Process Chamber Research Articles

Wafer Delay Analysis and Workload Balancing of Parallel Chambers for Dual-Armed Cluster Tools With Multiple Wafer Types

We examine a scheduling problem for a dual-armed cluster tool that processes multiple similar wafer types concurrently. It has been recently proved that the well-known swap sequence, which is widely used for single wafer type processing, also minimizes the cycle time for concurrent processing. In this article, we wish to minimize wafer delays in a process chamber, which are critical to wafer quality degradation, while maintaining the minimum cycle time. In particular, we show that concurrent processing of wafers with different processing times complicates the analysis of wafer delays significantly, and the wafer delays can be remarkably reduced by finding a proper cycle plan which is the release sequence of different wafer types. We first characterize wafer delays for a given cycle plan by analyzing the circuits of the timed event graph (TEG) model. From this, we prove that concurrent processing of wafers may cause a significant workload imbalance between parallel chambers of a process step, and hence the wafer delays increase substantially. We present that the wafer delays are minimized by a cycle plan that evenly balances workloads between parallel chambers. We also propose how wafer loading task at each process step has to be postponed to meet wafer delay constraints while maintaining the minimum cycle time. Note to Practitioners —Wafer quality control has become an essential fab operational problem in semiconductor manufacturing industry. In cluster tools, which are dominantly being used for diverse wafer fabrication stages, it has been proven that the wafer delays within process chambers have a crucial impact on the wafer quality. Accordingly, modern fabs have introduced stringent quality control to regulate wafer delays in cluster tools. In this research, we propose a scheduling strategy to minimize the wafer delays when a cluster tool concurrently processes multiple wafer types. We first show that the release sequence of different wafer types significantly impacts the wafer delays under concurrent processing, and we then propose how these wafer delays can be minimized by finding the optimal wafer release sequence.

Real-time stress evolution during laser surface line hardening at varying maximum surface temperatures using synchrotron X-ray diffraction

Multiple laser surface line hardening experiments on flat samples, made of tempering steel SAE 4140, were carried out under variation of the maximum control temperature in the range of 850 °C to 1150 °C at the sample surface. For each temperature, time resolved X-ray diffraction measurements using synchrotron radiation were performed at beamline P05@PETRA III at DESY (Deutsches Elektronen Synchrotron) in Hamburg, Germany. The samples were line hardened using a 4 kW High Power Diode Laser unit at a constant laser feed of 800 mm/min with a specific laser optics under pyrometer control of the maximum surface temperature. A special designed process chamber with 4 symmetrically attached fast silicon micro strip line detectors allows for stress analysis during the process according to the sin2ψ method in single exposure mode with a measurement frequency of 50 Hz. As a result of the time resolved analyses the elastic strains were separated from thermal ones and near surface local stress evolutions, longitudinal and transverse to the laser track direction in the center of the processed zone, were determined. The in situ experiments were complemented by high spatially resolved post-process residual stress analyses using conventionally generated X-rays and by metallographic investigations. The results are carefully discussed regarding mechanisms of local stress formation during the laser hardening process and their dependence on the maximum control temperature. The temporal stress course during laser surface line hardening is presented for the first time in both surface parallel directions, where temperature differences in the process zone lead to significant differences in the resulting residual stress profiles. In addition to the common validation of process simulations by comparison with experimentally determined data from the final state after processing, e.g. the resulting local residual stresses, the current work provides real-time data from in situ experiments during laser surface hardening for the validation of temporal stress courses of numerical process simulations.

Comparison of the EHLA and LPBF Process in Context of New Alloy Design Methods for LPBF

Additive Manufacturing (AM) processes are becoming more and more important for production of parts with increasing geometrical complexity and functionality. However, to draw on the full potential of AM technologies, alloys that exploit process inherent particularities such as extremely high cooling rates (ca. 106 K/s) have to be developed. One of most important AM-processes is Laser Powder Bed Fusion (LPBF), a batch-wise process. This complicates experimental alloy development and increases the use of powder resources since only one chemical composition can be tested within one test job and the process chamber has to be cleaned carefully in between. The process Extreme High-Speed Laser Material Deposition (EHLA) has been found to have similar cooling rates as LPBF, however it uses an in situ supply of powders which allows an easy switching between materials and has potential for rapid alloy development methods. Since the mechanical properties of materials primarily depend on chemical composition and microstructure, which in turn depends heavily on the cooling rates in the production process, the EHLA-process could be used as a means for an accelerated alloy development for LPBF. However, to explore this possibility, a thorough comparison of the two processes has to be performed.In this work, EHLA and LPBF processes are compared and evaluated regarding the following characteristics: process parameters, laser intensities and volume energy densities, resulting microstructure (primary dendrite arm spacing, DAS), melt pool size and shape. The reference samples were manufactured using one set of LPBF process parameters and EHLA samples were manufactured using three different sets of process parameters.The volume energy densities Ev [J/mm³] of the processes were found to differ by a factor 2.4 with higher Ev observed in LPBF. However, considering that approximately 2 to 3 layers are remelted with each pass of the laser beam, the introduced Ev per pass approximates the Ev introduced in the EHLA process. The melt pool size as seen in a cross section in the EHLA-manufactured samples is approximately 25 times larger than in the LPBF-manufactured samples and its depth to width ratio (d/w ratio) can be attributed to a heat conduction welding process while the d/w ratio observed in the LPBF-manufactured sample suggests a transition process between heat conduction welding and deep welding. The observed DAS is in the same order of magnitude for both processes ranging from 0.55 to 1.15 µm in EHLA-manufactured samples and 0.73 µm in the LPBF-manufactured reference sample. Since the resulting microstructures of samples manufactured with both processes show this common feature and EHLA process parameters can be adjusted to control the cooling rates, the transferability between EHLA- and LPBF-processes is supported in this first investigation. Research for a more efficient alloy development for LPBF using EHLA will be continued by e.g. examining chemical compositions and performing mechanical testing.

Optimization of Heat Treatment Cycles in Sub-atmospheric LiF-NaF-KF Based Fluoride Ion Cleaning for Removing Oxide Layers in Cracks of IN738-LC

An improved Fluoride Ion Cleaning (FIC) process required for removing all oxide layers by a molten mixture of alkaline fluoride salts under sub-atmospheric pressure was developed and applied to oxide layers on the cracks formed on the surface of Inconel 738-LC samples. This method is directly characterized by filling up the cracks with a molten mixture of alkaline fluoride salts (LiF-NaF-KF) overheated under sub-atmospheric pressure and subsequently, by injecting hot hydrogen gas into the process chamber. The effect of cleaning time on the microstructure of the finished surface was studied in time durations up to 120 min in intervals of 30 min using cross-sectional micrographs and elemental distribution maps. In accordance with the amount of mass loss and microstructural studies during the cleaning process, the optimum cleaning time was suggested to be 90-120 min. Perquisite microstructural outcome shows that in the suggested cleaning condition, all oxide scales in the cracks would be removed without any extra damage to the gamma prime depleted layer, which is a necessary layer for preventing sample oxidation before repair. In this regard, subsequent brazing operations need an oxide-free surface.

The Effect of Thermal Aging on Nano-Particle Removal

The cleanness of the wafer backside is vital in improving process quality and device yield for advanced technology nodes, 10, 7 nm, and beyond [1,2]. Defects such as particles and scratches on the wafer backside could be sources of local deformation of a wafer [2], causing the photolithography hotspot generation [3], the local variation of film thickness, wafer breakage, so on. The micron-size particle defects are easily generated during deposition, etch, CMP processes, and they can be easily transferred to subsequent processing tools. The particles will be exposed to various process conditions such as high compressive stress and high temperature, and it can be seen on the end-effector and wafer chuck in the process chamber. In order to make the wafer backside clean, we need to understand the behavior of particle adhesion and removal, especially at high temperature. In this paper, the adhesion behavior of aged particles at high temperature and the footprint of them were studied.

Control of gas concentration distribution in a semiconductor process chamber using CT-TDLAS measurement

Methane (CH4) concentration distribution in a semiconductor process chamber was controlled using the measurement of computed tomography-tunable diode laser absorption spectroscopy (CT-TDLAS) and the feedback control toward the feeding CH4 concentrations and flow rates. CH4 diluted with nitrogen was fed into the chamber through a shower head having three separate and concentric areas. Thirty-two laser paths were configured in the chamber to collect the infrared absorption spectra for the CT-TDLAS measurement. The computed tomography calculation using the 32 spectra reconstructed the two-dimensional CH4 concentration distribution in the chamber. The measured concentration distribution was updated once per second. Based on the measured concentration distribution, the feedback control algorithm determined the feeding CH4 concentration and flow rate of each shower head area. In this work, we set the target distribution as a ring shape. In the control algorithm, first the feeding CH4 flow rate of each shower head area was adjusted to match the concentration peak radius in the measured distribution to the radius in the target distribution. Then, the feeding CH4 concentration of each area was adjusted in sequence to match the measured average concentration of each area to the corresponding concentration in the target distribution. The algorithm worked successfully, and the concentration distribution reached the target distribution. The extension of the application and its limitations were also discussed.

Development and Studies on FBG Temperature Sensor for Applications in Nuclear Fuel Cycle Facilities

This paper presents the novel results on fiber Bragg grating temperature sensor deployment in high radiation and hazardous chemical environment of nuclear fuel recycle facility. The sensor is developed using thermally regenerated grating obtained from the thermal annealing of seed type-I grating inscribed in telecommunication fiber. The studies on the operating life time and average sensitivity of regenerated grating were carried out upto 900 °C. For the remaining reflectivity of $\sim 5$ % at 900 °C, the predicted lifetime of the grating was around 4 year. The radiation hardness of the packaged FBG sensor was tested in high radiation environment upto 1 MGy. Bragg wavelength shift of $\sim 35$ pm was observed after 1 MGy accumulated gamma radiation dose. Developed and characterized FBG sensor was finally deployed for temperature monitoring of nuclear fuel recycling chamber. Reliable and accurate temperature measurement helped to precisely control the process and to avoid the spill over of uranium, plutonium nitrate solution in the microwave heated process chamber.

Pulse length selection for optimizing the accelerated ion flux fraction of a bipolar HiPIMS discharge

The effect on the energy distributions of metal and gas ions in a bipolar high-power impulse magnetron sputtering (HiPIMS) discharge as the negative and positive pulse lengths are altered are reported. The results presented demonstrate that the selection of the pulse lengths in a HiPIMS discharge is important in optimizing the amount of accelerated ions. A short enough negative pulse is needed so that ions do not escape to the substrate before being accelerated by the positive pulse that follows the main negative HiPIMS pulse. The length of the positive pulse should also be long enough to accelerate the majority of the ions, but a too long positive pulse depletes the process chamber of electrons so much that it makes it difficult to initiate the next HiPIMS pulse. When pulse lengths of negative and positive pulses are properly selected, the fraction of ions, both metal and gas, accelerated by the positive pulse voltage is close to 100%.

Feasibility of Controlling Gas Concentration and Temperature Distributions in a Semiconductor Chamber with CT-TDLAS

The feasibility to control the gas concentration and temperature distributions in a semiconductor process chamber by measuring them was investigated. Gas concentration and temperature distributions for various flow rates were measured with the computed tomography-tunable diode laser absorption spectroscopy (CT-TDLAS). The infrared absorption spectra of multiple laser paths passing through the measured area were collected and the distributions of methane concentration and temperature in the chamber were reconstructed with the computed tomography (CT) calculations. The measured results indicated that the distributions can be independently controlled by measuring with the CT-TDLAS and adjusting the flow rates and the susceptor temperature.

Comparative Study of Titanium Nitride ALD using High Purity Hydrazine vs Ammonia

Emerging devices (Logic and Advanced Memory) require high quality thin (5-20 Å) electrode and barrier films. Difficult thermal budget constraints are now being placed on well-known materials such as TiN and TaN. Deposition temperature limitations have reached 350°C and below while very low resistivity (<100 micro-ohm/cm) must be achieved. Novel Ti precursors are being explored.1 Our approach has focused on hydrazine (N2H4) as an alternative nitrogen source to NH3.Hydrazine is generally difficult to handle from a safety perspective due to its flammability and toxicity. We have developed a novel method to store and generate hydrazine in situ as a high purity gas which is then delivered to ALD process chambers.2 In this study, we examine the effect that hydrazine has on low deposition temperature TiN ALD processes. We experimentally evaluate TiN ALD growth and film characteristics using TiCl4/ N2H4 in comparison to that using TiCl4/NH3.Titanium Nitride films were deposited using the hot wall type tubular furnace and a sequential gas supply system. Deposition temperature was set to predetermined value in the range from 250°C to 400°C. In the case of TiCl4/NH3, GPC was 0.10-0.27 Å/cycle at 250-400°C, where a decrease is noted as the deposition temperature is reduced. With the use of hydrazine, an increase in GPC is observed at each deposition temperature vs NH3. Moreover, the temperature dependence of GPC moved in the opposite direction of TiCl4/NH3. Specifically, GPC of TiCl4/N2H4 ALD was 19% higher at 400°C while an increase of 310% was seen at 250°C versus NH3. In the case of TiCl4/N2H4, the ALD window appears in the range of 250-400°C, where TiCl4 adsorption saturation curves are observed at 250°C and 400°C. Refractive index (R.I.) of TiCl4/N2H4 ALD was about 1.68-1.73 at 250-400°C. These values were approximately equal to R.I. of typical sputtered TiN which was about 1.66. In contrast, R.I. of TiCl4/NH3 ALD was larger than 2.00 in the 250- 300°C range. Because of R.I. of typical titanium oxide is about 2.4-2.5, it is assumed that the TiCl4/NH3 ALD film is contaminated by oxygen. In the ammonia case, background contaminants in the reactor appear to compete with low reactivity NH3 for reaction sites on the ALD surface. In the TiCl4/N2H4 case, higher temperature is not required to form high quality TiN films compared to TiCl4/NH3 ALD.From the above results, the use of high purity hydrazine is beneficial to maintaining high throughput and good film quality under the new constraints for emerging devices. Film composition by XPS and Resistivity measurements consistent with this conclusion will also be presented.

Optimization of Design Parameters Affecting the Performance of a Magnetic Fluid Rotary Seal

Three different design parameters of magnetic fluid seal are optimized for the maximum pressure differential capacity of the seal. The program finite element method magnetics (FEMM) based on finite element method is used to find the magnetic flux inside the gap for all the combinations of the design parameters. The data are analyzed using the MINITAB statistical analysis software. The multivariable regression analysis technique is used to find the percentage effect of different parameters and also the optimized dimensions. The effect of magnetic fluid volume on maximum pressure differential capacity of a seal is observed, and the optimized volume is found. The results provide enough data to design and fabricate the seal. This seal can be used in positive as well as negative pressure applications. For example, as vacuum rotary shaft seal in low-pressure chemical vapor deposition machines, pressure seal for stirrer shaft in chemical process chambers to seal toxic gases. This seal can also be used as an exclusion seal, such as a bearing protection seals in robotic arms used in manufacturing plants.

Plasma Etching Behavior of YOF Coating Deposited by Suspension Plasma Spraying in Inductively Coupled CHF3/Ar Plasma

Dense yttrium oxyfluoride (YOF) coating was successfully deposited by suspension plasma spraying (SPS) with coaxial feeding. After deposition for 6 min at a plasma power of 105 kW, the thickness of the YOF coating was 55 ± 3.2 µm with a porosity of 0.15% ± 0.01% and the coating rate was ~9.2 µm/min. The crystalline structure of trigonal YOF was confirmed by X-ray diffractometry (XRD). The etching behavior of the YOF coating was studied using inductively coupled CHF3/Ar plasma in comparison with those of the Al2O3 bulk and Y2O3 coating. Crater-like erosion sites and cavities were formed on the whole surface of the Al2O3 bulk and Y2O3 coating. In contrast, the surface of the YOF coating showed no noticeable difference before and after exposure to the CHF3/Ar plasma. Such high resistance of the YOF coating to fluorocarbon plasma comes from the strongly fluorinated layer on the surface. The fluorination on the surface of materials was confirmed by X-ray photoelectron spectrum analysis (XPS). Depth profiles of the compositions of Al2O3, Y2O3, and YOF samples by XPS revealed that the fluorination layer of the YOF coating was much thicker than those of Al2O3 and Y2O3. These results indicate that if the inner wall of the semiconductor process chamber is coated by YOF using SPS, the generation of contamination particles would be minimized during the fluorocarbon plasma etching process.

Real-Time Plasma Uniformity Monitoring via Selective Plasma Light Intensity Measurement Using Transparent-LCD-Module-Adapted Optical Emission Spectroscopy

A real-time selective plasma light intensity measurement technique involving optical emission spectroscopy (OES) is proposed for process uniformity monitoring and detection in various plasma regions during semiconductor processing. Among plasma diagnostic techniques, OES is a noncontact plasma measurement technique for process end point and plasma abnormality detection that does not cause any process damage. Therefore, it is mostly used in all plasma-based tools. However, conventional OES systems are incapable of real-time plasma uniformity detection at specific OES sampling times, and therefore, there is a need for improving them. In this study, we show that selective plasma light intensity, including those on the left and right sides of a process chamber, can be measured with transparent-LCD-module-adapted conventional OES for determining process uniformity. The proposed technique can be used to measure the selective plasma light intensity in specific regions of the process chamber, such as the center and edge regions of a wafer, and to thereby analyze the plasma process uniformity.

Predictive control of the plasma processes in the OLED display mass production referring to the discontinuity qualifying PI-VM

Metal target dry etching process applied for the organic light emitting diode display manufacturing is hard to control without the generation of the defect particles. A large amount of the metal-halide by-prodcucts with the non-volatile physical nature are produced in the large area plasma-assisted process chamber. To achieve high-density plasma-based throughput, the inductively coupled plasma type dry etchers were adopted for large-area display manufacturing processes. However, this type of plasma source causes the ion flux-driven damages on the chamber inner walls near the RF power supplied antenna. Sputtered Al atoms from the ceramic parts or etching targets were redeposited onto the chamber inner walls after they form the metal-halide compounds. Redeposited by-prodcucts have very high binding energies to decompose. Undecomposed layers were stuck on the chamber inner wall and flaked off later to form the defect particles. To control this undesired phenomenon, decomposition reaction activated—and plasma locality controlled—two types of ISDs (In Situ Dry cleanings) were designed. A more appropriate type of ISD had selected referring to the developed PI-VM (Plasma Information based Virtual Metrology) model, which qualifies the start of mass production after the discontinuities of the process. The big data set of equipment engineering system and optical emission spectroscopy, accumulated during the mass production, were parameterized to the PI parameters and were applied to the PI-VM modeling. Management of the mass production with the designed ISD and PI-VM model could reduce the 25% of defect particle driven yield loss.

Structure of RF fields in a twin-chamber radio-frequency inductive plasma source placed in a weak external magnetic field

An experimental study of the structure of waves excited in the low-pressure inductive radio-frequency discharge located in the magnetic field with inductance less than 70 Gs was carried out. A plasma source consisting of gas-discharge and process chambers was used in the work. The results of the wave structure study showed that a partially traveling wave was excited in the plasma source when a certain threshold value of the induction of an external magnetic field was exceeded and when a metal flange separating the gas discharge and processing chambers was used. It was also noted that the above-mentioned threshold value decreased with an increase in operating frequency. The amplitude and phase profile of the Bz component of the RF wave indicates the presence of waves reflected from the bottom flange and the local minimum of the electron concentration within the junction of the gas-discharge and process chambers.

EXCITATION OF NONLINEAR PRESSURE OSCILLATIONS IN LOW-PRESSURE FLUID FLOW USING A HIGH-PRESSURE HYDRODYNAMIC GENERATOR

The results of the study of the excitation process of nonlinear oscillations of finite amplitude pressure in a low-pressure (treated) fluid flow using a hydrodynamic oscillator of flow type, the working fluid of which has no direct contact with the fluid of the treated flow, are presented. It is shown that the oscillations from the hydrodynamic generator can be transmitted to the fluid flow through the interface (matching device) in the form of a disk with a certain acoustic resistance. It was found that resonant oscillations can be disturbed in the low-pressure flow processing chamber. The conditions of excitation of resonant oscillations in the processing chamber with a flowing liquid are found. Numerical values of the oscillation span and resonance frequencies are given.

Numerical Simulation of Vapor Deposition Process of Perovskite Solar Cells: The Influence of Methylammonium Iodide Vapor Flow to Perovskite Growth

Abstract This paper presents a two-dimensional (2D) transient numerical model for simulating the vapor deposition process for growing perovskite films. The diffusion process of methylammonium iodide (MAI) vapor through the processing chamber to react with the lead iodide (PbI2) substrate and grow the perovskite layer is analyzed with a diffusion coefficient that has been determined by measuring thicknesses of perovskite layers grown in a chemical vapor deposition (CVD) chamber. Innovations applied to the CVD chamber to improve the uniformity of layer thickness and offer control over the growth process are applied and computationally assessed. One is the addition of screens at various strategic locations in the chamber to improve flow uniformity. Another is changing the locations of MAI sublimation bowls and chamber outlet numbers and locations. The results show that adding screens makes the MAI vapor flow more uniform in the plenum while allowing a quicker purge of the N2 inert gas. This leads to a higher and more uniform growth rate of perovskite. The MAI vapor flow is influenced by the reaction plenum geometry, so the chamber is expected to allow good control of the process to achieve uniform surface deposition rate and controlled grain growth of the perovskite layer.

An Innovative Approach on Directed Energy Deposition Optimization: A Study of the Process Environment’s Influence on the Quality of Ti-6Al-4V Samples

Blown powder additive manufacturing technologies are not restricted to the use of a process chamber. This feature allows to build larger components with respect to conventional powder bed processes. This peculiarity is mostly promising for manufacturing large components or repairing/rebuilding parts of large systems. The main downside of using an open environment, even if a protective shielding gas system is adopted, is the lack of control of process atmosphere. This is particularly critical for titanium alloys which are very sensitive to oxygen/nitrogen pick-up; they have a detrimental effect on ductility, by causing embrittlement and possibly leading to the formation of cracks. It is then important to address how environmental factors, such as process atmosphere and platform temperature, impact not only on the processability but also on the final component properties, both from a compositional and mechanical point of view. The correlations between these environmental factors and microstructure, interstitials content, grain size, and hardness were investigated. Moreover, the Hall–Petch equation was then adopted to additive manufacturing microstructures, characterized by a columnar grain morphology, and used to further investigate the relationship intercurring between grains and hardness and how different microstructures might influence this correlation.

Synthesis of aluminum nitride coatings assisted by fast argon atoms in a magnetron sputtering system with a separate input of argon and nitrogen

To increase the deposition rate of aluminum nitride coatings on dielectric substrates argon is supplied directly to the surface of a magnetron target and then flows into the process chamber through an accelerating grid. The input of nitrogen into the chamber through separate fittings allows formation near the substrate of a plasma suitable for the synthesis of aluminum nitride coatings. The content of nitrogen near the target and grid surfaces is much lower than near the substrate, which limits the dielectric film deposition on the target and grid. It enables an increase in the coating deposition rate by more than three times and a significant improvement of the coating structure and adhesion due to the growth in the energy of fast atoms bombarding the coating.

Plasma resistance of sintered and ion-plated yttrium oxyfluorides with various Y, O, and F composition ratios for use in plasma process chamber

Yttrium oxyfluoride was developed for use in the plasma process chamber using various corrosive gases. In this paper, sintered yttrium oxyfluorides with various Y, O, and F composition ratios (YxOyFz) and ion-plated YOF and Y5O4F7 films were prepared, and the physical etching behavior due to Ar ion bombardment and NF3/Ar plasma resistance was investigated. It was found that the etching rate of the sintered yttrium oxyfluoride due to the bombardment of Ar ions with an energy of 500 eV decreased as the oxygen composition ratio in the samples decreased, i.e., F-rich yttrium oxyfluoride had better resistance against energetic-ion bombardment. It was also found that the surface roughness of sintered YOF and Y5O4F7, both of which had the stable phases, was much smoother after Ar ion bombardment than that of yttrium oxyfluorides without the stable phase. NF3/Ar plasma resistance was also investigated. For the sintered yttrium oxyfluoride, both YOF and Y5O4F7 showed good resistance against the NF3/Ar plasma, where the ideal stoichiometric atomic composition ratio could be kept even after plasma irradiation. For both as-deposited ion-plated YOF and Y5O4F7 films, the F composition ratio was slightly smaller than the ideal stoichiometric ratios, suggesting that some amount of fluorine was escaped from the starting materials of YOF and/or Y5O4F7 during the ion-plated film deposition processes. After NF3/Ar plasma irradiation, F composition ratios were increased for both ion-plated YOF and Y5O4F7 films, and the atomic composition ratio becomes closer to the ideal stoichiometric ratios for both films.