Ultra High Speed InP Heterojunction Bipolar Transistors

Ultra High Speed InP Heterojunction Bipolar Transistors

Ultra High Speed InP Heterojunction Bipolar Transistors Mattias Dahlstrm Trouble is my business, (Raymond Chandler) Ultra High Speed InP Heterojunction Bipolar Transistors Introduction to HBTs How to make a fast HBT Delay terms The graded base The base-collector grade Recent results Record fmax mesa DHBT* Record f DHBT *details regarding this to follow The transistor Small change in base current large change in collector current 30 I in 200 A steps 6 b 25 5 20 4 15

C 10 2 5 1 0 0 0 0.5 1 1.5 V CE Schematic of an HBT 2 (V) Typical common-emitter characteristics 2.5 2

3 J (A/um ) I (mA) 7 InP lattice structure Nearest neighbor: 2.5 A Lattice constant: 5.86 A InP InGaAs InP and InGaAs have -L separations of ~0.65 eV, vs ~0.4 eV for GaAs larger collector velocity InGaAs has a low electron effective mass lower base transit time Objectives and approach Objectives: fast HBTs mm-wave power, 160 Gb fiber optics desired: 440 GHz ft & fmax, 10 mA/m2, Ccb/Ic<0.5 ps/V better manufacturability than transferred-substrate HBTs improved performance over transferred-substrate HBTs Approach: narrow base mesa moderately low Ccb very low base contact resistance required, and good alignment carbon base doping, good base contact process high ft through high current density, thin layers bandgap engineering: small device transit time with wide bandgap emitter and collector Potential uses of InP HBT

Communication systems: wireless communication, fiber optics transceivers, digital processing in radar (ADCs, DACs) Types of circuits: broadband amplifiers, power amplifiers, laser/modulator drivers comparators, latches, fast logic Circuit characteristics 1-10 000 HBTs per IC Very high demands for speed (40-200 GHz) Fast logic with moderate power consumption (~20 mW/gate) Moderate Output Power mmwave power amps, optical modulator drivers ~6 V at Jc=4 mA/m2 , ~2 V at Jc=8 mA/m2 DHBT band diagram: under bias 1 0.5 emitter E (eV) 0 collector -0.5 E c -1 -1.5 -2 base

E v -2.5 -3 0 50 100 150 200 250 Distance () 300 350 400 High speed HBT: some standard figures of merit Small signal current gain cut-off frequency (from H21) 1 nk BT C je Cbc Rex Rc Cbc b c 2f qI c Maximum power gain ( from U) f max

f 8RbbCbci Collector capacitance charging time when switching : Ccb V Ic Scaling laws for fast HBTs for x 2 improvement of all parasitics: ft, fmax, logic speed base 2: 1 thinner collector 2:1 thinner emitter, collector junctions 4:1 narrower current density 4:1 higher emitter Ohmic 4:1 less resistive Challenges with Scaling: Collector mesa HBT: collector under base Ohmics. Base Ohmics must be one transfer length sets minimum size for collector Emitter Ohmic: hard to improvehow ? Current Density: dissipation, reliability Loss of breakdown avalanche Vbr never less than collector Eg (1.12 V for Si, 1.4 V for InP) .sufficient for logic, insufficient for power transferred-substrate base contact undercut collector junction

emitter narrow collector mesa InGaAs collector InGaAs base InP collector InGaAs subcollector InP subcollector SI substrate collector contact Contact resistance: tunneling through barrier Theory: idealized contact xd c 2 sVbi qN d 1 xd exp 2m * Ebarrier 1 1 2Vbi exp

qN d 2m * Ebarrier High doping: 1-9 1019 cm-3 Small bandgap: InAs

2f qI c c ,e Rex LeWe Emitter resistance: grades removed 0.4 E c electrons -0.4 InGaAs cap layer -0.8 InP emitter E v -1.2 light doping 0 50 100

150 200 250 300 350 19 10 18 10 17 10 16 10 15 10 14 10 13 10

12 n (cm-3) E (eV) 0 10 400 Distance () 10 19 -0.4 10 18 -0.8 10 17 10 16 10 15

0 electrons E Contact resistance: 50 m2 25 m2 15 m2 -1.2 E v n (cm-3) E (eV) c heavy doping -1.6 0 50 100 150 200 250 300

350 400 Distance () At degenerate doping levels grades are not necessary High doping 3 1019 cm-3 No InGaAs-InP grade necessary at very high doping Thin undepleted n- emitter Small emitter area increases Rex Base resistance f max f 8RbbCbci Rbb Rb ,cont Rgap Rspread Rb ,cont s c 2 Le Rgap sWgap 2 Le Rspread sWe 12 Le TLM measurement Rbb is a critical parameter for fmax, and in npn HBT the base contact resistance dominates. Rbb is minimized through high base doping and improved base contact metallization, small undercut Wgap, and long emitter Le Problems with very thin bases Etching and depletion effects reduce the effective base thickness Tb, and increases the base resistance. At 500 nm scaling generation, best base thickness is 30-40 nm better fmax , lower Rbb-related delay terms in gate delay ,

minimal improvement in f between 25 & 30 nm High resistance Increase of sheet resistance with thin base layers 1 E c 0 v 2 sVbi xd qN d -0.5 -1 0 50 100 150 200 Distance () 250 10 19

10 18 10 17 10 16 10 15 10 14 10 13 10 12 Rb,extrinsic=800-1000 /sq Rb,intrinsic=600-750 /sq -3 E ~ T 20

p (cm ) E (eV) 0.5 10 300 InGaAs base doped 6 1019 cm-3, surface pinned at 0.18 eV. Surface depletion decreases base thickness 40 . Tb ,extrinsic Tb ,intrinsic Rb ,intrinsic Rb ,extrinsic Base surface exposed : ~ T Tb ,extrinsic Tb ,intrinsic 51 A Base protected by E/B grade (contacts diffused through 160 grade) ~ T Tb ,extrinsic Tb ,intrinsic 17 A Surface depletion Wet etching Collector resistance 1 nk BT C je Cbc Rex Rc Cbc b c 2f

qI c Rc: access resistance between collector contact and the mesa. Minimized by large collector contacts, and low resistance subcollector Subcollector design Goals: minimize electrical resistance minimize thermal resistance limit thickness to improve manufacturability Thermal conductivity of InGaAs ~5 W/mK Thermal conductivity of InP ~68 W/mK Tsubc Etch stop layer provides collector undercut less Cbc Some still use all InGaAs subcollector Subcollector resistivity 500 A InGaAs + 2000 A InP ~ 11 /sq 125 A InGaAs + 3000 A InP ~ 9 /sq - 53 % of thermal resistance Etching selectivity of InGaAs vs. InP main limit 50 A InGaAs Contact resistance better to 125 A than 50 A after annealing Base-emitter capacitance 1 nk BT C je Cbc Rex Rc Cbc b c 2f qI c

Cje is the junction capacitance between the emitter and base Cje corresponds to ~100 depletion thickness Minimized by shrinking the emitter area at fixed or at increasing current Ic Base-collector capacitance 1 nk BT C je Cbc Rex Rc Cbc b c 2f qI c Cbc is the junction capacitance between the base and subcollector. Abc Cbc Tc Base-collector capacitance Collector thickness reduced due to speed requirements: Abc Cbc Tc Ccb increases ! Tc = 3000 A 2150 A 1500 A Abc must be kept small: narrow emitter narrow base contacts undercut of base contacts implant or regrowth Breakdown limits thickness Thickness (A)

Breakdown (V) 2150 7.5 1500 4-5 Thick collectors : Vbr Tc Thin collectors Vbr E gap , InP / q Very thin collectors Vbr limited by tunnelling Vbr decreases at high J destruction by heating, thermal instability Theory of the base If gain is limited by Auger recombination in the base: 2k BTA R e,base N a2Tb2 The base sheet resistance: s 1 q h ,base N a Tb Tb2 The base transit time: b 2 Dn ,base

Decreasing b increases . High Na and Tb for low s decreases Grade gives 30-50 % improvement is 10-50 ps is 400-900 /sq b (calc) is 100-250 fs Base Transit Time Kroemers double integral: d Drift-Diffusion equation for base current: Tb Tb 2 i J n q n Tb Fitting of relevant parameters of the form f ( x ) e n dx q J p 0 Q n ( x) N a ( z) 2 b b dx

dzdx T b Jc N a ( x) Dz ( z ) ni2 ( z ) 0 0 x Exit term ( Az B ) With doping as N a e ( N a 1 z N a 2 ) Intrinsic carrier concentration ni2 e ( Ni1z Ni 2 ) Diffusivity D e ( D1z D2 ) Tb exit N a (Tb ) ni2 ( x) N a ( x) ni2 (Tb ) 0 Tb vexit Ballistic injection: Solution used for evaluation of the base transit time:

b,int exit e (-D1 Wb - D 2 ) N a1 e (-D1 Wb - D 2 ) D1 e (-D1 Wb - D 2 ) N i1 ) N i1 e (-D2 ) N a1 e (-D2 ) D1 e (-Ni1 Wb - D 2 - D1 Wb N a1 Wb ) ( D1 N i1 - N a1 ) D1 (N i1 - N a1 ) N a (Tb ) (e (Ni1Tb Ni2 - N a1Tb - Na2 ) - e (Ni2 - N a2 ) ) v s n i2 (Tb )(N i1 - N a1 ) Tb Base grading Graded bandgap Change in In:Ga ratio InAs: Eg=0.36 eV GaAs: Eg=1.43 eV Graded doping Doping 8 5 1019 cm-3 Base grading: induced electric field Limits: strain Limits: Bandgap narrowing, needs degenerate doping Induced electric field accelerates electrons towards collector decreases base transit time and increases gain The effect of degenerate doping Strong variation in Fermi-level with doping at high doping levels

Evidence: Observed Vbe increase Von ~ bi , increases with Ev Nb=4 1019cm30.75 V Nb=8 1019cm30.83 V for graded base-emitter Base bandgap narrowing Bandgap grade Model after V. Pavlanovski Doping grade BGN provides an electric field opposing the doping-induced field. ~1:5 in magnitude Base Transit time 1 nk BT C je Cbc Rex Rc Cbc b c 2f qI c Base transit time Base transit time (ps) 1.5 Ballistic effects may arise when Tb<180-200 @5 1019 cm-3 (Tessier, Ito) bandgap grade constant doping grade

1 0.5 0 200 300 400 Results: DC gain ft 500 600 700 Base thickness (A) 800 900 Bandgap graded 25 250 GHz 1000 Bandgap grade and doping grade give same b Doping graded 18 282 GHz

Collector design 1 nk BT C je Cbc Rex Rc Cbc b c 2f qI c Grade Transit time: c No Grade Tc 2veffective Close inspection show velocity 1 c near base most important Tc Tc Tc x dx v( x) 0 -Use grade -Use setback Base-collector grade 1.5

E E (eV) 1 c 0.5 0 E -0.5 v -1 -1.5 -2 0 50 100 Distance () 150 200 Early grade designs: Too coarse No setback layer Gain: 7 f : 128 GHz (Tc=3000 A)

Jkirk: 1.3 mA/m2 1 E (eV) 0.5 E Recent grade designs: 15 A period 200 A setback layer c 0 E -0.5 v -1 -1.5 -2 0 50 100 Distance () 150 200

Gain: 27 f : 282 GHz (Tc=2150 A) Jkirk: 4 mA/m2 InAlAs/InGaAs super lattice Why super lattice? MBE is more suited for super lattice than quaternaries. InP/InGaAs gives poor quality material due to phosphorous-arsenic intermixing MOCVD growth InGaAsP grade GaAsSb base needs no grade Quantum mechanical trapping in grade Quantum well trapping Electron/hole in the InGaAs well 500 meV InAlAs potential barrier A rough approximation: the infinite potential well. 2 2 n 2 En ....n 1,2,3.... 2 2ma If En> 500 meV (InGaAs/InAlAs potential) no electron confinement ~31 A is the maximum allowed InGaAs width by this model The delta-doping H. Kroemer : a conduction band difference can be offset with a grade and a delta-doping r Ec N T 2

q Tgrade Vbc=0.3 V No delta-doping Vbc=0.3 V Delta-doping With this choice the conduction band will be smooth The setback layer An InGaAs layer beneath the base Margin for Base dopant diffusion Increases Electron speed at SL Vbc=0.3 V No setback Vbc=0.3 V Setback Collector design: doping Circuit designer will want a fully depleted collector (low Ccb ) at some minimum specified Vcb,min (e.g. 0.0 Volts in ECL) This specifies the maximum allowable collector doping N d ,max 2 (Vcb ) qTC2 Collector design: velocity and scattering Collector band profile designed for greatest

possible distance without -L scattering No -L scattering -L scattering possible Collector under current (simulation) 0.5 Current blocking E (eV) 0 J=0mA J=1mA J=2mA J=3mA J=4mA J=5mA J=6mA J=7mA J=8mA -0.5 -1 -1.5 -2 0 100 200 Position (A)

Nc reduced by Jc/q/vsat 300 400 Metal resistance Resistance of e-beam deposited metals higher than book values. Metal resistance increases when T<1000 A Au cm 3.4 TiPdAu 200/400/9000 A PdTiPdAu 30/200/400/600 A TiPdAu 200/400/4000 A 3.2 3 2.8 2.6 2.4 2.2 0 500 1000 1500 2000 2500

Gold thickness () 3000 3500 Reduces fmax Thermal stability? Problem for base contact (PdTiPdAu with 600 A gold) sm=0.5 /sq 3-8 added to Rbb Results 2150 A collector high fmax, high Vbr,CEO IPRM 2002, Electron Device Letters, Jul. 2003; M. Dahlstrm et al, ''Ultra-Wideband DHBTs using a Graded Carbon-Doped InGaAs Base'' 1500 A collector high f, high fmax , high Jc Submitted to DRC 2003; M. Dahlstrom, Z. Griffith et al.,InGaAs/InP DHBTs with ft and fmax over 370 GHz using Graded Carbon-Doped Base High fmax DHBT Layer Structure and Band Diagram Emitter Collector InGaAs 3E19 Si 400 InP 3E19 Si 800 InP 8E17 Si 100 InP 3E17 Si 300 InGaAs graded doping 300 Setback 2E16 Si 200 Grade 2E16 Si 240 InP 3E18 Si 30 InP 2E16 Si 1700 InP 1.5E19 Si 500

InGaAs 2E19 Si 500 InP 3E19 Si 2000 SI-InP substrate Base Vbe = 0.75 V Vce = 1.3 V 300 A doping graded base Carbon doped 8*10195* 1019 cm-2 200 n-InGaAs setback 240 InAlAs-InGaAs SL grade Thin InGaAs in subcollector High f DHBT Layer Structure and Band Diagram Emitter Collector InGaAs 3E19 Si 400 InP 3E19 Si 800 InP 8E17 Si 100 InP 5E17 Si 400 InGaAs graded doping 300 Setback 3E16 Si 200 Grade 3E16 Si 240 InP 3E18 Si 30 InP 3E16 Si 1030 InP 1.5E19 Si 500 InGaAs 2E19 Si 125 InP 3E19 Si 3000 SI-InP substrate Base

Vbe = 0.75 V Vce = 1.3 V Thinner InP collector Collector doping increased to 3 1016 cm-3 Thinner InGaAs in subcollector Thicker InP subcollector Results: DC High fmax DHBT High f DHBT emitter junction area: 0.44 m x 7.4 m I step = 50 uA A =0.6 x 7 m B 5 3 J (mA/m2) 2 1.5 1 0.5 0 2 0 1

2 3 4 5 6 7 8 c C 3 V =0V 10 2.5 4 I (mA) 2 jbe 1 cb

I 8 b step = 0.4 mA 6 4 2 0 0 0.5 1 1.5 V (V) 2 CE 2.5 3 0 0 0.5 1 1.5

V (V) ce Gain: 23-28 nb/nc: 1.05/1.44 Vbr,CEO: 7 V Gain: 8-10 nb/nc: 1.04/1.55 Vbr,CEO:4 V No evidence of current blocking or trapping 2 2.5 Results: RF High fmax DHBT High f DHBT 30 30 MAG/MSG f =282 GHz 25 f 20

H =400 GHz max Gains (dB) Gain (dB) U 21 15 A = 0.54 x 7.6 um 10 jbe I = 15 mA 5 2 20 10 5 2 ce 10 f =375 GHz

max 21 15 J = 3.6 mA/um , V = 1.7 V 10 H t A jbe = 0.6 x 7 um 2 I = 30 mA c 0 f = 370 GHz U 25 11 10 Frequency (Hz)

10 12 Highest fmax for mesa HBT c 2 J = 7.2 mA/um , V = 1.3V ce 0 10 10 11 10 Frequency (Hz) 10 Highest f for mesa DHBT Highest (f, fmax) for any HBT High current density 12 Results: Base width dependence Emitter junction 0.6 x 7 m, Vce=1.3 V Tb=300 A. Tc=1500 A f 330

320 320 J =7.2 mA/m 310 (GHz) J =7.2 mA/m 310 2 e max 300 f t f (GHz) f max e 300 290 290 2

J =5.9 mA/m 280 280 J =5.9 mA/m e 2 e 270 2 270 260 260 W =0.3 m b W =0.5 m b W =1.0 m b W =0.3 m b 1 k BT C je Cbc Rex Rc Cbc

b c 2f qI c f max f 8RbbCbci W =0.5 m b W =1.0 m b Results: RF - trends 300 300 250 V =1.7 V 1.25 V f 260 f t 150 max 0.9 V

f 240 t f (GHz) 200 max 280 ce 1.0 V f 100 220 50 V =0.75 V ce sdsd 200 0 0 1 2

3 4 5 6 2 J (mA/um ) e Variation of f vs. Ic and Vce , of an HBT with a 0.54 m x 7.7 m emitter, and a 2.7 m width basecollector junction. Need higher Vce for high current 1 1.2 1.4 1.6 1.8 V 2 2.2 2.4 2.6

CE Variation of f and fmax vs. Vce , of an HBT with a 0.54 m x 7.7 m emitter, and a 2.7 m width basecollector junction. Ic=20 mA. f drops at high Vce high Vce for full collector depletion Results: evolution f 400 450 400 350 Final grade max 250 150 100 100 opt 2 (mA/um ) DHBT3

8 10 7 10 5 6 10 5 5 10 5 4 10 5 3 10 5 2 10 5 1 10 5 250 200 New grade 150 5

300 f 200 J (GHz) 350 Old grade t f (GHz) 300 fmax DHBT6 DHBT9 DHBT18 DHBT3 DHBT6 DHBT9 DHBT18 Jopt Strong improvement in f and Jopt

f and fmax > 200 GHz at Jc >10 mA/m2 0 DHBT3 DHBT6 DHBT9 DHBT18 DHBT 19B Tc =1500 A Capacitance vs. current DHBT 17 Abrupt emitter base junction DHBT 20 Graded emitter base junction 18.5 16 Jmax~3 mA/m2 15 V =1.5 V 14 ce C (fF) 13

V =1.3 V 17.5 ce cb cb C (fF) Jmax~6.5 mA/m2 18 V =1.7 V 12 17 ce 11 V =1.5 V ce 16.5 V =1.7 V ce 16 10

0 1 2 3 4 5 6 2 3 4 5 6 J (mA/um ) e e Emitter junction 0.54x7.6 um and 0.34x7.6 um. Tc= 2150 A, Nc=2 1016 cm-3 J max J kirk 1 2

2 J (mA/um ) 1 2 Tc 0 48 % Emitter junction 0.5x7.6 um Tc= 1500 A, Nc=3 1016 cm-3 Jmax~3.2 mA/m2 for Tc=2150 A 7 8 Area dependence on capacitance reduction We E B B C We ratio Wbc Wbc 30

V =1.5-1.7 V ce DHBT 18 Vce=1.5 V DHBT 19 Vce=1.3 V 20 15 10 C CB reduction (%) 25 DHBT 19 5 DHBT 17 DHBT 20 DHBT 20 DHBT 20 DHBT 17 DHBT 19 DHBT 20 DHBT 20 0

0.1 0.15 0.2 0.25 0.3 Emitter to basemesa ratio Ccb from Y-parameters Ccb is reduced at 5 GHz Extrapolating with linear fit gives 55 % for r=1 0.35 0.4 where the current flows reduce extrinsic base Max current density vs. emitter size 300 The current at which Ccb increases (Jmax) as a function of emitter width for two different HBT e W =0.5 m f (GHz) e t 8

260 W =0.7 m e 240 V =1.5 V 7 ce 220 T =1500 A c 2 J C (mA/um ) W =0.6 m 280 6 200 2 2.5 3 3.5 4

4.5 5 5.5 2 5 J (mA/um ) e V =1.7 V 4 ce T =2150 A c 3 0.2 0.4 0.6 0.8 1 1.2 1.4

Emitter width (m) Narrow emitters have higher critical current density Not necessarily higher ft (due to Rex) - Current spreading 6 Calculation of current spreading Poissons equation with depth dependant current J(x) dE 1 J ( x) qN c dx r vc J ( x) J eWe Le 2L x We d Le Tc Solving double integral provides Kirk threshold correction term 2L We 2 We ln 1 d 2 Ld We Lateral diffusion Ld De,c c One-dimension Le We

Kirk condition at Jkirk E ( x 0) 0 2L 2 Ld ln 1 d We 2 Ld 1 J now has emitter width dependence 2 J kirk vc qN c 2r bi Vbc V Tc Summary of delay terms Tau_ec RexCcb RexClay tau_f kT/qI times Cje kT/qI times Ccb

kT/qI times Clayout 503.65 fs 31.4 11.27 385.62 48 26 9 fs fs fs fs fs fs 6.1 2.2 75.3 9.4 5.1 1.8 SUM 512 fs ft_corr ft_meas 311 GHz Rex-related 316 GHz % % % %

% % 100.0 % 8.3 % Emitter heat sinking Emitter interconnect metal 2 m to 7 m Process improvements: local alignment Machine alignment provides <0.2 m alignment in good weeks Process improvements: lift-off Improved hardening of top resist surface 0.4 x 8 m emitters, ~1 m thick What to do in the future: short term Have new material with InAs rich emitter cap less Rex increased f Doping grade and combined grade less b increased f ? Small scale circuits by Z. Griffith and others Write paper on Kirk effect / collector current spreading Hlls me slttern What to do in the future: long term Need a more SiGe like processing technology Lift-off Isolation Emitter regrowth Work on HBT design Emitter design Base grade

See circuits come out Summary of work Linear base doping grade New base-collector grade Pd based base ohmics Narrow base mesa HBT Record fmax Record f InAs HEMTs Conclusion Mesa HBT can achieve superior performance to T.S. InAlAs/InGaAs S.L. grade permits use of InGaAs for base and InP for collector Excellent transport characteristics in collector InGaAs setback layer improves b-c grade PdTiPdAu base ohmics can achieve ptype contact resistance as good as n-type in case of questions Results: base-collector capacitance Full depletion 70 60 cb

C (fF) 50 0.75 V 40 0.9 V 30 1.0 V 1.1 V 20 1.25 V 10 V =1.7 V ce 0 0 1 2 3 4 5 2

J (mA/um ) e Variation of Ccb vs. Ic and Vce. Note that Vbe=0.85-0.90 volts over the same bias range. Hole mobility extraction With measured base sheet resistance and doping level the 1 base hole mobility can be estimated s q h N bTb 90 C Be 80 70 60 h 2 u (cm / Vs) 2 80 Base holemobility (cm / Vs) 90

50 40 70 60 50 40 30 30 SHBT DHBT9 MHBT1 DHBT18 DHBT219 DHBT220 2 10 19 19 6 10 20 1 10 -3 Base doping cm 20 1.4 10

Collector velocity from Kirk threshold 4 V ) 2 (V J kirk qvsat r applied 2 bi qTc E 2Tc 2Tset Tgrade qvsat r 2 2c Tgrade q Tc 3.5 J mA/um 2 3 2.5 4.2 10 7 cm/s 2 1.5 1 1 T qvsat 2T N N c Tc Tset Tgrade 2 2 T

c 3.0 107 cm/s 0.5 200 nm collector, A =0.62 m x 7.7 m qvsat N c e 0 0 0.5 V 1 ce (V) 1.5 2 Slope corresponds to collector saturation velocity Collector velocity from bc 600

500 c t (fs) 400 300 200 vsat 3.05 105 m/s 100 0 0 1000 T (A) 2000 3000 c T c c vsat Tc 1 1 1 1 2vsat 2 c 2 d c 2 slope dTc vsat 4.46 105 m/s 3.05 105 m/s InP-InGaAs and InP-GaAsSb

Base-collector grades necessary Grades not necessary H21 at 5 GHz vs. current E0.7 B05 Emitter junction 0.5x7.6 um 27 26 25 V =1.25 1.5 1.75 V Gain 24 ce 23 22 21 20 19 0 1 2 3 4 2 J (mA/um ) 5

6 e Gain does not depend on Vce , but on bias. Max gain around 26.5 7 Current RF gain vs. voltage Heating likely cause Results: Gummel DHBT 20: Capacitance cancellation data Not max ft,fmax (current too low for that, but wanted to avoid blowing)cc 250 f t 150 100 f max t f and f max (GHz)

200 50 I=20 mA 0 0.5 1 1.5 2 V ce 2.5 3 (V) Theory: G-L scattering reduces collector transit time and heating Capacitance cancellation Previous slide -12 1 10 -12 8 10 -13 6 10

-13 4 10 -13 2 10 -13 t ec 1.2 10 c ec VCB VCE 1 ec 2f 4/I fF/A I=20 mA 0 0.5 1 1.5 V C cb,e

d c rAE Ic Tc dVCB 2 ce 2.5 3 (V) 4 fF reduction from ft vs. Vce relation, very close to measured Results: RF validity 25 f =370 GHz t f Gains (dB) 20 >370 GHz max U 15 H

10 21 5 0 10 100 Frequency (Hz) 1000 W-band measurements one week apart Re-measurements show similar ft and fmax. Roll-off is very close to -20 dB/decade in the 75-110 GHz band. Resistance vs. doping InGaAs and InP n-type doping : 1-3 1019 cm-3 InGaAs p-type doping 1.2 1020 cm-3: no p-InP with C doping Mesa HBT mask set: first iteration Emitters 0.4, 0.5, 0.6, 0.7, 1.0, 2.0 m wide, 8 m long for RF measurements Base extends 0.25, 0.5 and 1.0 m on each side of base Base plug in revision 1 Emitter ground metal 2 m wide Mesa HBT mask set: second iteration Emitters 0.4, 0.5, 0.6, 0.7, 1.0, 2.0 m wide, 8 m long for RF measurements Base extends 0.35, 0.5 and

1.0 m on each side of base Base plug now on smaller tennis-racquet handle Emitter ground metal extended to 7 m width RF measurements: CPW structure 230 m 230 m RF measurements: air bridges 120m 117 m New m: /4=137 um 120m RF measurements: calibration LRL calibration using on wafer Open, Zero-length through line, and delay line OLTS used to check U in DC-50 GHz band Probe pads separated by 460 m to reduce p-p coupling RF environment not ideal, need: thinning, air bridges, vias for parasitic mode suppression RF parameter extraction Emitter resistance 10 R =3.4 ex

n=1.45 21 1/Re(Y ) 8 6 4 Rex 2 Rbb 1 (Y21 ) 0 0 50 -1 -1 I (A ) 100 150 (Error page 101 eq. 5.4) Base collector capacitance

Base collector resistance Y12 1 2Ccbi Rbb jCcb Rcb Base collector delay time, ideality factor and capacitance 1 nk T b c B C je Cbc Rex Rc Cbc 2f qI c How do we get speed improvement Switching speed limited by output capacitance C.V I Design Specifications set V and RL sets I Formula simplistic insight Reduce C by decreasing AC Increase in J since I fixed J limited by Kirk Effect Increase in J increase dissipated power density Can we measure Rth (Method of Lui et al ) 0.010 Ramp IB for different VCE Measure VBE and IC I_Probe IC

VBE V_DC SRC1 Vdc=VCE IC.i 0.008 0.006 0.004 I_DC L8E7B21 SRC2X1 0.002 Idc=IB 0.000 0.48 Depends on current density 0.54 VBE 5000 4000 0.52 Large uncertainty in values. Fitting regression curves helps to reduce error 3000

RT VBE RT VCE I C 0.50 2000 1000 2 3 4 PAve 5 6 -3 x 10 Validation of Model 40 Caused by Low K of InGaAs Max T in Collector Temperature Rise (K)

35 center Edge 30 25 20 15 10 5 SC ES C B E E Metal 0 -0.2 0 0.2 0.4 0.6 0.8

1 Distance from substrate (m) Ave Tj (Base-Emitter) =26.20C Measured Tj=26C Good agreement. 1.2 Advice Limit InGaAs Increase size of emitter arm Ultra High Speed InP Heterojunction Bipolar Transistors Why this title? Some recent conference results show transistor f of 130 GHz InP is a brittle semiconductor with a metallic luster. We mix it with GaAs and AlAs. Use Si and C as dopants Heterojunction: contains junctions of different materials DHBT carrier profile 1 10 19 10 18 10 17 10

16 10 15 0.5 0 -0.5 -1 holes electrons -1.5 -2 -2.5 Emitter -3 0 50 Base 100 150 200 Collector Subcollector 250 300

Distance () quick comment that this is unbiased....under bias both DR will fill with E 350 400 -3 20 N (cm ) 10

Recently Viewed Presentations

  • The Value of Coins Dr. Beth McCulloch Vinson

    The Value of Coins Dr. Beth McCulloch Vinson

    When you get to the pennies say "twenty-one", etc. = Quarter One dime and 3 nickels equal a quarter. = Quarter Count the dime and nickels. 15 20 10 25 = 25 cents Quarter Two dimes and 1 nickel equal...
  • Vet Equipment ID Practice - montgomery.k12.ky.us

    Vet Equipment ID Practice - montgomery.k12.ky.us

    Clipper comb. Dehorner - Barnes. Dehorner - electric. Dehorner - scoop or tube. ... Scissors- Lister bandage. Scissors- Littauer suture removal. Scissors- Mayo dissecting. ... Vet Equipment ID Practice Last modified by: tdye Company:
  • Thermodynamics - FHI

    Thermodynamics - FHI

    Langmuir Hinshelwood Mechanism Both species are adsorbed, adsorption follows Langmuir isotherm (see class next week) Eley Rideal Mechanism How to Derive a Rate Equation I How to Derive a Rate Equation II How to Derive a Rate Equation III Structure...
  • ALOPECIA DATABASE - School of Engineering

    ALOPECIA DATABASE - School of Engineering

    ALOPECIA DATABASE Peter Claise & Barbara Visher Vanderbilt University Department of Biomedical Engineering Advisors: Dr. Lloyd King Dr. Jennifer Dempsey
  • Adlerian Therapy - Individual Psychology

    Adlerian Therapy - Individual Psychology

    Alfred Adler's Individual Psychology A phenomenological approach Social interest is stressed Birth order and sibling relationships are important Therapy as teaching, informing, and encouraging Basic mistakes in the client's private logic The therapeutic relationship—a collaborative partnership The Phenomenological Approach Adlerians...
  • Ladder Safety - Environmental Compliance Alert

    Ladder Safety - Environmental Compliance Alert

    When the 90-day storage limit for a hazardous waste container expires, that waste must be sent to a permitted TSD facility for treatment or disposal. Only approved hazardous waste transporters can carry our wastes. Before they can take our waste,...
  • Slides About Systems

    Slides About Systems

    Called smart contracts or chaincodes or dapps. Executes a sequence of ops as a deterministic RSM. add. jmp. mov. shl. Log. State. Machine. Revert. Apply. On a public blockchain, there is no commit: an op can always revert (e.g., to...
  • Bay Al Salam and Bay Al Istina Are Two Contracts Where the ...

    Bay Al Salam and Bay Al Istina Are Two Contracts Where the ...

    There are four main differences Bay al-istina consist of an agreement made in advance to pay a deffinate price for something that is to be made delivered at future date unlike the contract of Bay Salam which has been validated...