Schottky Diode RF-Detector and Focused Ion Beam Post-Processing MURI Annual Review Woochul Jeon, Todd Firestone, John Rodgers & John Melngailis University of Maryland. (consultations with Jake Baker Boise State & Michael Gaitan NIST)
Outline Operation and characteristics of Schottky power detector Mask layout for Schottky diodes Fabricated Schottky diodes with n+ substrate with n-epi layer on top Schottky diodes by CMOS process RF radiation test Schottky diodes by using Focused ion beam technology Schottky diodes designed for MOSIS standard CMOS process Conclusion and future work
Original Project Objectives: - Direct analog microwave level measurement on a chip using a) Schottky diodes b) Thermal detectors - Incorporation of RF detectors on chips, including FIB diode fabrication on existing chips - Focused ion beam diagnosis circuit restructuring and device diagnosis by burned out element sectioning Changes to Objectives: - Thermal detectors not pursued
Example of FIB Circuit Rewiring: Cut and Jumper
FIB-Milled Circuit Cross Section
Operation of RF Power Detector RF 50pF 1kΩ + - DC Output 100ns RF input
Key factors limiting maximum frequency Junction capacitance C j εqn d = A 2( V a + V d ) A: contact area, N d : doping concentration V a : applied voltage, V d : built in voltage R j Junction resistance: 1 R j = di dv x a = AI q nkt Series resistance = R n + R n+, R n =R o /A + R 1 /A1/2 s exp R n >> R n+ ( >> R j, after turning on) Series resistance mainly determined by R n (n layer resistance). RC time constant A 1/2 (due to spreading resistance), R n (N d ), etc. Objective Reduce junction capacitance(c j ) => decrease contact area Reduce series resistance => minimize n layer thickness qv nkt a R n R n+ C j Equivalent circuit
Schottky diode fabricated in MOSIS V. Milanovic, M. Gaitan, J.C. Marshall M. E. Zaghloul, IEEE Trans. Electr. Devices 43, 2210 (Dec. 1996)
Coplanar Schottky Diode Developed for Rectifying Antennas K.M. Strohm, J. Buecher, & E. Kasper, Daimler Benz Research, Ulm IEEE Trans. MTT Vol.46, 669, (May, 1998)
Proposed structure using n+ substrate with n-epi layer on top Reduce series resistance => use n+ substrate Reduce contact capacitance => decrease contact area 2 µm n-epi (0.3 µm) n+ Ohmic contact Schottky contact
Resistivity vs. Depth of n on n + Layer Resistivity Ω-cm Depth µm
Mask layout 9 mm 8.5 mm
Schottky diode with clock tree 600µm
Schottky diode with 150µm pitch pads 150µm G 4 25 µm Schottky diode S 130µm 100µm G 24 50 µm
Schottky Diode layout A Ohmic contact 2µm Metal (Al) Schottky Contact n+ SiO 2 A
Etch SiO2 and 0.3um n silicon layer with n-high mask (SiO2: wet etching, n-si: RIE) n+ Etch SiO2(wet etching) with n-low mask n+ Deposit 0.5um Al with E-Beam evaporator n+ Etch Al(wet etching) with metal mask n+
Measured result I (DC) DC Characteristics(2µm x 2µm by RIE, I-V curve) 6 2500 5 2000 4 1500 Iout 3 2 Iout(uA) 1000 1 500 0-1 -1-0.8-0.6-0.4-0.2 0 0.2 0.4 0.6 0.8 1 Vin Exponential change of contact resistance R j >> R s 0-500 -4-3.5-3 -2.5-2 -1.5-1 -0.5 0 0.5 1 1.5 2 2.5 3 3.5 linear series resistance R s >> R j Vin
Measured result II (RF) 2µm x 2µm contact area diodes are tested. These diodes worked at the power level from -10 dbm to 10 dbm DC output was linearly changed by changing power level. Observed diode response up to 5GHz These diodes could detect RF power level, but because of the direct capacitance connection between anode and cathode of diode, the output DC voltage substantially depended on frequency. This huge capacitance (1.2 pf by calculation) comes from pad structure for connecting diode to RF source. Al-Si Pad SiO 2 n n+ 1.2pF
Equivalent circuit R j C j R s C o Rj: Junction Resistance Cj: Junction capacitance Rs: Series resistance (Rs n + Rs n+ ) C o : Overlay capacitance between Al-Si pad and n+ layer Overlay capacitance gives direct path between anode and cathode of Schottky diode.
Schottky diode Design for CMOS process To remove the effect of Co, different substrate which has higher resistivity rather than n+ substrate should be used. Design new diode structure to minimize series resistance of n layer without using Silicon Molecular Beam Epitaxy(Si-MBE) Minimize contact area Schottky Contact Ohmic Contact SiO 2 Al 2um n+ Al n - Substrate
Measured result (DC) 2x2 patch I-V 30 Iout(mA) 25 20 15 10 5 Voltage range: -5V ~ 5V. (From 5V to 0V, current output was 0) Series resistance(between 4V and 5V) 83Ω 0 0 0.5 1 1.5 2 2.5 3 Vin(V) 3.5 4 4.5 5
RF direct injection test (50µm x 50 µm contact area) DC output vs RF Power level RF input: Cascade probe G S G DC output Output voltage(v) 4 3 2 1 0 0 5 10 15 20 25 RF power(dbm)
RF direct injection test (2µm x 2 µm contact area) DC output vs. Power level Output voltage(v) 2.5 2 1.5 1 0.5 0 DC output vs. Frequency 1 3 5 7 9 Frequency(GHz) Output Voltage [V] 2 1.5 1 0.5 0 5 GHZ 6 GHZ 0 5 10 15 20 Power [dbm] Flat response at high frequency range
RF direct injection test (50µm x 50 µm contact area) 10 Diode size 50X50 microns Rectified Voltage (V) 1 0.1 0.01 1 10 100 Frequency (GHz)
Output Voltage Pulse in Response to 20 GHz. RF Burst Output (V) 0.05 0.04 0.03 0.02 0.01 0-0.01 0 1 2 3 4 5 Time (µsec) Rectified Voltage 20 GHz RF Pulse Envelope
Output Voltage Pulse in Response to 2GHz.RF Burst Output (V) 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0-0.1 0 1 2 3 4 5 Time (µsec) Rectified Voltage 2GHz RF Pulse Envelope
RF radiation test on a patch antenna structure 12cm x 10cm x 6cm size box is used for radiation test RF radiation Horn structure Top view Ground plane Output port 6 cm 12 cm 10 cm Bonding to ground plane Patch antenna
RF radiation test result Rf_in vs. Vdc_out (frequency = 12 GHz) 120 100 80 60 40 20 0 25 27 29 31 33 35 37 39 41 43 45 47 49 Frequency vs. DC output (RF power = 40dBm) DC output(mv) DC output(mv) 20 18 16 14 12 10 8 6 4 2 0 0 18.8 6.6 14.6 12.6 17.6 1.8 1.8 0.4 2.8 2 0 1.4 0.4 0 1 0.8 0.8 0.4 0 0 0 RF radiation input(dbm) Frequency(GHz) Roll-off frequency 12 GHz
Fabricating Schottky diodes by FIB SiO 2 Al n+ n substrate FIB milling Al n FIB Metal(Pt) deposition Iout ( ua ) 3000 2500 2000 1500 1000 500 Measured result(iv curve) Schottky contact -500 n Al 0-2 -1.6-1.2-0.8-0.4 0 0.4 Vin ( V ) 0.8 1.2 1.6 2
Fabrication of Schottky diode by FIB N+ doped area FIB milling (2µm x 3 µm) Al SiO 2 FIB Pt deposition
RF direct injection test of FIB diode 10 Vout vs. Frequency at 15dBm RF power 2.5 2 8GHz 10GHz FIB power sweep at 8GHz and 10GHz Vout (V) 1 Vout (V) 1.5 1 0.5 0.1 1 3 5 7 9 11 13 15 17 19 21 Frequency(GHz) Vout vs. Frequency sweep 0 0 5 10 15 20 25 Injected RF power(dbm) Vout vs. RF power sweep
Schottky diodes with capacitor load and MOSFET amp for amplifying small output signal
One Schottky diode Ohmic contact 150 µm 100 µm Schottky contact Contact area: 2 µm x 2 µm - 40 µm x 40 µm
Al SiO 2 n+ n Fabricating Schottky diode by FIB (Future work) Al Al FIB milling n Al FIB milling(0.1-0.5 µm) and Metal deposition FIB SiO 2 deposition Al Al Al Al n Schottky contact n
500nm FIB Tungsten Vias Through FIB Deposited Oxide Plugs
Summary Schottky diodes on n-epi and n+ substrate were fabricated and tested CMOS process Schottky diodes were designed, fabricated and tested with RF radiation up to 12.5GHz (50X higher than previous CMOS result) and by direct injection up to 20GHz Schottky diodes were fabricated by FIB techniques and tested up to 17.5 GHz Various Schottky diodes have been designed and submitted to MOSIS for standard CMOS processing Paper will be presented at the 2003 International Semiconductor Device Research Symp. in DC
Future work MOSIS chips now being built will be tested by RF radiation and direct injection Post processing MOSIS chips for FIB diodes Diodes with in-situ amplifiers on chip Diodes with built in DC bias will be designed for MOSIS and built Diodes will be incorporated into test chips designed by colleagues to verify various RF propagation models Understand what limits frequency & push toward 100GHz without MBE