Effect of O/Si composition ratio on near-infrared photoluminescence (PL) of PECVD grown Si-rich SiOx after 1100°C annealing are analyzed by Rutherford backscattering (RBS) and Fourier-transformed infrared spectroscopy (FTIR) to show nonlinear relationship with strongest PL at 760 nm at optimized O/Si = 1.24, total Si concentration of 44.6 atom.%, and N2O/SiH4 fluence ratio of 4.5. A nearly Gaussian function of the normalized PL intensity vs. O/Si composition ratio has been observed due to the significant variation on the Si nanocrystals size with the density of the excessive Si atoms.

Amorphous silicon based solar modules are very attractive for the powering of various microsystems for both indoor and outdoor applications. This technology offers a lot of flexibility in terms of module design, output voltage, shape, size, choice of substrates and offers also the possibility to embed sensors such as photodiodes. This paper focuses on the development of micro-solar modules with area ≤0.15 cm2. Several micro-modules with output voltage of up to 60 V (form 0.1 cm2) were designed and fabricated. The performance limitation introduced by the segment monolithic interconnection and the design of the latter is presented and discussed. An example of a micro-module with a total size of 3.9×3.9 mm2 developed for a micro-robot with dual voltage outputs and embedded photodiodes is also presented.

Both CVD and ALD deposition techniques benefit from a detailed understanding of the reaction mechanisms of the precursor molecules with the surface. In this paper, the reactions of hexakis ethylaminodisilane (AHEAD™), hexamethoxydisilane and hexamethyldisilane were studied on high surface area silica granules at 200-375 °C. Silica was heat treated at 200-820 °C to control the number of surface Si-OH groups. The samples were characterized by FTIR and solid state NMR spectroscopy. After the chemisorption of the precursors with silica, Si-H bonds, not originally present in the molecules, were identified for AHEAD and hexamethoxydisilane, but not for hexamethyldisilane. It is suggested that with AHEAD and hexamethoxydisilane, cleavage of the Si-Si bond takes place during the chemisorption with Si-OH sites. Since no reaction for hexamethyldisilane at the studied temperatures was observed, a prerequisite for the reaction with Si-OH groups seems to be the presence of electronegative O or N atoms in the ligands. In the paper, possible reaction mechanisms with the various surface species are discussed.

In this work, hybrid atomistic-macroscale simulation is conducted to explore the crystallization process of Si surface in the situation of fast melting and solidification induced by ultrafast laser heating and heat conduction. Using the environment-dependent interatomic potential, samples containing 2,880 and 11,520 Si atoms are modeled to provide accurate details for the relationship between the finial crystal structure and the parameters of laser pulses. For different pulsed lasers, amorphous layers are found to form when the laser fluence exceeds a certain critical value. An empirical correlation Ec = 448.76 × (tg)0.56 is obtained to relate this critical fluence to the laser pulse width. It is found that the final thickness of amorphous layer is related to the fluence of the laser pulse with the same full width at half maximum (FWHM). Employing laser pulses with FWHM = 6.67 ns, the formation and recrystallization processes of a 12 nm thick amorphous layer is further investigated, which may have great potential in laser manufacture techniques for Si-associated structures.

This paper presents the fluorescence detection of DNA hybridization with a surface immobilized probe using an hydrogenated amorphous silicon photosensor. This sensor integrates a SiO2 layer for DNA probe immobilization, a p-i-n amorphous silicon (a-Si:H) photodiode for fluorescence detection and a fluorescence filter of amorphous silicon carbon (a-SiC:H) to cut the excitation light. With this integrated photosensor system, a five order of magnitude difference was obtained in the signal measured at the emission wavelength and that measured at the excitation wavelength for the same incident photon flux. The fluorophore Alexa Fluor 430 was used to label the DNA target molecules and a laser at 405 nm and a photon flux of 5.7×1016 cm−2.s−1 was used as the excitation light source. The detection limit achieved for fluorophores in solution in contact with the device and for fluorophores immobilized on the device surface is 5×10−9 M and 0.4 pmol/cm2, respectively. The fluorescence detection of the DNA target hybridization with a covalently or electrostatically immobilized probe was successfully detected at a surface density of ∼3 pmol/cm2.

We present a systematic study on the correlation of hydrogen dilution profiles to structural properties materials and solar cell performance in nc-Si:H solar cells. We deposited nc-Si:H single-junction solar cells using a modified very high frequency (VHF) glow discharge technique on stainless steel substrates with various profiles of hydrogen dilution in the gas mixture during deposition. The material properties were characterized using Raman spectroscopy, X-TEM, AFM, and C-AFM. The solar cell performance correlates well with the material structures. Three major conclusions are made based on the characterization results. First, the optimized nc-Si:H material does not show an incubation layer, indicating that the seeding layer is well optimized and works as per design. Second, the nanocrystalline evolution is well controlled by hydrogen dilution profiling in which the hydrogen dilution ratio is dynamically reduced during the intrinsic layer deposition. Third, the best nc-Si:H single-junction solar cell was made using a proper hydrogen dilution profile, which caused a nanocrystalline distribution close to uniform throughout the thickness, but with a slightly inverse nanocrystalline evolution. We have used the optimized hydrogen dilution profiling and improved the nc-Si:H solar cell performance significantly. As a result, we have achieved an initial active-area cell efficiency of 9.2% with a nc-Si:H single-junction structure, and 15.4% with an a-Si:H/a-SiGe:H/nc-Si:H triple-junction solar cell structure.

Our aim is the development of high-throughput production technology for high efficiency, amorphous and microcrystalline thin film silicon photovoltaics (PV) on flexible substrates. A roll-to-roll system is developed, with three deposition chambers for the continuous deposition of n-type, intrinsic and p-type Si. Novel MW sources for plasma enhanced chemical vapour deposition (PECVD) are introduced in the roll-to-roll coater as MW-PECVD combines good quality with large area, high rate deposition of intrinsic Si. A comparison is made between Raman and UV-reflection spectroscopy.

For the mobility gap of hydrogenated micro-crystalline silicon (μc-Si:H) a value near 1.1 eV is commonly found, similar to the bandgap of crystalline silicon. However, in other studies mobility gap values have been reported to be in the range of 1.48-1.59 eV. Indeed, for accurate modeling of μc-Si:H solar cells it is paramount that key parameters like the mobility gap are accurately determined. In this work we will discuss a method to determine the mobility gap of μc-Si:H using the dark current activation energy of μc-Si:H pin devices, and apply this method to μc-Si:H solar cells with varying crystalline volume fraction. We found the mobility gap is around 1.2 eV to 1.26 eV for μc-Si:H solar cells with a crystalline volume fraction between 50 % and 70 %. For a highly crystalline solar cell we found a mobility gap of 1.07 eV.

The energy distribution of gap states has been examined by means of transient photocurrent measurements in a series of 100°C VHF-deposited Si:H samples that spans the amorphous to microcrystalline transition. The ‘amorphous’ distribution, consisting of a continuous background and a prominent dangling-bond-induced peak, remains largely intact across the transition. The transport path located at the conduction band edge in a-Si:H, some 0.63 eV above the dangling bond D− energy, moves down to ∼0.55 eV above the corresponding D− level in the microcrystalline samples.

This paper presents the results of XRD and PL spectrum studies for Si nano-clusters embedded in amorphous silicon matrix. Investigated layers were deposited by the hot-wire CVD method on glass substrates at the wafer temperature 300°C and different filament temperatures from the range 1735-1885°C. It was shown that variation of filament temperatures allows producing the films with desirable parameters. Using of X-ray diffraction and photoluminescence methods the correlation between the intensity of some photoluminescence bands and the concentrations of Si nanocrystallites and amorphous Si nanoclusters has been shown. The nature of light emission is discussed.

An enhanced simulation model based on the carrier recombination through these states was developed to characterize the gap defect states in hydrogenated amorphous silicon materials (a-Si:H). The energy dependent density of electron occupied gap states, kN(E), was derived directly from Dual Beam Photoconductivity (DBP) measurements at different bias currents. Through Gaussian de-convolution of kN(E), the energy peaks of the multiple defect states, including both neutral and charged states, were obtained. These energy levels, together with the information on the capture cross sections, were used as known input parameters to self-consistently fit the subgap absorption spectra, the electron mobility-lifetime products over a wide range of generation rates, as well as the energy dependent density of electron occupied gap state spectra. Accurate gap state information was obtained and the nature of the defect states was studied. Simulation results on light degraded hydrogen diluted, protocrystalline a-Si:H show that the density of charged states is 2.3 times that of neutral states. The two states close to the midgap act as effective recombination centers at low generation rates and play key roles in photoconductivity studies.

We have fabricated the new top gate depletion mode n-type alternating magnetic field enhanced rapid thermal annealing (AMFERTA) polycrystalline silicon (poly-Si) thin film transistors (TFTs), which show the excellent electrical characteristics and superior stability compared with hydrogenated amorphous silicon (a-Si:H) TFTs and excimer laser crystallized (ELC) low temperature polycrystalline silicon (LTPS) TFTs. The fabricated AMFERTA poly-Si TFTs were not degraded under hot-carrier stress, and highly biased vertical field stress. The considerably large threshold voltage shift (ΔVTH) and trap state density reducing were occurred when the gate bias and drain bias were both large enough. The dominant mechanism of instability in the fabricated depletion mode AMFERTA poly-Si TFTs may be due to carrier induced donor-like defects reduction within the channel layer, especially near the drain junction.

A recent interest in thermal detectors is due to plasma deposited materials with large temperature coefficient of resistance (TCR) and micro-machining technology for thermal isolation. Both are compatible with dominating Si CMOS technology that makes it possible developing device-on-chip configurations. In our previous works we have reported on the study of fabrication and characterization of single cell micro-bolometers based on silicon-germanium thermo-sensing films deposited by low frequency plasma [1-3]. Noise measurements are very important part of device characterization, which have been poorly reported in literature. Models proposed by present time for noise description in non-crystalline samples (either in films or in device structures) are still debated and no one can be considered as the only accepted. The goal of this work is to study experimentally noise spectral density in several configurations of micro-bolometers with silicon-germanium (a-SixGey:H) as thermo-sensing films. We studied four configurations: planar structure with a) a-SixGey, y=0.5, b) a-SixGeyBz:H, x=0.5, y=0.45 and z=0.05 ; c) x=0.5, y=0.5, z=0 and d)sandwich structure with x=0.5, y=0.5. These samples were characterized by SIMS (composition), FTIR (H-bonding and H content), conductivity measurements (σ(T), activation energy, TCR), current-voltage characteristics in dark and under illumination (responsivity). Noise spectral density (NSD) versus frequency S(f) was studied in the range of frequency f=1 to f=103 Hz under IR illumination modulated with this frequency and constant bias. The measurements were performed in vacuum chamber with P=10 mTorr. Generally S(f) dependence demonstrated three regions separated by two corner frequencies: fc1 and fc2: 1) f ≤ fc1 S1 α f -γ γ =0.15 to 0.5, 2) fc1 ≤ f ≤ fc2, S2 α f-γ γ = 0.7 to 1.34 and 3) f ≥ fc2 S3 const (f). Different samples studied showed different values of fc1, fc2 γ and S3. These noise characteristics observed experimentally are analyzed in comparison with data reported in literature and possible mechanisms for noise in the frequency regions studied are discussed.

There has been an on-going effort to produce low cost radio frequency identification (RFID) tags as a replacement for traditional barcodes. One method to achieve low cost production is to integrate the manufacturing of the substrate, antenna and active devices into one single continuous process. Hydrogenated nanocrystalline silicon (nc-Si:H) is a suitable material for manufacturing the active devices in such a process.

We present a nc-Si:H diode suitable for use in rectifiers on RFID tags. It consists of a Cr bottom contact, an undoped layer of nc-Si:H, an n-doped nc-Si:H and an Al top contact. We demonstrate the current-voltage characteristics of the nc-Si:H diode are much improved over a-Si:H diodes. Current density of 10 A/cm2 and ON/OFF ratio greater than 106 was measured at 2 V forward bias. Output DC voltage of 2.6 V was achieved using four nc-Si:H diodes in a full-wave bridge rectifier. The input AC signal was a sine wave at 14 MHz and 2 VRMS amplitude.

Hydrogenated amorphous silicon (a–Si:H) n–i–p photodiodes are used as pixel sensor elements in large-area flat-panel detectors for medical imaging diagnostics. Accurate model of the sensor plays an imperative role in determining the performances of the detector systems as well as ascertaining design issues prior to production. This work presents the formulation of a compact model for segmented a–Si:H n–i–p photodiodes suitable for circuit-level simulation. The underlining equations of the model are based on device physics where the parameters are extracted from pertinent measurement results of previously fabricated a–Si:H n–i–p photodiodes. Furthermore, the implemented model allows photoresponse simulation with the addition of an external current source. Results of the simulation demonstrated excellent matching with measurement data for different photodiode sizes at various temperatures. The model is implemented in Verilog-A and simulated under Cadence Virtuoso design environment using device geometry and extracted parameters as inputs. The model formulation and parameter extraction process, as well as measurements and simulation results are presented.

We have investigated temperature dependence on the hysteresis phenomenon of SLS poly-Si TFT on a glass substrate, extremely at low temperature (213K). The p-type sequential lataral solidification (SLS) polycrystalline Silicon (poly-Si) TFT was fabricated on glass substrate. As the temperature was reduced, it was observed that hysteresis phenomenon was increased, whereas the hysteresis was suppressed at high temperature. This could be explained by a difference of initially electron and hole trapped charges into gate insulator is much larger in low temperature than in high temperature. And we have verified that drain current was changed with a different previous gate starting voltage even at same bias condition by experimental results due to the hysteresis phenomenon of SLS poly-Si TFT. Hysteresis of SLS poly-Si TFT should be improved for a pixel element of high quality AMOLED display.

Boron (B) doping of plasma deposited silicon films have been widely studied and applied in many devices, while B-doping of germanium has been poorly reported in literature. We have reported previously about Ge:H films with low density of localized states deposited by LF plasma with optimal hydrogen dilution.

This work is devoted to a study of boron incorporation and its effect on electronic properties in Ge:H films. The films were obtained by low frequency (LF) plasma deposition from GeH4+SiH4 +B2H6 mixture diluted with hydrogen. The deposition parameters were as follow: substrate temperature Ts = 300 oC, discharge frequency f= 110 kHz, pressure P= 0.6 Torr, power W= 300 W, germane flow QGeH4= 50 sccm, silane flow, hydrogen flow QH2=3500 sccm, diborane flow was varied in the range of QB2H6=0 to 20 sccm providing boron concentration in gas phase in the range of Y=0 to 4%. Composition of the films was characterized by SIMS profiling. Hydrogen bonding was studied by FTIR. Temperature dependence of conductivity measured in DC regime in vacuum thermostat was employed to study carrier transport. Optical measurements provided optical gap, sub-gap absorption and refraction index. Boron incorporation in solid film demonstrated fast increase in the range of Y = 0 to 1.4% and then increase became slower. Hydrogen concentration in the films was determined by absorption of Ge-H stretching mode at k ≈ 1870 cm−1 and it showed weak increase with change of Y from 0 to 4%. Activation energy of conductivity increased in the range of Y = 0 to 1.5% suggesting a compensation of electron conductivity, reaching maximum value Ea =0.5 eV (corresponding approximately to Eg/2) at Y= 1.5%. Then Ea reduced to minimum value Ea = 0.27 eV at Y= 3.5% showing a trend to saturation with further Y increase. This behavior is related to change of charge transport from electron to intrinsic at Y= 1.5% and further to hole transport.

Low temperature (150 °C) deposition of doped and undoped polycrystalline Si (poly-Si) as well as SiNX films on polyethylene terephthalate (PET) films has been achieved with practical deposition rates by using pulsed-plasma CVD under near-atmospheric pressure. The precursor is SiH4 diluted in H2 for poly-Si while N2 has been additionally used for SiNx. No inert gases such as He was used. A short-pulse based power system has been employed to maintain a stable discharge in the near-atmospheric pressures. With this technique, deposition of poly-Si thin film with virtually no incubation layer is possible, which in the case of P-doped poly-Si shows a Hall mobility (μH) of 1.5 cm2/V·s.

Mixed-phase hydrogenated amorphous silicon thin films containing nanocrystalline silicon inclusions have been synthesized in a dual chamber co-deposition system. A PECVD deposition system produces small crystalline silicon particles (3-5 nm diameter) in a flow-through reactor, and injects these particles into a separate capacitively-coupled plasma chamber in which hydrogenated amorphous silicon is deposited. Raman spectroscopy is used to determine the volume fraction of nanocrystals in the mixed phase thin films, while infra-red spectroscopy characterizes the hydrogen bonding structure as a function of nanocrystalline concentration. At a moderate concentration of 5 nm silicon crystallites, the dark conductivity and photoconductivity are consistently found to be higher than in mixed phase films with either lower or higher densities of nanocrystalline inclusions.

Large area digital imaging made possible by amorphous silicon thin-film transistor (a-Si TFT) technology, coupled with a-Si photo-sensors, provides an excellent readout platform to form an integrated medical image capture system. Major development challenges evolve around optimization of pixel architecture for detector fill factor, and manufacturability, while suppressing noise stemming from pixel array and external electronics. This work discusses the behavior and modeling of system noise that arises from imaging array operations. An active pixel sensor (APS) design with on-pixel amplification is studied. Our evaluation demonstrates that a 17 inch by 17 inch array can achieve system noise as low as 1000 electrons through proper design and optimization.