Alpha Particle Detection with a Boron Deep-Trenched Silicon PN Junction Diode

1. Alpha Particle Detection with a Boron Deep-Trenched Silicon PN Junction Diode Jonathan Marini Christopher Shing SDM 1 December 6, 2010

2. α α e- Ionization e- e- α α Excitation e- Alpha Particle Interactions • Two important parameters [1]: • Generation of carriers following an interaction event, dependent on ionization energy of interaction material • Acknowledgement of carriers following an interaction event, dependent on maximum electric field in material • Interacts by either excitation or ionization occurs Nucleus Figure 1: Model of alpha particle interaction • Solid state devices have both of these properties • Junction devices operate better at alpha particle detection

3. Neutron Detection • Two types [1]: • Slow Neutrons, significant interactions include elastic scattering, but small kinetic energy – not great for detection, but generates secondary particles • Fast Neutrons, significant interactions include inelastic scattering – detect neutron themselves N • Detection dominated by secondary radiation generation α Nucleus Figure 2: Model of neutron interaction • Many researchers interested in slow neutrons for study since induced by a wide range of events.

4. Boron Deep-Trenched Silicon PN Junction Diode • P+N junction to tailor WD to charged particle [2] • WD should be larger than interaction range of particle in question (α) • Particle generates e-h pairs in depletion region • Quantum efficiency v response time • A tailored depletion width prevents unwanted interactions • Ex: γ-particles • Planar Detector • Tradeoff • interaction in conversion layer • interaction in depletion region • Generally low efficiency Figure 3: Planar Neutron Detector [3]

5. Boron Deep-Trenched Detector • High aspect-ratio trenched structure • Manipulation of pillar width to maximize WD • Self-Powered • Detectors usually use reverse bias to increase e-h generation during operation • Geometry of device removes this requirement • Low leakage current Figure 4: a) Partially depleted and b) fully depleted pillar of boron deep-trenched silicon pn junction diode [3]

6. Boron Deep-Trenched Detector Applications • Neutron Detection • Neutron interacts with conversion layer (Boron) • Emits α-particle into Si Conversion reaction of an incident thermal neutron [3] Figure 5: Boron Deep Trenched Silicon Neutron Detector [3]

7. Boron Deep-Trenched Detector Parameters • Trenches are hex shaped, most efficient design for neutron detection • 1 µm pillar, 2.8 µm trench • NA: 1018 ND: 1014 • ~200 nm junction depth • 2 mm * 2mm device was used • Square shaped trenches also used in simulation for comparison • Hand Calc: Vbi=0.695V Figure 6: Hex geometry [3] Wtr = trench width Wpl = pillar width hpl = pillar height Xp = n-side Wd Capacitance/Area equation [4]

8. Alpha Particle Applications • Neutron Detection • Chip manufacturing, soft error detection • Smoke detection • Radon detection • Earthquake prediction • Alphavoltaics • Homeland Security Figure 7: Applications of alpha particle detectors [3, 5-8].

9. Neutron Detectors • Homeland Security • Nuclear Safety • Nuclear Reactors Monitoring • Particle Physics • Medical Sciences • Cosmic Rays • Protection against SNM Figure 8: Applications of neutron detectors [8-13].

10. Experimental Plans • GEANT4 • Open source simulation software • Determine penetration depth of alpha particle • MEDICI • Simulate depletion region width in pillars • Simulate C-V curve of PN photodiodes • Simulate I-V curve of PN photodiodes • I-V measurement • Make sure no large deviations from expected • C-V measurement • A large capacitance could interfere with detection • Time response measurement • Using LED, determine capability for alpha particle spectroscopy • Alpha particle detection • Determine alpha particle detection ability

11. GEANT4 Simulation Results • Very difficult to use. • Hand calculation provided instead. • Boron Ionization = 8.429eV, Z=5 • Silicon Ionization = 8.157eV, Z=14 • Aluminum Ionization = 5.985eV, Z=13 • Using the Stopping Power and the Bethe-Bloch equation [1]: • Estimate Range to be: • In Boron: • 12µm • In Aluminum/Silicon: • 25µm Figure 9: Estimated Range Depiction

12. MEDICI Simulation Parameters • Used same parameters as device • 1 µm pillar, 2.8 µm trench • NA: 1018 ND: 1014 • ~200 nm junction depth • Made assumptions as necessary • Aspect ratio of 16:1 for H:W of trench • Depth of boron layer on top calculated using trench width • Only 1 pillar simulated • Half of trench on either side • Used insulator to take the place of boron • Made entire device 100 µm deep • Assumed perfect electrodes • Results in Unit/Area so estimated number of pillars in device

13. MEDICI Simulation Result – Depletion Region b) • Depletion region under zero bias • Pillar is fully depleted • Grid tailored to doping and pillar geometry • Hand Calc: WD=3.024µm @ 0V a) Figure 10: a) Mesh and b) Depletion Region results of MEDICI simulation

14. MEDICI Simulation Result – I-V Characteristics a) b) Figure 11: a) Linear IV measurement and b) Log IV characteristics of the MEDICI simulation • Simulation variance is due to assumptions • Big difference in geometry of simulation to actual device • Multiplying by number of pillars makes current magnitude closer to measured results

15. MEDICI Simulation Result – C-V Characteristics a) b) Figure 12: MEDICI simulation results depicting a) C-V plot and b) 1/C2 • Proportional to 1/N • Similar to ideal resultsdepicted in Sze [2] • Hand Calc: C=0.251 µF/cm2

16. Measured I-V Characteristics b) a) Figure 13: a) Measured I-V plot versus Ideal I-V plot and b) Measured I-V plot versus MEDICI Simulation I-V plot • As identified in Sze, Boron Deep Trenched Silicon PN junction diode current is dominated by non-ideal defects such as high-injection current and series resistance [2]

17. Measured C-V Characteristics • Unable to get measured results • Used equipment in CII 4th Floor clean room annex • LabVIEW program caused errors in C-V plot • Equipment limitations

18. Time Response a) b) d) c) Figure 14: Time Response of the Boron Deep Trenched Silicon Device under illumination with a) -0.5V, b) -2.5V, c) -3.5V, d) -5V bias

19. Time Response • Applied Exponential Fitting • Device in series with 10kΩ • Measured Capacitance, series resistance not taken into account • LED Intensity = 112µW/cm2 Figure 15: Time Response Experimental Setup

20. Time Response • Since large deviation, calculations of series resistance provided • Used calculated capacitance. • Rseries = Rs1+Rs2+Rs3+… • τ=Cd x (Rseries+10kΩ) • Series resistance dominates, proves IV simulation and measured results Figure 16: Circuit model of BoDT device

21. Alpha Particle Detection a) b) • Used Po-210 source, 5.5MeV, an approximate 300µCi intensity • (11.1Mbq or 11.1x10-6 Interactions/second) • Very small Alpha/No Source Gain = 1.04 @ -5V • No distinguishable results Figure 17: a) Linear I-V and b) Log I-V plot of the Boron Deep Trenched Silicon PN junction diode under 300µCi alpha particle radiation source

22. Alpha Particle Detection a) • Alpha particles that penetrate boron layer are not captured by the silicon • This is due to lack of electric field in Boron as simulated in MEDICI. • The thickness of the boron layer is too great (>12µm) • Alpha particles cannot penetrate to silicon depletion region located on the substrate • Only alpha particle interaction is in area below contact or due to scattering from boron to depletion region • Not very effective Figure 18: Electric field in Boron Deep Trenched Silicon PN junction device as determined by MEDICI simulation a) zoomed and b) full pillars b)

23. Conclusion • The Boron Deep Trenched Silicon PN Junction Diode device does not work well as an alpha particle detector • Though it can detect some alpha particles • This is good for alpha sensitive neutron detection applications. Namely, the presence of alpha particles will not cause errors. Future Work • High series resistance, attempt to reduce. • Conduct tests on neutron detection capabilities • For neutron detector application • 3D fabrication techniques to increase efficiency • Exploration of effect area of device has on operation Figure 19: Imagined depiction of 3D fabrication

24. References [1] G.F. Knoll, Radiation Detection and Measurement, 4th Ed. Hoboken, NJ: John Wiley & Sons, 2010. [2] S.M. Sze and K.K. Ng, Physics of Semiconductor Devices, 3rd Ed. Hoboken, NJ: John Wiley & Sons, 2007. [3] N. LiCausi, “Design and Fabrication of A Novel Self Powered Solid-State Detector,” Doctoral Candidacy Paper, Dept. Elect. Eng., RPI, Troy, NY, 2009 [4] N. LiCausi, et. al. “A novel solid-state self powered neutron detector,” Proc. of SPIE, vol. 7079, 2008 [5] B. Crothers. (2008, Aug. 7) “IBM cuts chip plant pay, following job cuts.” Cnet News. [Photo] Available: http://news.cnet.com/8301-13924_3-10010267-64.html [6] C. Woodford. (2009, May 22) “Smoke Detectors.” Explainthatstuff.com. [Photo] Available: http://www.explainthatstuff.com/smokedetector.html [7] “Pro Series 3 Radon Detector.” OpAmerica.com. [Photo] Available: http://www.opamerica.com/pro-series-3-radon-detector-p-2038.html [8] Department of Homeland Security. [Logo] Available: http://www.dhs.gov/index.shtm [9] F. Chartrand. (2009, Dec. 1) “Nuclear Regulator’s Impartiality Questioned.” The Star. [Photo] http://www.thestar.com/business/article/732811--nuclear-regulator-s-impartiality-questioned [10] “Higgs Bosun” Wikipedia. [Photo] http://en.wikipedia.org/wiki/File:CMS_Higgs-event.jpg [11] “Radiation Therapy – Allied Health” Delgado Community College. [Photo] http://www.dcc.edu/campus/cp/ahealth/radiation_thera/ [12] K. Tuttle. (2009, June 30). “Researchers find evidence of the origin of cosmic rays.” Symmetry Magazine. [Photo] http://www.symmetrymagazine.org/breaking/2009/06/30/researchers-find-evidence-for-the-origin-of-cosmic-rays/ [13] A.M. Helmenstine. “Science Laboratory Safety Signs” About.com. [Image] http://chemistry.about.com/od/healthsafety/ig/Laboratory-Safety-Signs/Radiation-Symbol.htm

25. Questions?

26. Testing Images