The MIR regime is extremely interesting for hyperspectral imaging and chem-bio sensing because most chemical and biological toxins have their fingerprints in this range. Our MIR materials and devices create a planar, integrated, Si-CMOS-compatible microphotonics platform which enables on-chip imaging and sensing. Challenges and opportunities associated with planar integration of non-traditional glass materials with silicon photonic components such as light sources, sensors, detectors, and imagers are presented. Key attributes of glasses useful to integrated photonic device applications are presented. Technically important glass compositions are the â?onetwork glasses", in which atoms are bound to each other by covalent bonds to form a glass network. Devices are demonstrated on silicates (which include silica or silicon dioxide), heavy metal oxides, tellurites, as well as non-oxide glasses such as chalcogenides. For imaging, we demonstrate low-cost resonant-cavity enhanced narrow-band photon detectors for enabling large-area multi-spectral infrared digital cameras for advanced detection and intelligence applications. Novel IR sensitive material and resonant-cavity-enhanced (RCE) photodetector devices are demonstrated using polycrystalline PbTe films. A universal design theory has been established for multispectral detection, as our fully functional mid-IR RCE photodetector capable of monolithic integration with Si ROIC has demonstrated. For sensing, we demonstrate a planar glass-on-silicon-based highly sensitive and selective chem-bio sensor that integrates several key photonic and electronic functions. Optical resonator devices that dramatically amplify photon-molecule interactions enabling highly sensitive detection are demonstrated. Addition of polymer coatings is shown to enhance their selectivity. Lift-off patterning of high-index-contrast (HIC) chalcogenide glass devices is used to demonstrate the first micro-ring and micro-disk resonators in chalcogenide glasses, with a record loaded cavity quality factor (Q-factor) exceeding 2 Ã- 105. As the logical next step towards monolithically integrated imagers and sensors we demonstrate photonic components for the mid-IR regime, including emitters, couplers and waveguide-integrated detectors. The emitters are based on (a) Er-TeO2 sputtered films, and (b) quantum dot-doped chalcogenide glass hybrid films. Photonic crystal waveguide structures on these films not only manipulate light propagation within a plane, but also emit vertically upwards from within a plane into the air. Coupling between an external mid-IR Interband Cavity Laser (ICL) source and a chalcogenide waveguide has been demonstrated, paving the way for an integrated source and sensor element. A PbTe thin film integrated with a waveguide serves as an on-chip detector element. Our vision is to enable future technology adoption on the silicon-based photonic platform, relevant to the fields of mid IR imaging and sensing.

Several groups are investigating the use of subretinally-implanted silicon photodiode arrays to treat age-related macular degeneration (AMD) and retinitis pigmentosa (RP), diseases causing death of photoreceptors, while the inner retinal neurons survive to a large extent. The photodiode array aims to electrically stimulate these surviving neurons. Most approaches rely on wires and cables to deliver power to the stimulating array, requiring complex surgical procedures and creating a path for serious infections. In our approach the device operates in photovoltaic mode, utilizing light from video goggles as a source of information and energy. Each pixel in the two-dimensional array independently converts pulsed infrared light into electrical current that stimulates the nearby inner retinal neurons. To enhance the charge injection levels, each pixel contains three photodiodes connected in series to provide a photovoltage of ~1.8 V. An active and a return electrode in each pixel provide localized current flow. The wavelength of the stimulating light was chosen to be about 900 nm to avoid interference with any remaining vision of the patient. The weak absorption of silicon in this wavelength range dictates much of the device structure and process. The silicon thickness is chosen to be 30 μm â?" sufficient to absorb a significant fraction of the light, but still thin enough to allow practical fabrication and to fit underneath the retina. To obtain the complete electrical isolation required for series-connected photodiodes, deep trenches are etched through the 30 µm thick, lightly boron-doped, Si device layer of an SOI wafer. The trenches are insulated with oxide and filled with undoped polysilicon for planarization. n- and p- type dopant atoms are added to form the photodiode and contact regions, and the structure is metallized. The active and return electrodes are sputter coated with iridium oxide to provide high charge injection. The polysilicon is removed from the trenches between pixels to allow nutrients to flow through the structure to nourish the retina placed adjacent to the array. Finally, the arrays are released from the handle wafer. Three sizes of pixels (280 μm, 140 μm, and 70 μm) with active electrode diameters of 80 μm, 40 μm and 20 μm were fabricated. The reverse-bias dark current is adequately low (10-100pA), and the reverse breakdown voltage is sufficiently high (>20V). The turn-on voltages of one-, two- and three series-connected photodiode structures are approximately 0.6V, 1.2V and 1.8V, respectively. The measured photo-responsivities per diode at 880nm wavelength for large, medium and small pixels are about 0.33 A/W, 0.4 A/W and 0.36 A/W, at zero bias voltage. The measured photocurrent scales with exposed area at a rate of about 2 mA/mm2 at 6 mW/mm2 light intensity. The measured quantum efficiency is about 47%.

Mid-infrared wavelength region offers a plethora of possible applications ranging from sensing, medical diagnostics, and free space communications, to thermal imaging and IR countermeasures. Hence group IV mid-infrared photonics is attracting more research interest lately. Sensing is especially attractive area as fundamental vibrations of many important gases are found in the 3 to 14 μm spectral region. To realise group IV photonic mid-infrared sensors several serious challenges need to be overcome. The first challenge is to find suitable material platforms for the mid-infrared. In this paper we present experimental results for passive mid-infrared photonic devices realised in silicon-on-insulator (SOI), silicon-on-sapphire (SOS), and silicon on porous silicon (SiPSi). Although silicon dioxide is lossy in most part of the mid-infrared, we have shown that it has potential to be used in the 3-4 μm region. We have characterised SOI waveguides with < 1 dB/cm propagation loss. We have also designed and fabricated SOI passive devices such as MMIs and ring resonators. For longer wavelengths SOS or SiPSi structures could be used. An important active device for long wavelength group IV photonics will be an optical modulator. We present relationships for the free-carrier induced electro-refraction and electro-absorption in silicon and germanium in the mid-infrared wavelength range. Electro-absorption modulation is calculated from impurity-doping spectra taken from the literature, and a Kramers-Kronig analysis of these spectra is used to predict electro-refraction modulation. We have examined the wavelength dependence of electro-refraction and electro-absorption, and found that the predictions suggest longer-wave modulator designs will in many cases be different than those used in the telecom range.

We developed a monolithic sensing plate for a waveguide-mode sensor. For fabrication of the sensing plate, we used a silicon-on quartz (SOQ) substrate comprising a single crystalline Si (100) layer on a SiO2 glass substrate. The plate consists of a SiO2 glass substrate and a thin silicon layer the surface of which is thermally oxidized to form a SiO2 glass waveguide. We confirmed that the sensing plate is suitable for high-sensitivity detection of molecular adsorption at the waveguide surface. In addition, a significant enhancement of the sensitivity of the sensor was achieved by perforating the waveguide with holes with diameters of a few tens of nanometers by selective etching of latent tracks created by swift heavy-ion irradiation. As a next step, we have developed an optical system designed for detecting colored nano-materials in aqueous solutions, using the concept of evanescent-field-coupled waveguide-mode sensors. We found that the waveguide modes induced in the sensor are intrinsically sensitive to a change in optical absorption. The system detects less than one gold nano-particle with a diameter of 20 nm adsorbed per square micrometer. It is also demonstrated that significant signal enhancement due to adsorption of molecules is achieved using a dye. The system is expected to be applied as an effective sensing tool for metal colloids, nano-particles and colored bio-molecules in solution. Influenza viruses were detected with antibodies using a waveguide-mode sensor. Since hemagglutinins (HA) are the major protein existing at the surface of influenza viruses, we used anti-HA antibodies as the capturing molecules that were immobilized on the silica-based sensor chip. By applying this method, HA from avian influenza virus (H5N1) and human influenza virus H3N2 (A/Panama/2007/1999) were detected. An intact human influenza virus H1N1 (A/Brisbane/59/2007) was also detected. This kind of sensor development should pave the way for the development of a method to diagnose the presence of viruses at earlier stages, and therefore will be useful in the monitoring and forecasting of influenza epidemics. This study was partly supported by Industrial Technology Research Grant Program in 2009 from New Energy and Industrial Technology Development Organization (NEDO) in Japan.

Resonator-based optical biosensors have been the subject of long and successful research. However, while high sensitivity and selectivity have been demonstrated repeatedly, most of the designs studies so far rely on delicate, waveguide-based coupling schemes. Furthermore, in order to characterize the optical resonances, external optical devices such as a tunable laser or a spectrometer are usually required. Thus, the entire sensing apparatus is rather bulky and expensive, and difficult to employ for a low-cost, on-site diagnostic applications that are becoming more important. Here we report on results of integrating a self-luminescent microdisk resonator, a slab waveguide, and an array of diffraction gratings to realize a device that can integrate much of an optical biosensor in a simple, on-chip structure with a size that is less than 200 μm Ã- 200 μm. 300 nm thick Al2O3 and 100 nm thick SiO2 films were deposited on a 10μm â?" thick thermal oxide wafer with RF-sputtering system, and annealed at 1100 to remove the defects in the deposited layers and to crystallize the Al2O3 layer. Since the crystallized Al2O3 is transparent for visible light and has a higher refractive index than SiO2, it thus forms a buried slab waveguide. Afterwards, the SiO2 layer was patterned to form a circular array of 31 diffraction gratings to vertically deflect the light guided in the Al2O3 layer. The diffraction gratings are designed to provide resonances ranging from 600 nm to 700 nm with FWHM of 10 nm. Finally, a 10 μm diameter microdisk was created in the center of the array of diffraction gratings using CdSe/ZnS quantum-dot doped SU-8. The quantum-dot doped SU-8 microdisk supports whispering gallery modes when excited vertically using a broadband UV source. These modes are then coupled to the Al2O3 slab waveguides, spread out radially from the disk, and are eventually deflected vertically by the circular array of diffraction gratings. The degree of coupling is easily controlled by the SiO2 layer thickness. As a result, spectroscopy of the microdisk resonator, can be performed by simply directly observing the intensity distribution of the deflected light along the array of diffraction gratings, and any changes in the optical mode discerned by observing the changes in the intensity distribution. The application of using this integrated resonator/spectrometer for biosensing will also be discussed.

A silicon molecular microarray system is being developed at the National Research Council Canada, using photonic wire evanescent field (PWEF) affinity binding sensors. This microarray technology provides a low cost, disposable, label free sensor chip for multiplexed parallel real-time molecular binding measurements and multiparameter analysis for research, monitoring, and molecular detection. Detection is based on the interaction of light with captured molecules at the surface of a silicon photonic wire optical waveguide. Theoretical analysis shows that the sensitivity can be at least an order of magnitude better than comparable molecular binding tools, yet every sensor element is 100 um or less in diameter. The sensor chips are 6 Ã- 9 mm in size and versions with 16, 64 and 128 sensors per chip have been fabricated. The sensor elements on a chip can each be independently functionalized for specific molecular targets using commercial microarray spotting tools. A complete instrumentation system for interrogating silicon photonic wire microarray chips has been built. The chip can be inserted into the instrument, automatically aligned with input optics and fluid connections, and be ready for measurement within a few minutes. The system can read 16 or more photonic wire sensors on a single silicon chip simultaneously and in real time, while delivering sample fluid to the sensors through microfluidic channels that are monolithically integrated on the chip. Light is coupled on and off the chip by arrays of subwavelength gratings that are fabricated in the same step as the waveguide sensors. The sensor outputs are acquired in parallel by an imaging InGaAs detector array. This talk will review the optical elements integrated on the chip, describes design of the reader instrument, and will present selected results of application demonstrations in areas such as pathogen testing and chemical sensing.

A slot-structure that consists of regions of low-refractive index sandwiched between regions of high refractive index supports a slot-mode which provides a highly enhanced electric field in the low refractive index region. As the confinement volume is determined by the thickness of the low-index region and therefore can be nm-thin, the slot structure has been suggested to be a promising candidate for non-linear optical devices and sensors. So far, most of the work on slot structures focused on vertical slots. However, nm-scale slots with smooth sidewalls necessary for high performance can be difficult to fabricate in such a structure. Recently, we have shown that by using deposition with selective etching, horizontal slot microdisk resonators with air slots that are only tens of nm thick with smooth sidewalls can easily be fabricated, and applied for bio-sensing. However, the presence of air-slot complicates the optical properties of the resonator in a non-intuitive way. Thus, while the potential for highly sensitive biosensing using horizontal slot resonator has been demonstrated, the optimum design for bio-sensing applications has not yet been investigated. Here we report on the results of investigating the effects of slot geometry on sensitivity and Q-factors of horizontal slot microdisk resonators. Simulations are performed for 4um Si and 8um SiN disks in water and air circumstance with various disk thickness, gap thickness and number of slots. We find that in our case, the waveguide sensitivity is largely different from resonator sensitivity, since group index of waveguide does not affect the effective index shift of resonator. Only slot-mode overlap, defined by portion of electric field energy in a sensing region, and wavelength determines resonance peak shift. The slot-mode overlap has distinct local maxima as the disk thickness is increased due to pulling-in of the electric field from the top and bottom surface of the microdisk, but increases with increasing slot thickness. Furthermore, rapid oscillation of the slot-mode overlap can occur due to crossing of vertically second order TE modes. Surprisingly, a slot structure does not offer any advantage in bulk sensing over other evanescent field sensing methods. Only in the case of surface sensing does the large field enhancement result in sensitivity enhancement by a factor of about 10, and another factor of 3 enhancement is possible by employing multiple slots. Fabrication and characterization of such optimized slot sensors will also be presented

In recent years, Si-based optical biosensors have been studied intensely for compact integration on a chip. A key factor that determines the performance of any optical sensor is the overlap of the optical mode with sensing area. However, the need to guide the signal light can limit the mode overlap, and thus the performance, of such a sensor. Recently, â?oslot structuresâ? that rely on a thin â?oslotâ? of low refractive index sandwiched between regions of high refractive index have been shown to greatly enhance the E-field in the slot due to the requirement for the continuity of D-field across a dielectric interface, and thus lead to a far greater sensitivity than is possible with conventional, evanescent-field based sensors. Still, most of the works were done in the IR range due to the use of Si. Furthermore, tunable lasers were used as the signal source for the sensing process. Such an approach, while widely used, can be highly problematic for a practical optical biosensor. First of all, most of biosensing is done in aqueous environment where IR light is strongly absorbed. More significantly, the need to use an expensive, large tunable laser for the light source can render the cost- and size advantages of integrated, Si-based optical biosensor irrelevant. If such a Si-based optical biosensors can be made to emit visible light by itself, however, such problems can be solved completely. Recently, we have demonstrated that by using Tb3+ doped SiON microdisk resonators and monitoring its emission peaks, it is possible to achieve on-chip optical biosensing in the visible range without external signal source or delicate coupling setups. In this paper, we design and fabricate a self-luminescent horizontal slot microdisk resonator integrated with Si nanocrystals (Si-nc) to simultaneously achieve both high sensitivity of slot structure and the visible luminescence of Si-nc. Using the finite-difference time-domain (FDTD) method, we calculated the surface sensitivity air-slotted microdisk resonator to be ~2 nm/(nm*RIU) in the 700-800 nm range. Actual surface sensing of biomaterials using side-PL measurement will be presented.

The modern world is entirely dependent on the computer/microelectronic industry, which has relied on silicon to develop the integrated circuit and a multitude of other electronic devices. It has become apparent that this superb infrastructure could be used for other applications. This presentation will focus on label-free optical biosensing systems developed in the SOI platform and capable of detecting a single virus. The biosensing platform consists of 2-D photonic crystal (PhC) microcavities built in the silicon-on-insulator platform. The sensors are functionalized to capture the appropriate target with a negligible false positive response. This platform allows for the detection of minute amounts of biological matter (DNA, proteins) and also a single virus. We will discuss the design and mode of operation of various 2-D PhC microcavity biosensors and illustrate their performance. The presentation will conclude with a discussion of system-level considerations that are often neglected yet need to be taken into consideration before such devices can be deployed.

In this report, we propose a partial slotted one-dimensional silicon photonic crystal for biochemical sensing. An air slot is introduced into the one-dimensional photonic crystal that contains period air holes. To avoid extra light scattering from the slot, only the defect region of this photonic crystal is implemented by the slot. The waveguide width is 500 nm and the slot width is 100 nm. 2D-FDTD simulation result indicates that the TE-like cavity mode can be confined within the slot region, which could be expected to improve the sensitivity of the sensor. To theoretically further analyze the slot region in the one-dimensional photonic crystal sensor and check the performance of this kind of sensor, a finite element method is employed to solve the mode profile. The result shows that 50% of power confinement in the slot region can be achieved under the condition of TE-like mode. For the case of gas atmosphere for sensing, owing to the power confinement in the slot structure, the effective refractive index of the waveguide can vary as large as 41 % of the value of gas cladding refractive index shift, while the one for the waveguide without slot is only 11%. Therefore, due to the significant enhancement of electric field in the slot region for the TE-like mode, the sensitivity with refractive index change is 3 times higher for the slotted one compared with the waveguide without slot. According to our calculation, the sensitivity of about 400 nm/RIU can be expected by using this slotted one-dimensional photonic crystal. Our results predict that the partial slotted one-dimensional silicon photonic crystal is promising for the application of high sensitive biochemical sensor. In addition, since the footprint of one dimensional photonic crystal is within several microns, it can be achievable for mass integration of those sensing components within a small chip. The mass integration of sensors on-chip could meet the increasing demand of statistic results for biochemical application.

We will describe photonic and plasmonic crystal arrays in group IV elements that have applications to infrared sensing. These plasmonic crystal arrays have shapr surface Plasmon based absorption which leads to sharp tunable thermal emission mode used for spectroscopic gas sensors operating at mid-IR wavelengths[1]. We will discuss one class of hole arrays that are conformally coated with metal, that have been fabricated with lattice pitch of 4 micron. Another class of arrays consist of nano-cylinders or nano-cones, composed of metallic elements. The infrared absorption peaks and their dependence on structure will be described. Simulation results will be compared with experimental results on arrays fabricated with photo-lithography. We have also fabricated and simulated asymmetric hole arrays in silicon that have multiple absorption peaks. We will also describe the strong electric field enhancements in these plasmonic arrays and approaches for using these regions of large enhancements for novel sensing applications. [1] I. Puscasu et al, J. Appl. Phys. 98, 013531 (2005). R. Biswas et al , Appl. Phys. Lett. 93, 063307 (2008).

In this work we report our results on the study of a-GexSi1-x:H intrinsic films used as thermo-sensing element in microbolometers. These intrinsic films are attractive because of their relatively high activation energy (Ea â?^ 0.37 eV) and consequently high temperature coefficient of resistance (TCRâ?^ -0.047 K^-1), and as well their higher room temperature conductivity (ÏfRT â?^ 6x10^-5 (Ωcm)^-1), which is of around 3 - 4 orders of magnitude larger than that of the intrinsic a-Si:H films. Here we present a study of fabrication and performance characteristics of two different structures of microbolometers with a-GexSi1-x:H thermo-sensing films, labeled as planar and sandwich configurations. Metal electrodes were either planar providing current flow along the thermo-sensing layer or sandwich with current perpendicular to the thermo-sensing film surface. Current-voltage characteristics with and without IR illumination were performed and the responsivity of the devices was calculated. The noise spectra of the devices was studied, that allowed to determine the detectivity of the devices. The thermal response time was measured in the different microbolometer structures. These data are analyzed for the different micro-bolometer configurations and are compared with published data. As well, we present our latest results on the study of the next generation of thermo-sensing films based on Silicon and Geranium amorphous materials with embedded nanocrystals in the amorphous matrix. These materials have several advantages over amorphous, as a lower defect density, better stability, better transport properties and larger conductivity.

Semiconductor-based thermistors are very attractive sensor materials for uncooled thermal infrared (IR) bolometers. Very large scale heterogeneous integration of MEMS is an emerging technology that allows the integration of epitaxially grown, high-performance IR thermistor materials with pre-processed CMOS-based integrated circuits for the sensor read-out. The resulting bolometer structures have to be suspended to provide thermal isolation of the IR thermistor. Thermistor materials based on alternating silicon (Si) and silicon-germanium (SiGe) epitaxial layers have been demonstrated and their performance is continuously increasing. Compared to a single layer of silicon or SiGe, the temperature coefficient of resistance (TCR) can be strongly enhanced to about 3 %/K, by using thin alternating layers, which exhibit negative temperature dependence. In comparison it should be noted that doped silicon exhibits a weak positive TCR of less than 0.5 %/K. It has been observed that the TCR values in alternating Si/SiGe layers, sometimes referred to as quantum wells, are exponentially related to the bandgap offset between the silicon and SiGe layers. In this paper we present very large scale heterogeneous integration approaches for manufacturing uncooled IR bolometers consisting of suspended, epitaxially grown semiconductor thermistor membranes that have critical device dimensions of below 1 µm. We also report on the optimization of alternating Si/SiGe layers by advanced physically based simulations, including quantum mechanical corrections. Our simulation framework provides reliable predictions for a wide range of SiGe layer compositions, including concentration gradients. Finally, our SiGe thermistor layers have been evaluated in terms of low-frequency noise performance, in order optimize the bolometer detectivity. This work was partly supported by the EU-FP7-ERC grant M&M'S.

Plasmonics offers unique possibilities to confine and direct light at the nanoscale, with broad applications anticipated in sensing, imaging, and information technology. The combination of plasmonics and group IV photonics promises an opportunity to reach economies of scale in the production of innovative nanophotonic devices using the tool set of the semiconductor industry. Integrated Plasmonics Corporation was formed in early 2011 to create a new biochemical sensing platform based on the integration of plasmonic nanostructures with CMOS imaging sensors. We will discuss the progress of our proof-of-concept technology development program, which has largely focused on the modification of low-cost commercially-available CMOS imaging sensors with nanostructured gold films patterned using focused ion beam methods. Our technique allows us to simultaneously study the transmission properties of several hundred different patterns of nanohole arrays, efficiently reproducing in a single experiment a large body of past work in extraordinary optical transmission using an integrated silicon detector rather than far field optics. We have observed and classified a variety of interesting features related to the interaction of the nanostructured gold film and the underlying front-illuminated CMOS sensor. We will discuss how these transmission features can be used to implement biological and chemical sensors using several complementary optical sensing modalities.

In the recent past the generation and processing of multispectral data have had an immense impact on optical characterization systems. Typically carried out with a resolution of 10nm, the measurements are time-intensive, generate a lot of data, and are largely redundant. In this paper we present a band selection routine. To examine which bands provide a high information density, a virtual test environment is used. The photocurrent j = â^« E(λ)*Sabs(λ)*r(λ) dλ was calculated for a large number of different combinations of light sources E, spectral response curves Sabs (bands), and the reflectance r of whitish powder samples that were suspected to be dangerous or illegal. The complete measurement setup consists of several components that depend on a variety of parameters. Actually, there are more than 5000 measurement setup combinations, each of which considers 10 parameters that partially influence themselves. The underlying data set contains various a-SiGe:H multispectral photodiodes that were manufactured at the IMT by A. Bablich et. al. The outstanding feature of these bias sensitive photodiodes is the Ge doped intrinsic layer deposit in a ramp process for a continuous tunability. To determine whether we can gain knowledge from this the multivariate data set, we will have to analyze and simplify it. The employed factor analysis is a common method in the group of structure-discovering methods, and provides good results in the discovery of connections between parameters. It is used particularly if a variety of parameters must be reduced for a certain question. In the first step, an analysis of correlation is carried out to get rid of waste parameters. The second step is the binding of variables on two factors, considering the Kaiser-Guttmann-Criterion that leads to a clear decrease in the complexity. Both factor values can be rotated in the correlation pie diagram independently of each other to examine the influence of different parameters. The result after the factor analysis can be represented in a 2-D scattergram. To verify the different sensors and light source combinations and to separate those spectral response curves (bands) that provide a high information density, a dimension of the external separation must be defined. For this an n-dimensional vector P must be assigned to each measurement that is registered in the matrix M. All vectors together spread out in an n-dimensional room. To determine the volume V of this dot cloud, the dimension normalized volume is defined as Î"CL = sqrt (4^n-3/5^4 * sqrt (V = det MT * det M )), where n is the quantity of employed bands. As a result, the unequivocal characterization of whitish powder samples is made possible by using the VIS-spectrum. The use of such optimized multispectral photodiodes would simplify and accelerate the identification of potentially dangerous substances. Those devices open up new possibilities for material and chemical analysis, as well as civil security and defense applications.

Magnesium silicide (Mg2Si) is a semiconductor with an indirect energy gap of about 0.6eV at room temperature. One of the interesting applications of Mg2Si is an infrared photo detector since it makes ternary or quaternary alloy compounds mixed with Mg2Ge (Eg = 0.5eV) and Mg2Sn (Eg = 0.3eV) and covers from 2 to 5um in the wavelength[1]. Recently, we have succeeded growing high purity Mg2Si single crystalline substrate with low carrier concentration (n=1015cm-3) and fabricating Schottky diode with Au/n-Mg2Si interface[2]. In this paper, we report the Mg2Si pn-junction diode fabricated by thermal diffusion of Ag. A single crystalline n-type Mg2Si substrate was prepared from melt grown Mg2Si ingot[1,2]. Diffusion source of thin Ag layer was evaporated on as-polished surface of substrate through a metal mask with 0.8 mm diameter holes using a conventional resistive heating evaporator. Following the evaporation of Ag, thick Au layer was formed as a cap-layer and an electrode. Thermal diffusion of Ag acceptor was carried out in Ar-atmosphere using RTA furnace. Conditions of the thermal diffusion were varied between the temperature rang of 400°C and 550°C and the period of 10 to 60 min. An ohmic back-contact with n-type Mg2Si substrate was made by Ag adhesive (DuPont 4922). Thermal diffusion of Ag-acceptor was occurred depended on the annealing temperature and periods. We determined the diffusion coefficient D=0.8 x10-10 cm2/s at annealing temperature of 550°C. Clear rectifying J-V characteristics were observed from specimens annealed above 500°C. Current density in reverse bias region was about 0.1A/cm2 at 8V, which is less than 10% compared with that of forward bias region. We also observed that decrease of current density in reverse bias with decreasing the temperature. [1]D.Tamura et al., Thin Solid Films 515 (2007) 8272. [2]K. Sekino et al., Physical Procedia 11(2011) 171-173.

In this work we report on a double PIN photodiode for a compact polarimetric sensor. Using polarimetry the concentration of an optical active substance in solution can be measured based on the rotation of the polarization state of light. There is advantage in performing measurements at various wavelengths for reduction of noise, in particular the artifacts caused by time-variant birefringence due to sample motion or the presence of other interfering optical active substances. To allow measurements at two different wavelengths in a single compact device a double PIN photodetector is proposed. The photodetector was produced by Plasma Enhanced Chemical Vapour Deposition on a glass substrate and consists of a double piâ?Tn/pin a-SiC:H heterostructure (p(a-SiC:H)- í'(a-SiC:H)-n(a-SiC:H)-p(a-SiC:H)-i(a-Si:H)-n(a-Si:H). In this work we present an experimental and theoretical characterization of the photodetector, in terms of the wavelength selectivity and its dependence on the bias voltage, and demonstrate its applicability to the real-time polarimetric sensor proposed.

Silicon heterojunctions have been extensively studied for the understanding the physics of the device as well as their applications to majority carrier rectifier, photodetectors, solar cells and indirect gap injection lasers. Because of its indirect bandgap, silicon is highly inefficient material for light emitter. However, to overcome this problem different approaches were developed in this last decade for the fabrication of Si-based light emitting sources made of silicon nanoclusters (Sinc) embedded in silica or silicon oxynitride (SiO2-Sinc or SiOxNy-Sinc) matrix. In addition, IR light emitting properties were also reported in matrix embedding Sinc and rare earth. In such system, the emitting rare earth ions benefit from the quantum confinement properties of the carriers generated within Sinc to be efficiently excited by an energy transfer mechanism. Electroluminescence of silicon rich oxide based IR light emitting devices is limited by the difficulty in carrier injection. The aim of this work is to study the electrical and the optical properties of MOS diodes made of silicon rich oxide SiOx-Si-Nd thin layers for future electroluminescent devices, which benefit from the efficient sensitizing effect of Si nanoclusters towards the neodynium ions. In this way, Al/SiOx-Si-Nd/p-Si devices are fabricated are elaborated on p-type (111) oriented silicon substrates with resistivity in the range 0.001-0.005 ohm cm. First, a SiOx-Si-Nd active layer is deposited by reactive magnetron sputtering of a pure SiO2 target with Nd2O3 chips under a mixture of hydrogen/argon plasma, which was subsequently submitted to an optimized annealing treatment at 1050°C under N2 flux. Next, aluminum was thermally evaporated on the active layer. Both aluminum and active layers were patterned by wet etching to define the geometry of the device. A second thermal evaporation of aluminum on the back surface was carried out to ensure the ohmic contact with the p-type crystalline silicon. Finally, the devices were annealed into forming gas (H2:N2, 10%) at 390°C during an optimum duration to stabilize the electrical properties of the devices. The conduction mechanisms and the optical properties of the MOS structures are studied in relation with the silicon content (9, 11, 13, 14 or 16%) of the silicon rich oxide SiOx-Si-Nd layer.

Colloidal semiconductor quantum dots (QDs) are frequently called â?oartificial atomsâ?, and are undoubtedly one of the prominent light-emitting materials. In a colloidal QD, the electronic motion is confined in all spatial directions to give a discrete energy level. Hence, the precise size control of the QDs leads to highly-efficient light emission with narrow a full-width at half-maximum (fwhm). The optical use of colloidal silicon nanoparticles (Si NPs) has gained increasing attention for its possible contributions to building a sustainable society that ideally uses resources and energy with high efficiency without causing damage to the environment or human health. Si wafers (Eg â?^1.1 eV) dominate modern microelectronics as an impressive electronic material, but they exhibit relatively poor optical performance owing to an indirect bandgap structure. Interestingly, however, full control of the size distribution and surface chemistry of the NPs yields size-dependent light emission in a very wide range from near-ultraviolet through visible to near-infrared wavelengths. Because of the narrow size distribution, the photo excitation generates an intense peak with a narrow fwhm (~35 nm). The fwhm value was as narrow as that of a visible PL spectrum from a single Si dot. Interestingly, every PLE peak for the PL Maximum exhibits similar spectral shapes with the excitation maxima at a same photon energy position, suggesting that the UV PL emission originates from the monodispersed NPs. Surprisingly, the efficiencies of photoluminescence (PL) in the ranging from near-UV to blue regions reach 20% of PL quantum yield using organic dyes as standards. Time-resolved study of the fluorescence from Si NCs provides a better understanding about significant difference of electron-hole recombination process between a NC and a bulk. Specifically, the PL lifetime (~1.2 ns) is small comparable to those observed in direct bandgap semiconductor nanocrystals such as CdS and CdSe. We considered that this is due to the enhanced overlap of electron and hole wave functions in quantum confined Si. In this presentation, I note some parameters, including interfacial chemistry, that are emerging as important elements for increasing our understanding of the effect of quantum confinement in nanostructured Si and for realizing efficient fluorescence emission. This covers new aspects of derivatives of Si NPs in applications that utilize their optical absorption and emission features. REFERENCES; N. Shirahata, â?oColloidal Si nanocrystals: A controlled organic-inorganic interface and its implications of color-tuning and chemical design toward sophisticated architecturesâ?, Phys. Chem. Chem. Phys. 13 (2011) 7284-7294 (Perspective Article), N. Shirahata, T. Tsuruoka, T. Hasegawa, Y. Sakka, Small 6 (2010) 915-921; N. Shirahata, D. Hirakawa, Y. Sakka, Green Chem. 12 (2010) 2139-2141; N. Shirahata, Y. Sakka, J. Ceram. Soc. Jpn. 118 (2010) 932-939

The reactive monomers (RM) draw much attention due to many possible applications on the liquid crystal alignment in LCD. Poly(styrene-b-2-vinyl pyridine) (PS-b-P2VP) lamellar film with alternating hydrophobic block-hydrophilic polyelectrolyte block polymers (52 kg/mol-b-57 kg/mol) were prepared for the photonic gel. The alternating hydrophilic and hydrophobic lamellar stacks of PS-b-P2VP were obtained by exposing the spin coated homogeneous transparent films under the `chloroform vapor. The hydrophilic layer of lamellar films obtained the formal charge by quarternization with 5wt% of iodomethane diluted by n-hexane. The photophysical and Photochemical effects of RM to those photonic gel films were studied. The reactive monomer added to the photonic gel film increased the light scattering especially at low wavelength region compared to the PS-b-P2VP photonic gel films. Irradiation of photonic gel containing RM with UV light resulted clear color. The lamellar films were swollen by DI water, shifted to longer wavelength by calcinations with calcium carbonate solution. Swelling of films with ethanol exhibited shorter wave length reflection compared with those swollen with water. This study showed that the water induced more hydrophobic and hydrophilic contrast compared to the Ethyl alcohol. The irradiation of lamellar film with UV effectively controlled the reflecting color.

Ge has, besides high carrier mobilities, interesting optical properties. Ge efficiently absorbs infrared light compared to silicon. The absorption depth of light with a wavelength of 850 nm (1.46 eV) decreases from 17 μm in Si to 0.3 μm in Ge [1]. Thin Ge layers on Si substrates can therefore be used to integrate fast Ge photo detectors and modulators on Si CMOS devices for optical interconnects [2]. In view of this renewed interest in the optical properties of Ge, we have performed photoluminescence (PL) measurements on intrinsic and doped Ge substrates as a function of temperature (4 to 400 K) and excitation power. PL measurements are very well suited for the study of radiative recombination mechanisms in Ge. In addition to previously reported luminescence features, we observe luminescence at energies less than 700 meV for temperatures smaller than 50 K. A weak luminescence feature at 738 meV is observed at 6 K for Ge substrates with Al, Ga and Sb dopants. This feature can be explained by no-phonon (NP) assisted recombination of free excitons. At 6 K, the indirect band gap energy and free exciton binding energy of germanium are 742 meV and 4.2 meV, respectively. Besides the NP feature for doped germanium, we observe intense luminescence peaks at 729 meV, 709 meV and 700 meV, which can be attributed to phonon assisted exciton recombination of transverse acoustic (TA), longitudinal acoustic (LA) and transverse optical (TO) phonons in the (111) direction. The longitudinal optical (LO) phonon has an energy close to the LA phonon in the (111) direction, and could not be resolved at 6 K. The presence of dopants (Al, Ga, Sb) has no significant influence on the spectral shape for dopant levels of 1-4 x 1016 cm-3. Besides these known features, we observe luminescence at energy below 700 meV (1.77 μm). This luminescence shows much lower intensity (about 500 times) than the LA replica and is therefore difficult to observe. We explain the observed luminescence as a 2-phonon assisted recombination with the previously mentioned TA, LA, LO and TO phonons with momentum in the (111) direction (at the L point) and a TO phonon at the Î" point. [1] R. F. Potter, Handbook of Optical Constants of Solids II. (Academic, New York, 1985), p.465. [2] S. Assefa, F. Xia, S.W. Bedell, Y. Zhang, T. Topuria, P.M. Rice, Y.A. Vlasov, Optics Express 18, 4986 (2010) R.R. Lieten acknowledges support as Research Fellow of the Research Foundation â?" Flanders (FWO) and support of the Belgian American Educational Foundation (BAEF). Some of the authors (R.R. Lieten, J.W. Ager III and E.E. Haller) acknowledge support by the U.S. Department of Energy under Contract No. DE-AC02-05CH11231. K. Bustillo acknowledges support from the National Science Foundation under Contract No. DMR-0902179.

The data transmission in Si-based microelectronic device will reach a bottle-neck with their performance limited by metallic inter-connections. Optical inter-connection with newly emerged Si quantum dot light source is one of solutions to be developed for next-generation all-Si-based photonic chip is straightforward. However, the formation and control of Si quantum dot is still crucial for fabricating efficient Si-QD LEDs. In this work, the plasma power detuned synthesis of Si-QDs embedded in Si-rich SiOx with enhanced carrier injection rate is demonstrated for multi-color electroluminescence (EL) applications. The Si-rich SiOx films were grown on p-type Si substrate by using PECVD with SiH4/N2O fluence of 33 and 150 sccm, chamber pressure at 67 Pa and substrate temperature of 350oC, and RF plasma power was controlled from 20 to 50 W. After deposition, the SiOx was annealed with flowing N2 at 1100oC for 90 min. The evaporated Al and sputtered ITO on back and front of samples are employed to form the MOSLED for EL analysis. With plasma power increasing from 20 to 50 W, the normalized PL intensity increases by three times (from 25 to 70 count/nm) with its peak wavelength blue-shifted from 720 to 410 nm, indicating the increment of Si-QD density from 0.3Ã-1018 to 1.5Ã-1018 cm-3 and shrinkage of Si-QD size from 4 to 1.8 nm, respectively. The EL spectral peak with their color changed from red to blue also blue-shifts from 700 to 415 nm, whereas the turn-on current greatly decreases from 360 to 9 μA due to the varying Si/O composition ratio of the SiOx host matrix. The early turn-on current phenomenon correlated with the size shrinkage of Si-QD indicates the easier carrier tunneling behavior in Si-rich SiOx layer grown at higher RF plasma power. The fitting of Fowler-Nordheim plot determines the enlarging turn-on electric field from 2.6 to 9.2 MV/cm and the metal/SiOx barrier height from 1.0 to 3.6 eV. The band diagram shows that the intrinsic bending is more serious for larger Si-QDs to reduce the critical electric field for triggering F-N tunneling process. The maximum EL powers (for red, yellow, green, and blue EL colors) are 100, 350, 455, and 690 nW with corresponding power slope of 0.1, 0.4, 1.5, and 115.2 mW/A as the RF plasma power increases from 20 to 50 W at 10 W increment. The power conversion ratio of the Si-QD based MOSLED fabricated with enhanced carrier tunneling rate increases from 2.6Ã-10-6 to 4.7Ã-10-4, corresponding to the internal and external quantum efficiency enhanced from 0.01% to 0.4% and from 0.004% to 0.2 %.

Towards the optical sensing of molecules on a Si chip, a light emission with the wavelength as long as 1.7 μm, i.e., a large red shift of ~100 nm in the near-infrared (NIR) light emission peak is presented for Ge epitaxial layers on Si under a mechanical tensile stress. Although Ge is an indirect bandgap material, Ge photodetectors on Si have been realized for the optical communications (1.3 - 1.6 μm) due to the direct bandgap energy of 0.8 eV, corresponding to 1.55 μm in wavelength. Light emission at 1.55 μm has been also studied for the light sources on Si. A recent topic is on the lasing by optical pumping [1] obtained for a Fabry-Perot structure of heavily n-type doped Ge with a biaxial tensile strain of ~0.2%, which is induced by the thermal expansion mismatch between Ge and Si [2]. Further increase of tensile strain as large as 1.5% is theoretically anticipated to realize direct bandgap Ge, since the Î" valley becomes the lowest one in the conduction band. The bandgap energy is reduced as small as 0.6 eV, corresponds to 2 μm or longer in wavelength, where molecules show a strong optical absorption due to the excitation of vibrations. Thus, highly strained Ge should be useful for light emitters and detectors in the optical sensing of molecules on a Si platform. In this work, a microbeam structure of Ge-on-Si layer with a mechanical bending [3] is used to increase the tensile strain in the Ge layer. As a result, a large positive shift of ~100 nm in the NIR light emission peak is realized. In the experiment, Ge layer (380 nm) was grown at 600°C on a Si-on-insulator (001) wafer (250-nm-thick top Si and 3-µm-thick buried SiO2) by ultrahigh-vacuum chemical vapor deposition with the source gas of GeH4. After the growth, a biaxial tensile strain of ~0.1% was accumulated in the Ge layer at room temperature (RT). Microbeam (cantilever) structures of Ge-on-Si layer in the [100] direction (typically 5 µm in width and 20 µm in length) were patterned using electron-beam lithography and dry etching, followed by a wet etching of SiO2 in a HF solution. According to the finite element simulations, a uniaxial tensile strain as large as 0.5% is applied around the fixed edge by bending the structure. Micro-photoluminescence (PL) spectroscopy at RT was used to examine the effect of strain on the light emission from the Ge/Si beam. The PL peak, derived from the direct transition in Ge, was observed at ~1.58 µm without the bending, while a significant peak shift up to ~1.68 µm was found after the bending, i.e., a large red shift of ~100 nm was realized under a mechanical tensile stress. Further red shift is possible by optimizing the beam structure, suggesting that highly-strained Ge is promising for the application to light emitters and detectors in the optical sensing of molecules. [1] Liu et al., Opt. Lett. 35, 679 (2010). [2] Ishikawa et al., Appl. Phys. Lett. 82, 2044 (2003). [3] Lim et al., Opt. Express 17, 16358 (2009).

Si nanocrystal (Si-nc) colloids have attracted considerable research attention because they can be a key material to realize Si-based large-area optoelectronic devices by a low-cost printing process. In general, Si-ncs form agglomerates in solution. To avoid the agglomeration and to achieve high dispersibility, functionalization of the surface by large molecules is indispensable. However, the long-term stability of surface functionalized Si-ncs is still not satisfactory and the surface ligands often affect the transport properties of films made from Si-nc colloids. Development of surfactant-free Si-nc colloids is thus highly demanded. In this work, we demonstrate that Si-ncs become dispersible in polar liquids without surface functionalization by simultaneously doping n- and p-type impurities, i.e., phosphorus (P) and boron (B), respectively.[1] We prepare Si nanocrystals by simultaneously sputter-depositing Si, silica, phosphosilicate glass and borosilicate glass simultaneously and annealing the films at temperatures higher than 1100^oC. This process results in the growth of Si-ncs in glass matrices, and by dissolving the samples in hydrofluoric acid solution, Si-nc powder is obtained. We show that in methanol, intrinsic and either P or B doped Si-ncs form large agglomerates and the solution becomes muddy. On the other hand, co-doped Si-ncs do not form agglomerates and optically transparent colloids are obtained. In fact, no agglomerates are observed in TEM observations. The zeta potential of the co-doped Si-nc colloids is about -30 mV. This indicates that the dispersibility arises from the electrostatic force between Si-ncs. We show by TEM observations and optical transmittance measurements that the colloids are stable for more than 9 months. Co-doped Si-nc colloids exhibit PL in the near infrared (NIR) range with the lifetime of several hundreds of microseconds. The NIR luminescence suggests that the colloids also have applications in biological imaging of living cells because the PL wavelength is in the transparent window of human cells and they consist only of non-toxic materials. [1] M. Fukuda, et al., Opt. Lett. 36 4014 (2011).

Most of the research on silicon-on-insulator integrated circuits so far has been focused on applications for telecommunication. By using the large refractive index of silicon, compact complex photonic functions have been integrated on a silicon chip. However, the transparency of silicon up to 7 um enables the use of the platform for the mid infrared wavelength region, albeit limited by the absorption in silicon oxide from 4um on. This could lead to a whole new set of integrated photonics circuits for sensing, given the distinct absorption bands of many molecules in this wavelength region. These long wavelength integrated photonic circuits would preferably need broadband or widely tunable sources to probe these absorption bands. We propose the use of nonlinear optics in silicon wire waveguides to generate light in this wavelength range. Nonlinear interactions in just a few cm of silicon wire waveguides can be very efficient as a result of both the high nonlinear index of silicon and the high optical confinement obtained in these waveguides. We demonstrate the generation of a supercontinuum spanning from 1.53 um up to 2.55 um in a 2 cm dispersion engineered silicon nanowire waveguide by pumping the waveguide with strong picoseconds pulses at 2.12 um [1]. Furthermore we demonstrate broadband nonlinear optical amplification in the mid infrared of up to 50 dB [2] in these silicon waveguides. By using this broadband parametric gain a silicon based synchronously pumped optical parametric oscillator (OPO) is constructed [3]. This OPO is tunable over 70 nm around a central wavelength of 2080 nm. Finally, we also demonstrate the use of higher order dispersion terms to get phase matching between optical signals at very different optical frequencies in silicon wire waveguides. In this way we demonstrate conversion of signals at 2.44 um to the telecommunication band with efficiencies up to +19.5 dB [4]. One particularly attractive application of such wide conversion is the possibility of converting weak signals at the mid-IR to the telecom window after which they can be detected by a high-sensitivity telecom-band optical receiver. [1] B. Kuyken et al., "Mid-infrared to telecom-band supercontinuum generation in highly nonlinear silicon-oninsulator wire waveguides", Optics Express, 19, p.20172-20181, 2011 [2] B. Kuyken et al., "50 dB Parametric Gain in Silicon Photonic Wires", Optics Letters, 2011 [3] B. Kuyken et al., "Frequency conversion of mid-infrared optical signals into the telecom band using nonlinear silicon nanophotonic wires", OFC, 2011 [4] B. Kuyken et al., "Widely Tunable Silicon Mid-Infrared Optical Parametric Oscillator",Group IV Photonics, 2011

Si nanocrystals (Si-ncs) have attracted considerable research attention because of their potential applications in optoelectronics and biology. For the application of Si-ncs in optoelectronic devices, controlling the electric properties by impurity doping is indispensable. There has been a lot of theoretical work on doped Si-ncs which predicts strong modification of the energy level structures by n- and/or p-type impurity doping. These researches also demonstrate that location of dopants is a critical parameter to determine the energy level structures. Preferential doping sites of dopant atoms are also predicted from the formation energy calculation. In contrast to the progress of theoretical work, experimental work on doped Si-ncs is still limited. In particular, only a few works have been done on the preferential doping sites.[1] The purpose of this work is to experimentally determine preferential doping sites of boron (B) atoms in B-doped Si-ncs. B-doped Si-ncs are prepared by dissolving borosilicate glass films containing B-doped Si-ncs prepared by a cosputtering method in hydrofluoric acids (HF). We perform photoluminescence (PL) measurements of B-doped Si-ncs in HF solution and find that they exhibit PL around 1.3 eV with a long tail below the band gap energy of bulk Si crystal. The PL spectra are apparently different from those of intrinsic Si-ncs and are considered to arise from the transitions between the conduction band and the acceptor states introduced by B-doping. The PL spectra do not change significantly when B-doped Si-ncs are kept in HF solution, although the PL intensity decreases gradually due to etching of Si-ncs by HF. On the other hand, when B-doped Si-ncs are transferred from HF solution to methanol, the spectrum changes drastically. After two weeks, the PL peak shifts to higher energy for more than 0.15 eV, and the spectral shape becomes very similar to those of intrinsic Si-ncs, i.e., the tail below the bulk Si band gap almost disappears. The most plausible explanation of the drastic change of the PL spectra is that active B atoms in B-doped Si-ncs are located at the sub-surface and the formation of native oxide on the surface in methanol inactivates the B atoms. Our conclusion that B atoms are doped at the sub-surface of Si-ncs is consistent with some theoretical studies, but not consistent with a previous experimental research on B-doped Si-nc powder.[2] [1] X. D. Pi, et al., Appl. Phys. Lett. 92, 123102 (2008). [2] H. Sugimoto, et al., J. Appl. Phys. 110, 063528 (2011)

Si/SiGe superlattices are of great interest for the development of Group-IV THz quantum cascade lasers (QCLs). The main advantage of Group-IV over III-V materials such as GaAs is that the former do not feature polar phonon scattering, which significantly diminishes the efficiency of light emission. However, for Si/SiGe superlattices grown on bulk Si, the lattice mismatch between Si and Ge limits the critical thickness for dislocation formation and thus the number of periods that can be grown. Similarly, the use of composition-graded SiGe layers as a lattice-matched substrate leads to the transfer of dislocations from the graded buffer substrate into the superlattice, with a consequent decrease in light emission efficiency. Here we instead employ nanomembrane strain engineering to fabricate dislocation-free SiGe single crystals to use as the substrate, with lattice constants that match the average lattice constants of the superlattice. This procedure allows for the growth of many superlattice periods with excellent structural properties. We will present specifically designed superlattices grown by chemical vapor deposition (CVD) and characterized via high-resolution x-ray diffraction (showing their strain, average composition, and layer thicknesses) and via THz intersubband absorption spectroscopy. This research is supported by DOE, NSF, and Government of Thailand References 1. B. S. Williams. Terahertz quantum-cascade lasers. Nature Photonics, vol. 1, pp. 517-525, 2007. 2. S. A. Scott and M. G. Lagally, Elastically strain-sharing nanomembranes: flexible and transferable strained silicon and silicon-germanium alloys. J. Phys. D: Appl. Phys., vol 40, pp. R75-R92, 2007.

Ge epitaxial layers on Si have been studied for near-infrared (NIR) photodetectors and light emitters in Si photonics. The direct bandgap of 0.8 eV for Ge is reduced under the tensile lattice strain, leading to the device operation with the wavelength longer than 1.55 µm. According to the theoretical calculations, Ge (001) layers show a transition from the indirect semiconductor to the direct one under the in-plane biaxial strain (stress) as large as 1.4% (2 GPa) [1]. Because of the reduction of bandgap less than 0.6 eV, corresponding to the wavelength more than 2 µm, the devices are anticipated to operate in the mid-infrared (MIR) region with a high quantum efficiency. Such devices open up a new application to optical sensing having the functionality of molecule identification. This is because strong MIR absorption occurs due to the excitation of vibrations in the chemical bonds and the wavelength of absorbed light depends on the species. In this work, application of uniaxial stress on the order of 1 GPa is examined from the viewpoint of MIR device applications. Such a large uniaxial stress can be applied to Ge using microbeam (cantilever) structures with a downward bending, as proposed previously [2]. Here, we compare the [100] and [110] uniaxial stresses applied to the (001) Ge layer. As a result, the uniaxial stress in the [100] direction is more effective for the MIR devices in terms of the indirect-direct transition as well as the fracture of beam structures. Based on the deformation potential theory, the [100] uniaxial stress as large as 4 GPa (strain of ~3.5% in the [100] direction) was found to cause the indirect-direct transition, being similar to the biaxial stress of 2 GPa. On the other hand, under the [110] uniaxial stress, no such transition was found to take place. This is because the indirect L valleys along the [-1 1 1] and [1 -1 1] directions maintain the conduction band minimum, although the energy for the direct Î" valley is reduced with the gap energy of 0.67 eV at 4 GPa. These results suggest that the [100] uniaxial stress is more effective for the MIR device applications. The difference is derived from the shear stress generated in the case of [110] stress. The shear stress should more easily lead to the fracture/cleavage of beam structure. Finite element simulation on the stress distribution for Ge (300 nm)/Si (250 nm) beam (width of 5 µm and length of 25 µm) with a downward bending of 3 µm confirms the generation of shear stress around the fixed edge of [110] beam. In the next step, [100] and [110] Ge/Si beam structures will be fabricated, and the photoluminescence spectra will be compared. [1] Fischetti and Laux, J. Appl. Phys. 80. 2234 (1996). [2] Lim et al., Opt. Express 17, 16358 (2009).

Terahertz (THz) imaging is a powerful technique that exploits a nonionizing (i.e., safe) portion of the electromagnetic spectrum located between microwaves and the IR. Directed toward an appropriate sample, such as wood or ceramics, THz light passes through and is detected by a camera. Because of its nondestructive nature, THz imaging is applicable to many areas. We report a THz video rate imaging system consisting of a quantum-cascade laser light source, and a Si-technology based un-cooled micro-bolometer focal-plane array (an IR detector common in thermal cameras). We also describe two applications of our imaging system: stand-off imaging for search and rescue in a fire disaster, and label-free biomaterial detection. We performed stand-off (~5m distance) THz-imaging experiments under simulated fire conditions. We were able to obtain images in hot, black smoke (with an SNR of ~140 versus ~6000 with no smoke) that blocked visible light and induced signal saturation in the long-wavelength IR (LWIR) region. These results clearly show an advantage of THz compared with LWIR or visible imaging. Our camera is particularly adapted to search and rescue missions due to its 60Hz (real-time) image acquisition rate. In pharmaceutical drug discovery, molecular interactions are commonly tracked using chemical â?~labelsâ?T that are costly and prone to error. We applied our THz imaging system to label-free detection of small-molecule reactions with proteins. The THz waves are readily absorbed and enable sensing of very small changes in biomaterials.

Silicon nanocrsytals (Si-ncs) doped SiO₂ is a promising material toward the application of ultra-fast optical switching devices. Due to the large and ultra-fast nonlinear optical response and the compatibility with CMOS technology in its fabrication processes, it has been attracting great interest today. However, it is also becoming clear now that the further improvement of the nonlinear optical properties is necessary toward the practical applications. In this work, as a new approach to improve the nonlinear optical properties, doping of Si-ncs is suggested. We conduct comprehensive studies on the nonlinear optical properties of phosphorous (P)- or boron (B)-doped Si-ncs and demonstrate that the doping can improve the nonlinear optical properties. The samples were prepared by a co-sputtering method followed by thermal annealing. A z-scan method and a pump-probe optical kerr gate method were used to evaluate the nonlinear refractive index (n₂), two photon absorption coefficient (β), and the time response of the optical nonlinearity. The excitation source was a mode-locked Ti:sapphire femtosecond laser with the pulse width of 70 fs and the repetition frequency of 82 MHz, operating in the wavelength region of 750-840 nm. The values of n₂ and β of our undoped samples were the order of 10^-12 cm²/W and 10^-9 cm/W, respectively. The time response of the optical nonlinearity was faster than our time resolution of 100 femtoseconds. These results are consistent with previous reports. Regardless of P- or B-doping, n₂ and β were found to be enhanced by the doping.Interestingly, the mechanisms of the enhancement were different between P- and B-doping. In the case of P-doping, the enhancement was accompanied by the appearance of an infrared absorption band arising from excess carriers inside Si-ncs. This suggests that P donors in the substitutional site of Si-ncs are responsible for the enhancement. On the other hand, in the case of B-doping, no correlation was observed between the enhancement and the infrared carrier absorption. Probably, the enhancement arises from three-coordinated B at the surface of Si-ncs. In the case of B-doping, the enhancement of β was smaller than that of n₂, leading to the enhancement of the figure of merit (=n₂/βλ) for optical switching devices. The enhancement factor was about 2 at maximum. For all the B-doped samples, the time response was found to be faster than our time resolution of 100 femtoseconds. These results suggest that doping of Si-ncs may be a promising tool to improve the nonlinear optical properties of Si-ncs.

n+p photodiodes were fabricated from silicon hyperdoped with nearly 1% sulfur or selenium by ion implantation followed by nanosecond pulsed laser melting induced rapid solidification. These photodiodes exhibit gain (external quantum efficiency >3000% at 12 V of reverse bias) and substantial optoelectronic response to light of wavelengths as long as 1250 nm (Said, Recht, Sullivan, Warrender, Buonassisi, Persans, and Aziz APL 99, 073503 (2011)) . Our recent experiments have done much to elucidate the physical mechanism underlying these devices' remarkable performance. We will report these results and discuss how they enable the design of a next generation of hyperdoped silicon photodetectors.

We report on investigations of Terahertz detectors based on low-cost silicon technology field effect transistors. We show that detectors, consisting of a coupling antenna and a nMOS field effect transistor as rectifying element, are efficient for THz detection and imaging. We demonstrate that in the atmospheric window around 300 GHz, these detectors can achieve a record noise equivalent power below 10 pW/Hz0.5 and a responsivity above 5 kV/W without any on-chip amplification. We show also that they can be used in a very wide frequency range: from ~0.2 THz up to 1.1 THz. These results pave the way towards high sensitivity silicon technology based focal plane arrays for THz imaging.

Laser hyperdoped silicon has attracted interest in recent years because of its strong broadband sub-band gap absorption, sensitive photodetection, strong device responsivity, and appreciable photovoltaic conversion efficiency. Samples are fabricated by ion implantation of S, followed by pulsed laser melting with a XeCl excimer laser, wavelength 308 nm, with a pulse duration on the order of 30 ns, to produce a highly crystalline material supersaturated in S, up to four orders of magnitude above the equilibrium solubility limit. One unexplained feature of n+p junctions formed by implanting S into p-Si is that the absorption extends to wavelengths as long as 2000 nm and longer, but the device response decreases significantly beyond 1200 nm. A possible interpretation is that the dominant contribution to the device responsivity is due to absorption in the substrate layer. To examine this, we fabricated laser hyperdoped layers by implanting S into 500 nm thick Si80Ge20 films grown on p-Si (100). In this presentation, we report on the structural, optical, and optoelectronic properties of laser hyperdoped SiGe:S. We present SIMS analysis of the impurity depth profile as a function of laser fluence to span a range of melt depths, from just below the implant peak to down into the underlying substrate. We will also present the corresponding influence of laser fluence on the optical absorption and photoconductivity, in comparison to comparably prepared samples on pure Si substrates.

Graphene is a one-atom-thick planar sheet of a sp2-bonded honeycomb carbon crystal. Its gapless and linear energy spectra of electrons and holes lead to nontrivial features such as negative dynamic conductivity in the terahertz (THz) spectral range. In this presentation recent advances in terahertz-wave generation by stimulated emission of radiation in graphene under optical and/or current-injection pumping is reviewed. When graphene is pumped with an infrared photon energy electrons/holes are photogenerated via interband transitions. In case of room temperature environment and/or strong pumping, collective excitations due to the carrier-carrier (CC) scattering (e.g., intraband plasmons) heat up the photocarriers, resulting in ultrafast carrier redistribution along the energy. Then optical phonons (OPs) are emitted by carriers on the high-energy tail of the electron and hole distributions. This energy relaxation process accumulates the nonequilibrium carriers around the Dirac points. Due to a fast intraband relaxation (ps or less) and relatively slow interband recombination of photoelectrons/holes, one can obtain the population inversion under a sufficient pumping intensity. A graphene p-i-n junction under a weak dc current may reproduce a similar situation of population inversion, leading to a current-injection-type THz laser. The authors have first analytically found the possibility of THz gain in such systems under a cryogenic condition and have recently numerically verified the occurrence of the THz gain even at high temperatures. We measured the carrier relaxation and recombination dynamics in optically pumped graphene using THz time-domain spectroscopy based on an optical pump/THz-and-optical-probe technique. An exfoliated fine monolayer-graphene/SiO2/Si sample having a Dirac voltage close to zero and a surface potential distribution of 4.1 meV in variance was placed on the stage. A 0.12-mm-thick CdTe crystal was placed on the sample, acting as a THz probe pulse emitter as well as an electrooptic sensor. A 80-fs, 1550-nm fiber laser beam having 4-mW average power and 20-MHz repetition was split into two: one for optical pumping and generating the THz probe beam in the CdTe crystal, and one for optical probing. The pumping laser, which was linearly polarized, was simultaneously focused at close to the normal incidence from the back surface on graphene sample and the CdTe. The THz probe beam, being generated in the CdTe and reflecting back in part at the CdTe top surface, stimulates the THz emission in graphene, which is electrooptically detected as a THz photon echo signal. We have successfully observed coherent amplified stimulated THz emissions arising from the carrier relaxation/recombination dynamics of an exfoliated graphene. The results provide evidence of the occurrence of negative dynamic conductivity, which can potentially be applied to a new type of THz lasers.

Single-walled carbon nanotubes (SWNTs) thin films have already shown great promise in opto-electronic applications as IR detectors, emitters and optocouplers. Here we report a comprehensive study of the gate induced electromodulated transmittance of infrared light by single-walled carbon nanotube thin films. The observed electromodulation is significantly enhanced by utilizing large diameter SWNTs, increasing the ratio of semiconducting to metal SWNTs and by decreasing the SWNT film thickness. The amplitude of the effect reported herein is more than an order of magnitude larger than in previous SWNT thin film solid state devices.

Single-wall carbon nanotubes (SWCNTs) have an extremely small diameter about 1 nm, so they are attractive building blocks of small quantum dots (QDs) which are not realized with conventional lithography techniques. SWCNT QDs are easily formed by depositing two metallic contacts on them, and the charging energy of the QD falls in the millimeter to teraherz (THz) range. In this talk, we present experimental demonstration of two unique THz detection mechanisms. One is the quantum THz response of the SWCNT QD which includes the THz photon assisted tunneling. The other is the SWCNT QD used as a very sensitive charge sensor, where SWCNT QD is fabricated on the GaAs/AlGaAs 2- dimensional electron gas (2DEG). In this case, THz wave is absorbed in the 2DEG in high magnetic fields and the charge distribution change in the 2DEG due to the THz absorption is detected in the SWCNT single electron transistor on the 2DEG. We will also talk about our effort to extend the single quantum dots to the coupled quantum dots for expected better detector performance. For the needed array of the detector, we will present our effort to grow aligned SWCNTs and to transfer them on another useful substrate.

The introduction of large-area high resolution metrology of entire CVD-grown graphene sheets based on fluorescence quenching microscopy (FQM) opens the door for obtaining feedback on the growth mechanism and quality of large-scale fabricated graphene. This method utilizes the quenching of fluorescence by graphene via resonant energy transfer to increase the contrast between the graphene layers and the substrate. Fluorescence quenching has been visualized by coating a solution of polymer doped with fluorescent dye onto the graphene and imaging the sample under a fluorescence microscope. Consequently, a large-area fluorescence montage image of the dyed graphene sample is obtained and processed to identify the graphene and determine the graphene layer thickness throughout the entire graphene sample. Using this metrology method, the effect of different fabrication and processing techniques on the quality of the graphene sheet is studied. While small-area characterization is insufficient to truly evaluate the effect of the fabrication/processing technique on the graphene sample, through high-throughput nature of our metrology technique we are able to evaluate thickness, uniformity, and structure of the entire grown graphene layers. Using this powerful tool with other traditional characterization methods, we investigated the effects of different CVD growth parameters more comprehensively. Scalability and repeatability of this fluorescence metrology technique make it promising candidate to assess the quality of industrial-scale fabricated graphene materials. Adopting this novel approach, alongside utilizing principles of the gas flow dynamics in CVD, enabled us to gain deeper insight into synthesis mechanism of CVD grown graphene. Accordingly, based on the large-scale FQM feedbacks of the different graphene sample we defined novel quantitative evaluation criteria, which facilitate objective comparison of different graphene fabrication and processing techniques. Our methodology paves the way to establish reliable large-scale techniques for industrial fabrication and characterization of graphene materials.

We report on the Mid-infrared Photothermal response induced by a tunable Quantum Cascade Laser (QCL) in the neat liquid crystal 4-Octyl-4â?T-Cyanobiphenyl (8CB), without using any intercalated dye. The modulated pump QCL range spanned a weak combination absorption band centered at 1912 cm-1. The thermally induced modulation in the scatter of a probe laser operating at 805 nm was measured by lockin detection. Thus heterodyne detection using a Ti:Sapphire laser of the response in the solid, smectic, nematic(N) and isotropic(I) liquid crystal phases allows direct detection of a weak mid-infrared normal combination mode absorption using an inexpensive room temperature silicon photodetector. The sensitivity of the response exceeds that of a conventional FTIR spectrometer equipped with a liquid nitrogen cooled detector. At high pump power in the nematic phase close to the N-I phase transition, we observe an interesting peak splitting in the photothermal response. The advent of tunable lasers that can access still stronger modes suggests that the photothermal mid-infrared response has the potential to detect ultralow concentration of absorbers.