SHEET RESISTANCE AND MOBILITY MEASUREMENTS WITH

SHEET RESISTANCE AND MOBILITY MEASUREMENTSWITH MICRO-PROBES:FROM PLANAR TO 3D TECHNOLOGIESJ. BOGDANOWICZ, S. FOLKERSMA, S. SERGEANT, A. SCHULZE,H.H. HENRICHSEN, D. H. PETERSEN AND W.VANDERVORST

RESISTANCE MEASUREMENTS IN FRONT END OF LINE?IonIonmetalAdapted fromA. Schulze,PhD thesis(KULeuven, 2013)metalchannel improving transistor performance2

RESISTANCE MEASUREMENTS IN FRONT END OF talRcontactAdapted fromA. Schulze,PhD thesis(KULeuven, 2013)Rdrain improving transistor performance reduce RseriesRseries Rsource Rdrain 2 Rcontact Rchannel3

RESISTANCE MEASUREMENTS IN FRONT END OF talRcontactAdapted fromA. Schulze,PhD thesis(KULeuven, 2013)RdrainRcontact exp( 1 / N ) improving transistor performance reduce Rseries Today: Rcontact dominant increase carrier concentration N4

FROM A FOUR-POINT MEASUREMENT TO CARRIER CONCENTRATIONFrom sheet resistance Rsto resistivity IVdoped layertsubstrateRs t Rs resistivity 5

FROM A FOUR-POINT MEASUREMENT TO CARRIER CONCENTRATIONFrom resistivity to carrier concentration NFrom sheet resistance Rsto resistivity IV 1 /( q N )doped layersubstrateRs tt (ohm.cm)104Si10-41012N (cm-3)1021www.ioffe.ru Rs resistivity Assumption: mobility only depends on N N can be extracted6

MICROHALL-A300 TOOL OF CAPRES:FULLY AUTOMATED IN-LINE MICRO FOUR-POINT PROBEstandard 4pp: dinj 1 mm123424 md12 8 md23 8 md34 8 mDoped layerSubstratem4pp: dinj 24 mStrain gaugeDoped layerSubstrateAdvantages of micro-probes, i.e. with micron-scale pitch d: High-accuracy sheet resistance on ultra-shallow layers Sheet resistance on patterned wafers Mobility/carrier concentration (microHall) on patterned wafers Resistance of nm-wide conductive lines7

OUTLINEstandard 4pp: dinj 1 mm24 m1234d12 8 md23 8 md34 8 mDoped layerSubstratem4pp: dinj 24 mStrain gaugeDoped layerSubstrate High-accuracy sheet resistance on ultra-shallow layers Sheet resistance on patterned wafers Mobility/carrier concentration (microHall) on patterned wafers Resistance of nm-wide conductive lines

M4PP VS STANDARD 4PPSamples from Dr. A(undoped substrate) Samples from Dr. B(doped substrate) 9

M4PP VS STANDARD 4PPRs withstandard 4pp (ohm/sq)1200 1000800600400200Rs with standard 4pp (ohm/sq)Samples from Dr. A(undoped substrate)Samples from Dr. B(doped substrate)10000 1000?100101005001000Rs with m4pp (ohm/sq)110010000Rs with m4pp(ohm/sq)10

REQUIREMENTS FOR AN ACCURATE SHEET RESISTANCE MEASUREMENTIVD.C. Worledge et al,Appl. Phys. Lett. 83, 84 (2003)Rsource/drainWWWSubstrate of opposite typeWdepRdep dep WdepWWWRsubstrate Standard 4pp requires: opposite doping type (p vs n) in the substrate high dep lowly doped substrate wide depletion region Wdep11WWWdepletion regionWWWdoped layer

REQUIREMENTS FOR AN ACCURATE SHEET RESISTANCE MEASUREMENTIVdinjD.C. Worledge et al,Appl. Phys. Lett. 83, 84 (2003)Rsource/drainWWWSubstrate of opposite typeWdepWWWdepletion regionWWWdoped layerRdep dep Wdep/d2injWWWRsubstrate Standard 4pp requires: opposite doping type (p vs n) in the substrate high dep lowly doped substrate wide depletion region Wdep M4pp more relaxed on substrate doping12

M4PP VS STANDARD 4PPRs withstandard 4pp (ohm/sq)1200 1000800600400200R4ppdep 10 kOhmRm4ppdep 1 MOhm005001000Rs with m4pp (ohm/sq)Rs with standard 4pp (ohm/sq)Samples from Dr. A(undoped substrate)Samples from Dr. B(doped substrate)10000 1000standard4pp measuressubstrate10010R4ppdep 100 OhmRm4ppdep 10 kOhm1110010000Rs with m4pp(ohm/sq)M4pp is accurate also on doped substrates13

OUTLINEstandard 4pp: dinj 1 mm24 m1234d12 8 md23 8 md34 8 mDoped layerSubstratem4pp: dinj 24 mStrain gaugeDoped layerSubstrate High-accuracy sheet resistance on ultra-shallow layers Sheet resistance on patterned wafers Mobility/carrier concentration (microHall) on patterned wafers Resistance of nm-wide conductive lines

SAMPLE FROM DR. C: PATTERNED WAFER WITH SMALL PADoxide Standard 4pp: dinj 1 mmoxideDopedlayerp -GeSn80 mDopedlayer80 mTop viewoxidesubstrate 4pp does not fit into small pad limited sampling volume ( 24x20 m2) only small pads needed15

SAMPLE FROM DR. C: PATTERNED WAFER WITH SMALL PADm4pp: dinj 24 mDoped layeroxide80 moxide24µmD. H. Petersenet al.,JVST B 26, pp362-367(2008)substrateStandard 4pp: dinj 1 mmoxideDopedlayerp -GeSn80 mTop viewoxidesubstrate 4pp does not fit into small pad but m4pp does limited sampling volume ( 24x20 m2) only small pads needed16

SAMPLE FROM DR. C: PATTERNED WAFER WITH SMALL PADm4pp: dinj 24 mDoped layeroxide80 moxide24µmD. H. Petersenet al.,JVST B 26, pp362-367(2008)substrateStandard 4pp: dinj 1 mmoxideDopedlayerp -GeSnsubstrateoxide80 mB-doped Geepi layer 4pp does not fit into small pad but m4pp does limited sampling volume ( 24x20 m2) only small pads needed17

OUTLINEstandard 4pp: dinj 1 mm24 m1234d12 8 md23 8 md34 8 mDoped layerSubstratem4pp: dinj 24 mStrain gaugeDoped layerSubstrate High-accuracy sheet resistance on ultra-shallow layers Sheet resistance on patterned wafers Mobility/carrier concentration (microHall) on patterned wafers Resistance of nm-wide conductive lines

PROBLEM WITH DR. D’S SAMPLES: SATURATION OF CARRIER CONC.?1.E 21 1.E-031.E 20saturation@ 4x10-4 ohm.cm1.E-041.E 191.E 201.E 191.E 21total [B] (cm-3) Saturation of resistivity as total [B] increases?19Carrier conc. N (cm-3)n-Ge buffer1.E-02tresisitivity (ohm.cm)B-doped Ge epi with different [B]

PROBLEM WITH DR. D’S SAMPLES: SATURATION OF CARRIER CONC.?10-4 1 /( q N )Ge4x10-4 ohm.cm1014Carrier conc. N (cm-3)1020 2x1020 1.E 21saturation@ 2x1020 cm-31.E-031.E 20saturation@ 4x10-4 ohm.cm1.E-041.E 19cm-31.E 20Carrier conc. N (cm-3) (ohm.cm)102resisitivity (ohm.cm)n-Ge bufferwww.ioffe.ru1.E-02tB-doped Ge epi with different [B]1.E 191.E 21total [B] (cm-3) Saturation of resistivity as total [B] increases Saturation of carrier concentration ? Mobility not degraded when increasing [B] from 5e20 to 1e21 cm-3 ?20

FROM FOUR-POINT MEASUREMENTS TO CARRIER CONCENTRATIONFrom resistivity to carrier concentration NFrom sheet resistance Rsto resistivity resistivity resistivitySubstrateRs t 1 /( q N )104t (ohm.cm)IVSi10-41012N (cm-3)1021www.ioffe.ru also depends on defect density, layer thickness, inactive dopants etc measure 21

CONVENTIONAL HALL MEASUREMENTS-I- - Vhall mobilityFLorentz22BI Magnetic field B Van der Pauwconfiguration cm probe spacing Leakage

CONVENTIONAL HALL MEASUREMENTS-I- - Vhall mobilityFLorentzI Magnetic field B Van der Pauwconfiguration cm probe spacing Leakage Hall voltage with4 aligned probes?23B

REQUIREMENTS FOR A MICRO-HALL MEASUREMENT Insulating edge: non-zero Hall signal (asymmetry)Electrostatic potential (V)0.0insulatingedgeD.H. Petersen et al.,J. Appl. Phys. 104,013710 (2008)insulatingedgeInsulating edge asymmetry Hall signalCurrent injection pinsB-0.2STI

REQUIREMENTS FOR A MICRO-HALL MEASUREMENT Insulating edge: non-zero Hall signal (asymmetry)Electrostatic potential (V)insulatingedgeD.H. Petersen et al.,J. Appl. Phys. 104,013710 (2008)0.0B20 m 10 mpitch pitch12 asymmetry Hall signalCurrent injection pinsinsulatingedgeInsulating edge3456-0.2 7-point probe:m4pp at two different pitches(measurement of distance d to edge)7dSTI

MICROHALL VS HALL: THICK InGaAs LAYERS10,000Hall mobility 9r1824doping1e15 (cm-3)r2220MicroHall nicely agrees with conventional Hall on thick layers26

CARRIER CONCENTRATION KEEPS INCREASING BUT MOBILITY DROPSHall mobility (cm2 /V/s)2.50E 022.00E 02microHall1.50E 021.00E 025.00E 010.00E 001.E 191.E 20SIMS [B] (cm-3)1.E 21 0.0010.00011.E 191.00E 21m4pp1.00E 201.E 20Carrier conc. N (cm-3)resisitivity (ohm.cm)n-Ge buffermicroHall0.01B-doped Ge epi with different [B]1.00E 191.E 21SIMS [B] (cm-3)Mobility must be measured for an accurate determination of carrier concentration Carrier concentration in Dr. D’s samples does NOT saturate27

MICROHALL MEASUREMENTS ON PATTERNED WAFERSSi:B (5 KV, 3E15CM-2) SPIKE ANNEAL (1035OC, 1.5 S)Rs(ohm/sq)Hall mobility(cm2/V/s)Hall sheet carrier density(cm-2)Unique capabilities of Hall measurements on patterned wafers28

OUTLINEstandard 4pp: dinj 1 mm24 m1234d12 8 md23 8 md34 8 mDoped layerSubstratem4pp: dinj 24 mStrain gaugeDoped layerSubstrate High-accuracy sheet resistance on ultra-shallow layers Sheet resistance on patterned wafers Mobility/carrier concentration (microHall) on patterned wafers Resistance of nm-wide conductive lines

SAMPLE OF PROF. V: B-IMPLANTED SI FINSB (5 kV, 3x1015 cm-2)oxidep-Sioxidep-Sioxiden-Si10 fins of 20 nm width10 fins of 500 nm width24 mQuestions of Prof.V:1. Can you measureon nm-wide lines2. is Rs dimensiondependent?.n-Sin-Si80x80 led @ 450ºC, 15’.500nm300nm 200nm 150 nm 120 nm 100 nm 80 nm 60nm40 nm 20 nm

PROBES IN CONTACT WITH THE FIN?All 4 probes on fin: Low contact resistance short RC constantIFreq. 13 HzVSubstrate31timePhaseshift 0

PROBES IN CONTACT WITH THE FIN?All 4 probes on fin: Low contact resistance short RC constantIFreq. 13 HztimeVSubstratePhaseshift 01 probe on oxide: high contact resistance long RC constantItimeV PhaseSubstrateshift ?1st sanity check : PhaseI – PhaseV 0 4 probes on the conductive fin32

MICRO FOUR-POINT PROBE MEASUREMENT1 2 3 4IN CONFINED VOLUMEID.H. Petersen et al.,d23J. Appl. Phys. 104, 013710 (2008)Configuration A:W d23V1V4Potential distributioninside the fin fin resistance between pins 2 and 3 is measured33

MICRO FOUR-POINT PROBE MEASUREMENT1 2 3 4IN CONFINED VOLUMEID.H. Petersen et al.,d23J. Appl. Phys. 104, 013710 (2008)Configuration A:L d23IWW d23V1tV4Ra /t d23/WRsPotential distributioninside the fin fin resistance between pins 2 and 3 is measured34

MICRO FOUR-POINT PROBE MEASUREMENT1 2 3 4IN CONFINED VOLUMEID.H. Petersen et al.,d23J. Appl. Phys. 104, 013710 (2008)Configuration A:Configuration B:V1IV4123Ra /t d23/WRs4Id23W d23 V34 0tWW d23V1Potential distributioninside the finL d23V3Rb Rs d23/W Ra fin resistance between pins 2 and 3 is measured Same resistance measured in a and b configurations (Ra/Rb 1)35

Ra/Rb RATIO AND DIMENSIONALITY OF THE CURRENT FLOWD.H. Petersen et al.,1.3J. Appl.Phys. 104, 013710 (2008)D.H. Petersen et al.,J. Appl. Phys. 104, 013710 (2008)d23WRa/Rb 1.26Ra/Rb1.252D 23W dI I1.2IVSiO2V SiO2fins1.15blanketV1 2 3 4Substrate1.1W d23I1.05Ra/Rb 1Blanketlayerd231Substrate1D(2D0.95current flow):W d23Rb Rs d23/WRa 1.26 R0.01 b0.111001000V V10 0V RFin width/pin distancea Ra/Rb 1.00 1D current flow (fin) Ra/Rb 1.26 2D current flow (blanket) Information about leakage can be extracted based on Ra/Rb 114134336

SAMPLE OF PROF. V: B-IMPLANTED SI FINSB (5 kV, 3x1015 cm-2)oxidep-Sioxidep-Sioxide.n-Sin-Sin-Si80x80 led @ 450ºC, 15’10 fins of 20 nm width10 fins of 500 nm widthProbe scan diection.Rs 314 ohm/sq300nm 200nm 150 nm 120 nm 100 nm 80 nm 60nm10x 500 nmlines3740 nm 20 nm

B-IMPLANTED FINS : EXPERIMENTAL(Ra Rb)/2(ohm)3000003000030000Probe scan direction50100Position ( m).38150200

B-IMPLANTED FINS : EXPERIMENTAL300000(Ra Rb)/2(ohm)20 nm40 nm60 nm100 nm30000150 nm120 nm80 nm200 nm300 nm500 nm3000050100Position ( m)150 All widths captured Measured resistance increases with decreasing width39200

B-IMPLANTED FINS : EXPERIMENT VS THEORY(Ra Rb)/2(ohm)300000Rspad 314ohm/sqd23Resistor modelR Rspad d23/W30000WWRspad 314 ohm/sq30000200400Fin width (nm) Resistor model fits the measured data resistivity - independent fromdimension (mostly geometrical confinement)40600

B-IMPLANTED FINS : EXPERIMENT VS THEORY(Ra Rb)/2(ohm)300000T 450 C, 15’T 1150 C, 7 laser scansT 1250 C, 7 laser scansRspad 314ohm/sqd23R Rspad d23/W30000WWRspad 59 ohm/sq30000Rspad 135 ohm/sqRspad 314 ohm/sq200400Fin width (nm) Resistor model fits the measured data resistivity - independent fromdimension (mostly geometrical confinement) Annealing lowers the measured fin resistance41600

Sheet resistance(ohm/sq)SHEET RESISTANCE VS FIN WIDTH: B-IMPLANTED Si FINS600400T 450 C, 15’T 1150 C, 7 laser scansT 1250 C, 7 laser scansRspad200RspadRspad00200400fin width (nm)Rs (Ra Rb)/2/d23 WAnswer to Prof.V’s question: Long low-T anneal : dimension-dependent Rs Depletion effect, degraded mobility, defects, dopant diffusion? Laser anneal Rs independent from width60042