the side injection chamber pressure can be higher. In other parts of the ion implanter, such as the implantation station, the gas pressure changes, electrically neutralizing the ion beam and causing an error in the Faraday gauge's current measurement. [0025] Conventional ion implanters are provided with an opening in the support to expose the Faraday gauge to the ion beam for each revolution of the support. This Faraday gauge reads the ion beam current for each revolution of the support. The energy measuring device of the present invention can scan the energy and current of an ion beam faster than when using a Faraday gauge placed after the rotating support. [0026] In one operating mode of the energy measuring device of the present invention, the first and second sensors consist of inductive sensors. The inductive first and second sensors are placed adjacent to the beamline and at a predetermined distance. The first inductive sensor produces a signal as the pulse passes through the second sensor. [0027] In the second operating mode of the energy measuring device of the present invention, the first and second sensors are not inductive sensors but capacitive sensors. Like the inductive sensor, the capacitive sensor is also located at a predetermined distance along the ion beam line. [0028] These objectives, features, and effects of the invention will be better understood from the detailed description of preferred embodiments of the invention described in connection with the drawings. [0029] BEST MODE FOR CARRYING OUT THE INVENTION<u style="single">Structure of ion beam injection device 10</u>The ion beam injection device is shown by reference number 10 in FIG. The injection device 10 generates an ion beam 14 having a beam energy in the range of 10 to 5000 KeV. The injection device 10 includes an ion source 12 and supplies ions that form an ion beam 14 that passes through a beam path to the injection station or end station 16. Further, the injection device 10 uses the radio frequency ion accelerator 18 to accelerate the ions in the ion beam 14 at an appropriate high speed to achieve the desired ion beam energy. Suitable ion accelerators 18 for use in high energy injection devices are disclosed in US Pat. No. 4,667,111, granted to Glavish et al. And assigned to the assignee of the present invention. This U.S. patent is incorporated herein by reference. [0030] In a high energy injection device such as the injection device 10, the ion beam 14 is analyzed as one train, i.e. a series of individual pulses, where each pulse is 10.¹³It consists of a large number of ions having about one ion. The number of ion pulses of the ion beam 14 is outlined in FIGS. 3 to 5 as P0, P1, P3 ...... P (n-1), Pn, P (n + 1). [0031] In FIG. 2, a control electronic device (shown approximately 20) is provided to monitor and control the amount of ion implantation received by the plurality of semiconductor wafer workpieces 21 in the injection region of the injection station 16 or in the injection chamber 22. Has been done. The operator's input to the control electronic device 20 is executed via the user operation panel 67. [0032] The ion source housing 12 generates an ion beam that collides with the wafer workpiece 21 arranged on the disk 90 that can rotate and translate in the injection chamber 22. The ions in the ion beam 14 tend to diverge as the beam passes the distance between the ion source 12 and the injection station 16. The ion source 12 includes a plasma chamber 28 that forms an internal region into which the material of the ion source is injected. The material of the ion source may include a gas that can be ionized or a vaporized ion source material. [0033] The fixed-shaped ion source material injected into the plasma chamber 28 is placed in the evaporator. If a wafer material for an n-type impurity semiconductor is desired, boron (B), gallium (Ga), or indium (In) is used. Gallium or indium is a solidified ion source material, while boron, as a gas, is generally boron trifluoride (BF).₃) Or diborane (B₂H₆) Is injected into the plasma chamber 28. The reason is that the vapor pressure of boron is so low that it is not possible to bring it to a usable pressure simply by heating solid boron. [0034] When producing a P-type impurity semiconductor material, antimony (Sb), arsenic (As), and (P) phosphorus are selected as the solid ion source material. The energy is applied to the ion source material to generate positively charged ions in the plasma chamber 28. As can be seen in FIG. 2, positively charged ions exit the plasma chamber 28 through the oval arc slit 29 in the cover plate 30 that covers the opening side of the plasma chamber 28. [0035] During the manufacturing operation, i.e., when the semiconductor wafer workpiece 21 is collided by the ion beam 14 and thereby ion-implanted, the ion beam 14 extends from the ion source 12 to the injection chamber 22 and the beam forming and directing device 50. Passes through an evacuated passage through an internal region 52 (see FIG. 1) formed and evacuated by. Exhaust of the internal region 52 forming the beam passage during manufacturing operation is performed by a pressure regulating system 55 including a pair of vacuum pumps 31. [0036] The ions in the plasma chamber 28 are drawn out through the arc slit 29 of the plasma chamber cover plate 30 and formed in the ion beam 14 passing through the distance between the ion source 12 and the injection station 16 by the beam forming / directing device 50. To. The beam forming / directing device 50 includes a mass spectrometric or decomposing magnet 24, an rf (radio frequency) ion accelerator 18, a final energy magnet (FEM) 32 and a set of electrodes 34. A set of electrodes 34 draws ions from the inside of the plasma chamber and accelerates the ions within a defined region by the mass spectrometric magnet 32. [0037] The mass spectrometric magnet 24 is supported within the mass spectrometric magnet housing, and the ion beam passage is restricted by the aluminum beam guide 26 through this magnet region. Only ions with an appropriate mass-to-charge ratio to reach the ion implantation station 16 are produced. Ionization of the ion source material in the plasma chamber 28 produces a positively charged ionic species with a predetermined atomic mass. However, in addition to the desired ion species, the ionization process produces ions in proportions that include something other than the appropriate atomic mass. Ions with an atomic mass greater than or equal to an appropriate atomic mass are not suitable for ion implantation. [0038] The magnetic field generated by the mass spectrometric magnet 24 causes the ions in the ion beam 14 to move in a curved orbit. In the magnetic field achieved by the control electronic device 20, only ions having an atomic mass equal to the atomic mass of the desired ion species pass through the curved beam passage to the injection chamber 22 of the injection station. [0039] Ions with the desired atomic mass exit the mass spectrometric magnet 24 and the modules 18a, 18b, 18c, 18d, 18e of the rf ion accelerator 18. Accelerates to high speed. The acceleration of the ions to the desired energy level is controlled by the rf controller 19. What is important in the action of the accelerator 18 is that the ion beam 14 has P0, P1, P2, P3 ...... P (n-1), Pn, P (n + 1) in FIG. It is composed of a series of ion pulses outlined. [0040] For energy analysis, the beam 14 can be thought of as a series of separate ion pulses. Each ion pulse P0, P1, P2, P3 ...... P (n-1), Pn, P (n + 1) is 10¹³It is composed of a large number of individual ions. Further, from the characteristics of the ion accelerator 18, the frequency of the ion pulse is a predetermined and known value. [0041] A typical frequency value for an ion accelerator is f = 13.56 MHz. This means that the time interval T between successive pulses, that is, between P0 and P1, is T = 1 / f = (1 / 13.56) MHz = 73.75 (nsec.) Means. [0042] After being accelerated by the ion accelerator 18, the beam 14 passes through the magnetic field generated by the operating FEM 32 under the control of the FEM control circuit 40 (scheduled in FIG. 2). The FEM 32 is supported within the housing 33 and includes an aluminum beam guide 42 through which the ion beam 14 passes, such as the mass spectrometric magnet 24. [0043] The ions that make up the ion beam 14 move from the ion source 12 in the magnetic field set by the mass spectrometric magnet 32. The strength and direction of the magnetic field generated by the FEM32 is controlled by the FEM control circuit of the control electronics 20 to regulate the current through the magnetic field windings. FEM32 gives an approximation of the ion beam energy and the magnetic energy required to bend the beam along a suitable arched passage so that the energy of the beam 14 is directed to the target workpiece 21 in the injection chamber 22. The beam energy becomes proportional to. Therefore, FEM32 controls the ion energy. In general, the FEM energy control is accurately within ± 10% of the desired ion beam energy. [0044] Part of the beam forming / directing device 50 located downstream from the analysis magnet 32 is a quadrupole assembly 70, an analysis plate 80, a rotating Faraday flag or cup 72, and an ion beam neutralizer 74. The quadrupole assembly 70 includes a set of electrodes arranged around the ion beam 14 and is selectively activated by the control electronic device 20 to adjust the height position of the ion beam 14. The quadrupole assembly 70 is supported within the injection device housing 75. [0045] An analysis plate 80 (see FIG. 1) is connected to the end of the quadrupole assembly 70 facing the analysis magnet 32. The Faraday flag 72 is rotatably connected to the housing 75 and is rotatable to a position where the ion beams intersect to measure the characteristics of the ion beam 14, and when this measurement is satisfactory, the beamline is rocked. It is moved so that the injection chamber 22 does not interfere with the injection into the wafer work piece. [0046] The analysis plate 80 is composed of vitreous graphite and forms an elongated opening through which the ions in the ion beam 14 pass. Like the analytical magnet 24, the FEM 32 functions to remove unwanted ion species from the ion beam 14. [0047] As described above, the strength and direction of the FEM magnetic field is established by the FEM control circuit 40 and the control electronic device 20, whereby only ions having an atomic mass equal to the atomic mass of the desired ion species are predetermined. It reaches the injection station 16 through the desired beam passage. The unfavorable ionic species are greater than or less than the desired ionic atomic mass and are heavily deflected to collide with the slit boundaries formed by the aluminum beam guide 42 or analysis plate 80. [0048] The beam forming / directing device 50 also includes a beam neutralizer, which is usually referred to in the art as an electronic shower. The ions drawn from the plasma chamber 28 are positively charged. If the positively charged ions are not neutralized prior to wafer injection, the doped wafer will exhibit a net positive charge. Such a net positive charge on the wafer workpiece has undesired properties. The electrons generated by the beam neutralizer 74 are wiped out on the downstream side along the ion beam line, and a neutral space charge density is given to the downstream side of the ion beam neutralizer 74. [0049] Upstream of the ion beam neutralizer 74 is a magnetic repeller 82, which contains a permanent magnet to prevent backflow of electrons from the neutralizer 74. The neutralizer 74 and the repera 82 are formed as an integral unit and are supported by a common base 85 of the injection housing 75. [0050] Suitable ion neutralizers and magnetic repellers are disclosed in US Pat. No. 5,691,537, which was granted to Chen et al. And assigned to the assignee of the invention. This US patent is included herein for reference. [0051] The downstream end of the beam neutralizer 74 is adjacent to an injection chamber 22 in which ions are injected into the semiconductor wafer workpiece 21. A work piece support 90 of a disk-shaped semiconductor wafer is supported in the injection chamber. The wafer work piece 21 to be processed is located near the outer edge of the wafer support 90, which is rotated by a motor 92 at a constant angular velocity. The output shaft of the motor 92 is connected to the drive shaft 94 of the support by a belt 96. Alternatively, the drive shaft 94 of the wafer support 90 may be directly connected to the output shaft of the motor 92. The ion beam 14 collides with the wafer work piece when the wafer work piece rotates at a speed of about 1200 rpm in the circular passage. The step motor 98 drives the reed screw 99 to displace the support 90 in the vertical direction (as indicated by arrow A in FIG. 2). [0052] As a result, a large number of rows of semiconductor wafers can be ion-implanted during the manufacturing operation. The work piece 21 is carried in and out through the load lock chamber so that the beamline is kept in a vacuum state during the loading and unloading of the wafer. The injection station 16 is rotatable with respect to the beam neutralizer housing 75 by a telescopic bellows 100 (see FIG. 1). The ability to swivel the injection station 16 allows the angle of incidence of the ion beam to be adjusted when the ion beam collides with the wafer in the injection chamber 22. [0053] The Faraday gauge 110 is attached to the rear side of the workpiece support 90 and passes through a slot 112 formed in the support 90, and the ion beam current I in the Faraday gauge_f Is used to measure. [0054] The control electronic device 20 includes a dose amount control circuit 66 that controls the ion dose amount and the ion implantation depth received by the workpiece 21. The dose amount control circuit 66 applies the pressure P related to the input in the injection chamber from the ion gauge 114 in the injection chamber 22 and the ion beam current I from the Faraday gauge 110 (FIG. 2)._f Accept each. In order to control the dose amount of the workpiece 21, the dose amount control circuit 66 controls the motor control device 68 to adjust each speed and vertical movement of the support 90. In order to control the ion implantation depth of the workpiece 21, the dose amount control circuit 66 controls the rf control circuit 19 to adjust the energy of the ion beam 14. [0055]<u style="single">First Embodiment of Flight Time Energy Measuring Device 200</u>A novel first embodiment of the ion beam flight time energy measuring device or assembly in the ion beam injecting device 10 is outlined in FIG. 3 with reference number 200. The flight time energy measuring device 200 has a function of continuously measuring the energy of the ion beam 14. As a result, the ion implantation depth can be accurately controlled by the control electronic device 20. Generally, in the prior art, the "ion beam energy" is not actually measured, whereas the energy measuring device 200 has a function of constantly measuring the average kinetic energy of ions of a desired ion type in the ion beam 14. Have. [0056] The average kinetic energy of the ions is calculated based on the average velocity of the ion pulses when passing a predetermined distance d between the first and second sensors 210 and 220 of the device 200. The energy measuring device 200 of the present invention measures the average kinetic energy of the ion beam 14 about 20 times per second. [0057] The ion beam energy measurement provided in the energy measuring device 200 of the present invention can be updated almost instantly, and is timed with respect to an ion beam of 1 megaelectronvolt (MeV) boron in 1 nanosecond (nsec). The resolution is within ± 0.5% of the true energy value of one ion in the ion beam 14. Further, since the measurement of the ion beam energy measured by the energy measuring device 200 is performed in the internal region of the ion implanting device 10 in which the pressure hardly changes during the manufacturing operation, the signal of the flight time sensor is the ion beam current, that is, the ion beam current. Faraday Current I_f Greatly useful as a separate pressure monitor (scheduled in Figure 2). [0058] [0058] The ion beam measuring device 200 accurately models a high-energy ion beam (eg, 90 KeV or higher) as a series of pulses roughly represented as P0, P1, P2, P3, based on suitable time-of-flight (TOF) technology. Can be converted. This TOF technology determines the velocity of the ion pulse during the elapsed time for the pulse to pass a predetermined distance between the first and second sensors 210, 220. The sensors 210, 220 are arranged at a predetermined distance (for example, 25 cm) and are arranged adjacent to the ion beam line 14. [0059] Advantageously, the ion beam measuring device 200 accepts an approximate value of beam energy (within ± 10% of true beam energy) from the FEM control circuit 40. The measuring device 200 takes advantage of the fact that the ion accelerator 18 generates a series of beam pulses of constant frequency. For a 2.5 MeV boron ion beam, the typical frequency F is 13.56 MHz. Thus, the time interval between adjacent pulses in the ion beamline is the reciprocal of the frequency. That is, T = 1 / F = 1 / 13.56 MHz = 73.75 nsec. In addition to the first and second sensors 210 and 220, the measuring device 200 includes an energy measuring circuit 202. This circuit 202 includes a timing circuit 204 for determining the time t for the ion pulse to pass the distance d between the first and second sensors 210 and 220, and a conversion circuit 206 for determining the energy of the ions in the ion beam 14. And is included. [0060] The timing circuit 204 first determines the delay time Δt between the first pulse passing through the first sensor 210 and the second pulse passing through the second sensor 220. That is, looking at FIGS. 3 and 4, the time t₁ So, the pulse P0 is passing through the first sensor and the pulse Pn is approaching the second sensor 220. Time t₂ Then, the pulse Pn passes through the second sensor 220. The delay time Δt is equal to the time it takes for the pulse Pn to pass the labeled Δd distance in FIG. That is, Δt = t₂ -t₁ Is. [0061] Operationally, a dual-channel digital oscilloscope 230 is used, when pulses P0, P1, P2 pass the first sensor 210 (see Figure 8 (a)), and pulses P (n-1), P ( When n) and P (n + 1) pass through the second sensor, a digitized waveform of the format shown in FIGS. 8 (a) and 8 (b) is generated. Digital data is a function of the delay time Δt<u style="single">Cross-correlation</u>Value (cross correlation) Used to calculate C (Δt). [0062] [Number 1]<img file="JP4432003B2_D0001.tif" />Here, f1 (t) is digital data representing a waveform signal generated by the pulse P0 passing through the first sensor 210, and f2 (t) is the digital data represented by the pulse Pn passing through the second sensor 220. It is digital data representing the generated waveform signal. t'is the time when the waveform signal generated by the pulse P0 starts, and t "is the time when the waveform signal generated by the pulse P0 ends. [0063] To find the best estimate of the delay time Δt, different values of delay time from 0.0nsec. To time T (time interval between pulses 73.75nsec.) Are assigned to the integrated value by the timing circuit 204.<u style="single">Cross-correlation</u>Maximize the value C (Δt) 0.0 ~ 73.75nsec. The value of the delay time Δt between is selected by the timing circuit 204 as the best estimate of the delay time. The procedure for determining this delay time has the advantage that it is largely unaffected by the particular shape of the waveform signal generated by the sensors 210, 220. In addition, this procedure can accurately determine the delay time even when the waveform signal has a low signal-to-noise ratio. In contrast, time delay measurements using the level crossing method tend to spoil the waveform signal due to slow rise time or low signal to noise ratio. [0064] By knowing the time interval between consecutive pulses based on the approximate value of the pulse rate from the FEM control circuit 40, the elapsed time t of the selected pulse P0 passing the distance between the first and second sensors is appropriate. It is calculated by the timing circuit 204 using the above algorithm. The algorithm described below that does not include the calibration part is as follows. [0065] t = t (offset) + Δt = [N × (1 / F)] + Δt = [N × T] + Δt Here, t (offset) is the offset time for the selected pulse to pass a distance less than a predetermined distance d between the first and second sensors, and N is between the first and second sensors. The expected integer-th pulse, F is the frequency of the ion pulse, and T is the time interval between the ion pulses. [0066] As mentioned above, the time interval T between pulses is known (T = 73.75nsec.) Because the frequency of the pulse is known from the rf ion accelerator 18. ). The number N of pulses between the first and second sensors is known based on the FEM energy estimation. In FIG. 9, a chart based on the experience of arsenic (+) ion beam, phosphorus (+) ion beam, and boron (+) ion beam is shown. For example, when a 500 KeV phosphorus ion beam is generated, the number of pulses N is N = 2. The number found in the chart is always below the next integer value, that is, if the chart N value is 2.8, then N = 2 is used as the N value. [0067] As shown in FIG. 9, the data corresponding to the empirical chart of N values can be generated by the timing circuit 204 and input into accessible memory. The timing circuit 204 accesses the data to determine an appropriate value of N and the desired beam energy according to the composition of the ion beam 14. [0068] Instead, the timing circuit 204 calculates the expected number N of pulses between the first and second sensors, directly knows the time interval T between successive ion pulses, and the pulses are FEM energy estimates.<u style="single">value</u>It is possible to determine the estimated time to pass a predetermined distance d based on. That is, FEM energy estimation<u style="single">value</u>Based on, the timing circuit 204 can calculate the approximate velocity of the ion pulse, thereby calculating the approximate time it takes for the pulse to pass a predetermined distance d between the first and second sensors 210, 220. [0069] When the elapsed time t of the pulse P0 passing through the distance between the first and second sensors is accurately known, the conversion circuit 206 calculates the velocity of the pulse P0 as follows. v (pulseP0) = d / t Here, d is the distance between the first and second sensors 210 and 220. [0070] Finally, the conversion circuit 206 calculates the energy of ion E (ion) at pulse P0 using the following equation. E (ion) = (1/2) m (ion) v²Here, m is the mass of ions of the desired ion species in the beam 14. [0071]<u style="single">Calibration procedure</u>A calibration procedure is required to ensure that the energy measuring device 200 measures the accurate energy of the ion beam 14. There are four calibration procedures to make effective use of the measuring device 200 either separately or in combination. The calibration procedure is (a) Balanced signal technology, (b) Technology for checking using the speed of light, (c) Techniques for inserting known delay lines and / or (d) Technology using a DC beam of known energy, All of these techniques, with the exception of the DC beam, use external signals incident on the two flight time sensors 210,220 from a separate signal generator. [0072] (a) Balanced signal technology uses signals that are separated with a balanced splitter between the two sensors 210,220. This signal is sent in the two sensors 210,220 along a cable of the same length. As a result, two signals are output from the two sensors 210 and 220 without delaying each other. If there is a delay between the two sensor outputs, it indicates that there is a timing imbalance between the two channels used to measure the individual signals. Therefore, the result of this "balance" measurement is a constant timing offset included in the algorithm used to calculate the measured average ion pulse or beam energy from multiple velocities. [0073] (b) The technique of confirming using the speed of light uses a signal propagating along a rigid coaxial cable at a speed close to the speed of light. This cable is inserted through two sensors 210, 220 and generates a pulse signal each time a signal propagating on the cable passes through each sensor. In this case, the time delay to be measured is Δt = d / c, where c is the speed of light in the coaxial cable. [0074] If the measured time differs from this value, it indicates that the distance d between the two sensors 210,220 has changed from the reference value, and this new value has been found from the calibration procedure for subsequent energy measurements. use. This method provides a means of a bench top system and can confirm the effective distance between the sensors before the flight time measuring device 200 is installed along the beamline of the ion implanter 10. it can. [0075] (c) The technique for inserting a known delay line is similar to the technique for measuring the speed of light described above, in which case a cable with a known delay time is placed between the two probes. (The exact cable delay time is easily and accurately measured using a network analyzer.) A signal is input to a cable with a delay line and the delay time between the output signals of the two sensors is measured. To. [0076] If the measured delay time is different from the expected delay time, it indicates that there is a timing imbalance between the two analysis channels. Then, a correction element is incorporated in the algorithm to calculate the energy from the delay time. [0077] (d) The DC beam calibration technique uses an ion beam accelerated by the DC voltage applied to the extraction electrode 34 of the ion source 12. The energy of such a beam is identified within the accuracy of the voltage applied to the extraction electrode 34. A small RF voltage is applied to the DC beam by a small RF voltage applied to one electrode in modules 18a, 18b, 18c, 18d, 18e of the ion accelerator 18. Modulation of a small RF signal on a DC beam does not change the average energy of the beam. By measuring a large number of DC beam energies with the flight time measuring device 200, the estimated delay time is t (expected) = d / v Here, d is the distance between the first and second sensors 210 and 220, and v is the velocity of the ion in the DC ion beam. [0078]<u style="single">Operating mode of the first sensor (inductive sensor)</u>A preferred first mode of operation of the ion beam energy measuring device 200 includes first and second inductive sensors 210,220 (see FIG. 6), which are electrically connected to the energy measuring circuit 202 (see FIG. 3). It is connected and, in particular, electrically connected to the digital oscilloscope 230. [0079] The first sensor 210 and the second sensor 220 are located at predetermined, known distances, eg, 25 cm, adjacent to the ion beam 14 on the downstream side of the FEM 32 and on the upstream side of the injection chamber 22. [0080] [0080] The inductive sensor 210 includes a conductor 314 that surrounds a core material 312 with high magnetic permeability. Preferably, the sensor 210 has a ring shape and is arranged such that the ion beam line 14 passes through the sensor as shown in FIG. The sensor 210 is fixed to the support member 218. [0081] The support 218 is fixed to the internal portion of the FEM housing 33 using bolts 219 to support the sensor 210 in an appropriate position with respect to the ion beam line 14. The second inductive sensor (not shown) is similarly bolted to the internal portion of the beam forming / directing device 50 on the downstream side of the first sensor 210. [0082] The inductive sensor has the advantage of not responding to the space charge neutralization effect as compared to the capacitive sensor. [0083]<u style="single">Operation mode of the second sensor (capacitive sensor)</u>A preferred first mode of operation of the ion beam energy measuring device 200 includes first and second capacitive sensors 310,320 (see FIG. 7), which are electrically connected to the energy measuring circuit 202. The first and second sensors 310 and 320 are structurally the same, and only the first sensor will be described in detail below. [0084] In FIG. 7, the first sensor 310 is composed of a J-shaped glass tube 312 made of a highly insulating material such as glass and has an arched upper portion 313. The sensor 310 is arranged such that the arched portion 313 partially surrounds the ion beam 14. The tube 312 has a longitudinal central hole 314 through which the lead conductor (metal core) 316 penetrates. One end of the lead conductor 316 terminates at the far end 317 of the arched upper portion 313, and the other end of the lead conductor is connected to the oscilloscope 230. The lead conductor 316 functions as the negative side plate or cathode of the capacitor, and the ion beam 14 functions as the positive side plate or anode. [0085] When the current of the ion beam 14 changes, the charge on the negative plate also changes. The lead conductor 316 is appropriately discharged through the energy measuring device 202. The capacitive pickup on the negative side of the first sensor 310 responds to the electric field of the ion beam 14. [0086] As seen in FIG. 4, the first capacitive sensor 310 is fixed to the support member 318. The support member 318 is fixed to the internal portion of the FEM housing 33 using bolts 319 to support the sensor 310 at an appropriate position with respect to the ion beam line 14. The second capacitive sensor 320 is similarly bolted to the internal portion of the beam forming / directing device 50 on the downstream side of the first sensor 310. [0087] The sensor of this embodiment has the advantage of being more sensitive than the inductive sensor. [0088]<u style="single">Flight time energy measuring device 400 (second embodiment)</u>A preferred second embodiment of the energy measuring device of the present invention is shown by reference numeral 400 in FIG. The timing circuit 404 includes a delay circuit 408 for calculating the approximate delay period for the ion pulse to pass the distance d between the first and second sensors 210, 220. The first and second sensors have the same shape as that of the above-described embodiment. The delay period calculated by the delay circuit is used to ensure that the timer 470 of the timing circuit 404 clocks a single ion pulse. [0089] The first sensor 210 and the second sensor 220 are placed 25 cm apart and adjacent to the ion beam 14. The first sensor 210 is located downstream of the FEM32, while the second sensor 220 is 25 cm from it. When an ion pulse, such as pulse P0, passes through the first sensor, the sensor produces a reliable zero-tuned bipolar signal. [0090] The first timing diffifier 430 preferably includes an operational amplifier such as the Fairchild μA741, which operational amplifier is configured as a zero-cross detector. When the output of the first sensor 210 crosses zero volt, the first timing diffifier 430 generates a signal. In this way, each time each pulse P0, P1, P2, P3, etc. passing through the beam passage passes through the first sensor, the first timing differencer 430 generates a start signal (see FIG. 3). The start signal generated by the first timing differencer 430 is input to the timer 470 for the pulse height converter of the timing circuit 404, and is input to the gate / delay generator of the delay circuit 408. [0091] Similarly, the second timing diffifier 440 has the same structure as the first timing 430. Each ion pulse passing through the beam passage 14 passes through the second sensor, and the second timing diffifier 440 generates a stop signal. The signal generated from the second timing differencer 440 is input to the and gate 460 of the delay circuit 408. The gate / delay generator 450 receives data from the FEM control circuit 40 which gives an approximate energy value of the ion beam 14. [0092] The gate / delay generator 450 uses an approximate measurement of ion beam energy from FEM32 to calculate the approximate offset time t (offset) for the ion pulse. That is, the approximate energy of the ion beam 14 reduced by a predetermined factor is given by the approximate value of how long it takes for the ion pulse to pass the distance d between the first and second sensors 210 and 220. [0093] For example, the estimated energy measurements of the FEM can be used to determine the estimated elapsed time t (est.) For the ion pulse to pass the distance d between the first and second sensors 210, 220. The gate / delay generator 450 converts the estimated beam energy E (approx.) To the velocity v (pulse) of the ion pulse. When the distance d between the first and second sensors 210 and 220 is known, the pulse velocity v (pulse) is t (est.) = d / v (pulse) The gate delay generator 450 is used to determine the estimated elapsed time t (est.). [0094] At that time, the estimated time t (est.) Decreases the offset time t (offset) by, for example, 10%. The offset time t (offset) can be regarded as the elapsed time for the pulse to pass a distance less than the distance d between the first and second sensors 210 and 220. This t (offset) specific underestimation prevents the selected ion pulse from being lost as the pulse passes through the second sensor 220. [0095] For the selected ion pulse, i.e. P0, the first timing diffifier 430 allows the pulse to time the first sensor 210.₁ A start signal is generated when passing at (the position shown in Fig. 3). The start signal is time t₁ Is input to the start input of timer 470 and starts the elapsed time for the selected pulse. This start signal is also time t₁ Is input to the gate / delay generator 450. The gate delay generator 450 delays the start signal by a time equal to the calculated approximate offset time t (offset). Then, a logic high signal considered to be a delayed start signal is output. The delayed start signal is timed to the first input of the Andgate 260.₁ Input with + t (offset) and latch this input to logic high. [0096] The first input of Andgate 460 is time t₁ Time t that occurs after switching to logic high with + t (offset)₂ In, the next stop signal generated by the second timing diffifier 440 corresponds to the selected ion pulse P0 passing through the second sensor. This time t₂ The stop signal generated in the above shifts the second input of the Andgate 460 to logic high. Since both gate inputs are high, the logic high stop signal is time t from and gate 460.₂ Is output, and is further connected to the stop input of timer 470. This stop signal causes the timer 470 to stop the elapsed time for the selected pulse. The elapsed time for the selected pulse P0 passing through the distance d between the first and second sensors 210 and 220 is t = t.₂ -<u style="single">t</u>_{<u style="single">1</u>}be equivalent to. [0097] Another way for the gate-delay generator 450 to delay the start signal for the selected ion pulse is to determine the offset time t (offset) = N × T using an empirical chart as shown in FIG. is there. Where N is the number of pulses that must be input to the gate / delay generator 450 before the generator output switches to logic high, and T is the time interval between consecutive ion pulses. [0098] As mentioned above, the data corresponding to the empirical chart shown in FIG. 9 is stored in memory and using the timing circuit 404, N, i.e. the gate delay generator 450.<u style="single">But</u>It is used to determine how many ion pulses to ignore before switching to logic high. [0099] For example, when a 500 KeV phosphorus ion beam is generated, the gate / delay generator 450 finds that the continuous radio frequency (RF) and the selected ion pulse P0 generated by the first sensor 210 have passed through the sensor 210. Subsequently, as a result of the ion pulses P1 and P2 passing through the sensor 210, the continuous radio frequency (RF) generated by the first sensor 210 or the count number N = 2 is waited for. [0100] After two cycles have passed, i.e., the two start signals following the selected start signal are generated by the first timing diffifier 230. Then, the gate / delay generator 450 switches to the logic high output. [0101] Assuming that the period between continuous ion pulses (rf cycles) is 73.75 nsec., The first timing diffifier 430 is an approximation.<u style="single">value</u>Generates a start signal of 73.75ns. The N = 2 cycle delay results in an offset time t (offset) = N × T = 2 × 73.75 = 147.5 nsec. After receiving the ion pulse start signal from the first timing diffifier 430. The gate delay generator 450 receives the logic after receiving two start signals following the selected start signal (corresponding to the selected ion pulse).<u style="single">Yes</u>Produces the output of. This pulse, counted by the gate-delay generator 450, can output a logic high from the and-gate 460 when it receives the next stop signal from the second timing diffifier 440. [0102] The conversion circuit 406 (see FIG. 5) includes a multi-channel analyzer that converts the elapsed time t of the selected ion pulse P0 into the ion energy E (ion) of the ion pulse P0. [0103] The velocity v (pulse) of the selected pulse P0 of the ion beam 14 is v (pulse) = d / t Here, d is the distance between the first and second sensors, and t is the elapsed time for the pulse to move from the first sensor 210 to the second sensor 220. Once the velocity v (pulse) of the pulse is determined, the energy E (ion) of the ions of this pulse is determined as set in the first embodiment described above. [0104] Although preferred embodiments of the present invention have been described above, those skilled in the art can make other modifications without departing from the scope of the present invention, and the present invention is within the scope of the claims. It is intended that all modifications and modifications contained are possible. [Simple explanation of drawings] FIG. 1 is a schematic plan view of the ion beam injection device of the present invention. 2 is a schematic perspective view of selected components of the ion beam injection device of FIG. 1. FIG. FIG. 3 is a schematic block diagram showing a selected circuit according to the first embodiment of the energy measuring device of the present invention. 4 is a schematic view showing an ion pulse constituting a part of an ion beam generated by the ion beam injection device of FIG. 1. FIG. FIG. 5 is a schematic block diagram of a selected circuit according to a second embodiment of the energy measuring device of the present invention. FIG. 6 is a schematic view showing an inductive sensor of the energy measuring device of the present invention. FIG. 7 is a schematic view showing a capacitive sensor of the energy measuring device of the present invention. FIG. 8 (a) shows the trajectory of a digital oscilloscope showing the waveform obtained from the continuous pulse passing through the first sensor, and 8 (b) shows the waveform obtained from the continuous pulse passing through the second sensor. It is a figure which shows the trajectory of the digital oscilloscope which shows. FIG. 9 is a diagram showing the relationship between the number of cycles of radio frequencies counted and the ion beam energy before the elapsed time of the ion pulse is completed when the first and second sensors are separated by 25 cm. is there. [Explanation of symbols] 10 Ion beam injection device 12 Ion source 14 Ion beam 16 injection station 18 Ion accelerator 21 Work piece 22 Injection chamber 32 FEM 40 FEM control circuit 200,400 energy measuring device 204,404 Timing circuit 206,406 Conversion circuit 210,310 1st sensor 220,320 2nd sensor 230 oscilloscope 316 lead conductor 408 Delay circuit 430 1st timing diff 440 Second timing diff