{"title":"Particulate contamination control in plasma processing: building-in reliability for semiconductor fabrication","authors":"G. Selwyn","doi":"10.1109/IRWS.1995.493585","DOIUrl":"https://doi.org/10.1109/IRWS.1995.493585","url":null,"abstract":"Plasma processing is used for /spl sim/35% of the process steps required for semiconductor manufacturing. Recent studies have shown that plasma processes create the greatest amount of contaminant dust of all the manufacturing steps required for device fabrication. Often, the level of dust in a plasma process tool exceeds the cleanroom by several orders of magnitude. Particulate contamination generated in a plasma tool can result in reliability problems as well as device failure. Inter-level wiring shorts different levels of metallization on a device is a common result of plasma particulate contamination. We have conducted a thorough study of the physics and chemistry involved in particulate formation and transport in plasma tools. In-situ laser light scattering (LLS) is used for real-time detection of the contaminant dust. The results of this work are highly surprising: all plasmas create dust; the dust can be formed by homogeneous as well as heterogeneous chemistry; this dust is charged and suspended in the plasma; additionally, it is transported to favored regions of the plasma, such as those regions immediately above wafers. Fortunately, this work has also led to a novel means of controlling and eliminating these unwanted contaminants: electrostatic \"drainpipes\" engineered into the electrode by means of specially designed grooves. These channel the suspended particles out of the plasma and into the pump port before they can fall onto the wafer.","PeriodicalId":355898,"journal":{"name":"IEEE 1995 International Integrated Reliability Workshop. Final Report","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1995-12-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114347930","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"A new approach for predicting AC hot carrier lifetime","authors":"H. Kato","doi":"10.1109/IRWS.1995.493586","DOIUrl":"https://doi.org/10.1109/IRWS.1995.493586","url":null,"abstract":"The author focuses on how to predict AC hot carrier (MC) lifetime based on the DC HC stress data. As an introduction, he explains the AC transition waveform of an inverter. Then, conventional techniques to predict AC HC lifetime are described. The new technique is then detailed, including the acceleration model, the device parameter to be monitored, and the AC/DC ratio.","PeriodicalId":355898,"journal":{"name":"IEEE 1995 International Integrated Reliability Workshop. Final Report","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1995-10-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127536899","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
L. Anderson, S. Parikh, S. Nagalingam, C. Haidinyak
{"title":"A case study in a 100/spl times/reduction in sodium ions in a 0.8 /spl mu/m BiCMOS process using triangular voltage sweep","authors":"L. Anderson, S. Parikh, S. Nagalingam, C. Haidinyak","doi":"10.1109/IRWS.1995.493574","DOIUrl":"https://doi.org/10.1109/IRWS.1995.493574","url":null,"abstract":"This case study shows how to use Triangular Voltage Sweep (TVS) to reduce Na and K in the backend of a triple-metal BiCMOS process from 10/sup 12/ ions/cm/sup 2/ to 10/sup 10/ ions/cm/sup 2/. TVS is compared to Bias Temperature Stress (BTS) techniques. While Capacitance-Voltage plots are good monitors for bulk contamination (metal, deposited oxide, etc.), data is presented which shows that TVS is superior for detecting surface-introduced mobile ions (photoresist, solvent strip, etc.). Process integration techniques and issues in the reduction of mobile ions are discussed. When TVS structures are put on product-like wafers, the mobile ions can be measured accurately and repeatably within 10 minutes of completing the process step-no alloy is required! Finally, the use of Ammonium Fluoride solutions to reduce the surface mobile ions are discussed.","PeriodicalId":355898,"journal":{"name":"IEEE 1995 International Integrated Reliability Workshop. Final Report","volume":"97 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1995-10-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116793421","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"A design reliability methodology for CMOS VLSI circuits","authors":"L. Oshiro, R. Radojcic","doi":"10.1109/IRWS.1995.493572","DOIUrl":"https://doi.org/10.1109/IRWS.1995.493572","url":null,"abstract":"As integrated circuits become more complex, ensuring product reliability in a timely and efficient manner is becoming more challenging. Historically, the focus has been to manage reliability through robust manufacturing processes. This approach does not address reliability issues associated with IC design. 'Design-for-Reliability' (DFR) methodologies should be incorporated into chip design flow and CAD tools, to meet the submicron challenge. At Cadence Spectrum Design, a set of practical DFR tools and procedures have been developed and deployed to realistically manage product reliability throughout the design phase of a VLSI circuit. These procedures manage wear-out concerns through verifying compliance of every node with all the reliability design rules. Random failures (bottom of the bathtub curve) are addressed by analyzing the interaction between design and random process defects in order to determine a failure rate. Finally, infant mortality concerns are managed by insuring that the library is 'robust', by verifying the validity of test and/or burn-in screens.","PeriodicalId":355898,"journal":{"name":"IEEE 1995 International Integrated Reliability Workshop. Final Report","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1995-10-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129467315","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"WLR fast tests for characterization of a BiCMOS process and identification of mobile ionic contamination","authors":"M. Poulter, D. Brisbin","doi":"10.1109/IRWS.1995.493601","DOIUrl":"https://doi.org/10.1109/IRWS.1995.493601","url":null,"abstract":"Summary form only given. WLR can play a key role in process development. Reliability issues can be identified and fixed early in the development process. The WLR fast test self heated gate structure is ideally suited for monitoring and characterizing mobile ionic contamination. Correlation between the WLR self heated gate structure and traditional hot chuck bias stress measurement can be demonstrated. Activation of the drift process is easily obtainable from the plots. The throughput of the WLR test vastly exceeds the traditional hot chuck test allowing more comprehensive in line screening and more rapid data generation and problem solving. The WLR test can be used to reproduce all the experiments required when characterizing a mobile ionic problem. Control parameters include voltage bias, temperature, stress time, hysteresis, and bake recoverability.","PeriodicalId":355898,"journal":{"name":"IEEE 1995 International Integrated Reliability Workshop. Final Report","volume":"24 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1995-10-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125603600","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"A study of EEPROM endurance correlation with wafer level reliability data","authors":"D. Wilkie, M. Hensen","doi":"10.1109/IRWS.1995.493576","DOIUrl":"https://doi.org/10.1109/IRWS.1995.493576","url":null,"abstract":"The results of a study of EEPROM endurance correlation with various wafer level reliability tests are presented. Samples from multiple wafer lots of various EEPROM array sizes were cycled, and the data were compared to wafer level data taken from the same wafer lots. The results show that TDDB studies alone do not correlate well with endurance. Other data, such as yield studies and alignment measurements, also do not completely correlate with endurance. Despite this we believe that the wafer level tests such as TDDB being performed now are good indicators of overall oxide quality.","PeriodicalId":355898,"journal":{"name":"IEEE 1995 International Integrated Reliability Workshop. Final Report","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1995-10-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131751167","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Building reliability into an EPROM cell using in-line WLR monitors","authors":"G. Madson, D. Probst, L. Rawlins","doi":"10.1109/IRWS.1995.493573","DOIUrl":"https://doi.org/10.1109/IRWS.1995.493573","url":null,"abstract":"A systematic deprogramming failure on a small percentage of the wafers produced on our 0.8 micron EPROM line threatened profitability and reliability. The development of an in-line JEDEC Qbd test on scribe line capacitors lead to the finding of a direct correlation of the inter-poly oxide voltage at breakdown (Vbd) to deprogramming. Using the inline Vbd correlation enabled the discovery of the root cause of deprogramming-asperities growing from the first poly side wall creating a local area of a enhanced field allowing tunneling of electrons from the floating gates during write cycles of adjacent bits. Because the asperities were not created by one source, but from multiple sources, finding the root cause proved to be a formidable task. Before using the in-line Vbd test, wafer sort provided the only feedback. By using the in-line monitor, the evaluation time was reduced by 70%, reducing the time in our cycles of learning. With the reduced cycle time, a systematic approach to finding the root causes became possible. Screening experiments eliminated as many variables as possible. Finally, the contributing process steps were identified and further experiments lead to the development of a robust process. Descriptions of the methods followed and the resulting process changes are discussed.","PeriodicalId":355898,"journal":{"name":"IEEE 1995 International Integrated Reliability Workshop. Final Report","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1995-10-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130391280","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Passivation scheme impact on retention reliability of non volatile memory cells","authors":"R. Bottini, A. Cascella, F. Pio, B. Vajana","doi":"10.1109/IRWS.1995.493569","DOIUrl":"https://doi.org/10.1109/IRWS.1995.493569","url":null,"abstract":"Non Volatile Memory cells must retain the data (i.e. the charge stored in the floating gate) during the device lifetime, typically at least 10 years. In this work we study the impact of different passivation layers on the data retention of single polysilicon EEPROM cells, processed with an advanced 0.7 /spl mu/m process technology. Three passivation layers have been considered: (1) Phosphorus doped Silicon Glass (PSG), (2) Planarized (Oxynitride/SOG/Oxynitride/PSG), and (3) UV-Nitride. Accelerated tests were performed at high temperature (250-350/spl deg/C) up to 500 hours in order to monitor the threshold voltage shift of the floating gate transistor programmed either in the written or in the erased state. In the case of planarized passivation and of UV-nitride passivation the charge loss is small and it largely fulfils the data retention requirements; in the case of PSG passivation a much higher charge loss is observed. The effect of tunnel oxide degradation after extended cycling (1 Mcycles) has been investigated. No significant difference has been found after 200 hours at 250/spl deg/C between cycled and one time programmed cells, evidencing that the charge loss mechanism does not involve tunnel oxide degradation. The activation energy of the charge loss mechanism has been evaluated in the case of planarized passivation, using written cells. The measured value is 1.84 eV. The impact of different passivation schemes was studied with conventional techniques, the best results were obtained with the planarized passivation stack and with the UV-nitride layer.","PeriodicalId":355898,"journal":{"name":"IEEE 1995 International Integrated Reliability Workshop. Final Report","volume":"22 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1995-10-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122022726","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Corner point lot qualification technique","authors":"J. Gagnon, D. Potts, S.C. Park, R. Whitcomb","doi":"10.1109/IRWS.1995.493571","DOIUrl":"https://doi.org/10.1109/IRWS.1995.493571","url":null,"abstract":"This paper describes a new approach in the sample make up for qualifying a new process, a process change, and, or a material change. This new approach, which we will refer to as the corner point lot method, produces a qualification sample that better represents the process operating window. The corner point lot method uses a design of experiment approach to purposely force critical parameters, as identified through Failure Mode and Effect Analysis (FMEA), to the corners of their spec limits.","PeriodicalId":355898,"journal":{"name":"IEEE 1995 International Integrated Reliability Workshop. Final Report","volume":"31 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1995-10-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122478266","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"A new method to statistically monitor active area spread and oxide thinning in real devices","authors":"G. Ghidini, C. Clementi","doi":"10.1109/IRWS.1995.493588","DOIUrl":"https://doi.org/10.1109/IRWS.1995.493588","url":null,"abstract":"ULSI device scaling down continuously requires to improve the control over any key process parameter. In particular, the active dielectric thickness reduction and the use of innovative isolation schemes make it very difficult to have a direct monitor of oxide thickness and active area width at the same time. A new method is presented here which allows one to verify the electrical quality of the thin oxide, giving also information about the active area spread and the dielectric thickness. A standard technique to evaluate the quality of a thin oxide consists of the application of an exponential current stress up to the capacitor breakdown. In this work it is suggested to measure the capacitance and to start the current ramp not above a current density of 1E-5 A/cm/sup 2/ to avoid stress regimes in which the charge trapping heavily influences the measurements. In this way the procedure can be very effective in distinguishing between active area dimension fluctuations and real oxide thinning at the field oxide border, without adding any extra measurements.","PeriodicalId":355898,"journal":{"name":"IEEE 1995 International Integrated Reliability Workshop. Final Report","volume":"69 21","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1995-10-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120884872","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}