Back-illuminated Research Articles

Evolution of Image Sensor Architectures With Stacked Device Technologies

The evolution of CMOS image sensors and their prospects using advanced imaging technologies are promising candidates to improve the quality of life. With the rapid advent of parallel analog-to-digital converters (ADCs) and back-illuminated (BI) technology, CMOS image sensors currently dominate the market for digital cameras, and stacked CMOS image sensors continue to provide enhanced functionality and user experience in mobile devices. This article reviews the latest achievements in stacked image sensors with respect to the evolution of image sensor architecture for accelerating performance improvements, extending sensing capabilities, and integrating edge computing with various stacked device technologies.

Power conversion efficiency (PCE) performance of back-illuminated DSSCs with different Pt catalyst contents at the optimized TiO2 thickness

The Power Conversion Efficiency (PCE) of back illuminated dye sensitized solar cells (DSSCs) using Titanium Dioxide (TiO2) printed on Titanium foil photo-anode was investigated. It is essential to study the optimum volume of the Platinum (Pt) catalyst solution spin-coated on counter electrode and the TiO2 absorbance layer optimized thickness for optimum light harvesting in a back illuminated structure. In this study, we have concentrated on the optimized 11 μm thick of TiO2 film assembled on the photoanode using Doctor Blade printing method. The 11 μm thick films were tested in a constraint of Platinum contents obtainable from 30 μl, 50 μl and 70 μl volumes of Pt solutions. The same cells were also subjected to front illumination to compare the results. The back illuminated optimization was achieved with the 11 μm TiO2 film thickness which was at PCE = 2.99%, Jsc = 7.83 mA/cm2, Voc = 0.7V, FF = 0.55 with the minimum volume of the Pt solution. From the selected ideal thickness of 11 μm TiO2 film performances of the Fluorine Tin Oxide (FTO) glass as solid state photoanode during back illumination, a similar methodological structure was imposed on Titanium foil based anode cells where the achievement was indicated at PCE = 2.75%, Jsc = 4.62 mA/cm2, Voc = 0.71V, FF = 0.84 for a 70 μl volume of Pt solution. The results show that for a solid state photoanode substrate, 30 μl of Pt solution is sufficient to provide catalytic activities and at the same time allow light penetration to the active region, whereas the flexible photoanode requires a higher concentration of Pt solution for rapid catalytic capabilities.

Research on the Etalon Effect in Dispersive Hyperspectral VNIR Imagers Using Back-Illuminated CCDs

Dispersive hyperspectral visible and near-infrared (VNIR) imagers using back-illuminated (BI) charge-coupled devices (CCDs) suffer from interference fringes in the near-infrared band, known as the etalon effect. This effect causes a signal modulation which can become increasingly serious (±25% or more) when the spectral resolution gets higher than 5 nm, bringing huge difficulties for subsequent processing and quantitative applications of the hyperspectral data. A mathematical model to describe the fringe pattern is established by taking both the system parameters and the interference principle of multilayer thin films into account. Then, the model is used to simulate the distribution of interference fringes as a function of wavelength, and the simulated results are verified with the measured data from a VNIR grating-based hyperspectral imager. It turns out that the model is able to describe the interference fringes accurately. In addition, quantitative analysis of the influence of related factors on interference fringes is carried out, and a single-layer model is introduced for comparison, providing the theoretical foundation for the subsequent correction of interference fringes. This paper provides an important reference for the design and application of dispersive hyperspectral VNIR imagers using BI-CCDs.

X-ray response evaluation in subpixel level for X-ray SOI pixel detectors

We have been developing event-driven SOI Pixel Detectors, named “XRPIX” (X-ray soiPIXel) based on the silicon-on-insulator (SOI) pixel technology, for the future X-ray astronomical satellite with wide band coverage from 0.5keV to 40keV. XRPIX has event trigger output function at each pixel to acquire a good time resolution of a few μs and has Correlated Double Sampling function to reduce electric noises. The good time resolution enables the XRPIX to reduce Non X-ray Background in the high energy band above 10keV drastically by using anti-coincidence technique with active shield counters surrounding XRPIX. In order to increase the soft X-ray sensitivity, it is necessary to make the dead layer on the X-ray incident surface as thin as possible. Since XRPIX1b, which is a device at the initial stage of development, is a front-illuminated (FI) type of XRPIX, low energy X-ray photons are absorbed in the 8μm thick circuit layer, lowering the sensitivity in the soft X-ray band. Therefore, we developed a back-illuminated (BI) device XRPIX2b, and confirmed high detection efficiency down to 2.6keV, below which the efficiency is affected by the readout noise. In order to further improve the detection efficiency in the soft X-ray band, we developed a back-illuminated device XRPIX3b with lower readout noise. In this work, we irradiated 2–5keV X-ray beam collimated to 4μmϕ to the sensor layer side of the XRPIX3b at 6μm pitch. In this paper, we reported the uniformity of the relative detection efficiency, gain and energy resolution in the subpixel level for the first time. We also confirmed that the variation in the relative detection efficiency at the subpixel level reported by Matsumura (2015) has improved.

Front and Back Illuminated Dynamic and Active Pixel Vision Sensors Comparison

Back side illumination has become standard image sensor technology owing to its superior quantum efficiency and fill factor. A direct comparison of front and back side illumination (FSI and BSI) used in event-based dynamic and active pixel vision sensors (DAVIS) is interesting because of the potential of BSI to greatly increase the small 20% fill factor of these complex pixels. This brief compares identically designed front and back illuminated DAVIS silicon retina vision sensors. They are compared in term of quantum efficiency (QE), leak activity and modulation transfer function (MTF). The BSI DAVIS achieves a peak QE of 93% compared with the FSI DAVIS, peak QE of 24%, but reduced MTF, due to pixel crosstalk and parasitic photocurrent. Significant “leak events” in the BSI DAVIS limit its use to controlled illumination scenarios without very bright light sources. Effects of parasitic photocurrent and modulation transfer functions with and without IR cut filters are also reported.

$320 \times 240$ Back-Illuminated 10- $\mu \text{m}$ CAPD Pixels for High-Speed Modulation Time-of-Flight CMOS Image Sensor

A $320\times 240$ back-illuminated (BI) time-of-flight CMOS image sensor with 10- $\mu \text{m}$ current-assisted photonic demodulator (CAPD) pixels has been developed. The BI pixel structure maximizes fill factor, allows for flexible transistor positioning, and makes the light path independent of the metal layers. The BI-CAPD pixel, which has relatively thinner substrate than that of front-illuminated CAPD, makes it possible to generate steeper electric-potential gradient in the vertical direction of the substrate, and results in 80% modulation contrast at 100-MHz modulation frequency. A higher modulation contrast contributes to suppress system power consumption as the sensor integration time for reaching specific signal-to-noise ratio is reduced. Alternatively, it contributes to improve depth accuracy when integration time is fixed.

Table-top water-window soft X-ray microscope using a Z-pinching capillary discharge source

The development and demonstration of a table-top transmission soft X-ray (SXR) microscope, using a laboratory incoherent capillary discharge source has been carried out. This Z-pinching capillary discharge water-window SXR source, is a first of its kind to be used for high spatial resolution microscopy at λ = 2.88 nm (430 eV) . A grazing incidence ellipsoidal condenser mirror is used for focusing of the SXR radiation at the sample plane. The Fresnel zone plate objective lens is used for imaging of the sample onto a back-illuminated (BI) CCD camera. The achieved half-pitch spatial resolution of the microscope approaches 100 nm, as demonstrated by the knife-edge test. Details about the source, and the construction of the microscope are presented and discussed. Additionally, the SXR images of various samples, proving applicability of such microscope for observation of objects in the nanoscale, are shown.

Modeling the spectral response for the soft X-ray imager onboard the ASTRO-H satellite

The ASTRO-H satellite is the 6th Japanese X-ray astronomical observatory to be launched in early 2016. The satellite carries four kinds of detectors, and one of them is an X-ray CCD camera, the soft X-ray imager (SXI), installed on the focal plane of an X-ray telescope. The SXI contains four CCD chips, each with an imaging area of 31mm×31mm, arrayed in mosaic, covering the field-of-view of 38′×38′, the widest ever flown in orbit. The CCDs are a P-channel back-illuminated (BI) type with a depletion layer thickness of 200μm. We operate the CCDs in a photon counting mode in which the position and energy of each photon are measured in the energy band of 0.4–12keV. To evaluate the X-ray spectra obtained with the SXI, an accurate calibration of its response function is essential. For this purpose, we performed calibration experiments at Kyoto and Photon Factory of KEK, each with different X-ray sources with various X-ray energies. We fit the obtained spectra with 5 components; primary peak, secondary peak, constant tail, Si escape and Si fluorescence, and then model their energy dependence using physics-based or empirical formulae. Since this is the first adoption of P-channel BI-type CCDs on an X-ray astronomical satellite, we need to take special care on the constant tail component which is originated in partial charge collection. It is found that we need to assume a trapping layer at the incident surface of the CCD and implement it in the response model. In addition, the Si fluorescence component of the SXI response is significantly weak, compared with those of front-illuminated type CCDs.

Indium gallium arsenide antimonide photodetector grown by liquid phase epitaxy: Electrical characterization and optical response

Current–voltage (I–V) and R0A curves and spectral response as a function of bias voltage and temperature of p–n indium gallium arsenide antimonide (In0.14Ga0.86As0.13Sb0.87)/n-GaSb photodiodes are presented. InGaAsSb quaternary alloys with a bandgap energy of about 653meV were grown using the liquid phase epitaxy technique on top of (100) GaSb substrates. Device structure was fabricated using a process that includes passivation with sodium sulfide, thermal annealing and metallizations. The diode architecture was a back-illuminated (B-I) structure with a ring-shaped metallic contact in the GaSb substrate face. Photodiode spectral response showed good performance in the entire temperature range between 20K and 300K.

Measurement of the soft X-ray response of P-channel back-illuminated CCD

P-channel charge-coupled devices (CCDs) made from N-type silicon wafers were originally developed for ground-based optical and near-infrared telescopes. The thick depletion layer of these CCDs provides the significant advantage of high quantum efficiency (QE) for hard X-rays. On the other hand, high QE for soft X-rays is obtained with back-illuminated (BI) and fully depleted CCDs in which only a thin dead layer exists on the surface of incidence. Thus, P-channel BI CCDs can be applicable as superior wide band X-ray detectors. We have developed such a device specifically for the Soft X-ray Imager (SXI) on board the X-ray astronomy satellite ASTRO-H, scheduled to be launched in 2014. We previously reported that the depletion layers of our CCDs, a prototype of SXI-CCDs, have a thickness of more than 200μm. In this paper, we report a novel soft X-ray response of P-channel BI CCDs. First, we irradiate fluorescent X-rays of O, F, Na, Al, Si and K to the SXI prototype. This experiment reveals that our CCD has a significant low-energy tail structure in the soft X-ray response. Since the intensity of the low-energy tail is larger for lower X-ray energies, the tail is originated on the CCD surface layer. Then, we fabricate a new type of CCDs by applying an alternative treatment to its surface layer. The soft X-ray response of the CCD is measured by irradiation of monochromatic X-rays from 0.25keV to 1.8keV in a synchrotron facility, KEK-PF. The intensity of the low-energy tail for 0.5keV incident X-ray is one order of magnitude smaller than that for the previous CCD. The same treatment will be applied to the surface layer of the SXI flight model.

SUZAKUOBSERVATIONS OF NEAR-RELATIVISTIC OUTFLOWS IN THE BAL QUASAR APM 08279+5255

We present results from three Suzaku observations of the z=3.91 gravitationally lensed broad absorption line quasar APM 08279+5255. We detect strong and broad absorption at rest-frame energies of <2 keV (low-energy) and 7-12 keV (high-energy). The detection of these features confirms the results of previous long-exposure (80-90 ks) Chandra and XMM-Newton observations. The low and high-energy absorption is detected in both the back-illuminated (BI) and front-illuminated (FI) Suzaku XIS spectra (with an F-test significance of >99%). We interpret the low-energy absorption as arising from a low-ionization absorber with log NH~23 and the high-energy absorption as due to lines arising from highly ionized (2.75 < log xi < 4.0; where xi is the ionization parameter) iron in a near-relativistic outflowing wind. Assuming this interpretation we find that the velocities in the outflow range between 0.1c and 0.6c. We constrain the angle between the outflow direction of the X-ray absorber and our line of sight to be <36 degrees. We also detect likely variability of the absorption lines (at the >99.9% and >98% significance levels in the FI and BI spectra, respectively) with a rest-frame time scale of ~1 month. Assuming that the detected high-energy absorption features arise from Fe xxv, we estimate that the fraction of the total bolometric energy injected over the quasar's lifetime into the intergalactic medium in the form of kinetic energy to be >10%.

Advancement in Back Illuminated Technology

not Available.

Front-illuminated versus back-illuminated photon-counting CCD-based gamma camera: important consequences for spatial resolution and energy resolution

Charge-coupled devices (CCDs) coupled to scintillation crystals can be used for high-resolution imaging with x-rays and gamma rays. When the CCD images can be read out fast enough, the energy and interaction position of individual gamma quanta can be estimated by a real-time image analysis of the scintillation light flashes (‘photon-counting mode’). The electron-multiplying CCD (EMCCD) is well suited for fast read out, since even at high frame rates it has extremely low read-out noise. Back-illuminated (BI) EMCCDs have much higher quantum efficiency than front-illuminated (FI) EMCCDs. Here we compare the spatial and energy resolution of gamma cameras based on FI and BI EMCCDs. The CCDs are coupled to a 1000 µm thick columnar CsI(Tl) crystal for the purpose of Tc-99m and I-125 imaging. Intrinsic spatial resolutions of 44 µm for I-125 and 49 µm for Tc-99m were obtained when using a BI EMCCD, which is an improvement by a factor of about 1.2–2 over the FI EMCCD. Furthermore, in the energy spectrum of the BI EMCCD, the I-125 signal could be clearly separated from the background noise, which was not the case for the FI EMCCD. The energy resolution of a BI EMCCD for Tc-99m was estimated to be approximately 36 keV, full width at half maximum, at 141 keV. The excellent results for the BI EMCCD encouraged us to investigate the cooling requirements for our setup. We have found that for the BI EMCCD, the spatial and energy resolution, as well as image noise, remained stable over a range of temperatures from −50 °C to −15 °C. This is a significant advantage over the FI EMCCD, which suffered from loss of spatial and especially energy resolution at temperatures as low as −40 °C. We conclude that the use of BI EMCCDs may significantly improve the imaging capabilities and the cost efficiency of CCD-based high-resolution gamma cameras.

Wide-band imaging spectrometer with scintillator-deposited charge-coupled device

We report on a new photon-counting detector possessing unprecedented spatial resolution and moderate spectral resolution for 0.1– 100 keV X-rays. It consists of an X-ray charge-coupled device (CCD) and a scintillator. The scintillator is directly coupled to the back surface of the X-ray CCD. Low-energy X-rays below 10 keV can be directly detected by the CCD. The majority of hard X-rays above 10 keV pass through the CCD but can be absorbed by the scintillator, generating visible light photons. Since CCDs have a moderate detection efficiency, visible light photons can be detected by the CCD. We coupled the needlelike CsI(Tl) on the front surface of the back-illuminated (BI) CCD. We evaluated the spectroscopic performance for hard X-rays at the synchrotron facility, SPring-8, and found a good linear relationship between the incident X-ray energy and the pulse height up to 80 keV. The on-axis image of the hard X-ray telescope, supermirror, was measured by our device at 40 keV. A sharp core and the wing structure can be clearly imaged and high imaging capability of the SD-CCD can be demonstrated.

High resolution X-ray photon-counting detector with scintillator-deposited charge-coupled device

We report on a new photon-counting detector possessing unprecedented spatial resolution and moderate spectral resolution for 0.1-100 keV X-rays. It consists of an X-ray charge-coupled device (CCD) and a scintillator. The scintillator is directly coupled to the back surface of the X-ray CCD. Low-energy X-rays below 10 keV can be directly detected by the CCD. The majority of hard X-rays above 10 keV pass through the CCD but can be absorbed by the scintillator, generating visible photons. We coupled needlelike CsI(Tl) on the front surface of the back-illuminated (BI) CCD. The high detection efficiency of BI CCDs in the visible band enables us to collect visible photons emitted from the CsI(Tl) efficiently, leading to the moderate spectral resolution of 28.4% at 22.1 keV and 25% at 59.5 keV. We also investigated the imaging capability of our device and demonstrated high resolution imaging with an accuracy of 10 /spl plusmn/3 /spl mu/m at 17.4 keV.

Measurement of the Charge Cloud Shape Generated in the Fully Depleted Back-Illuminated Charge-Coupled Device

The mesh experiment enables us to specify the X-ray point of interaction with a subpixel resolution. Miyata et al. [Jpn. J. Appl. Phys. 41 (2002) 5827] measured the charge cloud shape produced in a back-illuminated (BI) charge-coupled device (CCD). They found that there are two components of the charge cloud shape: a narrow component and a broad component. The size of the narrow component is 2.8–5.7 µm in units of standard deviation and strongly depends on the attenuation length of incident X-rays in Si, suggesting that it originates in the depletion region. On the other hand, the size of the broad component is roughly constant at ≃13 µm and does not depend on X-ray energies, suggesting that it originates in the field-free region. We applied the mesh experiment with the fully depleted BI CCD and found that there is a single component of the charge cloud with a size similar to that of the narrow component in Miyata et al.'s. We thus confirmed that the narrow and broad components originate in the depletion region and the field-free region, respectively.

Mesh experiment for back-illuminated CCDs in improvement of position resolution

We have carried out a mesh experiment with back-illuminated (BI) CCDs. BI CCDs possess the same structure to those of front-illuminated (FI) CCDs. X-ray photons entering from the back surface of a CCD produce a primary charge cloud far from the electrodes. The cloud expands by diffusion process until it reaches the potential well just below the electrodes. Since the diffusion time of the charge cloud is longer than in the FI CCD, a larger charge cloud is expected. The mesh experiment enables us to specify the X-ray point of interaction with a subpixel resolution. We measured the charge cloud shape in the BI CCD. We found that there are two components of the charge cloud: a narrow component and a broad component. The size of the narrow component is 2.8– 5.7 μm and strongly depends on the attenuation length in Si of the incident X-rays. The shorter the attenuation length, the larger the charge cloud. This result is qualitatively consistent with a model of diffusion inside the CCD. On the other hand, the size of the broad component is independent of X-ray energy of ∼13 μm . We conclude that the narrow component is originated in the depletion region whereas the broad component is in the field-free region. Based on the charge cloud shape obtained, we determined the X-ray point of interaction within a pixel and found that the position accuracy of our method could be achieved up to 1.1±0.1 μm . Moreover, the fraction of events used to improve the position resolution can be increased by an order of magnitude. We can thus develop a high-throughput superfine X-ray resolution imager.

Application of the Mesh Experiment for the Back-Illuminated Charge-Coupled Device: I. Experiment and the Charge Cloud Shape

We have employed a mesh experiment for back-illuminated (BI) charge-coupled devices (CCDs). BI CCDs possess the same structure as front-illuminated (FI) CCDs. Since X-ray photons enter from the back surface of the CCD, a primary charge cloud is formed far from the electrodes. The primary charge cloud expands through a diffusion process until it reaches the potential well that is just below the electrodes. Therefore, the diffusion time for the charge cloud produced is longer than that in the FI CCD, yielding a larger charge cloud size than expected. The mesh experiment enables us to specify the X-ray point of interaction with subpixel resolution. We measured a charge cloud shape produced in the BI CCD. We found that there are two components of the charge cloud shape having different sizes: a narrow component and a broad component. The size of the narrow component is 2.8–5.7 µm in units of standard deviation and strongly depends on the attenuation length of incident X-rays in Si. The shorter the attenuation length of the X-rays, the larger the charge cloud. This result is qualitatively consistent with a diffusion model inside the CCD. On the other hand, the size of the broad component is roughly constant at ≃13 µm and does not depend on X-ray energies. Judging from the design value of the CCD and the fraction of each component, we conclude that the narrow component has its origin in the depletion region whereas the broad component originates in the field-free region.

Modeling charge transfer inefficiency in the Chandra Advanced CCD Imaging Spectrometer

The front-illuminated (FI) CCDs in the Advanced CCD Imaging Spectrometer (ACIS) on the Chandra X-ray Observatory ( Chandra) were damaged in the extreme environment of the Earth's radiation belts, causing charge traps that result in enhanced charge transfer inefficiency (CTI) during parallel readout. This causes row-dependent gain, event grade ‘morphing’ (spatial redistribution of charge) and energy resolution degradation. The ACIS back-illuminated (BI) CCDs also exhibit pronounced CTI due to their manufacturing. It is mild enough that position-dependent energy resolution is not seen, but it is present in both parallel and serial registers. This CTI also changes the gain and event grades, in a spatially complicated way as parallel and serial CTI interact. Given these realities, we have developed and tuned a phenomenological model of CTI for both FI and BI CCDs and incorporated it into our Monte Carlo simulations of the ACIS CCDs. It models charge loss and the spatial redistribution of charge (trailing), thus reproducing the spatially dependent gain and grade distribution seen in all ACIS CCDs and the row-dependent energy resolution seen in the FI devices. Here we explore the evidence for CTI, compare our simulations to data, and present a technique for CTI correction based on forward modeling.