GOALI: New Dielectrics for Development of Metal to Metal Antifuse for FPGA

本提案提出了一项使用a- 6，C1-x:H， a- c:H和a- c:H，N，F电介质开发用于现场可编程门阵列（fpga）的新型金属对金属防熔丝结构的计划，该技术将转让给Actel公司进行工艺集成。与工业界学术联络资助机会（GOALI）是一个合适的项目，因为该项目促进了大学与工业界在高风险/高收益研究领域的跨学科合作，而这些研究领域是工业界无法开展的。fpga的应用非常广泛。只要设备设计使用大量的中小型ic， FPGA就有可能提供更快、更紧凑和更可靠的解决方案。基本的防熔断器是夹在导体之间的一层薄薄的绝缘层，它会被高压改变（编程）。目前，有两种类型的抗熔断器被商业化生产：n+ Si/ONO/多晶硅抗熔断器（其中ONO代表氧化物/氮化物/氧化物）和金属/非晶硅/金属抗熔断器。虽然n+ Si/ONO/多晶硅的应用前景很好，但ONO表现出较强的时间相关介电击穿（TDDB）行为。a- si抗熔断器的最大问题是，当受到大电流脉冲的压力时，程控抗熔断器有时会恢复到高阻抗状态。在与德州仪器的合作中，PI在开发非晶碳（a-C:H）和非晶碳化硅（a-SiC:H）抗熔断器方面非常成功，具有非常有前途的性能。与目前使用的非晶硅（a-Si:H）金属对金属防熔断器相比，金属对金属a-C:H和a-C:H，N，F防熔断器具有泄漏电流低3个数量级、击穿电压低2倍、导通电阻相似、导通稳定性好等特点。Actel公司，我们的工业合作伙伴对德克萨斯理工大学a-C:H，N，F反熔断结构的开发非常感兴趣，可以在1999年的时间框架内转移并集成到他们的0.6微米CMOS工艺中。由于设备和工程资源的限制，没有德州理工的帮助，Actel很难研究这些材料。我们建议改变工艺条件，如CH4/NF3的流速、射频功率、偏置电压、压力，并在PECVD系统中沉积不同浓度的氢、氮和氟的a-C:H、N、F薄膜。为了优化器件的性能，将广泛研究薄膜的抗熔丝特性与N、F、H含量之间的关系。我们还将使用二乙基硅烷和甲烷沉积a-SixC1-x，：H。德克萨斯理工大学最有前途的反熔断结构将在Actel公司进行进一步的可靠性测试。选定的结构将转移到Actel公司进行进一步开发，并纳入0.6微米CMOS工艺流程。根据NSF的支持，Actel公司将提供204730美元的配套资金，其中包括6万美元的资金和144730美元的实物支持，包括样品和样品分析、处理、调查时间和捐赠的设备。德克萨斯理工大学提供了58,200美元作为配套资金，给予PI在Actel公司工作的休假，并免除Actel提供的6万美元资金的间接费用。*** mm1

9614593 Gangopadhyay mm1 This proposal presents a plan for the development of new metal to metal antifuse structures for the field programmable gate arrays (FPGAs) using a-Six,C1-x:H, a-C:H, and a-C:H,N,F dielectrics, technology of which will be transferred to Actel Corporation for process integration. Grant Opportunities for Academic Liaison with Industry (GOALI) is an appropriate program for the proposed work, because this program promotes an interdisciplinary university-industry collaboration in high-risk/high gain research areas which would not have been undertaken by industry. The applications of FPGAs are numerous. Wherever a device design uses a substantial amount of small and medium scale ICs, there exists the potential for an FPGA to provide a faster, more compact and more reliable solution. The basic antifuse is a thin insulating layer sandwiched between conductors that gets altered (programmed) by the application of a high voltage. At present, two types of antifuses are commercially produced: the n+ Si/ONO/poly Si antifuse (where ONO stands for Oxide/Nitride/Oxide) and the metal/amorphous silicon/metal antifuse. Although the applications of n+ Si/ONO/poly Si are quite promising, the ONO shows strong time dependent dielectric breakdown (TDDB) behavior. The biggest problem with a-Si antifuses is that a programmed antifuse sometimes reverts to a high impedance state, when it is stressed with a high current pulse. In collaboration with Texas Instruments, the PI has been very successful in developing amorphous carbon (a-C:H) and amorphous silicon carbide (a-SiC:H) antifuses with extremely promising properties. The metal to metal a-C:H and a-C:H,N,F antifuses showed three orders of magnitude lower leakage current, factor of two lower breakdown voltage, similar On-state resistance, and superior On-state stability compared to currently used amorphous silicon (a-Si:H) metal to metal antifuses. Actel Corporation, our industrial collaborator is extremely interested in the deve lopment of a-C:H,N,F antifuse structures at Texas Tech, which can be transferred and integrated into their 0.6 micron CMOS process within 1999 time frame. It is difficult for Actel to study such materials without the help of Texas Tech because of the constraints of limited equipment and engineering resources. We propose to change the processing conditions, such as CH4/NF3 flow rates, rf power, bias voltage, pressure and to deposit a-C:H,N,F films with various concentrations of hydrogen, nitrogen and fluorine in our PECVD system. The relationship between the antifuse characteristics and the N, F, H content of the films will be studied extensively for the optimization of the devices. We will also deposit a-SixC1-x, :H using diethysilane and methane. The most promising antifuse structures at Texas Tech will be further tested for reliability at Actel Corporation. Selected structures will be transferred to Actel Corp. for further development and inclusion into the 0.6 micron CMOS process flow. Contingent upon NSF support, Actel Corp. will provide $204,730 matching, which will include $60,000 funding and $144,730 in-kind support, including samples and sample analysis, processing, investigator time and donated equipment. Texas Tech has provided $58,200 as matching funds by granting the PI a sabbatical leave to work at Actel Corporation and by waiving the indirect cost for the $60,000 funding from Actel. *** mm1