Characterization and TCAD Simulation of 90nm Technology PMOS Transistor under Continuous Photoelectric Laser Stimulation for Failure Analysis Improvement (original) (raw)

Abstract

This study responds to our need to optimize failure analysis methodologies based on laser/silicon interactions, using the functional response of an integrated circuit to local laser stimulation. Thus it is mandatory to understand the behavior of elementary devices under laser stimulation, in order to model and anticipate the behavior of more complex circuits. This paper characterizes and analyses effects induced by a static photoelectric laser on a 90 nm technology PMOS transistor. Comparisons between currents induced in short or long channel transistors for both ON and OFF states are made. Experimental measurements are correlated to Finite Elements Modeling Technology Computer Aided Design (TCAD) analyses. These physical simulations give a physical insight of carriers generation and charge transport phenomena in the devices.

Figures (18)

Figure 1. Measured I-V curve of P+/Nwell junction vs. laser power. The Device Under Test (DUT) is embedded in a 90nm technology test structure and its surface is about 200um?. The experimentation was realized on an i-Phemos Hamamatsu system. The laser spot was positioned in the center of the junction and laser power was adjustable below 100mW. Current-voltage (I-V) characteristic was measured for different laser power, and presented on Figure 1. When the PN junction is reverse-biased under 0.6V, the photocurrent is almost constant independently of applied voltage since the SCR width does not vary significantly. The induced photocurrent is higher than the junction reverse leakage and depends on laser power. Then, when applied voltage raises over 0.6V, the SCR width starts to decrease and consequently the amount of induced photocurrent is also reduced. When the applied voltage reaches diode built-in potential (black cross on Figure 1), the SCR decreases, the induced photocurrent is therefore reduced and becomes negligible with respect to the junction forward current. Moreover, the so far negligible resistances of the P and N-type regions start to play a significant role in forward mode and get reduced under the illumination, which leads to a current increase with laser intensity.

Figure 1. Measured I-V curve of P+/Nwell junction vs. laser power. The Device Under Test (DUT) is embedded in a 90nm technology test structure and its surface is about 200um?. The experimentation was realized on an i-Phemos Hamamatsu system. The laser spot was positioned in the center of the junction and laser power was adjustable below 100mW. Current-voltage (I-V) characteristic was measured for different laser power, and presented on Figure 1. When the PN junction is reverse-biased under 0.6V, the photocurrent is almost constant independently of applied voltage since the SCR width does not vary significantly. The induced photocurrent is higher than the junction reverse leakage and depends on laser power. Then, when applied voltage raises over 0.6V, the SCR width starts to decrease and consequently the amount of induced photocurrent is also reduced. When the applied voltage reaches diode built-in potential (black cross on Figure 1), the SCR decreases, the induced photocurrent is therefore reduced and becomes negligible with respect to the junction forward current. Moreover, the so far negligible resistances of the P and N-type regions start to play a significant role in forward mode and get reduced under the illumination, which leads to a current increase with laser intensity.

Figure 3. Simulated I-V curve of P+/Nwell junction vs. laser power. I-V characteristics of the junction are simulated in function o: the intensity (from 100 W.cm? to 500 W.cm7”) of the 1064 nm smooth wave simulator (Figure 3). Results are correlatec to measurements, currents levels are not identical tc measurements because TCAD simulator is not calibrated to fit silicon results, it is only used to have trends.

Figure 3. Simulated I-V curve of P+/Nwell junction vs. laser power. I-V characteristics of the junction are simulated in function o: the intensity (from 100 W.cm? to 500 W.cm7”) of the 1064 nm smooth wave simulator (Figure 3). Results are correlatec to measurements, currents levels are not identical tc measurements because TCAD simulator is not calibrated to fit silicon results, it is only used to have trends.

Figure 2. Structure used for junctions TCAD simulations.

Figure 2. Structure used for junctions TCAD simulations.

Figure 4. Long channel transistor currents vs. laser power (transistor in OFF state). A ma aa OFF state: A 10um 3 x © 10m PMOS transistor is biased as follows: Gate, Source and Nwell are biased at 1.2V and the Drain is grounded. Evolution of these four electrodes currents in function of the laser power is given in Figure 4. Obviously the gate current is always equal to zero. The current conservation law is respected since the sum of all currents is always equal to zero. Moreover, the same quantity of current is approximately generated in Drain/Substrate and Source/substrate junctions.

Figure 4. Long channel transistor currents vs. laser power (transistor in OFF state). A ma aa OFF state: A 10um 3 x © 10m PMOS transistor is biased as follows: Gate, Source and Nwell are biased at 1.2V and the Drain is grounded. Evolution of these four electrodes currents in function of the laser power is given in Figure 4. Obviously the gate current is always equal to zero. The current conservation law is respected since the sum of all currents is always equal to zero. Moreover, the same quantity of current is approximately generated in Drain/Substrate and Source/substrate junctions.

Figure 5. Long channel transistor currents vs. laser power (transistor in ON state). decreases since the starting up in conduction of the channel eads an additional current which opposes to the circulation of Source photocurrent. On the other hand, |Igain| increases because it comes to be added to the normal circulation of electrons in the channel (due to its starting in conduction). Finally, the same quantity of photocurrent is induced in the transistor in both ON and OFF states (Figure 6): approximately 30 uA in Source-Nwell or Drain-Nwell junctions and 65 wA in Nwell-Psubstrate junction (at maximum laser power).

Figure 5. Long channel transistor currents vs. laser power (transistor in ON state). decreases since the starting up in conduction of the channel eads an additional current which opposes to the circulation of Source photocurrent. On the other hand, |Igain| increases because it comes to be added to the normal circulation of electrons in the channel (due to its starting in conduction). Finally, the same quantity of photocurrent is induced in the transistor in both ON and OFF states (Figure 6): approximately 30 uA in Source-Nwell or Drain-Nwell junctions and 65 wA in Nwell-Psubstrate junction (at maximum laser power).

Figure 6. Long channel transistor photocurrent generation difference between OFF and ON states (in this case subtraction between the current values obtained with and without laser stimulation).

Figure 6. Long channel transistor photocurrent generation difference between OFF and ON states (in this case subtraction between the current values obtained with and without laser stimulation).

Figure 7. Short channel transistor currents vs. laser power (transistor in OFF state). OFF state: A 10um x 0.09um PMOS transistor is used. Evolution of electrodes currents in function of the laser power is presented in Figure 7. The quantity of photocurrent induced is approximately three times smaller than in the long channel transistor case (for example Iprain-short channel= ~8-1WA and Ipyain-

Figure 7. Short channel transistor currents vs. laser power (transistor in OFF state). OFF state: A 10um x 0.09um PMOS transistor is used. Evolution of electrodes currents in function of the laser power is presented in Figure 7. The quantity of photocurrent induced is approximately three times smaller than in the long channel transistor case (for example Iprain-short channel= ~8-1WA and Ipyain-

Figure 8. Short channel transistor currents vs. laser power (transistor in ON state).

Figure 8. Short channel transistor currents vs. laser power (transistor in ON state).

Figure 9. Photocurrent induced (at maximum laser power) vs.transistor lentgh (W=10um).

Figure 9. Photocurrent induced (at maximum laser power) vs.transistor lentgh (W=10um).

Figure 10. Long channel transistor currents vs. laser power (transistor in OFF state). OFF state: The same experiment than in part A is done with the p-Substrate grounded. Results are presented in Figure 10. Gate current is always equal to zero, current conservation law is respected. When the p-Substrate is grounded, photocurrents collected at Drain and Source are small compared to the photocurrent induced in Nwell-Substrate junction. In this case we can consider that the laser stimulation has a dominating effect only on the Nwell-Substrate junction.

Figure 10. Long channel transistor currents vs. laser power (transistor in OFF state). OFF state: The same experiment than in part A is done with the p-Substrate grounded. Results are presented in Figure 10. Gate current is always equal to zero, current conservation law is respected. When the p-Substrate is grounded, photocurrents collected at Drain and Source are small compared to the photocurrent induced in Nwell-Substrate junction. In this case we can consider that the laser stimulation has a dominating effect only on the Nwell-Substrate junction.

Figure 11. Long channel transistor currents in function of laser power (transistor in ON state).

Figure 11. Long channel transistor currents in function of laser power (transistor in ON state).

Figure 12. Long channel transistor photocurrent generation difference between OFF and ON states (in this case subtraction between the current values obtained with and without laser stimulation).

Figure 12. Long channel transistor photocurrent generation difference between OFF and ON states (in this case subtraction between the current values obtained with and without laser stimulation).

Figure 13. Doping concentration mapping and SCR displayed in white lines (a) and doping profile of the mapping along the white dashed line (b)

Figure 13. Doping concentration mapping and SCR displayed in white lines (a) and doping profile of the mapping along the white dashed line (b)

Figure 14. Short channel transistor currents in function of laser power (transistor in OFF state). channel -OnA, Tywelt-short channel 1.6mA and TNwell-long channel— 1.48mA).

Figure 14. Short channel transistor currents in function of laser power (transistor in OFF state). channel -OnA, Tywelt-short channel 1.6mA and TNwell-long channel— 1.48mA).

Figure 15. Short channel transistor currents in function of laser power (transistor in ON state). In conclusion, the same magnitudes of photocurrent seem to be induced whether the transistor is short or long channel, in ON or OFF state.

Figure 15. Short channel transistor currents in function of laser power (transistor in ON state). In conclusion, the same magnitudes of photocurrent seem to be induced whether the transistor is short or long channel, in ON or OFF state.

Figure 16. Measurements of photocurrent induced (at maximum laser power) in function of transistor lentgh (W=10um). In conclusion, when the p-substrate of a long channel PMOS transistor is grounded, the same quantity of photocurrent is theoretically induced whether the transistor is short or long channel. In practice, more photocurrent is induced in transistors owning gate length smaller than the laser spot diameter.

Figure 16. Measurements of photocurrent induced (at maximum laser power) in function of transistor lentgh (W=10um). In conclusion, when the p-substrate of a long channel PMOS transistor is grounded, the same quantity of photocurrent is theoretically induced whether the transistor is short or long channel. In practice, more photocurrent is induced in transistors owning gate length smaller than the laser spot diameter.

Figure 17. Simulation of photocurrent induced (at maximum laser power) in function of transistor lentgh (W=10pm).

Figure 17. Simulation of photocurrent induced (at maximum laser power) in function of transistor lentgh (W=10pm).

Conclusions

Conclusions

Loading...

Loading Preview

Sorry, preview is currently unavailable. You can download the paper by clicking the button above.

References (17)

  1. K. Haraguchi, "Microscopic optical beam induced current measurements and their applications", Proceedings of 10th IEEE Instrumentation and Measurement Technology Conference (IMTC), vol. 2 pp.693-699, 1994
  2. R. Llido, et al., "Characterization and TCAD simulation of 90 nm technology transistors under continuous photoelectric laser stimulation for failure analysis improvement", Proceedings of IEEE International Symposium on the Physical and Failure Analysis of Integrated Circuits (IPFA), 2012
  3. A. Sarafianos, et al., "Building the electrical model of the Photoelectric Laser Stimulation of a NMOS transistor in 90nm technology", Proceeding of IEEE International Symposium for Testing and Failure Analysis (ISTFA), In press, 2012
  4. R. Minixhofer, "TCAD as an integral part of the semiconductor manufacturing environment", Simulation of Semiconductor Processes and Devices, pp.9-16, 2006
  5. T. Beauchêne, Ph.D Thesis, "Etude et localisation de défaults dans les circuits intégrés par stimulation photoélectrique laser", Université de Bordeaux 2004.
  6. J.L. Wirth and S.C. Rogers, "The Transient Response of Transistors and Diodes to Ionizing Radiation", IEEE Transactions on Nuclear Science, vol. 11 issue 5, pp.24-38, 1964
  7. V. Pouget, et al., "SPICE modeling of the transient response of irradiated MOSFETs", IEEE Transactions on Nuclear Science, vol. 47 issue 3, pp.508-513, 2000
  8. A. Douin, et al., "Influence of Laser Pulse Duration in Single Event Upset Testing", IEEE Transactions on Nuclear Science, vol. 53 issue 4, pp.1799-1805, 2006
  9. S.K. Brahma, Ph.D Thesis, "Functional Analysis of MOSFETs and HEMTs with Laser Stimulation and Photonemission", Berlin University of Technology, 2008.
  10. K. Nikawa and S. Inoue, "New laser beam neating methods applicable to fault localization and defect detection in VLSI devices", Proceedings of IEEE International Reliability Physics Symposium (IRPS), pp.346-354, 1996
  11. F. Beaudoin, et al., "Reliability Defect Monitoring with Thermal Laser Stimulation: Biased Versus Unbiased", Microelectronics Reliability, vol. 42 issue 9-11, pp.1729-1734, 2002
  12. R. Llido, et al., "Improving defect localization techniques with laser beam with specific analysis and set-up modules", Proceedings of IEEE International Reliability Physics Symposium (IRPS), pp.FA.2.1-FA.2.5, 2012
  13. R. Llido, et al., "Photoelectric Laser Stimulation applied to Latch-Up phenomenon and localization of parasitic transistors in an industrial failure analysis laboratory", Microelectronics Reliability, vol. 51 issue 9-11, pp.1658-1661, 2011
  14. C.B. T. Kiyan, C. Boit, "Timing Characterization of a Tester Operated Integrated Circuit by Continuous and Pulsed Laser Stimulation", Proc. of the 36th International Symposium for Testing and Failure Analysis (ISTFA), pp.211-216, 2010
  15. J.A. Rowlette and T.M. Eiles, "Critical timing analysis in microprocessors using near-IR Laser Assisted Device Alteration (LADA)", Proceedings of IEEE International Test Conference, pp.264-273, 2003
  16. A. Sarafianos, et al., "Building the electrical model of the Photoelectric Laser Stimulation of a PMOS transistor in 90nm technology", Microelectronics Relability, In press, 2012
  17. P. Perdu, et al., "Identification of some key parameters for photoelectric laser stimulation of IC: an experimental approach", Proceedings of the 12th International Symposium on the Physical and Failure Analysis of Integrated Circuits (IPFA), pp.21-26, 2005