Imaging thermal ion mass and velocity analyzer

Transcription

1 An Introduction to Space Instrumentation, Edited by K. Oyama and C. Z. Cheng, Imaging thermal ion mass and velocity analyzer Andrew W. Yau, E. Peter King, Peter Amerl, Kaare Berg, Greg Enno, Andrew Howarth, Ivan Wevers, and Andrew White Department of Physics and Astronomy, University of Calgary, 2500 University Dr NW Calgary, Alberta T2N1N4, Canada The aim of an imaging thermal ion mass and velocity analyzer is to apply imaging techniques to measure in-situ the mass composition and detailed velocity phase space distributions of a thermal plasma population in a planetary ionosphere or magnetosphere and use the measured distributions to derive the bulk plasma parameters and to detect the possible presence of non-thermal distributions. A hemispherical electrostatic analyzer (HEA) with a planar entrance aperture can sample simultaneously incident ions or electrons over an extended energy range and the full 360 range of incident azimuth, and disperse them by their energy-per-charge while retaining their incident azimuth, thus providing a means to image the 2-dimensional (2D) ion or electron energy-per-charge and angular (azimuth) distribution. Therefore an ion mass and velocity analyzer consisting of a HEA embedded with an ion-mass spectrometer is capable of imaging the 2-D detailed ion velocity distribution and measuring the 3D distribution on a spinning spacecraft if the planar entrance aperture is aligned along the spacecraft spin axis. For 3D velocity distribution measurements on a 3-axis stabilized spacecraft, an analyzer with electrostatic deflection capability will be required to deflect ions at arbitrary incident elevation angles into the planar entrance aperture for sampling. An imaging thermal ion mass and velocity analyzer is presented that combines a HEA, a time-of-flight ion mass spectrometer, and a pair of electrostatic deflectors, and is capable of sampling low-energy ions ( 1 to 100 ev/e) of all mass species (1 to > 40 AMU/e) from all incident directions on a non-spinning platform, at up to (10% energy resolution ( E/E) and 5 angular resolution. Using the HEA to measure the energy-percharge of each detected ion and the time-of-flight gate to measure the transit time of the ion inside the analyzer, this instrument can resolve all major ion species in the ionosphere including H +,He + and O +, and adjacent molecular ion species such as N + 2,NO+ and O + 2 under favorable conditions. In addition, it can image the 2D and measure the 3D velocity phase space distributions of each major ion species. Key words: Ion mass spectrometer, plasma analyzer, hemispherical analyzer. 1. Introduction The terms thermal, core and low-energy are often used interchangeably in the literature to refer to the lowest-energy, core component of a plasma population in a planetary ionosphere or magnetosphere. Both the mass composition of this thermal plasma and its detailed velocity phase space distribution often play an important role in the structure and the dynamics of a planetary ionosphere or magnetosphere. An imaging thermal ion mass and velocity analyzer aims to apply imaging techniques to obtain both of these quantities in-situ, by measuring the macroscopic ion mass composition distribution and the detailed mass-resolved velocity phase space distribution of each ion species. The goal is to use the measured distributions to derive the bulk plasma parameters ion composition, density, bulk velocity and temperature and to detect the possible presence of non-thermal distributions (deviations of the velocity phase space distribution from thermal distributions), for studying plasma acceleration and transport processes in different regions of a planetary ionosphere or magnetosphere. The structure and dynamics of the ionosphere on a weakly-magnetized or non-magnetized planet such as Mars Copyright c TERRAPUB, and Venus depends strongly on the nature, magnitude and topology of its intrinsic induced magnetic field, and its direct interaction with the solar wind is believed to result in a significant escape of the planetary atmosphere and play a key role in the latter s evolution over geological time. In the case of Mars, for example, thermal ion composition (Hanson et al., 1977) and cold ion measurements (Dubinin et al., 1993) have been key to our current knowledge on its ionospheric composition and dynamics and its solar wind interaction. In the case of the Earth, the thermal plasma plays an important role in the coupling between the ionosphere and the magnetosphere: As discussed in, for example, Chappell (1988) and Yau and Andre (1997), the magnetosphere often contains a significant component of cold ions originating as ion outflows from the ionosphere. In the polar ionosphere, both ion density and mass composition are important parameters in ionospheric ion acceleration and outflow; for example, the ion mass density affects the properties of Alfven waves, which carry field-aligned currents and plasma waves (Lysak and Lotko, 1996), and the plasma density is believed to control auroral acceleration and auroral kilometric radiation (AKR) (Morooka and Mukai, 2003). The ion bulk flow velocity is also an important plasma parameter due to its direct (E B) relationship with the convection electric field in the frozen-in and collision-less 203

2 204 A. W. YAU et al.: IMAGING THERMAL ION MASS AND VELOCITY ANALYZER Fig. 1. Schematic illustration of a cylindrical electrostatic analyzer (top), which measures ions at one ion energy-per-charge and one incident azimuth at a time using a channel electron multiplier detector, and the rotational cross-section of a top-hat electrostatic analyzer (bottom), which measures simultaneously ions at a fixed energy-per-charge over the full 360 range of incident azimuth (in and out of the page) using a bias voltage (V ) between the inner and outer analyzer surface and a micro-channel plate detector (denoted by the solid rectangle). regions of the magnetosphere and the ionosphere, where large-scale convective electric fields play a critical role in plasma circulation, redistribution and energization. Many of the existing techniques for measuring thermal ions are based on electrostatic or retarding potential analysis (RPA) and focused on resolving the energy and angular distributions of the measured ions. In many instances these techniques are also applicable to electron measurements. Some of the existing techniques also resolve the mass composition of the measured ions. The earliest ion composition and density measurements in space date back to those using Bennett or open ion mass spectrometers (Brinton et al., 1973) and retarding potential analyzers (Hanson et al., 1973) on the Earth-orbiting Isis and Atmospheric Explorer satellites, the Viking Mars Landers (Hanson et al., 1977) and the Pioneer Venus Orbiter (Knudsen et al., 1979a, b). A Bennett ion mass spectrometer achieves ion mass discrimination by selectively accelerating a particular ion mass species with a radio-frequency (RF) electric field of matching phase speed (to the speed of the ion mass species) (Bennett, 1950). A retarding potential analyzer sweeps its retarding potential grid through a range of potentials, to generate a characteristic curve of the measured ion current as a function of the retarding potential. The shape of the RPA curve is typically fitted to a model to infer the density, bulk velocity, and/or temperature of the measured ions; under certain conditions and assumptions, the relative ion composition of the major ion species can be derived also. Other early techniques for measuring thermal ion drift velocities in the ionosphere included the double probe (Mozer, 1973), which measures the potential difference between two spaced electric field probes to derive the ambient electric field and the corresponding E B drift velocity; artificial ion cloud, in which the motion of an ion cloud is tracked using ground triangulation (Wescott et al., 1969); and cylindrical electrostatic analysis, in which (as illustrated in Fig. 1 below) a cylindrical electrostatic analyzer is used to measure the differential ion energy-percharge spectrum of the observed ions (Green and Whalen, 1974). The cylindrical analyzer can be combined with ion time-of-flight analysis to resolve the mass composition of the ions also (Yau et al., 1981). Most early particle-counting thermal ion mass and/or energy analyzers used channel electron multipliers, and typically measured ions at one energy or retarding potential at one incident direction at a time; some measured ions over a wide and unresolved angular range, and the mass-resolving analyzers typically sampled one mass species at a time. Notable examples of such analyzers include the retarding ion mass spectrometer (RIMS) on Dynamics Explorer-1 (Chappell et al., 1981), which used a sector magnet behind a retarding potential analyzer for ion mass separation, and the suprathermal ion mass spectrometer (SMS) on Akebono (Whalen et al., 1990), which used a modified Bennett RF ion mass spectrometer behind a combination of retarding potential and electrostatic analyzers for ion energy selection and mass analysis. Both instruments measured ions at one retarding potential setting in each measurement. In the case of RIMS, the instrument had a large angle of acceptance, and sampled two mass steps of a fixed ratio simultaneously using two separate electron multipliers. In the case of SMS, which had 3 selectable apertures and corresponding angles of acceptance, the instrument sampled one mass step in each measurement and used a 16-pixel micro-channel plate detector to achieve limited ion energy or angular imaging in special operation modes (Yau et al., 1998a). The advent of micro-channel plate detector ushered in the era of imaging plasma analyzer designs, which aim to resolve the measured plasma in incident angle, energy, mass, or some combination thereof in a single measurement. Figure 1 compares the ion optics of a cylindrical electrostatic analyzer, which permits ion or electron entry only at a particular energy-per-charge and one incident azimuth, with that of the top-hat analyzer pioneered by Carlson et al. (1983), in which the rotational symmetry about the top-hat axis allows the simultaneous entry of ions or electrons at a particular energy-per-charge over the full 360 range of incident azimuth. On an attitude-controlled sounding rocket or orbiting satellite in which the axis of the top-hat analyzer is oriented perpendicular to the local magnetic field, the top-hat design provides an effective means to image ions or electrons over the full 180 range of pitch-angle. An excellent example of this is the thermal electron capped hemispheric spectrometer (TECHS), which measured the velocity distribution of ev electrons on the SCIFER sounding rocket (Pollock et al., 1998). In contrast with the top-hat design, the hemispherical electrostatic analyzer (HEA) design (Whalen et al., 1994) allows simultaneous entry of ions or electrons over an ex-

3 A. W. YAU et al.: IMAGING THERMAL ION MASS AND VELOCITY ANALYZER 205 (a) (b) Fig. 2. (a) Sensor cross section view, (b) sensor assembly exploded view. tended range of energy (instead of a single energy-percharge) and the full 360 range of incident azimuth. Figure 6 below depicts the principle of operation of a HEA: basically the analyzer accepts ion or electron entry through an entrance aperture plane and disperses the incident ions or electrons by their energy-per-charge while retaining their incident azimuth, effectively providing a means to image the 2-dimensional (2D) energy-per-charge and angular distribution. The basic HEA design was first flown in the cold plasma analyzer (CPA) on Freja (Whalen et al., 1994) which measured both electrons and ions in both thermal and the suprathermal energy ranges (Knudsen et al., 1998). Since then, the HEA design has been advanced to incorporate a charge coupled device (CCD) detector in the suprathermal ion or electron imagers (SII/SEI) on several sounding rockets, to provide much denser detector

4 206 A. W. YAU et al.: IMAGING THERMAL ION MASS AND VELOCITY ANALYZER Table 1. Sensor bias voltages. Voltage Definition Value/Allowed range (V) V ED Toroidal electrostatic deflector; ±at top and bottom rings 10 to +10 V ES Hemispherical electrostatic analyzer (HEA) outer dome 10 to +10 V EA Entrance aperture bias 0 V TF Time-of-flight gate electrode; at top and bottom electrodes +10 V SA Hemispherical electrostatic analyzer (HEA) inner dome 0 to 353 V FP Micro-channel plate (MCP) front surface bias 4 to 2487 V BP Micro-channel plate (MCP) back surface bias 0 to 356 V NC Anode 0 40 atomic mass units per charge (AMU/e) on both spinning and non-spinning spacecraft. An important part of the objective is to resolve the major ion species in the ionosphere (including H +,He + and O + ) and under certain conditions the minor molecular ions (N + 2,NO+ and O + 2 ), and to use the measured distributions to derive plasma density, drift velocity and temperature parameters. Fig. 3. Sensor bias voltages defined in Table 1. pixel density (Knudsen et al., 2003; Burchill et al., 2010). Both the original Freja CPA instrument and the CCD-based SII/SEI instruments were purely electrostatic devices and do not carry any mass-resolution capability. However, the HEA design has also been advanced to incorporate a timeof-flight gate in the thermal plasma analyzer (TPA) instrument on Nozomi (Yau et al., 1998b); the time-of-flight gate acted as an embedded ion-mass spectrometer, giving TPA the capability to resolve the mass identity of the detected ions and turning it into an ion mass and velocity analyzer capable of imaging the 2D detailed ion velocity distribution. On the spinning Nozomi spacecraft and on spinning sounding rocket payloads, successive images of 2D velocity distributions can be combined to produce a 3D distribution every half spin. To measure the 3D velocity phase space distribution of thermal ions on a non-spinning (3-axis stabilized) platform, an ion mass and velocity analyzer with electrostatic deflection capability is required to sample ions of arbitrary incident directions relative to the entrance aperture plane of the analyzer. In the following sections, such a 3D imaging thermal on mass and velocity analyzer is presented, which consists of a pair of toroidal electrostatic deflector and a time-of-flight ion mass spectrometer in addition to a HEA, and is therefore capable of measuring the mass composition and detailed 3D velocity phase space distributions on a non-spinning spacecraft. The basic measurement objective of this instrument is to measure the mass composition and detailed velocity phase space distributions of thermal- and suprathermal-energy ions in the energy-per-charge range of 1 to 100 ev/e and the mass-per-charge range from 1 to 2. Principle of Instrument Like its Nozomi TPA predecessor, the imaging thermal ion mass and velocity analyzer (thermal ion mass analyzer or TIM hereafter for brevity) consists of a sensor and an electronic module. The sensor is housed in a cylindrical enclosure and is to be mounted to the end of a deployable boom on a spacecraft, so that it is at a distance from the spacecraft body and outside of the spacecraft sheath. The electronic module is mounted inside the spacecraft, and provides all power, digital control, timing, and data interfaces to the sensor. Figure 2(a) shows a schematic cross section of the sensor through its axis of rotational symmetry, and identifies its key functional components, including the deflection rings that form the toroidal the electrostatic deflector, the time-of-flight gates, the inner and the outer electrostatic domes that form the HEA, micro-channel plates, and the anode printed circuit board (PCB). Figure 2(b) shows an exploded view of the internal sensor assembly. Figure 3 identifies the bias voltages at the respective surfaces or voltage grids inside the sensor, and Table 1 lists these voltages and their values or ranges of values. Figure 4 shows a photo of the sensor and the electronic module of the imaging and rapid-scanning ion mass spectrometer (IRM) on the Canadian CASSIOPE Enhanced Polar Outflow Probe (e-pop) satellite (Yau et al., 2006), which is based on the TIM analyzer design. The ion optics of the sensor is defined principally by its toroidal electrostatic deflector and HEA. As shown in Figs. 2(a) and 3, the electrostatic deflector consists of a pair of deflection rings. The top and bottom rings are biased to selected voltages of opposite polarity (V ED + and V ED ), to deflect incident ions of specific energy-per-charge and elevation combinations into the time-of-flight gate. Figure 5 illustrates the operation of the electrostatic deflector, and shows the simulated entry of ions of specific energy-percharge and elevation into the sensor in response to the voltage biases at the top and bottom rings ( 1.2 Vand +1.2 V, respectively). The deflector acts as an energy-elevation selector in that it accepts lower-energy ions at larger inci-

5 A. W. YAU et al.: IMAGING THERMAL ION MASS AND VELOCITY ANALYZER 207 Fig. 4. Sensor (foreground) and electronics module (background) of the imaging and rapid-scanning ion mass spectrometer (IRM) on the CAS- SIOPE/e-POP satellite. Fig. 5. Simulated trajectories of incident ions at 4-60 ev/e and 0-60 elevation, in response to the voltage biases at the top and bottom deflection rings of the sensor. dent elevations and higher-energy ions at smaller elevation. Note that when both the top and bottom deflection plates are set to ground, the sensor samples only ions at 0 elevation, and measures the 2D velocity distribution in the entrance aperture plane. The hemispheric electrostatic analyzer (HEA) consists of an inner and an outer hemisphere ( dome ). The inner dome is biased negative relative to the outer dome (V SA < V ES ), and the resulting central electric field disperses the incident ions by their energy-per-charge and incident azimuth onto the micro-channel plate detector: the landing radius and azimuth of each detected ion on the MCP surface maps to the ion energy and azimuth of arrival. Figures 6(a) and (b) illustrate the dispersion of ions through the HEA by their energy-per-charge and azimuth. In other words, the HEA focuses incident ions of a given azimuth and energy per charge onto a single point on its hemispherical plane, essentially irrespective of the ion elevation angles and entrance positions. The radial distance of this point increases with ion energy-per-charge. The maximum ion energy sampled depends primarily on the voltage difference (V SA V ES ). Figure 6(a) shows the side view of simulated trajectories of incident ions of 4 60 ev/e, when V SA was set to 200 V relative to V ES. The highest energy ions arrive at the outermost portion of the MCP. Figure 6(b) shows the top view of simulated trajectories of incident ions of lower energies (1 15 ev/e), when V SA was set to a correspondingly lower value. The highest energy ions again arrive at the outermost portion of the MCP, and all of the ions arrive at the opposite azimuth regardless of their energy. The time-of-flight (TOF) gate consists of a pair of fastswitching electrodes, which are controlled by the TOF gate driver, and it opens and closes repeatedly in a measurement to allow ion entry. Figure 7 shows an exploded view of the internal assembly of the TOF electrodes, which are made of a set of 3 concentric rings of foil, and Fig. 8 illustrates the

6 208 A. W. YAU et al.: IMAGING THERMAL ION MASS AND VELOCITY ANALYZER (a) (b) Fig. 6. (a) Side-view of simulated trajectories of incident ions at 4 60 ev/e dispersed by the HEA onto the MCP detector, with the higher-energy ions arriving at the outermost portion of the MCP. (b) Top-view of simulated trajectories of incident ions at 1 15 ev/e being dispersed by the HEA onto the MCP detector, with the ions arriving at the opposite azimuth of the MCP. operation of the TOF gate and its open-and-close sequences in a measurement. Figure 8 depicts the multiple TOF cycles in a measurement period (top panel) and the time-of-flight gate open and close sequence in each TOF cycle (bottom panel). Table 2 lists the ranges of timing parameters in the TOF gate. Each TOF cycle consists of 1024 time-of-flight bins, each of 40 ns duration (N BIN = 1024; T BIN = 40 ns). In other words, a TOF cycle is µs induration. A measurement consists of a number of TOF cycles, N TOF, which can have a value of 1 to 65535; corresponding to a measurement period of 0.4 to 25.6 ms. The default value of N TOF is 240 and corresponds to a measurement period of 9.83 ms (i.e. 100 measurements per second). In each TOF cycle, the TOF gate can remain open for up to 255 TOF bins (N OPN = 0 255); however, in practice, N TOF is typically between 10 to 50, which corresponds to a duty cycle of % and a gate-open period of 0.4 to 2 µs. The detector consists of a pair of micro-channel plates (MCP) and a discrete anode. An incident ion that is passing through the HEA produces a charge as it arrives at the front surface of the MCP. The charge is amplified through successive collisions with the channel walls inside the MCP into a charge cloud, which is proximity focused at the back of the MCP and then slightly defocused before being collected onto the anode. The front side of the anode contains 64 discrete pixels, which have radial widths from 0.33 to 1.6 mm and arc lengths from 2.8 mm to 1.0 mm, and arranged in 8 rows (rings) and 8 columns (spokes). The defocusing is to increase the spatial spread of the cloud, from tens of microns to hundreds of microns, to distribute the charge over a larger fraction of a pixel s surface area. On the backside of the anode, each pixel is connected to a pair of charge amplifiers, which detect the charge buildup ( particle hit ) at the pixel above a threshold value ( electrons).

7 A. W. YAU et al.: IMAGING THERMAL ION MASS AND VELOCITY ANALYZER 209 Fig. 7. Time-of-flight gate electrode assembly. Fig. 8. Top: TOF gate open and close sequences in a measurement period consisting of multiple TOF cycles; bottom: Time-of-flight gate (TOF) open-and-close sequence in a TOF cycle. In order to direct roughly half of the charge hitting a pixel to each of the two charged amplifiers, each pixel is subdivided into two interlaced, separated and electrically isolated halves, and each half is connected to one of the two amplifiers, so that statistically each amplifier is likely to detect half of the charge cloud. A consequence of the interlaced pixel subdivision is that whenever the split of a charge cloud is very unequal between the two amplifiers the charge may be detected above the threshold level at only one of the two amplifiers. On the adjacent pixel encoder board, the circuitry of preamplifiers identifies the row and column pixel address of each detected ion and its arrival time relative to the TOF gate opening. A field programmable gate array (FPGA) encodes the 6-bit pixel address and a 10-bit time-of-flight bin number into a 2-byte data word and subsequently writes it to a first-in-first-out (FIFO) data buffer. In addition, the FPGA tracks the number of such data words, which corresponds to the number of hits with definitive pixel row and column, as well as the number of detects, i.e. events of ion detection without definitive pixel address information such as those detected by only one preamplifier. In detailed laboratory testing, the hit-to-detect ratio was found to be in the range of 0.25 to 0.4, depending on the spatial spread of a charge cloud and its precise location on the pixel. At the end of each measurement period, the FIFO data words are sent to the digital signal processor in the electronics module, where the sensor data are processed, as will be discussed later in this section. For the given sensor voltage setting in a measurement, one can unambiguously determine the energy-per-charge, mass-per-charge, and incident azimuth and elevation of each detected ion. The optimum mounting orientation of the sensor on a non-spinning spacecraft is naturally dependent on the at- titude of the spacecraft. From the foregoing, it can be seen that on a nadir- or ram-pointing spacecraft, the entrance aperture plane of the sensor should ideally be aligned in the spacecraft ram-and-nadir plane (X-Z plane where the Z-axis is in the nadir direction and the X-axis is in the ram azimuth in the horizontal plane). When the spacecraft is nadir-pointing, the sensor will sample ions at all elevation angles at both the ram and anti-ram (wake) azimuth including the upward and downward vertical directions. Likewise, when the spacecraft is ram-pointing, the sensor will sample ions in the ram, anti-ram, and vertical directions. It should be noted here that the terms elevation and azimuth refer to the angle from the local horizontal (positive upward) and the angle from the east (positive anti-clockwise) in the local horizontal plane. On a spinning spacecraft, the entrance aperture plane of the sensor should ideally be aligned with the spacecraft spin axis, so that the sensor will sample all directions (4π of solid angles) every half a spacecraft spin. It is important to keep in mind here that in the case where the sensor entrance aperture plane is aligned with the local vertical, a one-to-one mapping exists between a sensor azimuth (an azimuth in the sensor entrance aperture plane) and the corresponding elevation and azimuth of its look direction with respect to the local vertical and horizontal directions. Figure 9 shows a schematic block diagram for the sensor, including the architecture of the time-of-flight (TOF) gate driver, anode, and pixel encoder card, and the series of high and low voltage grids. Figure 10 shows a schematic block diagram of the electronics module, which consists of a digital signal processor, a low voltage power supply, and a high voltage power supply. The digital signal processor acts as the command and data interface with the spacecraft, and controls the operations of both the high voltage power supply and the sensor. The low voltage power supply converts the spacecraft bus power (+28 Vdc) into regulated and isolated +3.3 Vdc, +5 Vdc and ±12 Vdc outputs. The high

8 210 A. W. YAU et al.: IMAGING THERMAL ION MASS AND VELOCITY ANALYZER Table 2. Time of flight gate timing parameters in a measurement. Parameter Definition Value/Range Typical value T BIN Duration of 1 TOF bin 40 ns N BIN #oftof bins in a TOF cycle 1024 N OPN #ofgate-open TOF bins in a TOF cycle N TOF #oftof cycles in a measurement Fig. 9. Sensor block diagram. voltage power supply provides the sensor voltages and their ground reference. Figure 11 shows the key dimensions of the sensor, which measures 94 mm in diameter and 164 mm in height and weighs 1.30 kg; the electronic module measures mm 3 and weighs 1.83 kg, and consumes 6.2 W of peak power and 4.7 W of standby power at aspacecraft bus voltage of 28 V. The instrument is capable of operating in a number of pre-programmed measurement modes in response to ground command. The control software for these measurement modes is stored in memory in the digital signal processor. In each measurement mode, the sensor can step repeatedly through a sequence of up to 4 sets of sensor settings (bias voltages and TOF gate timing parameters), and repeat each set of settings up to 255 times (measurements). Thus a measurement mode consists of a sequence of up to 1020 measurements. As can be seen in Table 2, a measurement period is typically 10 ms but can be as long as 2.68 s. In each measurement period, the sensor settings remain fixed and correspond to a fixed range of sampled ion mass-per-charge and a fixed region (subset) of the ion energy-per-charge and incident elevation phase space, and the sensor can transmit up to 65,536 bytes of data to the digital signal processor for processing. At the end of each sequence of measurements, the pixel data in each measurement period are packetized without any onboard data reduction and transmitted to ground whenever the telemetry bandwidth permits; at other times, the data are processed onboard: the pixel data are summed over all measurement periods of each sensor setting to create ion count histograms or moments at selected detector anode pixels and time-offlight bins to reduce the telemetry volume. Although the instrument is in principle capable of generating pixel data at a maximum rate of about 13 megabytes per second, the actual maximum data throughput rate is much lower, and is about 0.5 megabytes per second or 4 megabits per second. 3. Sensor Response A series of calibration measurements were made using the charged particle calibration facility at the University of Calgary to characterize the angular, energy, mass and sensitivity responses of the sensor to low-energy ions. Two different low-energy (1 200 ev) ion sources were used: the first one has a smaller beam area (2 cm 2 cm) and a beam width of ±2 and was used previously to calibrate the Akebono suprathermal mass spectrometer (Whalen et al., 1990), and the second one has a larger beam area (10 cm 10 cm) and a higher output beam current (up to 1

9 A. W. YAU et al.: IMAGING THERMAL ION MASS AND VELOCITY ANALYZER 211 Fig. 10. Electronics module block diagram. na). In both cases, a faraday cup was used for absolute flux calibration and for beam energy and width characterization. Both the sensor and the faraday cup were mounted on a manipulator table, to facilitate sensor rotation in 2 axes and translational movement in 2 directions with respect to the ion source. In all measurements the energy of the ion beams had a typical FWHM of 30 50%. Molecular nitrogen was used in most of the calibration runs because of its chemically inert properties; Argon was used in the remaining runs. The majority of the calibration runs were performed using ions between 10 and 50 ev, which in the case of N + 2 corresponds to a velocity of 8.3 to 18.6 km/s, to mimic the velocities of both cold and accelerated ionospheric ions seen by the sensor in orbit: since the spacecraft ram velocity in low-earth-orbit is 8 km/s, a co-rotating ion will have a relative velocity of 8 km/s in the sensor frame of reference. Ideally, it is desirable to characterize the detailed sensor response by using a highly collimated and energyselectable, mono-energetic ion source of a selectable mass species and systematically varying its orientation relative to the sensor (or vice versa) and the selected ion energy, to determine the geometric factor, G(E, ),ateach detector pixel as a function of incident ion energy E and angle for a given set of sensor (voltage and timing parameter) settings; = (θ, φ) and θ and φ are elevation and azimuth, respectively. However, because of the very large number of possible sensor settings (combination of voltage parameters in Table 1 and time-of-flight gate timing parameters in Table 2), it is not possible to calibrate each and every sensor setting in this manner. Moreover, as noted above, the energy spread (FWHM of 30 50%) of the ion sources was large compared with the intrinsic sensor energy response ( E/E of 10 15%) particularly at low energy. Therefore, calibration measurements were made only at a number of selected sensor settings and incident ion beam energies, angles and fluxes to determine the averaged sensor response at these settings over the broad energy range of the source ion beam. To determine the detailed sensor energy response function, i.e. G(E, ) at each pixel as a function of E and and the corresponding energy resolution ( E/E) and angular range (angle of acceptance; ), a particle-tracing code was used to compute the electric field distribution inside the sensor and in the vicinity of its entrance aperture as a function of the selected sensor settings, and the corresponding ion trajectory as a function of incident ion energyper-charge, mass-per-charge, elevation and azimuth. The simulated trajectories for the selected ion beam energies, angles and fluxes were then used to calculate the predicted response of the sensor at the selected sensor settings, for comparison with the calibration data. Based on the generally good agreement (a few percent) between the simulated and measured responses, the simulation code was then used for computing the sensor response function at all sensor settings, and the computed response will be used to convert inflight ion count data to ion fluxes and velocity phase space densities for detailed analysis. Figure 12 shows the measured dependence of the incident elevation of sampled ions on the toroidal electrostatic deflector voltage for nitrogen ions (N + 2 )at15ev. In this calibration run, the sensor was oriented so that the sensor axis was orthogonal to the ion beam and one of its anode pixel columns was aligned with the beam, and the toroidal deflec-

10 212 A. W. YAU et al.: IMAGING THERMAL ION MASS AND VELOCITY ANALYZER Fig. 11. Sensor dimensions of imaging thermal ion mass and velocity analyzer. tion voltage (V ED )wasset to 0 V; V SA was fixedto 69.5V, to ensure that the incident ions would arrive at the third outermost pixel ( pixel 6 ). The elevation angle was then varied at 5 steps. At each step, the number of ions detected in the pixel ( hits ) was measured as a function of the varying deflection voltage, and the measured profile fitted to a Gaussian to determine the corresponding optimum deflection voltage. Figure 12 shows a linear relationship between the elevation angle θ and deflection voltage amplitude V ED except at large elevation angles: i.e. θ = αv ED + β where θ is in deg, V ED is in volt, α = 11.1 /V and β = 0.9. Such a linear relationship was found to hold in other calibration runs at other ion energies or using argon ions (Ar + ) as well; in each case, the numerical value of α was found to be in good accord with the simulation and increase with decreasing incident ion energy, and the value of β was 1 and is attributed to the relative misalignment between the sensor and the ion source. Thus the measurements confirm the ability of the sensor to sample low-energy (1 15 ev) ions up to 60 off the sensor entrance aperture plane on a non-spinning platform. Figure 13 shows selected measured time-of-flight spectra of 50 ev N + 2 ions. The top and bottom panels show spectra at 1 and 0.1% TOF duty cycle, respectively; in each case, the right panel shows the data on an expanded TOF scale to reveal the detailed profiles of measured TOF peaks. In both cases, V SA was set to 300 V, resulting in the ions being detected at the second outermost pixel ( pixel 7 ) in the anode pixel column. In both cases, two TOF peaks were observed; in contrast, only one TOF peak was observed in the runs using argon ions. The larger peak starting near TOF bin 54 is attributed to N + 2 ions while the smaller peak starting near TOF bin 40 is attributed to N + ions, based on the analysis below. The TOF bin value of a measured ion is determined by Fig. 12. Measured dependence of the incident elevation of sampled ions on the torodial electrostatic deflector voltage for ions at 15 ev. the time delay between the start of the time-of-flight gateopen period in a TOF cycle and the time of arrival at the 6-bit pixel address of the detected ion at the FPGA. This time delay τ is comprised of 4 components: the time of ion transit from the outboard edge of the TOF gate to the outer dome, τ 1 ; the corresponding time from the outer dome to the front of the MCP detector, τ 2 ; the transit time of the electron charge cloud through the MCP to the anode pixel, τ 3 ; and the time delay of the pre-amplifier firing, τ 4. For agiven ion energy, τ 1 depends on V ES and is essentially proportional to the square root of the ion mass in the case of V ES = 0, and τ 2 depends on both the ion mass and the magnitude of V SA ; τ 3 is on the order of 1 ns and therefore negligible compared with τ 1 and τ 2. The quantity τ 4 is dominated by the triggering time of the preamplifiers, which is dependent on the rate of charge injection and the pulse am-

11 A. W. YAU et al.: IMAGING THERMAL ION MASS AND VELOCITY ANALYZER 213 Fig. 13. Measured time-of-flight (TOF) spectra of 50 ev N + 2 ions: top: 1% TOF duty cycle (TOF gate open period = 10 TOF bins = 0.4 µs); bottom: 0.1% TOF duty cycle (TOF gate open period = 1TOFbin = 0.04 µs); left: full TOF scale; right: expanded TOF scale to show the detailed profiles of observed TOF peaks. plitude of the injected charge, and was found to be about 220 ns under typical sensor operating conditions. Including the rise time of 25 ns and the fall time of 80 ns, τ 4 is about 320 ns or 8 TOF bins. Thus, for a given sensor voltage setting (V SA and V ES ) and ion energy, the starting TOF bin number of a measured ion is approximately given by (C(m/q) 1/2 + 8) where C is a proportionality constant, and m/q is ion mass-per-charge. In the case of 50 ev ions, V SA of 300 V and V ES of 0 V in Fig. 13, C is approximately 8.86 (AMU/e) 1/2 from simulation. Thus, the TOF peaks for mass 28 and 14 (N + 2 and N+ ) ions are predicted to start near TOF bin 55 and 41, respectively. In other words, the observed TOF peaks in Fig. 13 are consistent with the measured ions being N + 2 and N+ ions, respectively. In Fig. 13, the widths of the TOF peaks in the top panels are discernibly different from those in the bottom panels. This difference is attributed to the different TOF gate period in each case, as follows. As shown in Fig. 13, the TOF gate electrode has a width of 0.7 mm. Therefore, it would take a 50-eV N + 2 ion about µs and a corresponding N+ ion about µs totraverse the electrode. In comparison, the TOFgate remained open for 1% of the TOF cycle in the top panels, i.e. 0.4 µsor10tof bins of 40 ns each. Therefore, all of the ions arriving at the TOF gate within the first 9 of the 10 TOF bins after the gate opening will have sufficient time to traverse the electrode before the gate closes: this explains the relatively flat peaks spanning several TOF bins in the top panels. In contrast, in the bottom panel, the TOF gate remained open for only 0.1% of the TOF cycle or 0.04 µs, which is comparable to the transit time of the N + 2 ions through the TOF gate electrode. Therefore, a substantial fraction of ions arriving at the TOF gate during the gate open period failed to exit the TOF gate before the end of the gate-open period: this explains the much narrower TOF peaks in this case. Since cold ionospheric ions will appear as ions with a relative velocity of 8 km/s in the sensor frame of reference on a low-earth-orbit satellite, it can be seen from the results in this figure that they will have a transit time on the order of about 0.1 µs through the TOF gate. Thus, the TOF gate must remain open for at least 4 or 5TOF bin periods to measure such ions. The results in Figs. 12 and 13 and similar results from other calibration runs combine to demonstrate that the fact that as discussed above, the ion optics of the sensor is defined principally by the voltage settings at its toroidal electrostatic deflector and HEA (V ED, V ES and V EA ). The body of calibration results also confirms that as can be seen in Fig. 6, the overall ion optics is also affected by the voltage settings at the front and back of the MCP detector (V FP, V BP ), albeit to a lesser extent, while the mass response is controlled by the voltage settings and timing parameters of the time-of-flight gate (in particular V TF, N OPN ).

12 214 A. W. YAU et al.: IMAGING THERMAL ION MASS AND VELOCITY ANALYZER 4. Summary and Discussion The ion mass composition and detailed velocity phase space distributions of thermal (low-energy) plasma often play an important role in the structure and the dynamics of a planetary ionosphere or magnetosphere. The aim of an imaging thermal ion mass and velocity analyzer is to apply imaging techniques to measure in-situ both the mass composition distribution of a thermal plasma population and the mass-resolved velocity phase space distribution of each of its ion species, and to use the measured distributions to derive the bulk plasma parameters and to detect the possible presence of non-thermal distributions in a planetary ionosphere or magnetosphere. A hemispherical electrostatic analyzer (HEA) simultaneously samples incident ions or electrons over an extended energy range and the full 360 range of incident azimuth in its entrance aperture plane, and disperses the ions or electrons by their energy-per-charge while retaining their incident azimuth, effectively providing a means to image the 2D energy-per-charge and azimuth distribution. Thus a HEA with an embedded ion-mass spectrometer is capable of resolving the mass identity of the measured ions and imaging the 2D detailed velocity phase space distributions of the different ion species, and effectively becomes an imaging ion mass and velocity analyzer; on a non-spinning (3-axis stabilized) platform, an analyzer that has additional electrostatic deflection capability is capable of measuring the 3- dimensional velocity distributions also. This is in contrast to a top-hat analyzer, which also samples particles over the full range of azimuth and retains their individual azimuth but is restricted to sampling ions or electrons at only a single energy at a time. In the case of the imaging thermal ion mass and velocity analyzer presented above, a toroidal electrostatic deflector is used to sample ions at different elevations (angles to the sensor entrance aperture plane) and a time-of-flight gate is used to measure the individual ion time-of-flight. The instrument is designed to measure thermal-energy ions in the energy-per-charge range of 1 to 100 ev/e and the mass-per-charge range of 1 to >40 atomic mass units per charge (AMU/e) at up to 10% energy resolution ( E/E) and 5 angular resolution, and to resolve all major ion species in the ionosphere including H +,He +,O + as well as adjacent molecular ion species such as N + 2,NO+ and O + 2 under favorable conditions. Compared with the top-hat and other plasma analyzers, imaging analyzers are capable of measuring an ion or electron energy distribution at a much faster rate (up to 100 Hz), making them ideally suited for studying temporal structures or plasma processes down to time scales of tens of ms or small scale plasma structures down to spatial scales of tens of meters on a sounding rocket or hundreds of meters on an orbiting satellite. As noted in the Introduction section, both mass-resolving and non-mass-resolving imaging thermal plasma analyzers currently exist, and they often offer complementary advantages. For example, the analyzer above provides definitive ion mass species identification and velocity determination, which are crucial for certain scientific investigations such as the mass dependence of various ion acceleration mechanisms, but its detector has relatively limited pixel resolution (density). In comparison, the suprathermal ion analyzer flown on the GEODESIC and JOULE sounding rockets (Knudsen et al., 2003), which is also based on the HEA design, uses a CCD detector to provide much denser detector pixel density, thereby providing 2D ion energy distributions of much higher energy and angular resolution, while requiring assumptions on ion composition to be made in detailed velocity analysis due to the lack of any ion mass resolution capability. References Bennett, W. H. (1950), Radiofrequency mass spectrometer, J. Appl. Phys., 21(2), Brinton, H. C., L. R. Scott, M. W. Pharo, III, and J. T. Coulson (1973), The Bennett ion-mass spectrometer on Atmosphere Explorer-C and -E, Radio Sci., 8(4), Burchill, J. K., D. J. Knudsen, J. H. Clemmons, K. Oksavik, R. F. Pfaff, C. T. Steigies, A. W. Yau, and T. K. Yeoman (2010), Thermal ion upflow in the cusp ionosphere and its dependence on soft electron energy flux, J. Geophys. Res., 115, A05206, doi: /2009ja Carlson, C. W., D. W. Curtis, G. Paschmann, and W. Michel (1983), An instrument for rapidly measuring plasma distribution functions with high resolution, J. Adv. Sp. Res., 2(7), Chappell, C. R. (1988), The terrestrial plasma source: A new perspective in solar-terrestrial processes from Dynamics Explorer, Rev. Geophys., 26, 2,doi: /RG026i002p Chappell, C. R., S. A. Fields, C. R. Baugher et al. (1981), The retarding ion mass spectrometer on Dynamics Explorer-A, Space Sci. Instr., 5, Dubinin, E., R. Lundin, H. Koskinen, and O. Norberg (1993), Cold ions at the Martian bow shock: Phobos observations, J. Geophys. Res., 98(A4), Green, D. W. and B. A. Whalen (1974), Ionospheric ion flow velocities from measurements of the ion distribution function technique, J. Geophys. Res., 79, Hanson, W. B., D. Zuccaro, C. Lippincott, and S. Sanatani (1973), The retarding-potential analyzer on atmospheric explorer, Radio Sci., 8(4), doi: /rs008i004p Hanson, W. B., S. Sanatani, and D. R. Zuccaro (1977), The Martian ionosphere as observed by the Viking retarding potential analyzers, J. Geophys. Res., 82, Knudsen, D. J., J. H. Clemmons, and J. E. Wahlund, (1998), Correlation between core ion energization, suprathermal electron bursts, and broadband ELF plasma waves, J. Geophys. Res., 103, , doi: /97ja Knudsen, D. J., J. K. Burchill, K. Berg et al. (2003), A low-energy charged particle distribution imager with a compact sensor for space applications, Rev. Sci. Instr., 74, Knudsen, W. C., J. Bakke, K. Spenner, and V. Novak (1979a), Retarding potential analyzer for the Pioneer-Venus orbiter mission, Space Sci. Inst., 4, Knudsen, W. C., K. Spenner, R. C. Whitten et al. (1979b), Thermal structure and major ion composition of the Venus ionosphere: First RPA results from Venus orbiter, Science, 203, Lysak, R. L. and W. Lotko (1996), On the kinetic dispersion relation for shear Alfven waves, J. Geophys. Res., 101(A3), Miller, K. L., W. C. Knudsen, K. Spenner, R. C. Whitten, and V. Novak (1980), Solar zenith angle dependence of ionospheric ion and electron temperatures and density on Venus, J. Geophys. Res., 85A, Morooka, M. and T. Mukai (2003), Density as a controlling factor for seasonal and altitudinal variations of the auroral particle acceleration region, J. Geophys. Res., 108(A7), 1306, doi: /2002ja Mozer, F. S. (1973), Analyses of techniques for measuring DC and AC electric fields in the magnetosphere, Space Sci. Rev., 14, Pollock, C. J., V. N. Coffey, J. D. England, N. G. Martinez, and M. L. Adrian (1998), Thermal electron capped hemisphere spectrometer (TECHS) for ionospheric studies, in Measurement Techniques in Space Plasmas: Particles, Geophysical Monograph, 102, R. F. Pfaff, J. E. Borovsky, and D. T. Young, editors, p. 201, American Geophysical Union, Washington, D.C. Wescott, E. M., J. D. Stolarik, and J. P. Heppner (1969), Electric fields in

13 A. W. YAU et al.: IMAGING THERMAL ION MASS AND VELOCITY ANALYZER 215 the vicinity of auroral forms from motions of barium vapor releases, J. Geophys. Res., 74, Whalen, B. A., J. R. Burrows, A. W. Yau et al. (1990), The Suprathermal Ion Mass Spectrometer (SMS) onboard the Akebono (EXOS-D) satellite, J. Geomag. Geoelectr., 42, Whalen, B. A., D. J. Knudsen, A. W. Yau et al. (1994), The Freja F3C cold plasma analyzer, Sp. Sci. Rev., 70, Yau, A. W. and M. Andre (1997), Sources of ion outflow in the high latitude ionosphere, Space Sci. Rev., 80(1 2), Yau, A., B. Whalen, F. Creutzberg, M. Pongratz, and G. Smith (1981), Observations of particle precipitation, electric field, and optical morphology of an artificially perturbed auroral arc: Project Waterhole, J. Geophys. Res., 86(A7), doi: /ja086ia07p Yau, A. W., E. Drakou, M. J. Greffen, D. J. Knudsen, and E. Sagawa, (1998a), Radio-frequency ion mass spectrometer measurements of thermal ion composition, velocity, and temperature: the EXOS-D suprathermal mass spectrometer, in Measurement Techniques in Space Plasmas: Particles, Geophysical Monograph, 102, R. F. Pfaff, J. E. Borovsky, and D. T. Young, editors, pp , American Geophysical Union, Washington, D.C. Yau, A. W., G. P. Garbe, M. J. Greffen, W. Miyake, and E. Sagawa (1998b), Planned observations of thermal plasma drifts and solar wind interactions in the Martian ionosphere, Earth, Planets and Sun, 50, Yau, A. W., H. G. James, and W. Liu (2006), The Canadian Enhanced Polar Outflow Probe (e-pop) mission in ILWS, Adv. Sp. Res., 38(8), A. W. Yau ( yau@phys.ucalgary.ca), E. P. King, P. Amerl, K. Berg, G. Enno, A. Howarth, I. Wevers, and A. White