Aurrion’s heterogeneous integration process enables high performance active components such as lasers, modulators, and photodetectors to be elegantly integrated on a silicon photonics platform with high performance passive components. This platform also offers the unique capability to combine different types of active devices with separately optimized materials on the same wafer, die, and photonic integrated circuit. Similarly, devices and photonic integrated circuits operating in different wavelength bands can be formed within the same wafer and die. Experimental demonstrations show that these active components can achieve performance on par with commercially available discrete III-V components. In this paper we will discuss the advantages of Aurrion’s heterogeneous integration platform and discuss prototype demonstrations.
An InP-based tunable wavelength converter is investigated which monolithically combines a waveguide photodetector with a sampled-grating distributed Bragg reflector laser diode. We employ advanced device simulation to study internal physical mechanisms and performance limitations. Our three-dimensional finite-element model self-consistently combines carrier transport, optical waveguiding, and nanoscale many-body theory to accurately account for optical transitions within the quantum wells. Good agreement with measurements is achieved. The validity of several model simplification options is discussed.
We investigate the nonlinear response of an InP-based optoelectronic wavelength converter by three-dimensional device simulation including an advanced many-body model for gain and absorption in the InGaAsP quantum wells. The wavelength converter combines a pre-amplified receiver with a post-amplified sampled-grating distributed Bragg reflector tunable laser diode. Good agreement between simulation and measurements is obtained. The nonlinear signal transmission is mainly attributed to quantum well saturation effects in amplifier and photo-detector. Saturation related microscopic physical processes are analyzed in detail.
In this paper, we present three-dimensional (3D) simulation results for an integrated wavelength converter which monolithically combines a pre-amplifying receiver with a post-amplified sampled-grating distributed Bragg reflector tunable laser diode. The self-consistent physical model used in the simulation takes into account gain and absorption in the quantum wells, carrier drift and diffusion, and optical wave-guiding. In order to validate and calibrate the model, we compare the results to available experimental data. Microscopic physical processes inside the converter components are revealed and analyzed, such as receiver saturation effects.
In this work, we describe tunable wavelength converters based on a photodiode receiver integrated with a tunable laser transmitter. Devices are fabricated on a robust InP ridge/InGaAsP waveguide platform. The photodiode receiver consists of an integrated SOA pre-amplifier and a PIN diode to improve sensitivity. The laser transmitter consists of a 1550 nm widely tunable SGDBR laser modulated either directly or via an integrated modulator outside the laser cavity. An SOA post-amplifier provides high output power. The integrated device allows signal monitoring, transmits at 2.5 GB/s, and removes the requirements for filtering the input wavelength at the output. Integrating the SGDBR yields a compact wavelength agile source that requires only two fiber connections, and no off-chip high speed electrical connections. Analog and digital performance of directly and externally modulated wavelength converters is also described.
Intel is aggressively pursuing the use of 157 nm lithography for the 0.1 mm patterning node. Two areas of concentration have been in photoresist and reticle materials development. Over the six months, we have seen considerable progress in new materials development in both areas. In the photoresist area, the use of ultra-thin resists of currently used chemistries appear to be capable of providing short-term layer development and tool testing patterning capability. We have obtained imaging results using a 0.5 NA Schwartzchild optics system. Our best result to data show 70-80 nm lines printed on a pitch of 180 nm. While this small field system has considerably immature optics, it can be used effectively to do basic resist development. In the area of reticle materials development, we have seen considerable improvement in the reduction of OH in blank materials, resulting in higher transmission. We expect to see substrates with greater than 80 percent transmission within the next year at the current rate of accelerated progress. Furthermore, we are not seeing any major processing differences with these new blank materials. Overall, we have seen an accelerated pace of learning in materials development for both resist and new blank materials. Overall, we have seen an accelerated pace of learning in materials development for both resist and reticle materials for 157 nm lithography.
In this paper, we have demonstrated an electrical CD process capable of resolving linewidth swell below 100 nm compatible with a standard polysilicon patterning flow. Appropriate selection of dopant species combined with a reduction in anneal temperature were in the primary means for achieving a physical to electrical linewidth bias of 20 nm. These findings supported our hypothesis that dopant our-diffusion was the primary source of the bias. Also, ECD metrology is applied to quantifying poly CD variations in the presence of substrate topography.
A variety of different approaches were used in an effort to improve the photospeeds of single component TSI resists based on poly(4-hydroxystyrene). The variations included molecular weights, co-monomer partners, and selected substituents. The factors that were studied dramatically affected silylation rates, in one case by as much as an order of magnitude. However, when the silylation times were adjusted to compensate for the rate differences and silylation depths, only minimal differences in photospeed were observed. The apparent contrast measured by swelling upon silylation was very poor ((gamma) equals 1.5) while the contrast measured after etching was quite high, approximately ten times that of the silylation value.
The maturity and acceptance of top surface imaging (TSI) technology has been hampered by several factors including inadequate resist sensitivity and silylation contrast, defects and line edge roughness and equipment performance/reliability issues. We found that the use of a chemically amplified resist can improve the sensitivity by a factor of 1.5 - 2X, without compromising line edge roughness. While the post-silylation contrast of this chemically amplified material is poor ((gamma) < 1), the post-etch contrast is excellent ((gamma) >> 10) and the use of advanced silylation chemistries (disilanes) can further reduce the dose-to-size and increase the contrast. We have also demonstrated that using sulfur dioxide in the plasma etch process can improve the sidewall passivation of the resist lines, thus reducing the overall line edge roughness. Finally, we have been able to successfully use the TSI process to pattern deep sub-micron polysilicon and metal patterns.
Photospeed requirements for 193 nm and EUV lithography are approximately 10 mJ/cm2. As wavelengths are scaled, photon energy is increased, and the discrete photon events may become a fundamental contributor to resist edge roughness. A theoretical analysis of the shot noise impact on line edge roughness was performed. Based on the results, we estimate 1 nm of shot noise induced roughness at 10 mJ/cm2 resist sensitivity for 193 nm lithography. Therefore, discrete photon events are not expected to be a significant contributor to local CD control in 193 nm lithography. We conclude that edge roughness typically observed in 193 nm resists is therefore a process related effect.e However, at EUV, we are approaching a shot noise limit and edge roughness may be enhanced by photon counting effects. Without taking photo-acid diffusion into account, we estimate 8 nm of shot noise induced edge roughness at the EUVB wavelength and 10 mJ/cm2 sensitivity. Hence, local CD control may be compromised by the stringent 10 mJ/cm2 photospeed requirement. Edge roughness is improved by relaxing the resist sensitivity requirement. In addition, most resist technologies are likely to be chemically amplified, and thermally driven diffusion will improve local CD control but at the expense of image contrast and cross- wafer CD control.
The strong attenuation of extreme UV (EUV) radiation by organic materials necessities the use of a thin layer imaging (TLI) process for EUV lithography. Several TLI processes have been identified for potential use for EUVL, and the common theme in these approaches is the transfer of the aerial image to a thin layer of refractory-containing material, which is then used as a dry O2 etch mask during a subsequent pattern transfer to the device layer. One TLI process that has been extensively examined for EUVL is the silylated top-surface imaging (TSI) technology, which is discussed in this paper. Using a new disilane silylation reagent, dimethylaminodimethyldisilane (DMDS) and 13.4 nm exposure, the TSI process has been sued to print 100 nm lines and spaces at equal pitch and 70 nm lines and spaces at a higher 1:2 pitch. The line edge roughness for the printed lines has been determined using a custom image analysis program and, as expected, varies with the particular EUV exposure system and numerical aperture. Exposures done with 193 nm lithography and the TSI process using DMDS are also shown for comparison to the EUV results.
We investigated dry development of resist in the Lam Research TCP 9400SE plasma etcher to meet process specifications for tea 180 nm lithography generation. A full-wafer imaging interferometer was integrated onto the tool, and used to measure etch rates, uniformities and stability of same in-situ. Etch rates of greater than 5000 A/min and selectivities of greater than 15:1 of silylated to unsilylated resist can be obtained in oxygen plasmas in the TCP in the electrode temperature range studied. However, lateral etching (undercut) underneath patterned oxide islands was measured to be approximately 30 nm/min for a typical oxygen process and could not be eliminated in pure oxygen plasmas. To control the lateral etching, we investigated the use of SO2 addition to the oxygen discharge. SO2 addition was found to eliminate the lateral etching component of the resist etch and reduce the etch lag effect, while having a minimal reduction in the overall resist etch rate. We have used the SO2 process to minimize the effect of over-etch on developed resist profiles.
There is increasing interest in chemically amplified (CA) single-layer photoresist for 193 nm excimer laser lithography as a route to sub-quarter micron imaging. A quantitative understanding of the factors that limit the ultimate resolution of CA resists requires a detailed knowledge of both the kinetics of the acid-catalyzed chemical reaction and the diffusion properties of the photogenerated acid. Information of this type is key to the accurate modeling of all CA resists regardless of exposure wavelength. We have investigated the exposure, thermal processing and dissolution behavior of a methacrylate terpolymer-based 193 nm resist. The chemical reactions occurring during post-exposure bake were monitored by FTIR microscopy over a range of PEB temperatures and exposure doses. Using the FTIR data and dissolution contrast curves, parameters for a model of the exposure, the post-exposure bake and the development were extracted. The model was implemented in the SAMPLE lithography simulation tool to predict resist profiles and process latitudes of methacrylate resists on a 193 nm step and scan tool. Excellent agreement between the simulated photoresist profiles and SEM cross-sections was obtained.
Deep-UV chemically amplified (CA) resists are among the leading candidates for the manufacture of semiconductors at 0.25 micron ground rules. In systems of this type, a latent image of photogenerated acid is produced in the resist film on pattern-wise exposure to UV light. The subsequent post-exposure bake (PEB) step drives a thermal reaction, causing a change in the aqueous base solubility of the resist in the exposed regions. Due to the fact that the photochemical and thermal images are decoupled it is important to understand the details of the resist thermal chemistry in order to understand how process conditions affect properties such as linewidth control and resolution. We describe here in-situ, high data-rate, accurate measurements of the chemical kinetics that occur in CA resists during post-exposure bake (PEB). The experimental methodology employs IR or UV spectroscopic measurement under carefully controlled isothermal conditions to determine resist film composition as a function of time. The acid-catalyzed deprotection reactions of two candidate deep-UV resist materials, poly(t-butoxy carbony-loxystyrene)(PTBOC) and poly(t-butyl methacrylate)(PTBMA), were characterized. We propose a model for the acidolysis reactions for both polymer systems and extract coefficients using a stochastic kinetics simulator. This model explicitly addresses the effects of photo-acid strength on the efficiency of the deprotection step. Excellent agreement between the model and experimental data is obtained. The derived rate coefficients are shown to be useful for quantitative prediction of the chemical kinetics of related resist systems. Mechanistic implications of the values of the derived rate coefficients are discussed. The influence of chemical kinetics on the resist's lithographic properties is examined.
KEYWORDS: Deep ultraviolet, Data modeling, Signal processing, Semiconducting wafers, Process modeling, Photoresist processing, Picture Archiving and Communication System, Lithography, In situ metrology, Systems modeling
A study of the dissolution behavior of acid-hardened resists (AHR) was undertaken for spray and spray/puddle development processes. The Site Services DSM-100 end-point detection system is used to measure both spray and puddle dissolution data for a commercially available deep-ultraviolet AHR resist, Shipley SNR-248. The DSM allows in situ measurement of dissolution rate on the wafer chuck and hence allows parameter extraction for modeling spray and puddle processes. The dissolution data for spray and puddle processes was collected across a range of exposure dose and postexposure bake temperature. The development recipe was varied to decouple the contribution of the spray and puddle modes to the overall dissolution characteristics. The mechanisms involved in spray versus puddle dissolution and the impact of spray versus puddle dissolution on process performance metrics has been investigated. We used the effective-dose-modeling approach and the measurement capability of the DSM-100 and developed a lumped parameter model for acid-hardened resists that incorporates the effects of exposure, postexposure bake temperature and time, and development condition. The PARMEX photoresist-modeling program is used to determine parameters for the spray and for the puddle process. The lumped parameter AHR model developed showed good agreement with experimental data.
We explore the bulk and imaging properties of two commercially available resists, Shipley SAL-601 and AZ 5214, to 213 nm radiation operating in a liquid silylation mode. We use FTIR and thickness measurements to characterize the silicon uptake process, and explore the use of high frequency RIE etching of silylated resists to increase selectivity and reduce post- etch residues. We demonstrate sub quarter micron lithography using 213 nm exposure of a liquid silylation resist process etched in a 60 MHz O2 plasma.
Proc. SPIE. 1674, Optical/Laser Microlithography V
KEYWORDS: Lithography, Optical lithography, Data modeling, Deep ultraviolet, Signal processing, Semiconducting wafers, Statistical modeling, Systems modeling, Process modeling, Picture Archiving and Communication System
An investigative study of the dissolution behavior of acid hardened resists (AHR) was undertaken for spray and spray-puddle development processes. A unique tool, the Site Services DSM-100 End-point detection system, is used to measure both spray and puddle dissolution data for a commercially available deep ultra-violet AHR resist, Shipley SNR-248. The DSM allows in- situ measurement of dissolution rate on the wafer chuck and hence allows parameter extraction for modeling spray and puddle processes. The dissolution data for spray and puddle processes was collected across a range of exposure dose and PEB temperature. The development recipe was varied to decouple the contribution of the spray and puddle modes to the overall dissolution characteristics. The mechanisms involved in spray versus puddle dissolution and their impact on process performance metrics have been investigated. The PARMEX photoresist modeling program is used to determine parameters for the spray and for the puddle process. A lumped parameter AHR model developed at Intel was used in iPHOTO for simulation studies.