IV curves of different pixel cells

Transcription

1 IV curves of different pixel cells µm pitch, 10µm gap 100 µm pitch, 50µm gap current [pa] interface generation current volume generation current bias voltage [V] pitch wide gap small gap n-si n-si both pixels have identical pitch and length equal volume generation currents expected but wide gap pixel shows higher current! surface effect?? wide gap -> large surface area small gap -> small surface area

2 Crucial growth parameters goal : grow oxide films of well defined thickness and quality (low defect densities) control oxidation rate control ambient composition / contamination radiation hardness depends on the process water vapour increases oxidation rate high dopant surface concentrations (> 10^19/cm²) dopants diffuse into oxide film or pile-up at crystal surface increased oxidation rate local variations in film thickness crystal orientation density of available bonds activation energy depends on bonding angle defect density

3 chlorine additives (HCl,TCA, TCE) reduce mobile oxide charges reduce stacking faults lower defect densities increased minority carrier life-time at silicon surface (MOSFET)

4 3. Microscopic radiation effects nomenclature (a) (b) (c) gate contact spatial electrical oxide bulk oxide traps fixed states border region interface border traps + + switching states + interface traps silicon bulk oxide bulk consists of stoichiometric silicon dioxide border region is characterised by high bond stress it extends over ~30 Angstroem high defect densities interface: transition region ~5 Angstroem thickness 3 layers of Si, Si O and SiO

5 where do we stand? surface effects are correlated to the sensor design radiation induced surface damage occurs in the silicon oxide and at the interface surface effects are process related before and after radiation exposure the process related data is often not accessible by the customer quality assurance during large scale production process monitoring monitoring of radiation hardness set of macroscopic parameters required

6 4. Macroscopic surface parameters N ox density of positive oxide charges flat-band voltage U fb =U fb0 + ed ε ox ε 0 N ox oxide thickness permittivity of oxide S 0 surface recombination velocity interface generation current I ox =q 0 n i A gate S 0 intrinsic carrier concentration D it interface state density (midgap) determines surface recomb. velocity S 0 = σ eff v th πk B TD it,mg effective capture cross-section thermal velocity causes "stretch out" of high frequency CV curve of MOS devices

7 interface generation current (Iox) contributing surface region must be well defined I ox ~A gate switch Iox on and off to separate it from bulk contributions use gate-controlled diode Vbias K487 Vgate K487 A p+ oxide n-silicon n+ measure current through the diode by adjusting gate potential Iox can be turned on and off outer gates accumulated to define region contributing surface region well defined

8 current measurements with gate -controlled diodes -V SiO 2 gate V bias V gate V gate V gate V bias p + -Si SiO SiO inv.layer 2 n-si p + -Si 2 n-si max I vol,g I ox I vol,b inversion depletion accumulation p + -Si V bias I

9 flat-band voltage (Vfb) measure capacitance of a MOS device versus gate high frequencies capacitance C ox Cox Csi C fb Vfb gate voltage flat-band condition : C si (V fb )=A gate ε ox ε 0 λ λ extrinsic Debye length find gate voltage corresponding to Cfb Neff needed to determine flat-band voltage (correction for serial resistors and stray capacitances may be necessary)

10 interface state density use (for example) a combined high/low frequency CV measurement equivalent circuit for very low frequencies Csi Cox Cit capacitance C [pf] D it [10 10 cm -2 ev -1 ] voltage V bias [V] voltage V bias [V] D it = C lf -C hf q (1- C lf C ox ) -1 (1- C hf C ox ) -1 interface traps respond to "ac" gate voltage -> Cit (stretch-out) minority carrier response

11 schematic drawing of a test-field (CERN2, APIX...) gate controlled diode MOS capacitor 1mm 1.5mm 6mm 6mm topview diode 5 gate rings aluminium oxide p + n bulk n + µ m cross-section 5 50 the advanced test field provides current measurements as well as capacitance methods on the same device more details in Wunstorf et al., NIM A444 (2000)

12 Results before irradiation 350 Capacitance curves of P186-C-TF-2 MOS1, unirradiated Polovodice <111> DOFZ, Canberra 100 khz quasi static 300 capacitance [pf] capacitance [pf] gate voltage [V] CV measurement on diode : N = cm -3 ρ =0.6kΩcm eff CV curve of MOS : C ox = 330pF d ox = 160nm C si (V fb ) = 124pF C fb =90pF V fb =2.2V N ox = cm -2

13 Band bending versus gate voltage (P186-C-TF-2, MOS capacitors) MOS 1 (Ufb=2.12 V) MOS 2 (Ufb=2.18 V) 0.3 band bending [ev] gate voltage [V] band bending ψ(v gate )= V gate (1- C (V) lf )dv V fb C ox 4e e+10 Interface state density of P186-C-TF-2 (unirradiated MOS capacitors) MOS 1 MOS 2 3e+10 Dit [1/cm /ev] 2.5e+10 2e e+10 1e+10 5e band bending [ev] ψ midgap =0.17eV D it,mg ev -1 cm -1

14 0.08 Iox-curve of Polovodice <111> oxygen., Canberra (unirradiated) current [na] gate voltage [V] Iox = na S0 = 7 cm/s

15 comparison of unirradiated devices substrate / vendor N ox [10 11 /cm 2 ] S 0 [cm/s] D it,mg [10 10 /ev /cm 2 ] σ eff [10 16 /cm 2 ] Polovodice, DOFZ Canberra Topsil, <100>, DOFZ CiS Topsil, <100>, Std. FZ CiS Wacker, <111>, Std.FZ CiS Wacker <111> 16h DOFZ CiS Wacker <111> 24h DOFZ CiS Wacker <111> Std.FZ IRST Wacker <111> 24h DOFZ IRST Topsil <111> 72h DOFZ Sintef Wacker <111> 72h DOFZ Sintef ? ? <4.4 < value calculated input for calculation high quality oxidation effective capture cross-section in agreement with commonly used values the performance of <111> silicon is as good as for <100> crystals no difference between DOFZ and Std.FZ!

16 a) current IV curves of single pixels after 11 kgy ) IV curves of single pixels after 300 kgy current increase compared to current before irradiation exponential shape of the curves! gap dependent ) Wüstenfeld et al., Il Nuovo Cimento, Vol. 112A, N. (1999)