NASA/Glenn Research Center, 21000 Brookpark Road, Cleveland, OH 44135

D -- 2D/3D SEMICONDUCTOR SIMULATION SOFTWARE PACKAGE SOL RFQ3-130310 DUE 090399 POC Wayne R. Girard, Contract Specialist, Phone (216) 433-2799, Fax (216) 433-5489, Email Wayne.R.Girard@grc.nasa.gov WEB: Click here for the latest information about this notice, http://nais.nasa.gov/EPS/GRC/date.html#RFQ3-130310. E-MAIL: Wayne R. Girard, Wayne.R.Girard@grc.nasa.gov. Offerors have submitted the additional following questions concerning this RFQ: These questions refer to the P1 (1st tier) SiGe HBT simulation. Question 1) An Offeror has requested a cross-section drawing of the device to be simulated. Enough detail to to identify the topology and all lateral dimensions is also requested. Answer: 1) Dimensions are based on the general structures given in the article ("Enhanced SiGe HBT's with 160 GHz-fmax", A. Schueppen et al. IEDM, p 743 (1995) -- This is the original article referenced in the RFQ) and in references 3 and 6, therein. The most important dimensions for all three devices to be simulated is the emitter size given in the article: 2 fingers where each one is 8 microns long, 0.8 micron wide (unless the width is a variable), and the distance between the fingers is equal to the finger width. The emitter fingers are parallel, and each finger is on it's own mesa. The contact to the emitter covers all the emitter area and is connected with an air bridge with the same width as the emitter and half it's length to a large pad. The other parts of the structure are not as crucial. The base is also on it's own mesa. According to the scale on the published articles, the base size is roughly 4.5 by 12 microns. The long side is parallel to the emitter fingers, and continues roughly 4 microns beyond the end of the emitters. The short side is perpendicular and symmetrical to the emitters. The base contact covers all the area of the base and is connected on both sides of the long dimension through air bridges (with roughly 3 micron width and 5 micron length) to large pads. The collector is also on a mesa. It is located beneath the base, but extends roughly 8 micron beyond the long dimension of the base. Thus, we can approximate the collector dimension as 20 by 5 microns. The collector contact is on the part that is not under the base and it is on the edge further away from the base. The collector contact is separated by roughly 1 micron from the base and thus it's size is 7 by 5 microns. The collector contact is then connected at it's whole width of 7 micron and 10 micron long as an air bridge to a large pad. Question 2) Concerning Table 1 in Ref. 1.: For the middle "base: row three entries are given: "i_CB|w_B|i_BE[nm]" w_B refers to the base width, but what is the meaning of the other two quantities? Answer: 2) The base is made of two undoped spacers (called i for intrinsic, i.e. undoped) on top and bottom of the doped part. The i_CB is the width in nm of the spacer on the collector side of the base and i_BE is the width of the spacer on the emitter side. w_B is the width of the doped part of the base. The base is the part that is made of SiGe. Question 3) Concerning Table 1 in Ref. 1: Is the Germanium profile in the base graded or abrupt? If it is graded, what is the meaning of "x" in the bottom base row and please supply details about the grading profile. Answer: 3) The Ge is ungraded.The value of x is constant over the whole width of the base. For example, a value of x=0.28 means that the Ge concentration in the base is 28% at. %. Question 4) The requested simulations are for devices 2846, 2992, and 3400. The data shown in Fig. 3 of Ref. 1 displays results only for the device 3400 and 2846 (and other for devices which were not included in the request). Should a similar simulations be made for the device 2992? Answer: 4) Please simulate all 3 devices as required in the solicitation. Question 5) The requested simulations are for the devices 2846, 2992, and 3400. he data shown in Fig. 4 of Ref. 1 displays results only for the device 3400 and 2846 (and other for devices which were not included in the request). Should a similar simulations be made for the device 2992? Answer: 5) Please simulate all 3 devices as required in the solicitation. Question 6) What bias conditions are to be used for Fig. 5 (fT/fmax vs. Gummel Number)? Should the bias conditions given in Table 1 in Ref. 1 (i.e. "2846" Vce=1.8V, Ic=28mA; "2992" Vce=4V, Ic=5mA; "3400" Vce=5V, Ic=4mA) be used for each device? Answer: 6) The bias conditions mentioned above and in the article are the ones that should be used. Question 7) What bias conditions are to be used for Fig. 6 (fT/max vs. Collector doping number)? Should for each device the bias conditions given in Table 1 in Ref. 1 be used? Answer: 7) The bias conditions mentioned above and in the article are the ones that should be used. Question 8) What bias conditions are to be used for Fig. 6 (fT/max vs. emitter width)? Should the bias conditions given in Table 1 in Ref. 1 be used for each device? Answer: 8) The bias conditions mentioned above and in the article are the ones that should be used. Question 9) The Table in Ref. 1 also contains DC data. Should they be simulated also? Answer: 9) The DC data (beta, breakdown voltage between the collector and emitter V_CE-br) should be simulated. Posted 08/19/99 (D-SN369951). (0231)

Loren Data Corp. http://www.ld.com (SYN# 0023 19990823\D-0005.SOL)