Ultra-High Resolution Nano-Electrical Measurements for Semiconductors (Part 1) Peter De Wolf Dec. 3, 2014 AFM-based Nano-Electrical Modes 12/03/2014 Conductivity / Resistivity C-AFM, TUNA, Peakforce-TUNA, SSRM Electric Field EFM Charge EFM, SCM Surface Potential / Workfunction KPFM, Peakforce-KPFM Carrier Density SCM, SSRM Piezoelectric Properties PFM Specialty Modes Scanning Gate, Pyro-electric AFM, Photoconductive-AFM, 4-Point Probe, Non-linear Dielectric Microscopy… Bruker Confidential 2 AFM-based Nano-Electrical Modes Some examples Height SSRM on cross-sectioned MOS transistors 1.2x0.6 µm scan, log scale Current Tunneling-AFM (TUNA) on carbon nanotubes 4x2 µm scan, 4pA scale 12/03/2014 Peakforce-TUNA on PP/EPDM polymer blend 3x3 µm scan Bruker Confidential 3 AFM-based Nano-Electrical Modes Some examples KPFM, 2x1 µm scan, potential map on InP nanowire with 3V electrical bias between contacts n+ p n-- Topography dC/dV Phase dc/dV Amplitude SCM, 4x2 µm scan, carrier diffusion of cross-sectioned double-diffused SiC MOSFET 12/03/2014 Bruker Confidential 4 Practical Aspects: The Electrical AFM Probe • The electrical AFM tip should be: Conductive, Hard, Wear-resistant & Sharp. Good materials are Diamond, Pt, PtIr, Pt-Silicide PtSi covered Solid Metal (Pt) CVD Coated Diamond (doped) Solid Diamond (doped) 12/03/2014 Bruker Confidential 5 Practical Aspects: Environment • Many samples alter properties when exposed to Oxygen or Water • At ambient conditions, a thin water layer on the sample can degrade the spatial resolution & influence electrical properties • When applying voltages on some samples (metals, semiconductors,..), local anodic oxidation can take place and strongly influence the measurements SSRM on ITO layer, 10x10 µm scan 12/03/2014 Bruker Confidential 6 Practical Aspects: Environment - Glovebox Dimension Icon AFM inside Glovebox O2 ~ 0.1ppm H2O ~ 0.1ppm 12/03/2014 Bruker Confidential 7 Practical Aspects: Environment - Glovebox • Under ambient conditions, the conductivity map cannot observe the oxide layer. • Under a controlled environment of < 1 ppm O2 & H2O the layer can clearly be observed SSRM on thin oxide layer 12/03/2014 Bruker Confidential 8 Nano-Electrical Characterization of Semiconductor Devices • Lateral dimension of transistors continue to shrink • New architectures (SOI, FinFET, TFETS,..) and materials (SiGe, Ge, III-V,…) are introduced • ‘Classic’ 1D dopant profiling methods (SRP, SIMS,…) reach their limits ITRS 2016 Spatial resolution 1 nm Concentration precision 2% Dynamic range 1014 – 10 21 at/cm3 AFM-based 2D & 3D techniques with nm-resolution are needed 12/03/2014 Bruker Confidential 9 www.bruker.com © Copyright Bruker Corporation. All rights reserved. Ultra-high resolution nanoelectrical measurements for semiconductor applications. (Part 2) P. Eyben, K. Paredis, A. Schulze, A. Nazir, U. Celano, R. Chintala, T. Hantschel and W. Vandervorst IMEC AT A GLANCE Organic solar cell line 200mm pilot line NERF lab Silicon solar cell line Nano biolabs 300mm pilot line 450mm ready Established by state government of Flanders in 1984 with initial staff: ~70 Imec’s staff has grown to 2,086 people in 2013. Of these, 383 are residents - visiting researchers from partner companies & institutes and 289 are PhD researchers. 2 © IMEC 2014 / CONFIDENTIAL OUR MISSION World-leading research in nanoelectronics Scientific knowledge with innovative power of global partnerships in ICT, healthcare and energy Industry-relevant technology solutions International top talent in a unique high-tech environment committed to provide building blocks for a better life in a sustainable society 3 © IMEC 2014 / CONFIDENTIAL DRIVEN RESEARCH ELECTRONICS FOR HEALTHCARE & LIFE STYLE WIRELESS COMMUNICATION IMAGE SENSORS & VISION SYSTEMS LITHOGRAPHY DEVICES SENSOR SYSTEMS ENERGY HETEROGENEOUS INTEGRATION CORE CMOS DRIVEN RESEARCH TECHNOLOGY APPLICATION RESEARCH PROGRAMS INTERCONNECTS FLEXIBLE ELECTRONICS MEMS SENSOR PHOTONICS 4 © IMEC 2014 / CONFIDENTIAL MCA DEPARTMENT AT A GLANCE MCA-CA MCA-CSA TOFSIMS XPS MCA-NP Nanoprober, Raman, Conductive diamond SIMS RBS/ERD APT CAMS Materials & Components Analysis department MCA: ~50 people Researchers: ~15 R&D support: ~20 PhD students: ~10 MSc students: ~5 MCA-SA TEM FIB MCA-SPM SSRM KPFM C-AFM STM 5 © IMEC 2014 / CONFIDENTIAL RESEARCH IN THE SPM TEAM TCAD CALIBRATION (SSRM) DEVELOPMENT OF NEW AFM MODES (FFT-SSRM,...) TOMOGRAPHY (SSRM, C-AFM,...) ANALYSIS OF CBRAM (C-AFM) DEFECTS IN III-V SC (NC-AFM, STM, KPFM) ELECTRICAL PROPERTIES OF ORGANIC SC (C-AFM, KPFM) 6 © IMEC 2014 / CONFIDENTIAL ୫ ୣୟୱ ୱ୮୰ୣୟୢ୧୬ ୬ ୮୰୭ୠୣ+ ୱ୮୰ୣୟୢ୧୬ ୮ ୱୟ୫ ୮୪ୣ ୠୡ resistance (a.u.) SSRM CONCEPT TUNA 10-9 SSRM 10-6 force (N) 10-3 7 © IMEC 2014 / CONFIDENTIAL DEDICATED DIAMOND TIPS FOR SSRM Tips made of (boron) doped diamond in order to withstand the large pressures (and shear stresses) involved during SSRM Two tip technologies available : - Coated diamond tips (CDT) Full (molded) diamond tips (FDT) Nano-protrusions at the tip apex : Reduced contact size Possible multiple nano-contacts Ref: T. Hantschel et al., physica status solidi (a) 206 (2009) 2077. FDT present a lower aspect ratio vs. CDT no tip breakage FDT are sharper and present smaller diamond grains at their outer surface vs. CDT leading to better resolution 8 © IMEC 2014 / CONFIDENTIAL SSRM NANOCONTACT 4 5 6 7 8 Ref. : K. Mylvaganam et al. @ Nanotechnology 20 (30) p. 305705 (2009) 9 PIERRE EYBEN © IMEC 2014 / CONFIDENTIAL LIMITATIONS OF SSRM IN AIR • Commercial Coated Diamond Tips (CDT) can not be used for high resolution SSRM 1st scan SSRM in air 8st scan SSRM in air • Percentage of working homemade Full Diamond Tips (FDT) that can be used for high resolution SSRM is too low • Signal to noise ratio is too low • Repeatability and reproducibility are too low (Sample and tip damaged relatively quickly) SSRM ITRS 2016 Lateral/vertical steepness of dopant profile (nm/decade) 1-2 1.6 Lateral/depth spatial resolution for 2D dopant profile (nm) 1-3 1 Dopant profile concentration precision across concentration range (%) 5-20% 2% 1015 - 1020 1014 - 1021 3 Dynamic range (at/cm ) 10 © IMEC 2014 / CONFIDENTIAL IMPACT OF HUMIDITY In the presence of water, the growth of b-Si (6 coord.) is retarded In the presence of water, the volume of transformed silicon is higher Need to push more Lower resolution Ref. : C.Y.Tang et al. @ Nanotechnology 16, p.15 (2005) 11 © IMEC 2014 / CONFIDENTIAL IMPROVED SSRM PERFORMANCES SSRM SSRM HV-SSRM ICON-GB ITRS 2016 ITRS2016 Lateral/vertical steepness of dopant profile (nm/decade) 1-2 1 - 1.5 1.6 Lateral/depth spatial resolution for 2D dopant profile (nm) 1-3 0.6 - 1 1 Dopant profile concentration precision across concentration range (%) 5-20% 3-5% 2% Dynamic range (at/cm 3) 10 15 - 10 21 10 15 - 10 21 10 14 - 10 21 Quantified SSRM profile matches with SIMS Improved resolution (0.6-1nm) Improved signal to noise level ratio (noise <10%) 0.3-0.5 nm oxide Dopant/carrier concentration (cm-3) 1.E+22 SIMS P SIMS B SSRM 1.E+21 1.E+20 1.E+19 1.E+18 1.E+17 1.E+16 0 50 100 150 200 Depth (nm) 12 © IMEC 2014 / CONFIDENTIAL SAMPLE PREPARATION FOR SSRM Most measurements performed on cross-sections (XS) Protective SiO2 and Si/glass dummy used to protect the XS edge Back-contact (BC) employed to collect the spreading current : Performed manually using GaIn eutectic covered with Ag-paint On shorter structures FIB used to place local BC down to 1mm from the XS edge (= distance to avoid Ga contamination on the XS) 13 © IMEC 2014 / CONFIDENTIAL ADVANCED SAMPLE PREPARATION Back-contact pad Back-contact Back-contact angle of incidence FIB beam (1 º) FIB trench + BC reference metal line > 1 m FIB trench + BC below 1 m Oxide Metal coating FIB-lapping Oxide source Cross-section drain Cross-section Cross-section : Polishing using diamond and Al2O3 foils (RMS roughness < 0.6nm). Micro-cleavage SELA MC600. (Position XS determine with 300nm accuracy, less defects and extrinsic SS, RMS roughness < 0.4nm). Ga FIB lapping @ low incidence angle (1deg) and energy (3-5keV). (<10nm accuracy) FIB can also be used to mark area(s) of interest and/or to generate local BC pads. 14 © IMEC 2014 / CONFIDENTIAL APPLICATION : PROCESS CALIBRATION (DRAM) Predict the device performance and compare them to measured electrical data Device simulations 0.30 0.25 |Vth_sat @ Vd= -1V| (V) SSRM : measurement and quantification 0.20 experimental default_simu calibrated 0.15 0.10 0.05 0.00 -0.05 0.1 1 Lg (m) 10 350 300 250 DIBL (mV/V) Processing : implantation (dose, energy,...), annealing (diffusion, activation,...) experimental default_simu calibrated 200 150 100 50 0 0.1 Lg (m) 1 10 15 © IMEC 2014 / CONFIDENTIAL APPLICATION : CMOS IMAGE SENSOR Individual Pixels of Image Sensors photocathode SEM image courtesy of chipworks SSRM provides high resolution 2D carrier distributions inside the individual pixels of a CMOS image sensor. SSRM image with courtesy of www.imec.be/cams 16 © IMEC 2014 / CONFIDENTIAL APPLICATION : DRAM TRENCHES SEM image courtesy of chipworks Implant out-diffusion at the top and bottom of the DRAM trenches. SSRM image with courtesy of www.imec.be/cams 17 © IMEC 2014 / CONFIDENTIAL APPLICATION : POWER MOSFET metal n+ n+ n-poly SEM image courtesy of chipworks Dopant concentrations inside an integrated Schottky diode in a power MOSFET device. SSRM image with courtesy of www.imec.be/cams 18 © IMEC 2014 / CONFIDENTIAL APPLICATION : InP/InGaAs FINFET InP(Mg)/InGaAs(Si) in V-STI trenches A large in-diffusion of P into the Si substrate is measured. 19 © IMEC 2014 / CONFIDENTIAL APPLICATION : Ge FINFET SiGe / strained Ge in STI trenches P-doping level in SiGe and P diffusion into strained Ge Higher P concentration noticed in Ge 20 © IMEC 2014 / CONFIDENTIAL APPLICATION : Vertical hetero TFET 21 © IMEC 2014 / CONFIDENTIAL Tomography : How to realize 3D-SSRM ? Use of dedicated staggered test-structures (implemented on FinFETs) Use of slice and view concept (implemented on CNT interconnects) 100nm 0nm COMPARATIVE STUDY Staggering Digging 2-3nm 3-5nm Dedicated test-structure needed YES NO Versatility (different materials present,...) YES More limited Resolution in 3rd direction 22 © IMEC 2014 / CONFIDENTIAL 3D-SSRM ON NW-TFET (1) (2) (3) (4) (5) (6) (7) 0 0 100 100 200 200 300 300 400 400 500 500 600 100nm 23 100nm 100nm 100nm 100nm 600 100nm 100nm 23 © IMEC 2014 / CONFIDENTIAL 3D-SSM ON CNT-BASED INTERCONNECTS : Tomogram of individual CNTs in contact hole Ref: A. Schulze et al., Nanotechnology 23 (2012) 305707 24 © IMEC 2014 / CONFIDENTIAL 3D-SSRM ON FINFET Possible short between gate and plug due to poor aligment G S D Possible leaky junction of the drain 25 © IMEC 2014 / CONFIDENTIAL 3D C-AFM ON CBRAM Filament evolution from top to bottom electrode [Schindler ,et al.. APL 2009, 94, 072109] 20 nm Conductive (filament) Bridging Device (Rem. Rate 0.7 nm/scan) Anode Cathode (~ 2nm) U. Celano et al., Nanoletters, 2014, 14, 5, 2401 26 © IMEC 2014 / CONFIDENTIAL C-AFM ON ORGANIC SOLAR CELLS Operation principle P3HT:PC[60]BM P3HT:PCBM PEDOT:PSS ITO P3HT:PCBM 1:1 ratio 200 nm active layer η~4% hv Cross section view of spray coated P3HT:PCBM (1:1) BHJ solar cell PCBM P3HT Craters with varying depth were made using Ar cluster beam Depth profile using SPM techniques based on electrical properties of P3HT and PCBM 27 © IMEC 2014 / CONFIDENTIAL C-AFM-BASED DEPTH PROFILING : Vertical phase segregation &efficiency of exciton dissociation On degraded sample, no more P3HT plateau poorer efficiency Hypothesis : 100% optical absorption Equal absorption for P3HT and PCBM 10 nm exciton diffusion length ηfresh = (87 ± 3)% Red P3HT Blue PCBM ηdegraded= (65 ± 5)% 28 © IMEC 2014 / CONFIDENTIAL QUESTIONS ? Ultra-high resolution nanoelectrical measurements for semiconductor applications. P. EYBEN, J. MODY, A. NAZIR, K. PAREDIS, A. SCHULZE,T. HANTSCHEL AND W. VANDERVORST
© Copyright 2024 ExpyDoc
