BACKGROUND STATEMENT FOR SEMI DRAFT DOCUMENT #5944 REVISION TO SEMI F , GUIDE FOR ULTRAPURE WATER USED IN SEMICONDUCTOR PROCESSING

Transcription

1 BACKGROUND STATEMENT FOR SEMI DOCUMENT #5944 REVISION TO SEMI F , GUIDE FOR ULTRAPURE WATER USED IN SEMICONDUCTOR PROCESSING Notice: This background statement is not part of the balloted item. It is provided solely to assist the recipient in reaching an informed decision based on the rationale of the activity that preceded the creation of this Document. Notice: Recipients of this Document are invited to submit, with their comments, notification of any relevant patented technology or copyrighted items of which they are aware and to provide supporting documentation. In this context, patented technology is defined as technology for which a patent has issued or has been applied for. In the latter case, only publicly available information on the contents of the patent application is to be provided. Background This document is an update of the existing SEMI F63 ultrapure water guide. The objective is to align it with the suggestions coming from the International Technology Roadmap for Semiconductors (ITRS) when building or retrofitting today s <65nm fabs. The guide supplies both recommended limits as well as supplemental information about important ultrapure water parameters. Associated analytical methodology references are also contained within. Since this document is a guide it is not intended to tell the reader what quality of water must be used, rather it suggests which parameters are important as well as listing the values that can be achieved. This document is also intended to continue to be updated every two years, keeping it in alignment with the ITRS development pace. As well, the intent is to continue working closely with the ASTM D19 team (authoring ASTM D5127) so that all three documents, F63, ITRS and D5127, are complementary to each other. The methodology used in setting the UPW quality parameters for the guide is based on the risk definitions of the ITRS as well as based on the best available UPW treatment and analytical technology. The resulting water quality guide recommends the UPW characteristics that would mitigate the risks to advanced semiconductor manufacturing with a reasonable level of investment. When risks are not fully addressed (per ITRS), specific comments have been added in the supplemental information section (see Section 10) that would support the semiconductor fab experts in setting their UPW quality specifications. In some cases, additional investment may be chosen due to specific factory needs. Therefore, it is important that the UPW characteristics table will be reviewed together with the supplemental information section. Main changes in the updated revision of the document are as follows: New parameter H 2O 2, has been introduced and explained in the document. There is no value provided at this point, due to insufficient amount of data from the end user and ITRS. The intent is to add the value in the next 2017 revision as more data is obtained. Particle limits have been thoroughly reviewed and modified to undefined, suggesting that the end users should make their informed decision at later time, when additional data is available and based on the project specific considerations. Detailed reviewed of the current state-of-the-art particles metrology was provided in Section 10. Silica values were specified with the focus on colloidal silica control. Bacteria sampling was changed from using 100ml and 1000ml sample volumes to 100ml only. Use of different sample volume sizes is still recommended in the supplemental information section for the case, if troubleshooting is required. Pressure was changed to the parameter requiring consideration, rather than the one that needs to be specified in this guide. Points of monitoring of the UPW quality have been updated specifically to each quality parameter.

2 This document is a guide and therefore complete analytical methods are not contained within. A selected list of optional methods is found in the related sections. Reporting Limits from various analytical test methods are used when ITRS suggestions challenged analytical capabilities. The reporting detection limits have been provided in the parameters table to support the idea of the feasibility of detection of the respective parameter level. In some cases judgment was applied by F63 task force to increase the value of the reporting limit to strengthen the confidence in metrology capability. Please refer to the limitations section and supplementary portion for additional clarification regarding the analytical test methods. Review and Adjudication Information Task Force Review Committee Adjudication Group: High Purity Water TF/F63 TF NA Liquid Chemicals Committee Date: Monday, April 4, 2016 Tuesday, April 5, 2016 Time & Timezone: 8:00 AM to 9:00 AM, Pacific Time 1:00 PM to 4:00 PM, Pacific Time Location: SEMI Headquarters SEMI Headquarters City, State/Country: San Jose, California San Jose, California Leader(s): Slava Libman (Air Liquide)/ slava.libman@airliquide.com Frank Flowers (FMC)/ Frank.Flowers@fmc.com Standards Staff: James Amano (SEMI NA) jamano@semi.org James Amano (SEMI NA) jamano@semi.org This meeting s details are subject to change, and additional review sessions may be scheduled if necessary. Contact the task force leaders or Standards staff for confirmation. Telephone and web information will be distributed to interested parties as the meeting date approaches. If you will not be able to attend these meetings in person but would like to participate by telephone/web, please contact Standards staff. Check for the latest meeting schedule. Notice: Additions are indicated by underline and deletions are strikethrough.

3 SEMI DOCUMENT 5944 REVISION TO SEMI F : GUIDE FOR ULTRAPURE WATER USED IN SEMICONDUCTOR PROCESSING 1 Purpose 1.1 This Guide may be used: To establish performance criteria when purchasing ultrapure water (UPW) purification equipment. To set the process control parameters for UPW-system operation. To establish quality expectations for the supplied UPW. NOTE 1: These suggested guidelines are published as technical information and are intended for informational purposes only. 2 Scope 2.1 UPW is used extensively in the production of semiconductor devices for all wet-processing steps (including wafer rinsing). UPW purity is therefore critical to the manufacture of semiconductors. This Guide provides UPW quality parameters and background information for the decision-making process related to new or retrofit facilities that manufacture semiconductors with line widths of 65 nm and smaller. 2.2 The goal of members of the International Technology Roadmap for Semiconductors (ITRS) UPW committee is to look years ahead on a continual basis to assess the semiconductor industry s future UPW technology requirements. In addition to ITRS input, the SEMI committee contributed its knowledge of the feasibility and cost viability of the technical solutions required to support UPW quality values specified in this Guide. 2.3 The information in this Guide has been developed from the following sources: ITRS risk assessment with input from facility- and manufacturing-experts and technology providers in the semiconductor manufacturing industry. The results of UPW testing at semiconductor manufacturing sites as measured by independent laboratories that test high purity water for the semiconductor industry. Specifications from water-system equipment manufacturers. Input from UPW producers and users at SEMI Standards committee meetings, and also through the balloting process. NOTICE: SEMI Standards and Safety Guidelines do not purport to address all safety issues associated with their use. It is the responsibility of the users of the Documents to establish appropriate safety and health practices, and determine the applicability of regulatory or other limitations prior to use. 3 Limitations 3.1 This Guide is not intended to define UPW system performance requirements. It only provides a reference set of parameters and their values, considered to be satisfactory for the advanced semiconductor manufacturing needs. In some cases the values are limited by the feasibility of the available treatment technology or metrology. It is expected that semiconductor company experts will define the performance requirements for the UPW system based on specific manufacturing process sensitivities, with reference to the information provided in this Guide. 3.2 The purity of water generated in water systems differs. These guidelines were developed for properly maintained, state-of-the-art UPW systems as found in high-end semiconductor manufacturing plants and may not reflect the needs of lesser demanding end-users. 3.3 This Guide is not intended to be used to define the criteria needed for selecting process tools. 3.4 This Guide defines parameters that apply to the quality of UPW required in semiconductor processing when proper analytical hookup and tool installation is made. Page 1 Doc SEMI

4 3.5 This Guide may include suggested values that are at the detection limit of today s available, cost-viable, analytical instrumentation or techniques. UPW system owners, engineers, and technicians are expected to decide whether the benefits of meeting the suggested values outweigh the associated time and costs. The quality of the data measured will depend on which testing method and calibration techniques are used. Laboratory or metrology vendors may be requested to provide additional analytical validation data to confirm detection limits before internal specification values are set. 4 Referenced Standards and Documents 4.1 SEMI Standards and Safety Guidelines SEMI C1 Guide for the Analysis of Liquid Chemicals SEMI C10 Guide for the Determination of Method Detection Limits SEMI F61 Guide For Ultrapure Water System Used in Semiconductor Processing 4.2 ASTM Standards 1 ASTM D859 Standard Test Method for Silica in Water ASTM D4327 Standard Test Method for Anions in Water by Chemically Suppressed Ion Chromatography ASTM D4517 Standard Test Method for Low-Level Total Silica in High-Purity Water by Flameless Atomic Absorption Spectroscopy ASTM D5127 Standard Guide for Ultra Pure Water Used in the Electronics and Semiconductor Industry ASTM D5173 Standard Test Method for On-Line Monitoring of Carbon Compounds in Water by Chemical Oxidation, by UV Light Oxidation, by Both, or by High Temperature Combustion Followed by Gas Phase NDIR or by Electrolytic Conductivity ASTM D5391 Standard Test Method for Electrical Conductivity and Resistivity of a Flowing High Purity Water Sample ASTM D5544 Standard Test Method for On-Line Measurement of Residue After Evaporation of High-Purity Water ASTM D6317 Standard Test Method for Low Level Determination of Total Carbon, Inorganic Carbon and Organic Carbon in Water by Ultraviolet, Persulfate Oxidation, and Membrane Conductivity Detection ASTM D7126 Standard Test Method for On-Line Colorimetric Measurement of Silica ASTM F1094 Test Method for Microbiological Monitoring of Water Used for Processing Electron and Microelectronic Devices by Direct- Pressure Tap Sampling Valve and by the Pre-Sterilized Plastic Bag Method 4.3 Other Documents International Technology Roadmap for Semiconductors (ITRS) 2 NOTICE: Unless otherwise indicated, all documents cited shall be the latest published versions. 5 Terminology 5.1 General terms and acronyms used in this standard are listed and defined in SEMI F Points Of Measurement (POM) of the specified parameters follow commonly accepted terminology in semiconductor industry (UPW ITRS 2015, Table YE1): POD Point of Distribution = Outlet of final filtration in UPW system POC- Point of Connection = Outlet of lateral take off valve 1 American Society for Testing and Materials, 100 Barr Harbor Drive, West Conshohocken, Pennsylvania , USA. Telephone: ; Fax: ; Page 2 Doc SEMI

5 5.2.3 POE Point of Entry = Inlet of wet bench or sub-equipment POU Point of Use = Inlet of wet bench bath, spray nozzle, or connection point to piping, which is also used for other chemicals POP Point of Process = Wafer in production 6 Use of the Guidelines 6.1 Supplemental information, including recommendations on how to use this Document, is provided at the end of the guide. The supplemental information is critical to understanding the intent of this Guide. 6.2 The supplemental information should be considered when making decisions on UPW performance requirements. Any additional data about manufacturing process sensitivities should also be considered when making a final decision about appropriate UPW quality. 7 Units 7.1 Parts per million (ppm) is equivalent to µg/ml or mg/l, where 1 L approximately equals 1 kg. 7.2 Parts per billion (ppb) is equivalent to ng/ml or µg/l, where 1 L approximately equals 1 kg. 7.3 Parts per trillion (ppt) is equivalent to pg/ml or ng/l, where 1 L approximately equals 1 kg. 7.4 Micrometer (µ) is a unit of length equal to one millionth of a meter, or one thousandth of a millimeter. 7.5 Colony forming units (CFU) is a measurement of bacteria organisms as referred to in ASTM F Description of Parameter Tests NOTE 2: SEMI Guides do not require analytical data to support analytical methods, therefore the recommendations of specific analytical methods are only for informational purposes. Alternative methods may also be applicable (see SEMI C1 for analytical validation methodology). 8.1 Resistivity (megohm centimeters or MΩ cm) Resistivity (reciprocal of electrical conductivity) is measured accurately only with only on-line instrumentation MΩ cm is the theoretical upper limit for pure water at 25 C. 8.2 Temperature Temperature stability (K) is defined as a range of the temperature deviation from the facility-specified temperature target. The temperature is measured with an on-line sensor installed inline Temperature gradient (K per time interval) is defined as the rate of change of temperature within a specific interval of time. 8.3 Non-volatile Residue (NVR) NVR or a residue after evaporation (usually expressed in units of ppt) is measured using an online, NVR monitor. 8.4 Total Organic Carbon (TOC) (usually expressed in units of ppb) also known as Total Oxidizable Carbon Conductivity cells or infrared photometry can be used to measure the carbon dioxide created by the oxidation of organic materials. 8.5 Dissolved oxygen (usually expressed in units of ppb) is measured accurately with only on-line instrumentation. 8.6 Particulate Matter (Particles/L) Currently, accurate trend analysis requires online methods using optical particle counters. Since no method is currently available (although several promising methods are under development) for accurate monitoring particles at the target size (see Table 1), no specific recommendation for the metrology is provided (see also additional clarification in Section #10). 8.7 Bacteria (CFU/L) Triplicate samples are cultured based on the ASTM F1094 using a minimum sample size of 1 L. Page 3 Doc SEMI

6 8.8 Silica Total Silica (usually expressed in units of ppb) is measured by graphite furnace atomic absorption spectrophotometry (GFAAS), inductively-coupled plasma atomic emission spectroscopy (ICP-AES), or high resolution-inductively coupled plasma-mass spectroscopy (HR-ICP-MS) Dissolved silica (usually expressed in units of ppb as SiO 2) is measured by heteropoly blue photometry, or by ion chromatography. 8.9 Ions and Metals (usually expressed in units of ppt) Anions and cations such as ammonium, bromide, chloride, fluoride, nitrate, nitrite, phosphate, and sulfate, can be detected using ion chromatography The 21 elements included in this guide are listed in Table 1. NOTE 3: Up to 68 elements, including the 21 of interest can be determined by GFAAS, ICP-AES, or ICP-MS. 9 Recommended Ultrapure Water Quality 9.1 Table 1 lists each parameter with its range of performance. 9.2 Table 1 should be considered together with the supplemental information ( 10). Table 1 Recommended Ultrapure Water Quality Typical Linewidth <0.065 MICRONS PARAMETER PERFORMANCE Limit of Detection (RL) #1 Resistivity 25 C (MΩ cm) >18.18 ±0.2 Temperature Stability (K) ±1 Not Applicable #2 Temperature Gradient (K/10 min.) <0.1 Not Applicable #2 Pressure Stability (bar) To Be Determined see TOC On-line (ppb) < Dissolved Oxygen On-line (ppb) <10 ±0.2 Dissolved Nitrogen On-line (ppm) 8 to 18 ±0.3 Dissolved Nitrogen Stability (ppm) ±2 ±0.3 Residue After Evaporation On-line (ppt) <100 <20 On-line Particles > 0.05 µ size (#/L) <200 <100 Bacteria (CFU/L) #3 1 L Sample < ml Sample <1 1 Silica Silica Total (ppb) <0.5 <0.5 Silica Dissolved (ppb as SiO2) <0.5 <0.1 Ions and Metals (ppt) Ammonium <50 10 Bromide <50 10 Chloride <50 5 Fluoride <50 5 Nitrate <50 5 Nitrite <50 5 Phosphate <50 10 Page 4 Doc SEMI

7 Typical Linewidth <0.065 MICRONS PARAMETER PERFORMANCE Limit of Detection (RL) #1 Sulfate <50 10 Aluminum <1 0.5 Antimony < Arsenic < Barium <1 0.5 Boron <50 15 Cadmium < Calcium <1 0.5 Chromium < Copper <1 0.5 Iron <1 0.5 Lead <1 0.5 Lithium <1 0.5 Magnesium <1 0.5 Manganese < Nickel < Potassium <1 0.5 Sodium <1 0.5 Tin < Titanium < Vanadium < Zinc <1 0.5 #1 Reporting limits (RL) as reported by the analytical vendors. #2 The specified value is relative to the absolute temperature target chosen by a semiconductor manufacturing facility and therefore the accuracy or detection limits are considered to be not applicable. The sensor suppliers state that temperature stability is within 0.1% of the reading for >2 years operation, which is well within the expectations implied by the specified values. #3 See 10.8 for additional clarification. Point of Measurement Typical Linewidth <0.065 MICRONS PARAMETER of the Specified Parameters PERFORMANCE Limit of Detection (RL) #1 On-line Particles (#/L) #3, see 10.7 TBD TBD TBD Silica Silica Total (ppb), 10.9 Feed to the Final <0.5 <0.5 Filters Silica Dissolved or Reactive (ppb Feed to the Final <0.5 <0.5 as SiO2), 10.9 Filters On-line TOC (ppb), see 10.3 POD < On-line Dissolved Oxygen (ppb), POD <10 ±0.2 see 10.4 On-line Dissolved Nitrogen (ppm) for megasonic cleans only. see 10.5 POP 8 to 18 ±0.3 Page 5 Doc SEMI

8 Dissolved Nitrogen Stability POP ±2 ±0.3 (ppm), for megasonic cleans only. see 10.5 Residue After Evaporation Online POD <100 <20 (ppt), see 10.5 Hydrogen Peroxide, ppb, see POD TBD TBD Bacteria (CFU/L) #43 see 10.8 POD 100 ml Sample POD <1 1 Ions (ppt) Ammonium POD <50 10 Bromide POD <50 10 Chloride POD <50 5 Fluoride POD <50 5 Nitrate POD <50 5 Nitrite POD <50 5 Phosphate POD <20 10 Sulfate POD <50 10 Metals and B (ppt) Aluminum POD <1 0.5 Antimony POD < Arsenic POD < Barium POD < Boron POD <50 15 Cadmium POD < Calcium POD <1 0.5 Chromium POD <1 0.5 Copper POD <1 0.5 Iron POD <1 0.5 Lead POD < Lithium POD <1 0.5 Magnesium POD <1 0.5 Manganese POD < Nickel POD < Potassium POD <1 0.5 Sodium POD <1 0.5 Tin POD < Titanium POD <1 0.5 Vanadium POD < Zinc POD <1 0.5 Resistivity 25 C (MΩ cm), see 10.1 POD not less than18.18 ±0.20 Temperature Stability (K) (for immersion litho) see 10.2 POP ±1 Not Applicable #2 Temperature Gradient (K/10 min.) (for immersion litho) (See 10.2) POP <0.1 Not Applicable #2 Page 6 Doc SEMI

9 Pressure see Consideration Only #1 Reporting limits (RL) as reported by the analytical vendors. #2 The specified value is relative to the absolute temperature target chosen by a semiconductor manufacturing facility and therefore the accuracy or detection limits are considered to be not applicable. The sensor suppliers state that temperature stability is within 0.1% of the reading for >2 years operation, which is well within the expectations implied by the specified values. #3 Particles size is specific to the dimensions of the manufacturing node (see Section 10 for additional clarifications). The ITRS definition of the "killer" particle size uses the roadmap s definition of DRAM ½ Pitch (nm). #4 See 10.8 for additional clarification 10 Supplemental Information for the Specification Table 10.1 Resistivity Resistivity is commonly used to indicate UPW purity because it provides an instant response to only significant change in water quality. It is a well-established, relatively inexpensive method, but the absolute values of resistivity readings are affected by other variables such as probe setup and calibration, sample flowrate, and temperature variation/compensation. Therefore, resistivity measurements cannot provide an indication of the ionic contamination at the ppt range, as indicated in Table 1. Reported resistivity-measurement accuracy is usually within the ±0.2 MΩ cm range, but in some applications the accuracy may be optimized to a more narrow range and still be insufficient to indicate trace ionic contamination. The limitations of the resistivity instrumentation should be taken into account, when specifying the UPW system design Temperature for Immersion Lithography Based on ITRS analysis, this guide offers a temperature-stability requirement for immersion photolithography tools, using UPW as an immersion fluid. It represents the maximum rate of change (or gradient) of the temperature of the cold UPW supplied to the photolithography tool in order for the tool to maintain temperature stability. In some cases, additional temperature stability requirements, such as <0.1 K/10 min., may be used Depending upon the immersion photolithography tool used, the UPW temperature-range requirement for immersion photolithography can be set within a relatively wide value around room temperature. However, the temperature gradient is very important for immersion photolithography Temperature stability requirements for UPW uses other than immersion photolithography are generally less stringent than those for immersion photolithography TOC For the purpose of this Guide, TOC is defined as equivalent amount of CO 2 produced from UV oxidation at 185 nm wavelength inside of the UV reactor of a TOC monitoring device. The CO 2 is measured using conductivity cells, which results in a calculated TOC level. It is either assumed that the change in conductivity is a result of only the CO 2 produced from the oxidation of organic compounds present in UPW or the nonco 2 contributions to conductivity are eliminated or subtracted from the final conductivity measurement and TOC calculation Organic carbon compounds can be broken into the following major groups: Natural Organic Matter (NOM) NOM includes species that either exist naturally in source water or are byproducts of drinking water sterilization. Synthetic Organic Matter (SOM) SOM includes chemical pollutants in the source water, contamination from components within the UPW system, or trace chemicals introduced via manufacturing rinse-water reclaim. Some of these organic species can be difficult and expensive to remove using conventional treatment processes Benchmarking studies conducted by the UPW ITRS Committee at several semiconductor manufacturing plants (located throughout the world) have shown that when the TOC levels are controlled to 1 ppb (as recommended in this specification), no known related defects occur The limit of the current metrology for measuring TOC online is 50 ppt, therefore the recommended level of less than 1 ppb is easily achievable with currently available online instrumentation. Page 7 Doc SEMI

10 Critical Organic Compounds UPW ITRS is currently working on a better definition of the critical organic species that may have higher impact on semiconductor manufacturing. This definition will be based on the failure mode and effect analysis (FMEA) taking into account device impact, occurrence, and detection probability (see additional explanation of FMEA methodology used in 10.10). Currently used online, TOC analyzers are non-specific in their measurement of organic contaminants As features of semiconductor devices are scaled down, the oxide thickness is reduced and becomes more vulnerable to voltage leakage. There is evidence that oxide layers such as native oxide, gate oxide, and tunnel oxide can be affected by the presence of organic compounds in UPW. Organic compounds containing polar (-OH) groups form strong bonds with the oxygen of the oxide layer, leading to oxide breakdown and voltage leakage. The presence of organics on the oxide layers can also lead to poor adhesion of photoresist causing undercutting during the wetetching process. This undercutting creates larger windows, or vias, which may short-circuit adjacent metal lines Dissolved Oxygen Dissolved oxygen may affect native gate oxide growth and, in higher concentrations, enhance corrosion of metals such as copper. Although this Guide specifies 10 ppb of dissolved oxygen, very critical processes (such as SiGe device cleaning) may require lower specifications (down to 1 ppb of dissolved oxygen). This can be achieved only if the cleaning bath is closed to the environment Oxygen can be used as a dissolved gas for megasonic cleaning when oxidation is not a concern. The detection accuracy of the on-line analytical instruments is within ±0.2 ppb, adequately supporting the specified value of 10 ppb. The UPW section in the ITRS report (Table YE3) considers some parameters as process variables rather than contaminants. Some semiconductor manufactures now treat dissolved oxygen (DO) as a process variable while others consider it a contaminant Dissolved Nitrogen This Guide defines the level of nitrogen necessary for efficient megasonic cleaning and for the lithography scanner. Some variability of concentration can be tolerated when process engineers determine the specific range for their facility. Maintaining a 10 ppb dissolved oxygen level in the water implies a dissolved nitrogen level of approximately 1 ppm, therefore gasification with nitrogen may be required to achieve the desired dissolved nitrogen concentration. Regassification at the point of use is more economical and easier to control when the incoming level of dissolved nitrogen is stable. The detection accuracy of on-line analytical instruments is within ±0.3 ppm, adequately supporting the specified dissolved nitrogen value of 8 to 18 ppm. Dissolved nitrogen stability is significant for megasonic cleaning and for UPW consumption NVR Measuring NVR online, as described by ASTM D5544, provides an additional, operational, water-quality parameter. For example, measuring NVR can provide a clear indication of silica (dissolved and colloidal) breakthrough in ion-exchange beds at a level not detectable by on-line silica analyzers (see 10.9). ITRS studies indicated that colloidal silica may adsorb metals and act as a transport mechanism for metals. The online measurement of NVR has shown that a change in water quality of as little as 20 ppt of NVR is significant. However, interference from noncritical organics has been observed. Most current semiconductor UPW systems operate at <100 ppt of NVR. Given the current limitations for both particle measurement and monitoring high-boiling point organics the measurement of NVR serves as a usual indicator of a stable UPW system. Each semiconductor facility should evaluate its own need for NVR testing and target values based on existing experience with UPW systems Particles The ability to measure particles online is a critical parameter for semiconductor manufacturing. Particles can cause any of the following problems: Blocked vias, resulting in chemical damage and epitaxial defects. Deposits on the wafer when the ph is low (<3) because particles can overcome the electrokinetic layer around the wafer. Significant damage during the etching process when the thickness of the semiconductor layer is a few Angstroms (1 Å = 0.1 nm). Page 8 Doc SEMI

11 Deposits on wafers caused when residual liquid layers evaporate in the wafer-drying process and deposit any particles they contain. The maximum allowed concentration of particles in UPW is a function of the allowed defect level on the wafer as defined in the FEP Surface preparation table and particle deposition form UPW to the wafer surface. The concentration per wafer will remain constant and will only change with the wafer size (300 mm to 450 mm). The reference point will be the critical particle size. The relationship between acceptable particle concentrations on the wafer and that in UPW is based on experimental data and theoretical modeling. The user of this part of the table is encouraged to refer to the supplemental materials for specific details of the calculations. Significant metrology issues for particles at the critical size still exist. Given the fact that the metrology gap is growing and has not been addressed for several years, the risk management for particle control is being shifted towards the effective particle filtration and reduction of the particle challenge to the final filters. SEMI C079) is a guide for filter performance validation, which is a way to mitigate the risk of particle contamination impacting yield. There is a new method under development for evaluation of the cleanliness of the virgin ion exchange resin (expected to be published early 2016). Existing UPW metrology is not sensitive enough to allow the detection of silica particles at the specified size as a chemical. Silica particles are believed to have the ability to adsorb metals and act as a transport mechanism to potentially deposit metals on the surface of a wafer. Such metal deposition could seriously impact yield. There is no metrology that can effectively detect colloidal silica directly. However, the difference between total silica (measured by ICP-MS) and reactive silica, also called dissolved silica, is thought to give an accurate measurement of colloidal silica. The EUV mask production requires even tighter particle control than the wafer manufacturing. Critical particle size remains equal to the DRAM ½ Pitch (nm). However the allowed concentration is an order of magnitude lower than that specified for high volume manufacturing facilities (refer to supplemental information for additional details). The particles specification is considered at POP. However at the POE to the tool, additional POU filtration may be required. Therefore an even tighter specification may be necessary at the POE Particle metrology has not kept pace with the decreasing line width of semiconductor manufacturing. Current line widths require the ability to monitor 20 nm particles. However, existing optical particle counters (OPCs) are only capable of detecting 50 nm particles with a counting efficiency of <5%, and a background count (noise level) of less than 100 particles per liter (the noise level has been confirmed with leading OPC suppliers). Particle counting statistics become important as count levels approach the noise level, therefore the OPC setup and performance must be optimized. The particle value listed in Table 1 of this document considers data averaging over a period of one hour. Particle levels must consistently be at or near the noise level of any OPC (regardless of any specified level). down to 10 nm particles. The particles specification in this guide is not addressed due to limitations of the existing metrology and growing risk of the defect due to the following: ITRS/SEMI data indicate high risk of 10-20nm particles due to ion exchange resin leach-outs Final filters (cross flow UF state of the art) do not remove 100% of the particles in those ranges UFs are susceptible to fiber breakage Existing particle counters (including the most recent/advanced) either do not cover the target range of most advanced semiconductor manufacturing technology or are too new and require better understanding of the technology and the data produced. It is also difficult to specify UPW quality for particles relying on non-proprietary technologies only (per SEMI guidelines) as the most advanced are all proprietary The following is the summary of the known for 2015 particle monitoring methods and the pros/cons analysis for the choice of the methods. Page 9 Doc SEMI

12 # Particle Monitoring Options 1 No change use 50nm LPC counter 2 New LPCs Laser particle counter Rationale Limitations Mitigation Relative Rating (1 highest) No good solution available, any change is to worse > 20nm(closer to target) 50nm >> 10nm (most advanced killer particle) Very high cost, partial solution, counting efficiency, optical properties of the particle Suggest putting more emphasis on a) particle challenge reduction and b) ensuring final filter performance. Specific recommendations for both will be included in the notes to this line in the table Recommend as viable and optional solution, define range 20-30nm to allow all 3 major players to compete; add above comments (See #1) SEM Scanning electron microscope (capture filter) Add robustness to existing monitoring at relatively low added cost Long sampling time, methods for <50nm counting are not readily available Consistent with ASTM D5127; improved version of #1 (include the comments) 2 4 New generation 10nm counters (Example: Scanning TPC threshold particle counter) Target size of the particles can be monitored; commercially available Not fully proven, single source; high cost Could be used one per fab in compliment to the existing 50nm (and others) LPC. Appropriate comments required in the notes The challenges associated with the particle monitoring: Existing and commonly used 50nm LPC technology is outdated as it does not address the need. Hence this method was removed from this version of the guide. Newer nm smallest particles counting LPC technology provides counting particles at sizes closer to the target size (all be it with very low counting efficiency at the rated size) of the most advanced semiconductor fabs, but it does not address their need, while requiring substantial investment. This investment will be even more difficult to justify for the longer period of the factory life cycle. SEM is good solution to provide an indication of the UF fiber issues and characterize elemental composition of the particles. However, SEM filters are readily available at the pore sizes of 50nm and larger. SEM technology is available for counting 10nm particles, but methods for those sizes are not readily available and will be proprietary. SEM based particle counting method also requires very long sample collection period of time (2 to 3 weeks is not uncommon). Therefore this method is considered not being able to address the need. The new methods such as Scanning TPC are being developed and commercialized. They will require better understanding of their capability Bacteria Bacteria can cause the same types of defects as other particles. Bacteria are particularly important during start up and commissioning, commonly as a result of contamination from newly installed parts. Bacteria cannot be detected by OPCs because they have a similar refractive index to water. The ASTM F1094 is a reliable, cost-effective, and representative method of bacteria detection. Its limitations include long bacteria cultivation time and possible sample Page 10 Doc SEMI

13 contamination. Although critical bacteria monitoring point is defined at POD, it is highly recommended to consider monitoring at 2 additional locations: POC this will provide an indication of the quality of UPW near the tools and the quality of the design of the distribution system. A well designed UPW system will not allow bacteria the opportunity to multiply. Point of Return sampling port of UPW Return line close to the UPW tank ASTM F1094 provides for running cultures (in triplicate) by concentrating samples on acceptable filters; sample sizes are 0.1 and 1 L or larger. Total count of the media is conducted after incubation. Incubation temperature ranges from 25 C to 28 C and incubation time ranges from 48 hours to 7 days; choice of both parameters depend on the within-company specifications. The common industry standard is 28 ± 2 C for 48 hours Two different Although only 100ml sample volume is specified in Table 1,both sizes (1 and 0.1 L) of the sample are recommended to address the concern of sample contamination. Only one type of the sample (0.1 L or 1 L) is required. Two different volumes are recommended to be used for the when a bacteria analyses, as they can help to reconfirm whether excursion is reported. For real UPW contamination (if identified) is related to the UPW or sample contamination using both volumes should result in appropriately different bacteria counts The recommended sampling is minimum in triplicates for each set of analyses at the given sample volume (five samples are recommended for new system commissioning) It is accepted that if a majority of the samples is nondetect, the result is considered to meet the spec requirement. The positively detected bacteria in some samples will be viewed as sample contamination during sampling Silica Incoming city or raw water is the main source of silica in UPW. Silica may also dissolve from the wafer surfaces and then be deposited back on the wafers, causing water spotting. Silica is detected by measuring either reactive silica or total silica. Reactive silica (also known as dissolved silica) is a frequently used online UPW operational parameter. Total silica (dissolved plus colloidal) is an offline measurement. Silica and boron are the first two anions to break through a mixed bed. Therefore silica is an indicator of anion exchange resin removal efficiency. If the chemistry within an ion exchange is ideal, reactive silica may transform into colloidal silica, and become a source of colloidal silica particles. Colloidal silica can also affect the wafer by scavenging metal ions (much like ion exchange). The specification values for both total and reactive silica are provided at the limit of detection, indicating that no detectable silica is recommended. Both levels, reactive and total silica, are used to indicate that colloidal silica is not present Typical limits of dissolved silica detection for commercially available instruments are around 0.1 ppb. Similar levels can be achieved using grab sample analysis Silica may occur in UPW in a form of reactive (dissolved) or non-reactive (colloidal silica). Colloidal silica is considered particularly detrimental for semiconductor manufacturing. The killer particle size has become so small, that it is impractical to expect that the colloidal silica might be measured as delta between total and reactive silica in UPW. Hence, colloidal silica is considered to be controlled as particles down to non-detectable (as silica) level. Nevertheless, as long as effective particles monitoring for the killer size of the particles of advanced semiconductor industry is not available, for practical purposes, measuring both total and reactive silica at the lowest feasible levels of detection is highly recommended to confirm no presence of colloidal silica In order to improve detectability of colloidal silica and allow for more effective risk mitigation, the location of the point of monitoring colloidal silica was recommended at the feed to the final filters. Final filters will further reduce colloidal silica, thus making the risk to production even lower In order to allow lower cost of silica monitoring, Table 1 recommends 0.5 ppb of total and colloidal silica values to be controlled. The user of the guide may consider testing colloidal silica down to lowest possible detection limit, thus reducing the risks of colloidal silica occurrence even further Reactive Silica is commonly measured by Heteropoly Blue Photometry either online using ASTM D7126 or by preconcentrating grab samples using ASTM D859. Total silica is measured by HR-ICP-MS, graphite furnace Page 11 Doc SEMI

14 atomic absorption spectroscopy (GFAAS) ASTM D4517, by digestion with Heteropoly Blue Photometry using ASTM D7126 or pre-concentration inductively coupled argon plasma optical emission spectroscopy (ICP-OES) Colloidal silica is measured indirectly, subtracting reactive silica from total silica. Although a large percentage of UPW particles may be colloidal silica, detection limits for total silica are too high to be useful as a substitute method of detecting colloidal silica particles. The measurement of NVR can provide an on-line indication of the present of colloidal silica. However, the current technology for the measurement of NVR cannot discriminate between the various sources of NVR Ions and Metals Metal ions have a significant effect on semiconductor yield for the following reasons: Alkali metals can cause the breakdown of dielectrics. Transition and heavy metals can lead to crystal defects, minority carrier lifetime issues, and even to surface sputtering around metal particles. Transition metals can impact the p-n-junction. Silicide-forming metals can cause dielectric breakdown Metals cause gate oxide breakdown and changes in the substrate resistivity Metals are removed from the wafer by mechanisms such as the etching of the underlying layer, or the replacement with H+ ions on the surface at low ph Anions can influence metal adsorption behavior The ITRS has chosen a methodology similar to a failure mode-and-effect analysis (FMEA) to determine the acceptable concentration of metals and ions in wafer-cleaning solutions and rinse water. A risk priority number (RPN) is calculated as a product of the four weighted risk factors for contaminants with values above a critical limit. Acceptable concentrations are listed in the critical ion lists per chemical and UPW. A lower, permissible, maximum concentration is also specified. The complete FMEA table can be found in supplementary materials of UPW ITRS The following risk factors have been defined as important for the successful manufacturing of semiconductors: Device Impact Factor Characterizes the severity of the impact on device performance, device reliability or yield or both, when a specific contaminant is deposited on the wafer surface. Occurrence Risk Factor Characterizes the probability that a specific contaminant occurs in the chemical or UPW in concentrations that could affect devices on the wafer. Deposition Risk Factor Characterizes the probability that a specific contaminant does deposit on the wafer, calculated either by a deposition model (diffusion or Pourbaix), or determined in experiments at ph of 7 or both. Non-detection Risk Factor Characterizes the probability that a specific contaminant is not detected either in the chemical, UPW, or on the wafer Analytical Detection Limits Analytical limits of detection must be considered when setting specifications because the specifications should not be lower than those which can be accurately measured Detection limits fall into one of the three following categories: Instrument Detection Limits (IDLs) IDLs are statistically based on multiple analytical runs and calculated from instrument noise or background equivalent concentrations under optimum conditions. Impacts from sampling and bottle variation are not taken into account. Therefore specifications should not be set to IDLs. Method Detection Limits (MDLs) SEMI C10 defines MDL as the level at which errors in the measurement method become large enough such that the preset maximum acceptable risk of seeing the quantified level, when none of the contaminant in question is present in the sample, is exceeded. False positives, but not false negatives, Page 12 Doc SEMI

15 are a concern. MDLs are statistically based and include variations from sampling, bottles, matrix effects, and sample preparation. Reporting Limits (RLs) RLs are generally set above the MDLs. When possible, as in the case of ions, total silica, and metals, they are verified by spiking with known concentrations of the analytes of interest to eliminate false negatives. RLs are set to be achievable on a regular basis in a production analytical environment. NOTE 5: NOTE 4: Overall, RLs > MDLs > IDLs The following instrumentation is used for ICP-MS analysis of trace elements: Quadrupole ICP-MS (typically collision/reaction cell based), with samples analyzed as received. This is the least expensive method, but it has the highest MDLs. High-resolution ICP-MS, with samples analyzed as received, or with some matrix modification. The instrumentation is more expensive than quadrupole ICP-MS, but spectral interferences are reduced and MDLs are lower overall. Quadrupole ICP-MS with sample pre-concentration. This has a higher cost and longer turnaround time than the previous instrumentation, but MDLs and RLs are lower than the ICP-MS methods without pre-concentration. Not all 2010 ITRS specifications for critical elements can be met by sample analysis without pre-concentration. All 2010 ITRS critical element specifications are at, or above, the RLs for quadrupole ICP-MS with pre-concentration. The pre-concentration method was used for reference of the reporting limits provided in Table Pressure Stability (±0.3 bar or ±4.4 PSIG) This Guide offers a pressure stability requirement consideration for UPW system design. It is defined as a range of the pressure deviation from the facility/tool specified pressure target. This is to ensure that the design and operation of the UPW delivery system to the manufacturing tools is capable of maintaining the pressure within the specified range The pressure is measured with an on-line sensor installed at the critical point of connection (POC) at the distribution system as required critical tools to provide target pressure to specific tools The Absolute pressure values including specification stability over time and consistency across the distribution system will be determined by specific tool requirements or site facilities specs UPW using tools are becoming increasing sensitive to the actual delivery pressure and variations of the delivery pressure. Specifically manufacturing tools that have, but are not limited to: Critical timed UPW bath fills. Sensitive UPW spraying operations that require consistent and stable pressures and flows. Exact chemical dilution procedures utilizing UPW line pressure Fluctuations in UPW delivery pressures to the tool POC (even minor) are capable of causing differences in process recipes that will and can reflect an out of control quality indicator Due to the increasing size and complexity of modern UPW delivery piping systems, there is frequent risk of having measurable delivery pressure differentials/variations between different POC locations throughout the Fab that have a propensity to cause UPW using tools to perform differently. Stable and reliable UPW pressure is a key variable that is easy and important to minimize to ensure improved and consistent quality Hydrogen Peroxide, H 2O This parameter was added to the table with no recommended values for the following reasons: Hydrogen Peroxide was reported to exist in UPW at concentrations higher than what is specified for dissolved oxygen. Given that hydrogen peroxide has an even stronger oxidation potential, it is expected that this parameter is going to be controlled in future UPW systems At this point there is limited information about presence of hydrogen peroxide in most UPW systems as well as it is unclear what the impact is going to be to UPW systems performance and cost from determining specific values for this parameter. Page 13 Doc SEMI

16 It is also unclear what method of measurement can provide accurate and reliable measurement. NOTICE: SEMI makes no warranties or representations as to the suitability of the Standards and Safety Guidelines set forth herein for any particular application. The determination of the suitability of the Standard or Safety Guideline is solely the responsibility of the user. Users are cautioned to refer to manufacturer s instructions, product labels, product data sheets, and other relevant literature, respecting any materials or equipment mentioned herein. Standards and Safety Guidelines are subject to change without notice. By publication of this Standard or Safety Guideline, SEMI takes no position respecting the validity of any patent rights or copyrights asserted in connection with any items mentioned in this Standard or Safety Guideline. Users of this Standard or Safety Guideline are expressly advised that determination of any such patent rights or copyrights, and the risk of infringement of such rights are entirely their own responsibility. Page 14 Doc SEMI