Al Etch Research Articles

Fabrication of long range surface plasmon waveguide biosensors in a low-index fluoropolymer

The authors report the fabrication of long-range surface plasmon polariton biosensors consisting of thin narrow Au stripes embedded in a low refractive index fluoropolymer with etched fluidic channels. The fabrication process incorporates a sacrificial SiO2 channel etch stop layer to avoid waveguide damage and prevent channel over-etching, and an Al etch mask to minimize thermal cracking of the fluoropolymer during channel etching. Process details are reported along with fabrication results, and the optical and biosensing performance of the fabricated devices are demonstrated. The biosensors were tested for bulk sensitivity and bovine serum albumin (BSA) protein adsorption producing notable changes for solutions with different refractive indices (increment of 2 × 10−3) and a change in insertion loss of 0.1 dB as a monolayer of BSA forms on the device.

Confocal sputtering of conformal α-β phase W films on etched Al features

The authors report on thin-film processing improvements in the fabrication of superconducting quasiparticle-trap-assisted electrothermal-feedback transition-edge sensors used in the design of cryogenic dark matter search detectors. The work was performed as part of a detector upgrade project that included optimization of a new confocal sputtering system and development of etch recipes compatible with patterning 40 nm-thick, α-β mixed-phase W films deposited on 300–600 nm-thick, patterned Al. The authors found that their standard exothermic Al wet etch recipes provided inadequate W/Al interfaces and led to poor device performance. The authors developed a modified Al wet-etch recipe that effectively mitigates geometrical step-coverage limitations while maintaining their existing device design. Data presented here include scanning electron microscope and focused ion beam images of films and device interfaces obtained with the new Al etch method. The authors also introduce a method for quantitatively measuring the energy collection efficiency through these interfaces.

Clean Mode Al Etch Process Development for Defect Reduction

Clean mode chamber condition controlling provides a good method to improve Al etch defect performance. In the same time, it brings significant difference of physical performance such as CD, side wall, corrosion etc. In this study, in order to maintain clean chamber condition, Wafer-less Auto Clean(WAC) is introduced to clean out by-products deposition and optimized to achieve stable particle performance during full MTBC(mean time between clean). To compensate process shift because of chamber condition change, Al etch process and PR strip process are optimized. Cl2/BCl3 ratio, bias power and Cl2 gas flow in DARC step were found having significant impact on CD controlling. PR strip process adjusting has been proved to have great help to enlarge corrosion process window. One question that different Cl2/BCl3 ratio affect plasma sustaining when etching through TiN/Al interface was found during this study and need further efforts to find out the detail mechanism.

Novel etch process to tune crater size on magnetron sputtered ZnO:Al

AbstractMagnetron sputtered aluminum‐doped zinc oxide (ZnO:Al) is used as a window layer in silicon‐based thin‐film solar cells due to its high transparency, high conductivity, and ability to provide effective light trapping after etching in a dilute HCl solution. A challenge with this method is the strong influence of sputtering conditions on the density and shape of the ZnO:Al etch features. Here we present a novel chemical etch process based on HF that enables the modification of surface features through the etch process itself, thus allowing the separate optimization of ZnO:Al deposition and texturization. The different etch characteristics of HCl and HF are studied on single crystal ZnO. Using the different etch characteristics of HF we effectively texturize polycrystalline ZnO:Al films which previously exhibited only poor light trapping in silicon thin‐film solar cells. The light trapping improvement is seen by an 18% increase in short circuit current density when used in microcrystalline silicon solar cells. Additionally, using mixtures of HCl and HF we are able to tune the feature size and shape on a given ZnO:Al sample.magnified image Scanning electron microscope images of high‐rate deposited ZnO:Al etched in HCl and HF.

MODELING OF THIN FILM PROCESS DATA USING A GENETIC ALGORITHM-OPTIMIZED INITIAL WEIGHT OF BACKPROPAGATION NEURAL NETWORK

Artificial neural network, particularly the backpropagation neural network (BPNN), has been used to construct a prediction model of plasma processes. In building a BPNN model, many training factors are typically involved and the most difficult factor is the initial weight distribution (IWD). In this study, a technique to optimize the IWD effect on BPNN prediction performance is presented. This was accomplished by using genetic algorithm (GA). The experimental data were collected from the etching of silica thin films in a CHF3-CF4 inductively coupled plasma. The etch process was statistically characterized and the etch responses to model include silica etch rate, Al etch rate, Al selectivity, and silica profile angle denoted as anisotropy. The effect of GA parameters (mutation and crossover probabilities) was also evaluated by conducting a 42 full factorial experiment. The performances of GA-BPNN models were compared to those for conventional models. The comparison revealed an improved prediction of GA-BPNN models for all etch responses. The improvement was even more than 15% for all but the Al etch rate data. The proven improvements support the finding that the presented technique is effective in optimizing an IWD effect on BPNN modeling.

Modeling of plasma process data using a multi-parameterized generalized regression neural network

For characterization or optimization process, a computer prediction model is in demand. A new technique, to improve the prediction performance of conventional generalized regression neural network (GRNN) of plasma process data was presented. Genetic algorithm (GA) was applied to optimize multi-parameterized training factors of GRNN. To evaluate the technique, two data sets were collected from the etchings of silica and silicon carbide (SiC) thin films in inductively coupled plasmas. Both data sets called Data I and Data II were statistically characterized by means of 2 3 and 2 4 full factorial experiment plus one center point. The GRNN models trained with these data were tested with additional six and 16 experiments. A total of eight etch outputs were modeled and compared with conventional GRNN and statistical regression models. The five etch outputs comprising Data I include silica etch rate, aluminum (Al) etch rate, Al selectivity, profile angle, and DC bias. Data II consisted of three etch outputs, including SiC etch rate, surface roughness, and profile angle. Compared to GRNN models, GA-GRNN models yielded more than 40% and 15% improvements for all etch outputs comprising Data I and Data II, respectively. Similar improvements were also demonstrated with respect to statistical regression models. All these results reveal that a multi-parameterization of training factors and GA optimization is an effective technique to considerably improve the prediction performance of conventional GRNN model.

Advanced neural network model of plasma-driven integrated circuit process data

Plasma processes are key means to deposit or etch thin films during the manufacture of integrated circuits. An advanced model of plasma process data was constructed by applying genetic algorithm to the set of typical training factors combined with multi-parameterized gradients of neuron activation functions. The presented technique was evaluated with plasma etch data, collected during silica etching in CHF 3-CF 4 inductively coupled plasma. The etch responses measured include Al etch rate, Al selectivity over silica etch rate, silica profile angle, and DC bias. Constructed models demonstrated better predictions over conventional models for all etch responses. The improvement measured by the relative percentage error was even more than 30%. This was also demonstrated over previous models. In consequence, the presented modeling technique can be effectively applied to construct accurate prediction models with a limited set of experimental data.

학습과 예측의 유전 제어: 플라즈마 식각공정 데이터 모델링에의 응용

A technique to model plasma processes was presented. This was accomplished by combining the backpropagation neural network (BPNN) and genetic algorithm (GA). Particularly, the GA was used to optimize five training factor effects by balancing the training and test errors. The technique was evaluated with the plasma etch data, characterized by a face-centered Box Wilson experiment. The etch outputs modeled include Al etch rate, AI selectivity, DC bias, and silica profile angle. Scanning electron microscope was used to quantify the etch outputs. For comparison, the etch outputs were modeled in a conventional fashion. GABPNN models demonstrated a considerable improvement of more than 25% for all etch outputs only but he DC bias. About 40% improvements were even achieved for the profile angle and AI etch rate. The improvements demonstrate that the presented technique is effective to improving BPNN prediction performance.

Prediction of plasma processes using neural network and genetic algorithm

Using genetic algorithm (GA) and backpropagation neural network (BPNN), computer models of plasma processes were constructed. The GA was applied to optimize five training factors simultaneously. The presented technique was evaluated with plasma etch data, characterized by a statistical experimental design. The etching was conducted in an inductively coupled plasma etch system. The etch outputs to model include aluminum (Al) etch rate, Al selectivity, silica profile angle, and DC bias. GA-BPNN models demonstrated improved predictions of more than 20% for all etch outputs but the DC bias. This indicates that a simultaneous optimization of training factors is more effective in improving the prediction performance of BPNN model than a sequential optimization of individual training factor. Compared to GA-BPNN models constructed in a previous training set, the presented models also yielded a much improved prediction of more than 35% for all etch outputs. The proven improvement indicates that the presented training set is more effective to improve GA-BPNN models.

Modelling of plasma etching process using radial basis function network and genetic algorithm

Computer prediction models are crucial to control complex plasma processes. A new plasma model was constructed by using a radial basis function network (RBFN) and genetic algorithm (GA). The GA was used to search for an optimized set of training factors. This technique was evaluated with the plasma etching data. The etching of silica thin film was conducted in an inductively coupled plasma. The etch responses modelled include aluminum (Al) etch rate, silica etch rate, Al selectivity, silica profile angle, and silica sidewall roughness. For comparison, conventional RBFN models as well as four types of statistical regression models were constructed . Compared to conventional RBFN models, GA-RBFN models exhibited improved predictions of more than 20% for Al etch rate, Al selectivity, and silica sidewall roughness. For the remaining two etch responses, both GA-RBFN and RBFN models were almost comparable. Compared to statistical regression models, GA-RBFN demonstrated improved predictions for nearly all etch responses. The improvement was even more than 35% for the Al selectivity and silica sidewall roughness. The comparisons revealed that the presented method can be effectively used to construct improved prediction models for plasma control.

Patterning of thick polymeric substrates for the fabrication of microfluidic devices

Plasma etching is investigated as an alternative technology to soft lithography for the fabrication of microfluidic devices. First, soft lithography was employed for the fabrication of microchannels in poly-dimethylsiloxane (PDMS), for use in a biosensor for protein and DNA detection. Then, a plasma etching process was investigated and optimized with respect to etch rate maximization for PDMS and PMMA substrates. Preliminary results for deep pattern transfer to these polymeric substrates with an Al etch mask are shown.

Prediction of Plasma Etching Using a Classification-Based Neural Network

Qualitative models of plasma etching are essential for understanding physical etch behaviors as well as plasma control. Artificial neural networks (ANN) have been widely used in constructing predictive etch models. In most applications, the ANN prediction performance has been examined only in terms of the training factors. A technique for building a predictive model is presented here. This is accomplished by applying a neural network to classification data while optimizing the effect of the data boundary. The technique was evaluated with plasma etch data, characterized with a statistical experimental design. The etch response modeled included the etch rates for silica and aluminum (Al) as well as Al selectivity. Compared to the models for nonclassified data, the classification-based model demonstrated improvement in the prediction errors of 36.3 and 15.8% for the Al and silica etch rates, respectively, while exhibiting a poorer prediction for Al selectivity. However, by controlling the data boundary the Al selectivity model was improved significantly by about 43.8%. Thus, the classification-based modeling technique in conjunction with the control of the data boundary is effective in improving the prediction ability of neural network models. © 2004 The Electrochemical Society. All rights reserved.

Silicon anisotropic etching without attacking aluminum with Si and oxidizing agent dissolved in TMAH solution

We study silicon anisotropic etching using tetramethyl ammonium hydroxide (TMAH) containing Si and several oxidizing agents. We focus on the Al etch rate (ER) and the formation of micropyramids. In addition to the previously reported TMAH with Si and ammonium persulfate, Si anisotropic etching without Al etch can be achieved by dissolving Si and ammonium nitrate in TMAH. By surface analysis of the etched aluminum, we have found that thin oxide layers, which cannot be etched by TMAH, form on the aluminum surface. Formation of micropyramids is dependent on the (1 1 1)/(1 0 0) ER ratio. We demonstrate that micropyramids formation can be prevented by increasing the (1 1 1)/(1 0 0) ER ratio. For TMAH with Si and ammonium persulfate, the Si (1 0 0) ER and the occurrence of micropyramids changes according to the sequence that the materials are dissolved in. Consequently, dissolving Si before ammonium persulfate is important.

플라즈마 식각장치내 노즐의 위치에 따른 희박기체유동 및 알루미늄 식각률의 변화에 관한 연구

플라즈마 식각장치내 노즐의 위치에 따른 희박기체유동 및 알루미늄 식각률의 변화에 관한 연구

Semiempirical profile simulation of aluminum etching in a Cl2/BCl3 plasma

A semiempirical profile simulator to predict topographic evolution during Cl2/BCl3 plasma etching of photoresist patterned Al lines has been developed. Given incident flux distributions, the profile simulator uses a combination of a particle based Monte Carlo algorithm and analytic ray-tracing algorithm for solving feature-scale ion and neutral flux transport, respectively. We use angular and energy distributions for reflected ions that are consistent with experimental observation and molecular dynamic simulations. Etch yields with energy and angular dependence are experimentally determined for physical sputtering and ion-enhanced etching. The spontaneous etch rate of A1 by chlorine and the spontaneous desorption rate of Cl from photoresist are estimated from experimental results. Sticking coefficients for etchant, chlorine, and depositor, CClx, and depositing flux are determined by fitting simulated profiles to experimental data. A semiempirical site-balance model is developed to compute the surface coverage of etchant. The reaction probability of neutrals at surfaces is self-consistently determined from the surface coverage at incident location. Competition between etching and deposition on feature sidewalls is modeled. A shock-tracking method is used to advance the profile using computed etch/deposition rates. Simulation results are presented which demonstrate that facet formation, aspect ratio dependent etching, and critical dimension gain, are captured accurately by the calibrated profile simulator. In addition, the calibrated profile simulator along with results of a 23 design of experiments in which photoresist and Al etch rates were measured on open frame wafers have been used to create a plasma reactor model. The reactor model relates the operational parameters including inductively coupled power, rf bias and gas flow ratio to the flux variables, i.e., Cl flux, ion flux, ion energy, and depositor flux, that are used as inputs to the profile simulator. In this manner, calibration of the profile simulator requires a minimum of high resolution, expensive, patterned wafers. The reactor model so created is shown to be in quantitative agreement with results from the hybrid plasma equipment model (Ref. 1).

Selective etching of Al/AlN structures for metallization of surface acoustic wave devices

Metallization is a critical step in the fabrication of high frequency AlN thin film based surface acoustic wave devices. Both state-of-the-art lithography as well as high selectivity and anisotropy during etching of Al with respect to AlN are required for low loss and high performance devices. In this work, the etch rates of reactively sputtered AlN, sputtered Al, thermal SiO2, and Shipley 1813 photoresist as well as the selectivity among Al/AlN, Al/SiO2, and resist/Al have been systematically studied during inductively coupled plasma (ICP) etching. Emphasis is focused on obtaining high Al etch rates, while at the same time keeping the etch rate of AlN and that of the resist sufficiently low. High anisotropy is obtained by passivating the sidewalls by the addition of oxygen. The recipe developed is based on a modified Al etch using a mixture of BCl3, Cl2, O2. The parameters varied were gas composition, and substrate bias. Generally it is found that the Al etch rate exhibits a maximum with the O2 flow, while the AlN etch rate decreases monotonically. Substrate bias is found also to be an important parameter with respect to both etch rate and selectivity. At optimized conditions with respect to process selectivity (500 W ICP power, 83 V bias, 50 sccm BCl3, 25 sccm Cl2, 10 sccm O2, pressure 10 mTorr) the Al etch rate is 1000 nm/min with a selectivity of 1000 toward AlN and 9 toward the resist. The same recipe, slightly modified, has also shown similar Al etch rates when etching Al over SiO2 with a selectivity of up to 200. High anisotropy of the Al etch rate is observed with increasing O2 flow. The former has been determined from cross-sectional scanning electron microscope observations.

Negative ion etching of Al−Si−Cu by employing a magnetic filter in halogen plasmas

Al etching was studied employing negative ions generated in downstream Cl2 plasma. The Al etch rate by the alternate irradiation of positive and negative ions was two times higher than that by usual positive ions. In order to etch the Al film practically on an insulator covered electrode coupled with RF power, reduction of the negative self bias voltage (Vdc) was examined using a magnetic filter which traps electrons. The reason for the high etching reactivity is that the dominant negative ions F in SF6 and Cl− in Cl2 plasmas are atomically very similar to the radicals. Addition of SF6 and H2 to a Cl2/BCl3 mixture reduced Vdc significantly. However, fluorine and hydrogen atoms inhibited Al etching in spite of sufficient ion energy.

Modeling and simulation of feature-size-dependent etching of metal stacks

Aspect ratio dependent etching (ARDE) has often been observed in various etching processes. During etching of metal films in a high density plasma reactor, this phenomenon is more prominent. With a high amount of ARDE present, the narrow spaces will not be cleared unless there is enough overetch, which removes an excessive amount of the underlying layers. The main focus of this article is on the understanding of the mechanisms behind metal ARDE. The results from an extensive design of experiments on the subject were utilized for this study. SPEEDIE, the Stanford etching and deposition profile simulator, was used to develop and test appropriate models. The Langmuir adsorption model with added surface recombination of the adsorbed species, both etchants and inhibitors, was used to model the phenomenon. The added surface recombination acts as the major species loss mechanism on the feature sidewalls. The simulation results indicate that a process with inhibitors, which are highly adhesive to the metal surface and which do not like to recombine with the etchants to form volatile products, will lead to a low degree of ARDE. This is in agreement with experimental results which showed that the addition of CHF3 to the Al etch process helps to reduce the ARDE in narrow space regions.

Study on the Properties of Interlayer Low Dielectric Polyimide during Cl-Based Plasma Etching of Aluminum

The properties of polyimide for interlayer dielectric applications are investigated during plasma etching of aluminum on it. Chlorine-based plasma generally used for aluminum etching results in an increase in the dielectric constant of polyimide, while SF6 plasma exhibits a high polyimide etch rate and a reducing effect of the dielectric constant. The leakage current of polyimide is significantly suppressed after plasma exposure. An optimal combination of Al etch with Cl2 plasma and polyimide etch with SF6 plasma is expected to be a good tool for realizing multilevel metallization structures.

Investigations of rate coefficients in the Cl model

Investigations of ion distributions in Cl plasmas in the Applied Materials Decoupled Plasma Source chamber have been undertaken through the use of simulation. The Hybrid Plasma Equipment Model plasma simulation software was employed. These simulations were performed to investigate which reactions may be considered significant in the DPS tool according to the current Cl model. In addition, new reactions that have been proposed internally to explain some experimental results were also tested to determine their significance. One of the reactions that was determined to be significant is Cl/sup +/ to Cl/sub 2//sup +/ charge exchange reaction which influences the ratio of Cl/sup +/ ions to Cl/sub 2//sup +/ ions in the DPS chamber. This reaction may influence the ability of Al etch processes to remove Cu residue.