Detection and resolution enhancement of laser induced fault localization techniques 1

DETECTION AND RESOLUTION ENHANCEMENT OF LASER INDUCED FAULT LOCALIZATION TECHNIQUES ALFRED QUAH CHENG TECK A THESIS SUBMITTED FOR THE DEGREE OF DOCTOR OF PHILOSOPHY DEPARTMENT OF ELECTRICAL AND COMPUTER ENGINEERING NATIONAL UNIVERSITY OF SINGAPORE 2010 _ ACKNOWLEDGEMENTS I would like to thank my supervisor, Prof Jacob Phang for his guidance for the last years, beginning from my undergraduate final year In spite of his busy schedule, he never fails to allocate time every week to review my progress and gives valuable academic advice It is also under his supervision that I am given the opportunities to collaborate with local and overseas industry partners I am also thankful for his great efforts for getting grants or even out of his sponsorship for me to attend overseas conference whenever a paper is accepted He has trained me to be an independent and confident researcher Another person whom I wish to express my heartfelt gratitude is the CEO of SEMICAPS, Mr Chua Choon Meng He has been a great industrial mentor to me throughout my graduate years He never fails to engage me in valuable technical discussions and to enlighten me with the needs of the industry He and his company have given me great equipment and facilities support in the course of our collaboration For these, I would like to take this opportunity to thank SEMICAPS and the staff In addition, I would like to thank Dr Lap Chan and Dr Ng Chee Mang from GLOBALFOUNDRIES, Singapore for including me into the company postgraduate special project team They have provided me with invaluable training related to semiconductor wafer fabrication and the company has given additional top-up to my research scholarship This program has also given me great opportunities to interact and learn from other postgraduate students in the team who are researching on other semiconductor fields, from process, materials, reliability to novel transistor devices i _ I would also like to thank Mrs Ho Chiow Mooi, the principal laboratory officer and the staff of the Centre for Integrated Circuit Failure Analysis (CICFAR) for having provided excellent administration and logistics support throughout my PhD candidature Finally, I would like to thank my fellow peers who are in the same research team under the supervision of Prof Phang They have been a continual source of encouragement, moral support and of course fun in the midst of my postgraduate years Part of this work has also resulted from the collaboration with this team Not forgetting my wife, Jeanette who has been my greatest source of love and one who never fails to lift me up when the going gets tough Thank you ii _ TABLE OF CONTENTS Page ACKNOWLEDGEMENTS i ABSTRACT ix LIST OF ABBREVIATIONS xi LIST OF SYMBOLS xiii LIST OF TABLES xvi LIST OF FIGURES xvii Chapter 1: Failure Analysis and Fault Localization Challenges 1.1 Failure Analysis 1.2 Failure Analysis Challenges 1.3 Active and Passive Fault Localization Techniques 1.3.1 Photon Emission Microscopy 1.3.2 Laser Induced Techniques 1.3.3 Frontside and Backside Failure Analysis 1.4 Fault Localization Challenges 1.5 Project Motivation 11 Chapter 2: Physics & Literature Review of Laser Induced Detection Systems 14 2.1 Physics of Laser Induced Phenomena 14 2.1.1 Light Transmittance in Doped Silicon Substrate 14 2.1.2 Thermal Stimulation 16 2.1.2.1 Temperature Change 17 iii _ 2.1.2.2 Effect on Metal Interconnects and Silicon 17 2.1.2.3 Effect on PN Junction 18 2.1.2.4 Effect on Field Effect Transistors 20 2.1.3 Carrier Stimulation 2.1.3.1 Carrier Stimulation on CMOS Transistor 2.2 Scanning Optical Microscope System 21 22 23 2.2.1 SEMICAPS SOM 1005 Hardware System 23 2.2.2 Fault Localization Procedures 25 2.3 Fault Localization Techniques 26 2.3.1 Power Alteration Techniques 28 2.3.1.1 OBIRCH and TIVA 28 2.3.1.2 TBIP/XIVA 30 2.3.1.3 SEI 32 2.3.1.4 Defects Localized with Power Alteration Techniques 33 2.3.2 Tester-Based Techniques 34 2.3.2.1 Implementation of RIL/SDL 35 2.3.2.2 Soft Defects 37 2.3.2.3 Development of Dynamic Laser Stimulation Techniques 38 2.4 Defect Characterization Techniques 39 2.4.1 OBIC and SCOBIC 40 2.4.2 LIVA 41 2.5 Resolution Enhancement with Solid Immersion Lens Technology 2.5.1 Refractive Solid Immersion Lens (RSIL) 2.6 Project Objectives 43 45 47 iv _ Chapter 3: AC-Coupled Laser Induced Techniques 49 3.1 Circuit Modeling of AC-Coupled Detection Systems 49 3.1.1 TIVA Circuit Model 50 3.1.2 SRM Circuit Model 51 3.1.3 TBIP Circuit Model 53 3.2 Sensitivity Comparison 56 3.2.1 SNR Calculation 56 3.2.2 SRM Sensitivity 59 3.2.3 TBIP Sensitivity 61 3.2.4 SNR Comparison of TIVA, OBIRCH & TBIP 63 3.3 Summary 67 Chapter 4: DC-Coupled Laser Induced Technique 68 4.1 AC-Coupled Detection System Limitations 68 4.2 Differential Resistance Measurement Detection System 69 4.3 AC-Coupled and DC-Coupled Detection Systems Comparison 70 4.3.1 Scanning Artifacts 75 4.3.2 Scan Speed 76 4.3.3 Point Averaging 80 4.3.4 Amplifier Bandwidth 81 4.3.5 Thermal Banding Effects in DReM Configuration 83 4.4 Case Studies 85 4.4.1 Meander Structure 85 4.4.2 Serpentine Resistive Test Structure 86 4.4.3 Diode in Microprocessor 87 v _ 4.5 DReM Sensitivity Optimization 89 4.5.1 Detection Sensitivity at Constant Arm Ratio 89 4.5.2 Detection Sensitivity at Varying Arm Ratio 92 4.6 Measurement of Laser Induced Resistance Change 93 4.7 Sensitivity Comparison of AC-Coupled and DC-Coupled Detection Systems 94 4.8 Improvements on DReM Circuit 97 4.9 Summary 98 Chapter 5: Pulsed Laser with Lock-In Detection 99 5.1 Pulsed Laser Induced Signal Response Model 99 5.1.1 DC-Coupled Pulsed Laser Induced Signal Response 100 5.1.2 AC-Coupled Pulsed Laser Induced Signal Response 105 5.1.3 Experimental Verification of Model 107 5.2 Lock-In Detection 5.2.1 Narrowband Lock-In Detection 111 115 5.2.1.1 DC-Coupled Detection Systems 115 5.2.1.2 AC-Coupled Detection Systems 117 5.2.2 Wideband Lock-In Detection 120 5.2.2.1 DC-Coupled Detection Systems 120 5.2.2.2 AC-Coupled Detection Systems 121 5.2.2.3 DC-Coupled and AC-Coupled Comparison 122 5.3 Experimental Results 5.3.1 Narrowband Lock-In Detection 123 124 5.3.1.1 Experimental Setup 125 5.3.1.2 Laser Scan Speed 125 vi _ 5.3.1.3 Lock-In Time Constant 128 5.3.1.4 Phase Difference 130 5.3.1.5 Detection Sensitivity Variation with Pulsing Frequency 131 5.3.2 Wideband Lock-In Detection 136 5.3.2.1 Experimental Setup 136 5.3.2.2 Averaging of N Pulses 138 5.3.2.3 Time Delay 139 5.3.2.4 Detection Sensitivity Variation with Pulsing Frequency 141 5.3.3 Sensitivity Comparison 145 5.3.4 Impact of Device Scaling on Lock-In Applications 148 5.4 Summary 150 Chapter 6: Application of Pulsed Laser Induced Technique for Localization of Cu/low-k Interconnect Reliability Defects 152 6.1 Reliability for Cu/Low-k Interconnects 152 6.2 Test Structure and Electrical Stress 153 6.3 Sensitivity Comparison of TIVA & Pulsed-TIVA 154 6.4 Physical Failure Analysis Results 158 6.5 Summary 162 Chapter 7: Effect of Refractive Solid Immersion Lens Parameters on the Enhancement of Laser Induced Fault Localization Techniques 164 7.1 Refractive Solid Immersion Lens Technology 164 7.2 RSIL Experimental Setup 166 vii _ 7.3 TIVA Enhancements from RSIL 168 7.4 Power Loss from RSIL and Objective Lens 170 7.5 Objective NA Matching 172 7.6 Effect of RSIL Parameters on Imaging and TIVA Enhancement 174 7.7 Combining RSIL and Pulsed Laser Induced Techniques for Effective Defect Localization on 65 nm Microprocessors 178 7.7.1 Case Study : TIVA with RSIL 179 7.7.2 Case Study 2: Pulsed-TIVA with RSIL 180 7.7.3 Case Study 3: Pulsed TIVA with RSIL 184 Chapter 8: Conclusions 187 Chapter 9: Recommendation for Future Work 193 9.1 Detection Sensitivity Enhancement with Double Lock-In Detection 193 9.2 Beyond Fault Localization to Defect Characterization 195 References 196 Appendix A 204 List of Publications 207 Award & Patents 208 viii Abstract _ Abstract Failure analysis is an integral step for the development and manufacturing of semiconductor integrated circuits Fault localization is the most crucial step in failure analysis as successful or unsuccessful fault localization determines the success or failure of the entire FA cycle The key limiting factors for the application of laser induced fault localization detection systems on advanced technology node below 65 nm are detection sensitivity and resolution The objective of this work is to develop techniques and enhancements to improve both the detection sensitivity and localization precision of existing laser beam techniques This would extend the applicability of the techniques and allow them to remain compatible for optical fault isolation for 65 nm technology node and beyond To achieve this objective, a comprehensive theoretical and experimental study was firstly carried out to evaluate the sensitivity of existing fault localization detection systems Following then, a new dc-coupled laser induced fault localization detection system called Differential Resistance Measurement was developed for accurate measurement of laser induced phenomena It enhances sensitivity by eliminating the artifacts inherent in existing ac-coupled detection systems and allows precise localization of the resistive sites without artifacts, within a single scan Next, a systematic approach was taken to assess and optimize the detection sensitivity of pulsed laser with narrowband and wideband lock-in detection on both ac-coupled and dc-coupled detection systems By developing an analytical model based on heat transport mechanism to describe the pulsed laser induced phenomena of a biased ix Chapter _ line scan, the algorithm realized the offset by calculating the average grayscale level within the non-signal pixels in the line It is then compared with the default grayscale intensity of 127 corresponding to zero DC offset condition The difference is then compensated by addition or subtraction This background leveling procedure is repeated line by line for the image Fig 4.14 shows the DReM images before and after the correction where the corrected image clearly shows better distinction between the signal and the background (a) (b) Fig 4.14 DReM signal images (a) before and (b) after the image correction These results prove that DReM is more effective in providing an overall laser induced resistance change mapping of the DUT and its sensitivity is independent of defect structure, laser scan direction, laser scan speed and averaging mode The advantages and disadvantages of ac-coupled and dc-coupled detection systems are summarized in Table 4.1 84 Chapter _ Table 4.1 Summary of the advantages and disadvantages of ac-coupled and dc-coupled detection systems 4.4 Case Studies In this section, several case studies on the applications of DReM on semiconductor devices are presented These case studies highlight the advantages of a dc-coupled detection system 4.4.1 Meander Structure The test structure is a meander test structure, with a resistance of 590 Ω Figs 4.15(a), (b) and (c) show the frontside reflected-DReM overlay image, TIVA image and DReM image, respectively The TIVA image shows significant artifacts in the highlighted areas In the left area, the “tails” from the ac-coupled amplifier response are mixed with the laser induced signal At certain parts, indicated by the arrows, the strong tail has almost masked out the signal In the right area, the signal tails are clearly visible The DReM image, on the other hand, provides a consistent contrast corresponding to the laser induced signal throughout the entire structure where the 85 Chapter _ brighter intensity represents an increase in the laser induced signals while the darker intensity indicates a decrease in the laser induced signals (a) Reflected-DReM overlay (b) TIVA image (c) DReM image Fig 4.15 (a) Frontside reflected-DReM overlay image, (b) TIVA and (c) DReM images of 590 Ω meander test structure 4.4.2 Serpentine Resistive Test Structure a (i) a (ii) TIVA (ac-coupled) a (iii) DReM (dc-coupled) b (i) b (ii) TIVA (ac-coupled) b (iii) DReM (dc-coupled) Fig 4.16 a (i) DReM reflected-overlay image, a (ii) TIVA image and a (iii) DReM image when the laser scan is perpendicular to the 1.7 kΩ resistor structure; b (i) DReM reflected-overlay image, b (ii) TIVA image and b (iii) DReM image when the laser scan is parallel to the resistor structure 86 Chapter _ TIVA and DReM are applied to a large area 1.7 kΩ resistor structure Figs 4.16a (i) (iii) show the reflected-DReM overlay and the TIVA and DReM images, respectively when the laser scan direction is perpendicular to the resistor structure Figs 4.16b (i) (iii) are the corresponding images when the laser scan direction is parallel to the resistor structure When the laser scan is perpendicular to the resistor structure, both TIVA and DReM images show clear localization of the resistive lines of the structure However, when the laser scan is in parallel to the structure, TIVA image in Fig 4.16b (ii) fails to map out the same resistive lines Only the initial portion or the edges of the lines are evident Thus, to achieve thorough defect localization using ac-coupled detection system, multiple scans at various orientations are required The DReM image in Fig 4.16b (iii), on the other hand, is able to detect the resistor structure This shows that the ac-coupled system is generally edge sensitive and is inadequate in detecting large area defects The DC-coupled system gives a consistent laser induced signal, independent of laser scan orientation relative to the structure 4.4.3 Diode in Microprocessor This case study demonstrates the use of DReM and TIVA on a diode structure in a microprocessor 87 Chapter _ a (i) Reflected-TIVA overlay a (ii) TIVA image at pt average a (iii) TIVA image at 512 pt average b (i) Reflected-DReM overlay b (ii) DReM image at pt average b (iii) DReM image at 512 pt average Fig 4.17 TIVA and DReM images of a diode structure in a microprocessor Fig 4.17a (iii) shows that 512 point averaging rate has removed most of the TIVA signal The DReM results in Figs 4.17(b) again show a much more precise indication of the thermally sensitive region on the diode without the artifacts These results show that the voltage bias, voltage detection, dc-coupled detection mode of DReM allows precise localization of the resistive sites without artifacts, within a single scan It is also less dependent on laser acquisition parameters like scan speed, point averaging, amplifier bandwidth, defect structure and scan orientation These results were presented at the International Symposium of Physical & Failure Analysis of Integrated Circuits ( IPFA 2006 ) at th July 2006 [82] The US patent for DReM was issued on 25th Nov 2008 [86] 88 Chapter _ 4.5 DReM Sensitivity Optimization In this section, optimization of DReM detection sensitivity is described The objective is to determine the value of the balancing resistors for a given resistive sample that yields the best detection sensitivity 4.5.1 Detection Sensitivity at Constant Arm Ratio DReM bridge setup is similar to SRM setup implemented with a series resistor described in Fig 3.2 except that DReM has an additional balancing resistor arm for establishing the reference quiescent DUT voltage for common mode subtraction Substituting eqn (3.1a) and eqn (3.5b) into eqn (4.3) gives eqn (4.11a), which describes DReM voltage sensitivity as follows: v d  d F  Vd 1 F v d Vd  d (4.11a) , (4.11b) F  Eqn (4.11a) shows that DReM voltage sensitivity is the same as SRM voltage sensitivity described by eqn (3.9a) It is proportional to  d and dependent on F It also shows that DReM voltage sensitivity at a specific value of arm ratio is independent on the actual resistance value of the balancing arm as long as the ratio is attained Eqn (4.11b) shows that optimized DReM voltage sensitivity is achieved at a large arm ratio and it is comparable to the optimum voltage sensitivities of TIVA, SRM and TBIP 89 Chapter _ A series of experiments is done using the same Al line structure in Fig 3.4 with the DReM setup at F  Resistor R3 is fixed at 15.4 Ω to match Rd Varying R1 and R2 are used to achieve F  in the balancing arm and Fig 4.18 shows the DReM images of the line structure with varying balancing resistances The images seem to show similar signal contrast independent of resistance R1 = R2 = 3.6 Ω R1 = R2 = 15.3 Ω R1 = R2 = 150 Ω R1 = R2 = 1.48 kΩ R1 = R2 = 15 kΩ R1 = R2 = 150 kΩ Fig 4.18 DReM signal images with different balancing resistance, R1 and R2 , maintaining F  The SNR and sensitivity parameters of these images, plotted against the resistance ratio R1 / Rd , are then tabulated in Figs 4.19(a) and (b), respectively The results in Fig 4.19(b) validate the DReM circuit model showing that DReM voltage sensitivity remains constant, independent of the resistance value for constant F However, due to the increase in noise with a higher resistance, DReM SNR starts to decrease sharply when R1 / Rd  100 as shown in Fig 4.19(a) The higher Johnson noise from using large resistors can be reduced by connecting a capacitor in parallel with R2 to filter off the noise from the reference node Although in this case, DReM detection sensitivity is not affected for smaller balancing resistors where R1 / Rd  , it should be noted that smaller resistors result in larger current drawn in the balancing arm which 90 Chapter _ may affect the stability of the reference voltage due to resistance variation from nonuniform heating Thus, to attain good symmetry for a fixed F , R1 is chosen to match R3 and R2 is chosen to match Rd (a) Detection Sensitivity, SNR (b) Voltage sensitivity and noise standard deviation Fig 4.19 DReM (a) detection sensitivity, (b) voltage sensitivity and noise standard deviation F  with varying R1 and R2 91 Chapter _ 4.5.2 Detection Sensitivity at Varying Arm Ratio (a) DReM SNR (b) DReM voltage sensitivity and noise standard deviation Fig 4.20 (a) DReM detection sensitivity, (b) voltage sensitivity and noise variation at varying arm ratio Next we investigated DReM sensitivity variation at different arm ratios Figs 4.20(a) and (b) show the DReM SNR and sensitivity parameters plotted against varying arm ratio R1 is chosen to match R3 and R2 is chosen to match Rd F varies by using different values of R1 and R3 , and keeping R2 fixed to balance the bridge circuit The 92 Chapter _ simulated results of DReM voltage sensitivity, described by eqn (4.11a), are overlaid onto Fig 4.20(b) showing good correlation with the experimental results Fig 4.20(a) shows that for F  10 , DReM SNR increases with F due to increased voltage sensitivity and it peaks at F  10 For F  10 , the voltage sensitivity saturates and the SNR drops due to increased noise as shown in Fig 4.20(b) The gradual increase in DReM noise with arm ratio could be attributed to higher broadband noise at a higher bridge biasing voltage [87] From the theoretical analysis and experimental results, the ideal case of infinite F would not achieve the maximum detection sensitivity due to increased noise at higher resistance Optimum DReM detection sensitivity is achieved at F ~ 10 The choice of resistance for the balancing arm should match the DUT arm for good symmetry 4.6 Measurement of Laser Induced Resistance Change Since DReM voltage sensitivity is less dependent on the scan speed and averaging mode, it can be applied to quantify the thermally induced resistance change The signal mean from Fig 4.20 is converted from grayscale average to average voltage change After normalization with gain, rd can be expressed as follows: rd  v d F  ( ) Id F (4.12) rd is tabulated with eqn (4.12) and shown in Fig 4.21 The result shows consistent laser induced resistance change of ~ 68 mΩ, independent of DReM arm ratio, on an 15.4 Ω Al line structure biased at 15.4mV under 26.7mW laser irradiation measured from the 20x objective The laser spot size is measured to be around 2.8 μm [80] and the estimated power density is 4.3 mW/μm2 93 Chapter _ Fig 4.21 DReM resistance change measurement of a 15.3 Ω Al line structure biased at 15.4 mV at varying arm ratio under 26.7 mW laser irradiation 4.7 Sensitivity Comparison of AC-Coupled and DC-Coupled Detection Systems A sensitivity comparison was made between ac-coupled and dc-coupled detection systems under the same laser power, biasing conditions and scanning parameters shown in Table 4.2 The systems are also optimized to an identical r.m.s noise level ~50 – 55 mV measured across DUT at 10 kV/V gain with detection bandwidths of 0.03Hz – 10kHz and DC – 10 kHz for ac-coupled and dc-coupled detection systems, respectively 94 Chapter _ Table 4.2 Experimental parameters for sensitivity comparison of ac-coupled and dccoupled detection systems AC-Coupled TIVA, SRM & TBIP Detection Systems DC-Coupled DReM DUT Biasing Conditions Vd 15.4 mV 15.4 mV Id mA mA Magnification Objective 20x 20x Digital Zoom 1 Laser Parameters Wavelength 1340 nm 1340 nm Power ~26.7mW ~26.7mW Scan Parameters Digital Averaging None None Scan Speed 50 μs/pix 50 μs/pix Amplifier Settings Input Coupling Mode AC DC Gain, Av 10 kV/V 10 kV/V Band Pass Filter 0.03 Hz-10 kHz DC-10 kHz Figs 4.22 show the SNR comparison for TIVA, SRM, TBIP and DReM and their corresponding sensitivity parameters respectively plotted against varying arm ratio 95 Chapter _ Fig 4.22 (a) SNR comparison of ac-coupled and dc-coupled detection systems Fig 4.22 (b) Voltage sensitivities of ac-coupled and dc-coupled detection systems Fig 4.22 (c) Noise standard deviation of ac-coupled and dc-coupled detection systems 96 Chapter _ For F  , the detection sensitivities of TBIP, SRM and DReM are weaker than TIVA due to compromised voltage sensitivities At this operating range, the noise of DReM is observed to be approximately times smaller than SRM and TBIP Although the same characteristic voltage supply noise, due to low voltage sourcing as described in Fig 3.12(a), is also observed in both the DUT arm and balancing arm of DReM setup This noise is significantly reduced via the high common mode rejection ratio (  85dB ) during dc-coupled amplification [85] At F  10 , better noise rejection allows DReM SNR to peak at 1.4 times higher than TIVA At F  10 , the voltage sensitivities of the detection systems are comparable This is in correlation to the theoretical analysis from their circuit models The detection sensitivities of the detection systems eventually converge to a comparable level when their noise levels become comparable 4.8 Improvements on DReM Circuit The disadvantages in the DReM bridge circuit are that a large voltage source is needed to achieve high resistive arm ratio, significant Johnson noise is observed at large arm ratio and thermal banding effects occur at long dwell time Fig 4.23 Improved setup for dc-coupled detection 97 Chapter _ Fig 4.23 shows an improved setup for dc-coupled detection with some analog circuits to eliminate the disadvantages of the bridge circuit Instead of a resistor, R3 can be replaced by a transistor whose bias is adjusted to give a useful differential impedance, ' R3  VDS , for the purpose of giving large voltage signal In addition, the reference I DS arm in the bridge circuit can be replaced by a low pass filtered copy of the DUT voltage which eliminates the noise in the reference node and allows it to track any drift to reduce the thermal banding effects 4.9 Summary In conclusion, a dc-coupled laser induced detection system has been developed for fault localization in the voltage bias, voltage detection configuration The advantages and disadvantages of ac-coupled and dc-coupled detection systems have been compared The dc-coupled detection mode of DReM allows precise localization of the resistive sites without artifacts, within a single scan It is also less dependent on laser acquisition parameters like scan speed, point averaging, amplifier bandwidth, defect structure and scan orientation DReM can be used for quantitative measurement of laser induced signals and can be optimized to achieve higher detection sensitivity than ac-coupled detection systems due to better noise rejection 98 ... lock-n detection assuming T1 /   DUT 16 9 17 1 17 2 17 3 17 3 17 4 17 5 17 9 18 0 18 1 18 2 18 2 18 3 18 5 19 3 19 4 xxii Chapter _ Chapter 1: Failure Analysis and Fault Localization. .. Narrowband Lock-In Detection 11 1 11 5 5.2 .1. 1 DC-Coupled Detection Systems 11 5 5.2 .1. 2 AC-Coupled Detection Systems 11 7 5.2.2 Wideband Lock-In Detection 12 0 5.2.2 .1 DC-Coupled Detection Systems 12 0... image (a) and reflected-TIVA overlay image (b) from a polysilicon structure using Mitutoyo 20x objective and RSIL application 14 2 14 7 14 9 15 3 15 5 15 6 15 8 15 8 15 9 16 0 16 1 16 1 16 2 16 6 16 6 16 8 16 9 xxi