Effects of hydrogen on device total ionizing dose and dose rate response

Transcription

1 Effects of hydrogen on device total ionizing dose and dose rate response Jie Chen 1, Hugh Barnaby 1, Bert Vermeire 1 Ron Pease 2, Philippe Adell 3, Harold Hjarmalson 4 1 Electrical Engineering, ASU, Tempe, Az 2 RLP Research, Las Lunas, NM 3 JPL, Pasadena, CA 4 Sandia National Labs, Albuquerque, NM 1

2 Acknowledgements NASA Electronic Parts and Packaging Program (NEPP) Jet Propulsion Laboratory MURI 2

3 Topics Effect of hydrogen on dose rate sensitivity ELDRS mechanism and modeling Results of TID exposure at different dose rates in H environment Modeling the effect of H on ELDRS Effects of hydrogen on TID and annealing response (Covered in Appendix) Previous TID experiment in hydrogen Un-biased annealing in air/hydrogen 3

4 ELDRS in bipolar devices Enhanced Low Dose Rate Sensitivity in bipolar devices After Witczak et al, IEEE Trans. Nucl. Sci.,

5 Existing ELDRS Models Space Charge Model After Rashkeev, Cirba, Fleetwood, Schrimpf, Witczak, Michez, Pantelides, TNS After Boch, Saigné, Schrimpf, Vaillé, Dusseau, Lorfèvre, TNS Trapped e - Model? Other competing reaction models After Freitag and Brown, TNS H 2 cracking model After Hjalmarson, Pease, Witczak, Shaneyfelt, Schwank, Edwards, Hembree, Mattsson, TNS

6 Some Key ELDRS Processes ELDRS really is: Reduced High Dose Rate Sensitivity Low E-field HDR High space charge induced localized HDR Increases recombination of n - /p - Increases probability of n - trapping, recombining with free holes - Si-substrate - Electrons (free or trapped) Holes (free or trapped) Increases probability of trapped p annihilation by n - - 6

7 Modeling ELDRS ELDRS mechanism simulation Using 2D COMSOL Reduced hole yield at HDR due to increased n - /p recombination caused by space charge Hole yield directly affects H and N it generation Hole fraction yield Hole yield dependence on dose dose rate (rad 7

8 Modeling ELDRS Hole flux is affected by applied E-field Appied E-field directly affects ELDRS response Effect of external field on dose rate response 1.4E11 1.2E11 1.0E11 8.0E10 6.0E10 4.0E10 2.0E10 0.0E Dose rate (rad/s) Bias Applied at the Interface 0.05V 0.075V 0.1V 0.15V After Witczak et al, IEEE Trans. Nucl. Sci.,

9 Effect of H 2 on ELDRS ELDRS/H 2 Experiment on bipolar parts Gated bipolar test structures (GLPNP) with P-glass passivation (ASU). Nit (x 10^11) Zero bias ASU-air ASU-1.3%H Crane Air 1.3%H 100%H? Dose rate (rad/s) Presenting at NSREC

10 Effect of H 2 on ELDRS ELDRS/H 2 Experiment on bipolar parts (LM193) Delta Ib (na) Zero bias Nitride-0% NP-0% NP-1% NP-100% Nitride-0% NP-0% NP-1% NP-100% Dose rate (rad/s) Presenting at NSREC

11 Effect of H 2 on radiation response " Z1 P1! Nit (cm -2 ) GLPNPs exposed to 30krad Co-60 gamma rays at 20rad/s Enhanced buildup in N it as H2 concentration increases due to H 2 cracking reactions " Z0, P H 2 (cm -3 ) ASU data Crane data H 2 2D 2DH After Chen et al, IEEE Trans. Nucl. Sci.,

12 Impact of hydrogen on ELDRS ELDRS response also changes with H HDR Shift in dose rate curve due to H 2 : H 2 2D 2DH p DH D H ASU-air ASU-1.3%H Crane Air 1.3%H 100%H? Competing with Nit (x 10^11) n - D D p n - X p D D p leaving oxide Dose rate (rad/s) Becoming more competitive with high hydrogen concentration 12

13 Modeling ELDRS with Hydrogen Comsol 2D ELDRS simulation with 4 different H concentrations Saturation Effect not considered of hydrogen on dose rate response 2.5E11 2.0E11 1.5E11 1.0E11 5.1E10 DH1 (3e18 cm^ -3) DH2 (4e18 cm^ -3) DH3 (6e18 cm^ -3) DH4 (8e18 cm^ -3) DH uniformly distributed within 100n of interface 1.0E Dose rate (rad/s) 13

14 Ongoing experiments ELDRS Experiment is planned for April and May, 08 All test devices are P-glass passivated devices showing ELDRS Devices will have either air and 100% H 2 in radiation environment Data points will be taken at 0.1 rad/s, 150 rad/s, and 20 rad/s ASU-air ASU-1.3% H Crane Air 1.3%H 100% H? Nit (x 10^11) JPL ASU Dose rate (rad/s)

15 Summary Effect of hydrogen on ELDRS Hole yield is dependent on dose rate High hydrogen concentration in low-field oxide shifts the ELDRS response of bipolar devices. Preliminary simulations shows good qualitative response. Annealing experiments after TID exposure in H 2 No signification annealing of N it. Large differences in N ot annealing slopes. Differences in annealing slope maybe due to spatial distribution of H-induced trapped charge (p or H ). 15

16 Future work Isochronal annealing and study the effect of E- field and spatial distribution of H-induced trapped charge. REOS (Sandia) and COMSOL(ASU) modeling of ELDRS and TID response. FTIR and ERD (Elastic Recoil Detection) measurements on H-defect properties in samples irradiated in H 2. 16

17 Appendix Effects of hydrogen on TID and annealing response Previous TID experiment in hydrogen Un-biased annealing in air/hydrogen Biased annealing in air 17

18 Enhanced TID degradation in H 2 Increase in N ot BJTs exposed to 30krad Co-60 gamma rays Enhanced buildup in N it Increase in N it Enhanced buildup in N ot After Chen et al, IEEE Trans. Nucl. Sci.,

19 Annealing in air! Not (cm -2 ) 5.E11 4.E11 3.E11 2.E11 100% H2 (2nd test) air (2nd test) 100% H2 (1st test) 10% H2 (1st test) 1% H2 (1st test) 0.1% H2 (1st test) 0.01% H2 (1st test) air (1st test) 1.E11 0.E Time (hours) Anneal with no bias N ot anneal slope greater for higher H 2 Annealing experiments on N it showed little change at room temperature 19

20 Annealing in 100% H 2 Anneal with no bias Sudden drop of N ot corresponds w/ increase of N it * H 2 D DH H * After Mrstik & Rendell, IEEE Trans. Nucl. Sci.,

21 Brief Summary of Results Post Annealing Annealing Irradiation in Air in 100% H 2 Samples irradiated H 2 N ot N ot N ot N it N it N it Samples irradiated in N ot N ot N ot Air N it N it N it 21

22 Mechanism of N ot Annealing Oxide H 2 samples Air samples x x = 0 ρ ot,0,h2 Si e - tunneling ε field ρ ot,0,air "! ot ( x, t) " t # ( x, t) = # ot = # g( x, t)! ( x, t) ot,0 ( ) ot x e If assuming uniform spatial distribution near the Si/SiO 2 interface and low E-fields in the oxide: $ " te $! x 5.E11 4.E11! Not (cm -2 ) 3.E11 2.E11 100% H2 (2nd test) air (2nd test) 100% H2 (1st test) 10% H2 (1st test) 1% H2 (1st test) 0.1% H2 (1st test) 0.01% H2 (1st test) air (1st test) $ q# % N t & t " ot,0 ot ( ) ln(! ) 1.E11 0.E Time (hours) After Mclean, TNS,

23 Biased annealing results Biased annealing behavior of rad-induced N ot With applied bias, change in annealing rate. 100% H 2 irradiated Usually the tunneling equation is solved numerically with a E-field dependent tunneling probability: g( x, t) =!( E)exp( " Z( E)) After Mclean, TNS,