This paper reports the development of aluminum-gallium nitride (AlGaN or Al_xGa_1-xN) photodiode technology for high-operability 256×256 hybrid Focal Plane Arrays (FPAs) for solar-blind ultraviolet (UV) detection in the 260-280 nm spectral region. These hybrid UV FPAs consist of a 256×256 back-illuminated AlGaN p-i-n photodiode array, operating at zero bias voltage, bump-mounted to a matching 256×256 silicon CMOS readout integrated circuit (ROIC) chip. The unit cell size is 30×30 μm². The photodiode arrays were fabricated from multilayer AlGaN films grown by MOCVD on 2" dia. UV-transparent sapphire substrates. Improvements in AlGaN material growth and device design enabled high quantum efficiency and extremely low leakage current to be achieved in high-operability 256×256 p-i-n photodiode arrays with cuton and cutoff wavelengths of 260 and 280 nm, placing the response in the solar-blind wavelength region (less than about 280 nm) where solar radiation is heavily absorbed by the ozone layer. External quantum efficiencies (at V=0, 270 nm, no antireflection coating) as high as 58% were measured in backilluminated devices. A number of 256×256 FPAs, with the AlGaN arrays fabricated from films grown at three different facilities, achieved response operabilities as high as 99.8%, response nonuniformities (σ/μ) as low as 2.5%, and zero-bias resistance median values as high as 1×10¹⁶ ohm, corresponding to R₀A products of 7×10¹⁰ ohm-cm². Noise Equivalent Irradiance (NEI) data were measured on these FPAs. Median NEI values at 1 Hz are 250-500 photons/pixel-s, with best-element values as low as 90 photons/pixel-s at 1 Hz.

Recently there has been a great deal of interest in the growth of dilute nitride quaternary alloys, such as InGaNAs, on GaAs substrates for the fabrication of GaAs-based components and optoelectronic integrated circuits. The addition of indium to the binary compound GaAs produces a ternary with a lower bandgap and larger lattice constant. The incorporation of nitrogen in this ternary further decreases the bandgap while reducing the lattice constant. This makes it possible to grow material lattice-matched to a GaAs substrate but with a narrower bandgap offering the possibility of growing materials suitable for opto-electronic devices on a GaAs substrate while operating at wavelengths used in long-distance optical communications. These devices can then be integrated with mature GaAs device technologies (MESFET, HBT) in photoreceivers and receivers/transmitters for improved functionality and reliability, lower cost, reduced size, etc. We have fabricated metal-semiconductor-metal (MSM) photodetectors on 1-μm thick In _.1Ga_.9N_.03As_.97 epilayers, a composition that results in a bandgap in the 1.3 μm region. We report on the DC characteristics, frequency dependence and wavelength dependence of the photoresponse. The results are compared to MSMs fabricated on GaAs. The temporal response is not as fast as that of GaAs MSMs and may be related to low carrier mobility. This shortcoming has been reported as the cause for the lower-than-expected efficiency of solar cells fabricated using this quarternary. The effect of growth conditions and thermal processing on detector characteristics such as bandwidth and dark current were investigated. The challenges associated with the use of InGaNAs in photodetectors (such as defects, response speed, requirement for thermal anneal) will be discussed.

The performances of a pin versus a pn structure from GaInAsSb materials operating at room temperature are compared both from a theoretical point of view and experimentally. Theoretically, it is found in materials limited by generation-recombination currents, pn junctions have a higher D* than pin junctions. The thinner depletion region of pn junctions results in a lower responsivity but a higher dynamic resistance, giving an overall higher D* compared to a pin structure. A series of five p+pn+ Ga_0.80In_0.20As_0.18Sb_0.82 detector structures latticed matched to GaSb substrates and with 2.37 μm cut off wavelength were grown by molecular beam epitaxy and processed into variable size mesa photodiodes. Only the doping of the absorbing (p) region was varied from sample to sample, starting with nominally undoped (~1x10¹⁶ cm^-3 pbackground doping due to native defects) and increasing the doping until a p+n+ structure was attained. Room temperature dynamic resistance-area product R0A was measured for each sample. A simple method is presented and used to disentangle perimeter from areal leakage currents. All five samples had comparable R0A's. Maximum measured R0A was 30 Ω-cm² in the largest mesas. Extracted R0A's in the zero perimeter/area limit were about ~50 Ω-cm² (20-100 Ω-cm²) for all samples. Within uncertainty, no clear trend was seen. Tentative explanations are proposed.

In this study, we examine processes limiting the performance of 4 micron superlattice pin photodiodes for different temperature and mesa size regimes. We show that the performance of large mesa photodiodes at low temperature is most severely limited by a trap-assisted tunneling leakage current (x300), while small mesa sizes are additionally limited by perimeter leakage (x20). At room temperature, large mesa photodiodes are limited by the diffusion current, and small mesa photodiodes are further limited by the perimeter leakage (x100). To reduce or eliminate the impact of perimeter leakage, we have tried passivating the mesa sidewalls with SiN, an approach that was only minimally successful. We have also laid the groundwork for another approach to elimination of perimeter leakage currents, namely, elimination of the sidewalls altogether through planar processing techniques. Planar processing schemes require the deposition of a thick, wide bandgap semiconductor or "window layer" on top of the homojunction. We compare the performance of two otherwise identical InAs/GaSb superlattice homojunction detectors, except one with a GaSb window layer, and one without. We show that inclusion of the thick GaSb window layer does not degrade detector performance.

The noise characteristics associated with dark current, photoconductive gain (PC), capture probability in doped InAs dots embedded in In_0.1Ga_0.9As/GaAs spacer layer have been proposed. The photoconductive and photovoltaic behaviors of the InAs/GaAs quantum dot infrared photodetector (QDIP) from the intersubband transition measurements are also clearly observed. Through noise measurement in dynamic signal analyzer (HP35670A) ¹, the electronic bandpass filter frequencies are set up ranging from 3 to 10 KHz in a low noise current preamplifier (SR570) ². The lock-in amplifier (SR830) ³ can be also used to measure and calibrate the noise density by means of the mean average deviation (MAD) contrast with noise spectra from HP35670A. The InAs/GaAs QDIP studied in this work belongs to n⁺-n-n⁺ structure with the top and free blocking barrier layers. It is observed that the owing blocking layer of QDIP not only suppress dark current successfully but also probably reduce the photocurrent ^4-6. By systematically optoelectronic measurements and simulations, the modified model of noise current, photoconductive gain, and capture probability in the quantum devices have been proposed. It is shown that photoconductive gain is almost independent of bias under the lower bias, then increasing exponentially under higher bias and below the temperature of 80K. In contrast to quantum well infrared photodetector (QWIP), a higher photoconductive gain of the quantum dot infrared photodetector has been demonstrated and attributed to the longer lifetimes of excited carriers in quantum dots ^7-10. At 80K, a photoconductive gain of tens of thousand is shown in the regions of higher biases. It is clear to note that the highest detectivity of the QDIP surprisingly approach to 3.0×10¹² cmHz^1/2/W at -0.6V under measured temperature 20 K. Under 80K, the average D* is obtained ~10¹⁰ cmHz^1/2/W. To our knowledge, this is the one of highest D* data in the world.

We report the development of a photon-counting solid-state photomultiplier that consists of an array of Geiger mode CMOS avalanche photodiodes (APDs). The detector is based on the design described by Buzhan et. al.¹ in which the individual outputs of an array of Geiger APDs are coupled together to drive a common output signal. The total output signal is a sum of the Geiger outputs of each individual pixel in the array. For a large array, the sum of the signals from the discrete pixels producess an analog representation of the flux on the detector. In this report we describe our most recent measurements of the spectral response and noise characteristics of the individual detector elements. We present results for a 14 element array of Geiger mode pixels that is used as a solid state photomultiplier (SSPM). We use this SSPM to create a prototype radiation detector that can identify the source based on the energy of the emitted radiation.

Until very recently Single Photon Avalanche Diodes (SPAD), which yield high detection efficiency in the visible spectrum, provided poor timing performance. This paper will review the current state of the SPAD technology and review new SPAD developments that provide: sub 50ps-timing resolution, are stable with count rate, and yield high detection efficiency. Examples will be provided; comparing timing resolution of PMT's and solid-state photon counting modules, effect of count rate on timing resolution, thus illustrating the stability of these newly developed SPAD's. In addition, the paper will review the basics of photon counting using SPAD's and illustrate how these SPAD's are used in Time-Correlated Single Photon Counting (TCSPC) and the results from these experiments.

Previous generation low light detection platforms have been based on the photomultiplier tube (PMT) or the silicon single photon counting module (SPCM) from Perkin Elmer¹. A new generation of silicon CMOS compatible photon counting sensors are being developed offering high quantum efficiency, low operating voltage, high levels of robustness and compatibility with CMOS processing for integration into large format imaging arrays. This latest generation yields a new detector for emerging applications which demand photon counting performance providing high performance and flexibility not possible to date. We describe a 4-channel photon detection platform, which allows the use of 4 separate photon counting detectors in either free space or fibre-coupled mode. The platform is scalable up to 16 channels with plug in modules allowing active quenching or Peltier cooling as required. A graphical user interface allows feedback and control of all device parameters. We show a novel ability to integrate separate detection modules to extend the dynamic range of the system. This allows a PIN or APD mode detector to be used alongside sensitive photon counting detectors. An advanced FPGA and microcontroller interface has been designed which allows simultaneous time binning of counting rates and readout of the analog signals when used with linear detectors. This new architecture will be discussed, presenting a full characterization of count rate, quantum efficiency, time binning and sensitivity across the broad spectrum of light flux applicable to PIN diodes, APDs and Geiger-mode photon counting sensors.

The operation and performance of multi-pixel, Geiger-mode APD structures referred to as Silicon Photomultiplier (SPM) are reported. The SPM is a solid state device that has emerged over the last decade as a promising alternative to vacuum PMTs. This is due to their comparable performance in addition to their lower bias operation and power consumption, insensitivity to magnetic fields and ambient light, smaller size and ruggedness. Applications for these detectors are numerous and include life sciences, nuclear medicine, particle physics, microscopy and general instrumentation. With SPM devices, many geometrical and device parameters can be adjusted to optimize their performance for a particular application. In this paper, Monte Carlo simulations and experimental results for 1mm² SPM structures are reported. In addition, trade-offs involved in optimizing the SPM in terms of the number and size of pixels for a given light intensity, and its affect on the dynamic range are discussed.

The many advantages of silicon such as low cost, abundancy and a level of maturity that allows for very large scale integration, means that silicon is the most commonly used semiconductor in microelectronics and optoelectronic devices. Silicon, however, has one disadvantage, this being that it is unable to absorb light greater than 1100 nm. The two primary telecommunications wavelengths, 1300 nm and 1550 nm, can therefore not be detected. An interesting method used to extend silicon's wavelength range is the formation of black silicon on the silicon surface. Black silicon is formed when gases that are passed over the silicon react and etch the silicon surface, forming a dark spiky pattern. When light is shone on such a pattern, it repeatedly bounces back and forth between the spikes thus reducing surface reflection and trapping the light. This reduced reflectance and light trapping increases the sensitivity of the silicon to long wavelengths and makes it viable for use in a wide range of commercial devices such as infrared detectors and solar cells. This paper presents novel black silicon PIN photodiodes of various sizes (25 mm², 4 mm² and 1 mm²). The diodes have been extensively characterized at wafer level, with breakdown voltage, dark current, shunt resistance, threshold voltage and junction capacitance measurements being made. Extensive responsivity measurements were also performed and it was established that the black silicon surface resulted in responsivity increases of greater than 50 % at long wavelengths (≈ 1100 nm).

Noise characteristics of the backlit, pin photodiode arrays having different vertical structures were studied. We showed that in many cases, the non-optical crosstalk between adjacent elements determines the noise performance and detectivity of the array pixels. For the arrays with the structure described in our recent works, the crosstalk always remained well below 0.01%, which allowed reaching the minimum noise level of ~ 10^-15 A/&sqrt;Hz determined by the thermal noise. In contrast, for the arrays built applying conventional structures the crosstalk was two orders of magnitude higher, which noticeably decreased the sensitivity of the pixels increasing their noise and switching their operation towards background-limited performance. The background signal originated from the non-optical crosstalk and produced a noise level significantly higher that the thermal noise. We also compared the temperature coefficients for different arrays. For the structures with improved electrical crosstalk, the measured value of the shunt resistance temperature coefficient was typically below 8 %/C and the responsivity temperature coefficient value did not exceed +0.02 %/C within the spectral range from 450 through 800 nm. The advantages and drawbacks of application of the reported in this work photodiode arrays in high quality imaging systems are discussed.

The charge injection device, CID25, is presented. The CID25 is a color video imager. The imager is compliant with the NTSC interlaced TV standard. It has 484 by 710 displayable pixels and is capable of producing 30 frames-per-second color video. The CID25 is equipped with the preamplifier-per-pixel technology combined with parallel row processing to achieve high conversion gain and low noise bandwidth. The on-chip correlated double sampling circuitry serves to reduce the low frequency noise components. The CID25 is operated by a camera system consisting of two parts, the head assembly and the camera control unit (CCU). The head assembly and the CCU can be separated by up to 150 meter long cable. The CID25 imager and the head portion of the camera are radiation hardened. They can produce color video with insignificant SNR degradation out to at least 2.85 Mrad of total dose of Co⁶⁰ γ-radiation. This represents the first in industry radiation hardened color video system, based on a semiconductor photo-detector that has an adequate sensitivity for room light operation.

We are developing an ultrahigh-speed, high-sensitivity broadcast camera that is capable of capturing clear, smooth slow-motion videos even where lighting is limited, such as at professional baseball games played at night. In earlier work, we developed an ultrahigh-speed broadcast color camera¹) using three 80,000-pixel ultrahigh-speed, highsensitivity CCDs²). This camera had about ten times the sensitivity of standard high-speed cameras, and enabled an entirely new style of presentation for sports broadcasts and science programs. Most notably, increasing the pixel count is crucially important for applying ultrahigh-speed, high-sensitivity CCDs to HDTV broadcasting. This paper provides a summary of our experimental development aimed at improving the resolution of CCD even further: a new ultrahigh-speed high-sensitivity CCD that increases the pixel count four-fold to 300,000 pixels.

X-ray Photoelectron Spectroscope (XPS) was applied to investigate the fresh surface of CdTe crystal surface, prepared by traditional bromine methanol acid etching process. A thin layer of tellurium precipitate had been discovered covering the etched CdTe surface, which is very harmful for the final X-ray detector performance. A special kind of solution had been developed to remove this tellurium precipitates layer for a real fresh CdTe based crystal surface achievement. Further XPS experimental results confirmed that by the application of this solution, the surface tellurium precipitate had been eliminated. Traditional photolithograph process had been employed to fabricate X-ray two-dimensional detector arrays on the tellurium precipitate free CdZnTe surface. Various kinds of two-dimensional CdZnTe X-ray detector arrays, such as 3×3, 4×4 and 32×32 arrays had been fabricated. 10μm width guard rings are precisely achieved for some detector elements. The high-resolution microscope inspection proves that the pattern precision of 0.1μm is achieved on CdZnTe bulk crystal surface. Edge effecting of the photolithograph has been eliminated.

The birefringence of the yttrium vanadate (YVO₄) crystals have been measured for the first time in the middle wavelength infrared (MWIR), i.e., 3-5μm. A Fourier transform spectrometer has been used in the channel spectra technique to achieve a quick measurement. Large birefringence over 0.21 has been observed in the YVO₄ crystals. The transmission over 2.5-17μm of the YVO₄ (0.7mm thick) has been measured, which showed a transparent range up to 5.3μm. These unique features illustrate the good potential of this material for the MWIR polarization applications.

We have designed and developed a new family of photodetectors with Internal Discrete Amplification (IDA) mechanism They operate as solid state photomultiplier devices at room temperature and may be used in numerous applications where high bandwidth of the detector is necessary in combination with maximum sensitivity and low excess noise. The photodetectors can operate in linear detection mode with gain-bandwidth product up to 5 times 10¹⁴ as well as in photon counting mode with count rates up to 10⁸ counts/sec. The key performance characteristics exceed those of Photomultiplier Tube (PMT) and Avalanche Photodiode (APD) devices. The detectors have gain > 10⁵, excess noise factor as low as 1.03, photoresponse rise/fall time < 300 ps, and timing resolution (jitter) < 200 ps. The combination of low excess noise at high gain and wide bandwidth, as well as scalability to large active areas, presents the main advantages of this technology over conventional photodetector solutions. Ultra low excess noise is one of the main features of the internal Discrete Amplification Detector (DAD), and in this paper its nature has been investigated more comprehensively. We investigated the behavior of the noise-factor and afterpulsing, and conclude that both have the same physical nature. Optical cross-talk between channels is shown to be responsible for the afterpulsing phenomenon, and, in turn, is the main source of excess noise. Thus, the noise characteristics of an DAD device and its timing resolution may be significantly improved as they are limited not by the discrete amplifier channel properties itself, but by the cross-talk, which strongly depends on the device design.

A novel top-illuminated In0.53Ga0.47As p-i-n photodiodes, with the partially p-doped photoabsorption layer, grown on GaAs substrate by using a linearly graded metamorphic In_xGa_1-xP (x graded from 0.49 to 1) buffer layer is reported. The dark current, optical responsivities, noise equivalent power, and operational bandwidth of the MM-PINPD with aperture diameter of 60 μm are 13 pA, 0.6 A/W, 3.4 times 10^-15 W/Hz^1/2, and 8 GHz, respectively, at 1550 nm. Under the illumination of 1.2-ps pulse-train, the measured impulse response is 41 ps and the frequency bandwidth is up to 8 GHz with heterodyne beating measurement. The low cost InGaAs photodiode with high current bandwidth product (350 mA times GHz, at 10 GHz) and bandwidth-efficient product (4.8 GHz times A/W) have been achieved.