Enhancement Mode Gallium Nitride (eGaNTM) FET Characteristics under Long Term Stress Alexander Lidow J. Brandon Witcher Ken Smalley Chief Executive Officer Efficient Power Conversion Corporation El Segundo, California 90245 Senior Member of the Technical Staff Sandia National Laboratories Albuquerque, New Mexico 87185 Product Engineer Microsemi Corporation Lawrence, Massachusetts 01841 Abstract: Enhancement mode HEMT transistors built with Gallium-Nitride-on-silicon (eGaN) have been in the commercial marketplace since 2009 as a replacement for silicon power MOSFETs. Superior conductivity and switching characteristics allow designers to greatly reduce system power losses, size, weight, and cost. Military and space applications could benefit from using eGaN FETs, but the parts need to operate reliably under harsh environmental conditions. In this paper we present results demonstrating the stability of these devices at temperature and under radiation exposure. FET. Aluminum gallium nitride (AlGaN) is then grown on top of the GaN to form a highly conductive two dimensional electron gas layer (2DEG) immediately underneath . Additional layers are formed to create an enhancement mode gate electrode as well as drain and source electrodes (see Figure 1). This structure is repeated many times to form a power device. Three layers of top metal, isolated from each other by planarized insulators, are then deposited to complete the circuit and bring the terminals to the outside world. The end result is a simple, cost-effective solution for power switching. This device looks and behave similarly to a silicon MOSFET. Keywords: Gallium nitride; GaN; eGaN; MOSFET; FET; Radiation tolerance; SEE; Gamma radiation; Proton radiation Introduction: The basic requirements for power semiconductors are efficiency, reliability, controllability, and cost effectiveness. Recent breakthroughs by Efficient Power Conversion Corporation (EPC) in processing gallium nitride have produced enhancement mode devices (eGaNTM) with conductivity and switching speeds much higher than Si power FETs. These improvements enable power converters with higher efficiency and switching frequency, as well as greater input voltage range, leading to simpler, smaller power systems. Ranging in voltage from 40 V to 200 V, and on-state resistance from 4 mΩ to 100 mΩ, this new class of devices has also proved very capable in a high radiation environment. This paper is divided into four sections. The first discusses the basic eGaN FET structure. The second section shows the result of heavy ion testing with linear energy transfers (LETs) ranging from 28.8 to 87.2 MeV*cm2/mg. The third section gives pre-and post results for total dose testing up to 500 kRad (Si). The final section discusses the successful multi-thousand hour stress testing under a variety of bias and temperature conditions. Structure: EPC’s eGaN FET process begins with silicon wafers. A thin layer of Aluminum Nitride (AlN) is grown on the silicon to isolate the device structure. On top of the AlN, a thicker layer of highly resistive GaN is grown. This layer provides a foundation on which to build the eGaN Presented March 24, 2011 GOMACTech11 Figure 1. Simplified cross section of an eGaN FET SEE Testing: Heavy-ion testing of EPC eGaN FETs was performed at the Texas A&M cyclotron in September, 2010 following MIL-STD-750E, METHOD 1080. EPC’s entire family of eGaN FET products was tested (See Table 1 for device parameters). EPC1005 (60 V, 7 mΩ), EPC1014 (40 V, 16 mΩ), and EPC1015 (40 V, 4 mΩ) did not exhibit any gate ruptures (SEGR) or drain ruptures (SEB) at the highest LET available. For the remaining part numbers, drain-ruptures were the primary mode of failure and were typically observed as being a gradual transition into failure. In figure 2 is shown the applied drain-source voltages up to which the parts remained within data sheet limits (dark grey); the voltages up to which the devices exceeded data sheet limits but remained functional (light grey), and the voltage beyond which the devices failed catastrophically (black). In general, the eGaN FETs demonstrated SEE capabilities that exceed similar silicon MOSFETs currently listed on the Qualified Military Listing (QML). Table 1: EPC’s eGaN FET Electrical Characteristics 250 200 200 200 150 150 150 100 100 50 50 0 0 Voltage 250 Voltage Voltage SEE Heavy Ion Testing - Kr SEE Heavy Ion Testing - Xe SEE Heavy Ion Testing - Au 250 100 50 0 Figure 2. SEE Results for Au, Kr, and Xe bombardment. Dark grey bars represent the voltage range within which devices remained inside data sheet limits. Light grey bars represent the range where devices continued to function but drain-source leakage exceeded data sheet. Black bars represent regions where devices catastrophically failed Total Dose Testing: Utilizing the “Gamma Cave” at the University of Massachusetts, Lowell, six EPC1014 (40 V, 4 mΩ) were subjected to a total gamma dose of 500 kRads (Si) at a dose rate of 96 Rads (Si)/sec. A 60Co source was used and all testing was according to MIL-STD-750, Method 1019. Two different test conditions were used. The first test condition biased the drain-source at 80% of rated VDS(MAX) (32 V in the case of the EPC1014). The second test condition biased the gate-source at 5 V. Table 2 shows the pre and post electrical characteristics of these devices. Very little change is seen in any of the characteristics under either bias condition. With 32 V from drain to source during irradiation, the threshold voltage (VTH) changed less than 18% percent; RDS(ON) changed less than 8 percent, and all parameters remained well within the data sheet limits. With 5 V from gate to source during irradiation, the threshold voltage (VTH) changed less than 4 percent; RDS(ON) changed less than 3 percent, and all parameters again remained well within the data sheet limits. Presented March 24, 2011 GOMACTech11 Whereas we can’t quantify how long the parts will last in a given orbit since that depends on factors such as the placement of the part within the payload and within the spacecraft (This can make an order of magnitude to the dose rate), we can make relative comparisons. The problem designers run into with silicon MOSFETs is that they must choose between radiation tolerance and electrical performance. Commercial MOSFETs have thick gate oxides and trap a lot of charge, resulting in large shifts in the threshold voltage and eventual failure at relatively low total-dose. Meanwhile, the rad-hard MOSFETs available have on-resistance and device capacitance several times higher than their commercial counterparts, leading to either low efficiency or large size (due to the low switching frequency). Now we have a new capability; eGaN FETs, with electrical performance superior to the cutting edge Si MOSFETs, and radiation tolerance at least as high as the best rad-hard power MOSFETs available. These eGaN FETs bring a combination of electrical and radiation performance that cannot be matched. Table 3. Reliability test results for eGaN FETs Additional radiation testing is ongoing using different part numbers, different bias conditions, and different dose rates. These results will be reported in a later publication and are expected to reinforce the conclusion that eGaN FETs are resistant to very high doses of radiation. Table 2. Pre and post irradiation data shows little change after 500 kRad (Si) EPC1014 kRAD 0k VDS 0k VGS Group D Limit 500k VDS 500k VGS Test Bias1 Bias2 Min Max 1 2 3 4 5 6 Min Max 1 2 3 4 5 6 IGSSr 5.00 V IGSSf 5.00 V IDSS 32.0 V IDSS 40.0 V 500.0uA 12.60u 13.23u 11.27u 11.35u 11.49u 12.47u 2.000mA 100.0uA 100.0uA 38.87u 29.91u 29.06u 58.39u 26.26u 27.16u 34.95u 31.80u 30.24u 27.85u 31.42u 30.63u 27.90u 30.04u 29.69u 23.76u 30.29u 29.55u 500.0uA 5.119u 4.698u 3.860u 9.059u 6.500u 11.61u 2.000mA 100.0uA 100.0uA 24.95u 10.58u 11.93u 46.89u 8.679u 11.07u 28.35u 8.187u 10.75u 22.15u 33.16u 30.80u 32.39u 22.82u 23.51u 19.52u 35.05u 35.64u VTH 2.00mA 700.0m 2.500 V 1.624 1.583 1.552 1.487 1.548 1.606 700.0m V 2.500 V 1.714 1.868 1.879 1.482 1.611 1.568 RDON 5.00 A 5.00 V VDSON 500 mA 0.00 V 16.00mR 1.800 V 14.18m 2.304 14.64m 2.283 13.83m 2.254 13.68m 2.227 13.54m 2.251 13.81m 2.302 16.00mR 1.800 V 14.71m 2.401 15.83m 2.449 14.55m 2.386 13.47m 2.252 13.58m 2.243 13.41m 2.292 Long Term Reliability Testing: The key reliability considerations for power transistors include: (a) device stability in the on-state when the FET is fully enhanced with voltage applied to the gate; (b) device stability in the off-state when the FET is in voltage blocking mode withstanding up to its rated drain-source voltage; and (c) device stability in switching operation. Device stability is impacted by device design, packaging technology, and operating environment. Good reliability results have previously been reported for depletion mode GaN FETs for RF [2,3] applications and power switching applications . EPC’s eGaN FET reliability test results are summarized in Table 3, in which the type of test, stress conditions, part numbers, sample size, stress hours, and number of fails are listed. JEDEC standards were followed where applicable. Even very high levels of environmental stress applied to hundreds of parts generated no failures thus demonstrating the basic capability of the technology and the product to survive over many years in terrestrial or space environments. Presented March 24, 2011 GOMACTech11 # of fails 1000HR 3000HR Stress Test Test Condition Part Number Sample Size HTRB 100Vds, 125oC EPC1001 45 0 - HTRB 40Vds, 125oC EPC1014 50 0 - o HTRB 200Vds, 125 C EPC1012 50 0 HTRB 200Vds, 125oC EPC1010 with underfill 50 0 - HTRB 200Vds, 150oC EPC1010 50 0 0 Stress Test Test Condition Part Number Sample Size HTGB 5V 5Vgs, 125oC EPC1001 45 0 0 HTGB 5.4V 5.4Vgs, 125oC EPC1001 45 0 0 HTGB 5V 5 Vgs, 150oC EPC1010 45 0 - HTGB -5V -5Vgs, 125oC EPC1001 50 0 - Stress Test Test Condition Part Number Sample Size TC -40C to 125oC EPC1001 - # of fails 1000HR 3000HR # of fails 1000cys 45 0 TC -40C to 125 C EPC1014 50 0 TC -40C to 125oC EPC1012 45 0 - TC -40C to 125oC EPC1012 with underfill 45 0 - Stress Test Test Condition Part Number Sample Size THB 85oC/85RH, 40Vds EPC1014 45 0 - THB 85oC/85RH, 40Vds EPC1015 45 0 - THB 85oC/85RH, 100Vds EPC1010 25 0 - EPC1010 with underfill 25 0 - THB o o 85 C/85RH, 100Vds Stress Test Test Condition Part Number Sample Size MSL1 85oC/85RH, 168HR EPC1001 50 Stress Test Test Condition Part Number Sample Size Power Supply Life Test 10A, 250 kHz, 30oC EPC1001 10 - # of fails 1000HR # of fails 168HR 0 - # of fails 1200HR 0 - Conclusions: EPC’s eGaN FETs have been tested under heavy ion bombardment, gamma irradiation, and various environmental stress factors. These devices demonstrate their ability to be used in the most stringent of environmental as well as radiation environments and exceed the capabilities of silicon power MOSFETs. References 1. M. Asif Khan, A. Bhattarai, J.N. Kuznia, and D.T. Olson, “High Electron Mobility Transistor Based on a GaN-AlxGa1-xN Heterojunction,” Appl. Phys. Lett., vol. 63, no. 9, 1993, pp. 1214-1215. 2. S. Singhal, J.C. Roberts, P. Rajagopal, T. Li, A.W. Hanson, R. Therrien, J.W. Johnson, I.C. Kizilyalli, K.J. Linthicum, “GaN-on-Si failure mechanisms and reliability improvements,” in Proc. IEEE Int. Rel. Phys. Symp., 3. P. Saunier, C. Lee, A. Galistreri, D. Dumka, J. Jimenez, H.Q. Tserng, M.Y. Kao, P.C. Chao, K. Chu, A. Souzis, I. Eliashevich, S. Guo, J. del Alama, J. Joh, M. Shur, “Progress in GaN Performances and Reliability,” in Proc. 4. M. Briere, “GaN on Si Based Power Devices: An Opportunity to Significantly to Impact Global Energy Consumption,” CS MANTECH, May, 2010.