rus
Issue #5/2015
A.Dolotov, P.Konovalov, R.Nurtdinov
High-Current Photomultiplier Tube Based on Microchannel Plate for Detection of Subnanosecond Light Pulses
High-Current Photomultiplier Tube Based on Microchannel Plate for Detection of Subnanosecond Light Pulses
The results of the development of photomultiplier tube (PMT) based on microchannel plate (MCP) having subnanosecond time resolution and high range of anode current linearity are given. PMT is intended for detection of poor radiation pulse fluxes in UV, visible and near IR band (150 to 900 nm).
Теги: photocathode photomultipliers spectral response спектральная чувствительность фотокатода фотоэлектронные умножители
Introduction
Photoelectronics regularities form the basis of experimental equipment for the detection of nano – and subnanosecond light pulses, plasma study, identification of pions, kaons and protons and many other nuclear and physical experiments. The special requirements are specified for the photomultiplier tubes (PMT), which are the part of Cherenkov and scintillation detectors. Mainly, it applies to the time characteristics of PMT: transit time spread of the signal through PMT (jitter) δt, duration of rising edge τф, duration at half-height τ0,5 of pulse response characteristic (the pulse response characteristic is PMT response to the infinitely short pulse of light emission). Fast time characteristics allow operating with short pulses and this is necessary during the operation with fast scintillators, continuous deexcitation time of which is not more than 1 ns, with Cherenkov scintillators, which generate light pulses with the duration of about 10–11 s, with sensors, which are applied in broadband laser communication systems. Development of certain areas of experimental physics and wide implementation of physical research techniques in various fields of science and engineering stimulate the development of PMT manufacturing process. The constant improvement of time characteristics is typical for all constructions of developed PMTs of domestic and foreign manufacture.
Fast-response high-current PMTs based on discrete dynodes developed at FSUE "All-Russian Research Institute of Automatics named after N.L. Dukhov" have the time characteristics which meet the requirements of modern detection equipment – duration of pulse response characteristic at half-height is equal to 1 ns at the linear current of not less than 0.5 A (SNFT18M) and 0.5 ns at the linear current of not less than 0.15 A (SNFT20) [1]. However, their large overall dimensions, high cost, impossibility to operate in magnetic fields due to the distortion of electron motion paths, metal-glass frame and, as a result, low mechanical strength significantly narrow down the capabilities of their use.
PMTs based on microchannel plates (MCP) in comparison with PMTs based on secondary emission discrete dynodes have evident advantages. Firstly, they have much better time resolution. It is known that the shape of output pulse is determined by the transit time spread of photo – and secondary electrons connected with the spread of values and directions of their initial velocities, difference of path lengths and transit velocities. Decrease of the time spread of electron transit in PMT with MCP can be explained by the significant decrease of the total length of electron paths during their multiplication. The signal delay time in regular PMTs is close to 30 ns and in PMTs with MCP is less than 1 ns.
Secondly, PMTs based on MCP have smaller mass-dimensional characteristics: MCP diameter varies from several millimeters to 10 cm and their thickness is less than 1 millimeter.
Thirdly, such PMTs can operate in strong magnetic fields. In PMTs based on discrete dynodes the path of electron motion in inter-electrode spaces under the influence of magnetic field is noticeably distorted which results in the distortion of output pulse. Therefore, during the operation in strong magnetic fields one has to screen PMT. In the PMTs with MCP the influence of magnetic fields on electron motion is not significant. The reason is that the electrons will move strictly inside the plate channel in which they were generated despite the external magnetic field.
The fourth distinctive feature is the capability to create gate pulse. Most often, the interval "photocathode – MCP entrance" is used for pulse control. The cutoff of power supply of this interval transfers it to the close state and the voltage supply to the deenergized pulse space with the relevant polarity and amplitude results in PMT opening for the period of time which is equal to pulse duration. Of course, in PMTs based on discrete dynodes the attempts to use the pulse control with the help of control grid in cathode region were made; however, the photocathode itself had grid-type structure and it had negative impact on its properties.
The next property of PMT based on MCP consists in the reproduction of information concerning the spatial distribution of incident radiation with high time resolution. MCP has important advantage in comparison with the systems based on discrete dynodes – the amplification of photocurrent is performed in it and its spatial distribution is maintained. This property forms the basis for position-sensitive sensor operation.
This article represents the PMT based on microchannel plates which is produced in several variations: with one or two MCP, with multialkali photocathode with the spectral sensitivity range of 200 to 900 nm or CsI-photocathode which has the spectral sensitivity range of 150 to 230 nm. The main advantages of the devices in the series of analogous PMTs include the low duration of anode pulse at half-height in linear range of light characteristic of PMT upon photocathode exposure to light flash with the duration of about 0.1 ns (not more than 0.65 ns for PMT with two MCPs and not more than 0.5 ns for PMT with one MCP) as well as high value of linearity limit of light characteristic under pulse condition upon photocathode exposure to light flash with the duration at half-height of about 20 ns (not less than 0.3 A).
PMT Structure
The entrance window of PMT vacuum unit (Fig. 1) is made of quartz glass or MgF2, and photocathode with the diameter of 25 mm is formed on its interior surface. The input surface of microchannel plate (type MKPO 25–8) is located near the cathode window. If necessary, depending on the required amplification one or two MCP (in such case they are joined in chevron) is used in the construction. From MCP exit the electron flow gets on the anode connected with coaxial-line output, the wave impedance of which is 50 Ohm.
The vacuum unit is placed into the frame, inside of which the resistive divider is also placed (it ensures the supply of required difference of potentials to electrodes), and encapsulated in compound. The connector with the wave impedance of 50 Ohm for output signal release (N-type of SMA), connector for the supply of feed voltage (SHV-connector, 5 kV) and terminal for the connection of ground wire are located on the frame end surface (Fig. 2).
Main Parameters and Characteristics of PMT
Light Sensitivity of Photocathode and Range of Spectral Sensitivity
Light sensitivity of multialkali photocathode lies within the range 350–500 µA/lm. The light source of A type (photometric incandescent electric lamp of type SIS-40–100) was used during the measurements. The typical spectral sensitivity of photocathode Sλ (mA/W) within the range of incident radiation wavelengths of 200–950 nm is shown in Fig. 3; the test results are specified in Table.
Limit of Linearity Characteristic Light
When determining the limit of light characteristic linearity, the oscilloscope TDS-3052В, laser DTL-419QT (produced by Laser-Compact), photodetector SE518 and set of light filters were used. The laser pulses were supplied to the entrance window of PMT and photodetector placed into the light-proof chamber with the interval of 2–5 s through the semitransparent beam-splitting mirror. The electric pulses entered on the entrances 1 and 2 of oscilloscope through attenuators SDNR14–02 (1:5) from PMT and photodetector exits and the amplitudes of output pulses were registered (F1 is the amplitude of voltage pulse from photodetector exit; H1 is the amplitude of voltage pulse on PMT exit; k is the attenuator division ratio). The radiation flux was varied with the help of light filters. In order to determine the value of light characteristic deviation from linearity δ, the radiation flux attenuation factor n1, which can be calculated as follows:
,
and factor of anode current pulse amplitude attenuation n2, which can be calculated as follows:
.
were used. The indices 1 and 2 correspond to the different levels of radiation flux.
Then, the deviation from linearity was calculated according to the following formula:
.
The amplitude of PMT anode current (linearity limit) Iа for the obtained value δ depended on the obtained values according to the expression:
,
where Rl is the load resistance.
The results of measurements of light characteristic linearity limit in pulse condition of PMT with two MCPs upon the photocathode exposure to single light pulse with the duration at half-height of 6 ns are given in Fig. 4. As is seen, upon the deviation from linearity reaching –15% the amplitude of PMT anode current was 3.7 A. Fig. 5 demonstrates the results of measurements of light characteristic linearity limit in pulse condition of PMT with two MCPs upon the photocathode exposure to single light pulse with the duration at half-height of 20 ns. With the same deviation from linearity (–15%) the amplitude of PMT anode current was 1.15 A.
Time Resolution
In order to measure the duration of anode pulse at half-height in the linear range of PMT light characteristic, the photocathode area was completely exposed to the flux of picosecond generator of light pulses PLP-10 (Hamamatsu Photonics) with the duration at half-height of ~ 0.1 ns. In order to register the pulses, the oscilloscope DSO7052В was used. The oscillograms of output signals from both devices – PMT with one and two MCP are shown in Fig. 6 and 7 respectively.
The time resolution Т05 can be calculated according to the formula:
,
where Твых is the duration at half-height of PMT output pulse; Ткр is the duration at half-height of pulse characteristic of electric signal recording channel; Ти is the duration at half-height of radiation pulse; Тосц is the duration at half-height of pulse characteristic of oscilloscope which can be calculated as follows (taking into account the dimensional ratio (ns·GHz) which is assumed to be equal to 0.35) for the oscilloscope pass band f:
.
The time resolution for PMT with MCP Т05 estimated as a result of tests is equal to:
for PMT with one MCP –0.16 ns at the output current of 0.26 A;
for PMT with two MCPs –0,25 ns at the output current of 0.31 A.
Dependence of Anode Pulse Duration on the Voltage in the Interval "MCP Exit – Anode"
The oscillograms of output signal of PMT with MCP at various voltages in the interval "MCP exit – anode" (1500, 2000 and 2500 V) were obtained. Analysis of the measurement results (Fig. 8–10) showed that the duration of anode pulse of PMT with MCP does not depend on the voltage in the interval "MCP exit – anode".
Amplification Factor
The measurements of dependence of amplification factor on MCP voltage were carried out using the laser (wavelength 527 nm; pulse duration 0.56 ns; pulse repetition rate 1 Hz). PMT photocathode was exposed to the laser radiation in the entire area. The pulse from PMT anode was detected by the first channel of oscilloscope DSO7052В. The second channel of oscilloscope detected the pulse from photodetector SFDE6, exposure to the light of which was performed together with the exposure to the light of PMT through the semitransparent beam-splitting mirror.
MCP voltage was selected so that initially the spectral anode sensitivity of PMT at the wavelength of 527 nm was equal to the spectral sensitivity of photocathode at the same wavelength. In other words, such conditions were selected when the amplification factor of microchannel plate was equal to one. Then, the ratio of amplitudes of the signals from PMT and photodetector was calculated upon the voltage increase on microchannel plate:
,
where i is the measurement number at the relevant voltage on MCP; Нi is the amplitude of signal from PMT anode; Fi is the amplitude of signal from photodetector SFDE6.
PMT amplification factor at the relevant voltage on MCP was calculated according to the formula:
.
As it was mentioned above, when i =1 we have the PMT amplification factor М1 = 1.
Dependence of the amplification factor М on the voltage on microchannel plate obtained for the PMT with two MCP is shown in Fig. 11.
Conclusions
When designing the individual elements of PMT by the specialists of FSUE "All-Russian Research Institute of Automatics named after N.L. Dukhov", the original engineering solutions were used. It allowed creating the PMT based on the microchannel plate which has high level of time and current characteristics. Obtained results showed that technical characteristics of the designed photomultipliers are at the same level with the characteristics of foreign analogs (PMT produced by Photek Limited) [2–4], and by some parameters (linearity limit of light characteristic) they even excel them.
Photoelectronics regularities form the basis of experimental equipment for the detection of nano – and subnanosecond light pulses, plasma study, identification of pions, kaons and protons and many other nuclear and physical experiments. The special requirements are specified for the photomultiplier tubes (PMT), which are the part of Cherenkov and scintillation detectors. Mainly, it applies to the time characteristics of PMT: transit time spread of the signal through PMT (jitter) δt, duration of rising edge τф, duration at half-height τ0,5 of pulse response characteristic (the pulse response characteristic is PMT response to the infinitely short pulse of light emission). Fast time characteristics allow operating with short pulses and this is necessary during the operation with fast scintillators, continuous deexcitation time of which is not more than 1 ns, with Cherenkov scintillators, which generate light pulses with the duration of about 10–11 s, with sensors, which are applied in broadband laser communication systems. Development of certain areas of experimental physics and wide implementation of physical research techniques in various fields of science and engineering stimulate the development of PMT manufacturing process. The constant improvement of time characteristics is typical for all constructions of developed PMTs of domestic and foreign manufacture.
Fast-response high-current PMTs based on discrete dynodes developed at FSUE "All-Russian Research Institute of Automatics named after N.L. Dukhov" have the time characteristics which meet the requirements of modern detection equipment – duration of pulse response characteristic at half-height is equal to 1 ns at the linear current of not less than 0.5 A (SNFT18M) and 0.5 ns at the linear current of not less than 0.15 A (SNFT20) [1]. However, their large overall dimensions, high cost, impossibility to operate in magnetic fields due to the distortion of electron motion paths, metal-glass frame and, as a result, low mechanical strength significantly narrow down the capabilities of their use.
PMTs based on microchannel plates (MCP) in comparison with PMTs based on secondary emission discrete dynodes have evident advantages. Firstly, they have much better time resolution. It is known that the shape of output pulse is determined by the transit time spread of photo – and secondary electrons connected with the spread of values and directions of their initial velocities, difference of path lengths and transit velocities. Decrease of the time spread of electron transit in PMT with MCP can be explained by the significant decrease of the total length of electron paths during their multiplication. The signal delay time in regular PMTs is close to 30 ns and in PMTs with MCP is less than 1 ns.
Secondly, PMTs based on MCP have smaller mass-dimensional characteristics: MCP diameter varies from several millimeters to 10 cm and their thickness is less than 1 millimeter.
Thirdly, such PMTs can operate in strong magnetic fields. In PMTs based on discrete dynodes the path of electron motion in inter-electrode spaces under the influence of magnetic field is noticeably distorted which results in the distortion of output pulse. Therefore, during the operation in strong magnetic fields one has to screen PMT. In the PMTs with MCP the influence of magnetic fields on electron motion is not significant. The reason is that the electrons will move strictly inside the plate channel in which they were generated despite the external magnetic field.
The fourth distinctive feature is the capability to create gate pulse. Most often, the interval "photocathode – MCP entrance" is used for pulse control. The cutoff of power supply of this interval transfers it to the close state and the voltage supply to the deenergized pulse space with the relevant polarity and amplitude results in PMT opening for the period of time which is equal to pulse duration. Of course, in PMTs based on discrete dynodes the attempts to use the pulse control with the help of control grid in cathode region were made; however, the photocathode itself had grid-type structure and it had negative impact on its properties.
The next property of PMT based on MCP consists in the reproduction of information concerning the spatial distribution of incident radiation with high time resolution. MCP has important advantage in comparison with the systems based on discrete dynodes – the amplification of photocurrent is performed in it and its spatial distribution is maintained. This property forms the basis for position-sensitive sensor operation.
This article represents the PMT based on microchannel plates which is produced in several variations: with one or two MCP, with multialkali photocathode with the spectral sensitivity range of 200 to 900 nm or CsI-photocathode which has the spectral sensitivity range of 150 to 230 nm. The main advantages of the devices in the series of analogous PMTs include the low duration of anode pulse at half-height in linear range of light characteristic of PMT upon photocathode exposure to light flash with the duration of about 0.1 ns (not more than 0.65 ns for PMT with two MCPs and not more than 0.5 ns for PMT with one MCP) as well as high value of linearity limit of light characteristic under pulse condition upon photocathode exposure to light flash with the duration at half-height of about 20 ns (not less than 0.3 A).
PMT Structure
The entrance window of PMT vacuum unit (Fig. 1) is made of quartz glass or MgF2, and photocathode with the diameter of 25 mm is formed on its interior surface. The input surface of microchannel plate (type MKPO 25–8) is located near the cathode window. If necessary, depending on the required amplification one or two MCP (in such case they are joined in chevron) is used in the construction. From MCP exit the electron flow gets on the anode connected with coaxial-line output, the wave impedance of which is 50 Ohm.
The vacuum unit is placed into the frame, inside of which the resistive divider is also placed (it ensures the supply of required difference of potentials to electrodes), and encapsulated in compound. The connector with the wave impedance of 50 Ohm for output signal release (N-type of SMA), connector for the supply of feed voltage (SHV-connector, 5 kV) and terminal for the connection of ground wire are located on the frame end surface (Fig. 2).
Main Parameters and Characteristics of PMT
Light Sensitivity of Photocathode and Range of Spectral Sensitivity
Light sensitivity of multialkali photocathode lies within the range 350–500 µA/lm. The light source of A type (photometric incandescent electric lamp of type SIS-40–100) was used during the measurements. The typical spectral sensitivity of photocathode Sλ (mA/W) within the range of incident radiation wavelengths of 200–950 nm is shown in Fig. 3; the test results are specified in Table.
Limit of Linearity Characteristic Light
When determining the limit of light characteristic linearity, the oscilloscope TDS-3052В, laser DTL-419QT (produced by Laser-Compact), photodetector SE518 and set of light filters were used. The laser pulses were supplied to the entrance window of PMT and photodetector placed into the light-proof chamber with the interval of 2–5 s through the semitransparent beam-splitting mirror. The electric pulses entered on the entrances 1 and 2 of oscilloscope through attenuators SDNR14–02 (1:5) from PMT and photodetector exits and the amplitudes of output pulses were registered (F1 is the amplitude of voltage pulse from photodetector exit; H1 is the amplitude of voltage pulse on PMT exit; k is the attenuator division ratio). The radiation flux was varied with the help of light filters. In order to determine the value of light characteristic deviation from linearity δ, the radiation flux attenuation factor n1, which can be calculated as follows:
,
and factor of anode current pulse amplitude attenuation n2, which can be calculated as follows:
.
were used. The indices 1 and 2 correspond to the different levels of radiation flux.
Then, the deviation from linearity was calculated according to the following formula:
.
The amplitude of PMT anode current (linearity limit) Iа for the obtained value δ depended on the obtained values according to the expression:
,
where Rl is the load resistance.
The results of measurements of light characteristic linearity limit in pulse condition of PMT with two MCPs upon the photocathode exposure to single light pulse with the duration at half-height of 6 ns are given in Fig. 4. As is seen, upon the deviation from linearity reaching –15% the amplitude of PMT anode current was 3.7 A. Fig. 5 demonstrates the results of measurements of light characteristic linearity limit in pulse condition of PMT with two MCPs upon the photocathode exposure to single light pulse with the duration at half-height of 20 ns. With the same deviation from linearity (–15%) the amplitude of PMT anode current was 1.15 A.
Time Resolution
In order to measure the duration of anode pulse at half-height in the linear range of PMT light characteristic, the photocathode area was completely exposed to the flux of picosecond generator of light pulses PLP-10 (Hamamatsu Photonics) with the duration at half-height of ~ 0.1 ns. In order to register the pulses, the oscilloscope DSO7052В was used. The oscillograms of output signals from both devices – PMT with one and two MCP are shown in Fig. 6 and 7 respectively.
The time resolution Т05 can be calculated according to the formula:
,
where Твых is the duration at half-height of PMT output pulse; Ткр is the duration at half-height of pulse characteristic of electric signal recording channel; Ти is the duration at half-height of radiation pulse; Тосц is the duration at half-height of pulse characteristic of oscilloscope which can be calculated as follows (taking into account the dimensional ratio (ns·GHz) which is assumed to be equal to 0.35) for the oscilloscope pass band f:
.
The time resolution for PMT with MCP Т05 estimated as a result of tests is equal to:
for PMT with one MCP –0.16 ns at the output current of 0.26 A;
for PMT with two MCPs –0,25 ns at the output current of 0.31 A.
Dependence of Anode Pulse Duration on the Voltage in the Interval "MCP Exit – Anode"
The oscillograms of output signal of PMT with MCP at various voltages in the interval "MCP exit – anode" (1500, 2000 and 2500 V) were obtained. Analysis of the measurement results (Fig. 8–10) showed that the duration of anode pulse of PMT with MCP does not depend on the voltage in the interval "MCP exit – anode".
Amplification Factor
The measurements of dependence of amplification factor on MCP voltage were carried out using the laser (wavelength 527 nm; pulse duration 0.56 ns; pulse repetition rate 1 Hz). PMT photocathode was exposed to the laser radiation in the entire area. The pulse from PMT anode was detected by the first channel of oscilloscope DSO7052В. The second channel of oscilloscope detected the pulse from photodetector SFDE6, exposure to the light of which was performed together with the exposure to the light of PMT through the semitransparent beam-splitting mirror.
MCP voltage was selected so that initially the spectral anode sensitivity of PMT at the wavelength of 527 nm was equal to the spectral sensitivity of photocathode at the same wavelength. In other words, such conditions were selected when the amplification factor of microchannel plate was equal to one. Then, the ratio of amplitudes of the signals from PMT and photodetector was calculated upon the voltage increase on microchannel plate:
,
where i is the measurement number at the relevant voltage on MCP; Нi is the amplitude of signal from PMT anode; Fi is the amplitude of signal from photodetector SFDE6.
PMT amplification factor at the relevant voltage on MCP was calculated according to the formula:
.
As it was mentioned above, when i =1 we have the PMT amplification factor М1 = 1.
Dependence of the amplification factor М on the voltage on microchannel plate obtained for the PMT with two MCP is shown in Fig. 11.
Conclusions
When designing the individual elements of PMT by the specialists of FSUE "All-Russian Research Institute of Automatics named after N.L. Dukhov", the original engineering solutions were used. It allowed creating the PMT based on the microchannel plate which has high level of time and current characteristics. Obtained results showed that technical characteristics of the designed photomultipliers are at the same level with the characteristics of foreign analogs (PMT produced by Photek Limited) [2–4], and by some parameters (linearity limit of light characteristic) they even excel them.
Readers feedback
