ABSTRACT

The optoelectronic reliability of representative radioactivity-exposed nanophotodetectors and the degree of functionally tolerable radioactivity-induced responsivity de-emphasis, against increasing cumulative radioactivity-dose, is notionally considered and modelled, with a view towards experimental findings concerning p-i-n photosensors being exposed to regulated successive (α, β)-particle bombardments.

1. Introduction

The communication-technology importance of optoelectronic semiconductor photodetectors has been being righteously discussed [1-3], with the ambient-sensitive functionality of their nanointerface interestingly focused upon.

Unfavourable functional alteration is, still, known to result from exposure of semiconductor optical-signal sensors to a particle-irradiation environment, as in the case of optoelectronic nanodevices carried by Space vehicles. And yet, appropriate assessment of such performanceconsequences may allow for the emergence of a reliability interval regarding the optoelectronic responsivity of the photodetector, despite its ambient-fatigue, like its radioactivity-induced photoresponse de-emphasis, in particular.

The non-ionising energy loss of the incoming bombarmsnt particle has been monitored, for example, for GaAs and Si devices as leading to displacement damage. Such lattice disruptions are believed to invoke deep-level defect-states modifying the electrical properties of the device active material and even onsetting relaxation-like and semi-insulating behaviour.

In the present, work the optoelectronic reliability of representative radioactivity-exposed nanophotodetectors and the degree of functionally tolerable radioactivity-induced responsivity de-emphasis, against increasing cumulative radioactivity-dose, is notionally considered and modelled. Systematic experimental findings concerning commercial p-i-n photodetectors being exposed to regulated successive α-particle bombardments appear in compliance with the analysis and reveal noteworthy trends of radiation-effected optoelectronic-yield alteration.

2. Modelling Scheme

Choosing to be monitoring the optoelectronic reliability of the photodetector nanointerface ultimately through the device detection yield Y, we employ as its exact definition in this study the number of photogenerated charge carriers per incident illumination photon. Thus, it may be expressed as the ratio of the time rate of creation of flowing photocarriers (1/e) I (with e being the elementary electron-charge and I the detection photocurrent responding to the illumination beam impinging upon the photodiode under study) over the temporal rhythm of incidence of illumination photons AΦ (with A being the exposed photodiode-surface area and Φ the illuminating-beam photonic flux (photons/( cm²·s))):

Y = (1/e) I/(AΦ)(1)

On the other hand, the number (1/e) I of charge carriers photogenerated per unit time is given by that part of the number AΦ of illumination photons striking the exposed nanodevice per temporal unit which having escaped reflection at the illuminated surface (by a probability of (1 – R), R being the reflectivity of the exposed-surface semiconductor at the specific illumination wavelength) inhabit mainly [4] the photodiode depletion-zone (by a cumulative occupation probability of [1 – exp (–αW)] with α being the depletion-zone material absorption-coefficient for the specific illumition wavelen -grh and W the depletion-zone width valid for the value of reverse bias applied to the sensor under testing, deriving as the difference between probability of photonic entrance into and probability of photonic exit from the depletion-zone extension under the assumption of shallow depletion-zone, materialising for the technologically conventional photodetectors, like the commercial p-i-n photodiodes) and, furthermore, succeed (by a quantum efficiency of F corresponding to the specific illumination-wavelength) in being absorbed within the depletion zone, ultimately energetically liberating initially bound (in the semiconductor valence-band or at impurity or lattice-defect trap-levels) charge carriers:

(1/e) I = F [1 – exp(–αW)] (1 – R) AΦ. (2)

It is, then, obvious that the photodetector nanointerface detection yield Y, as defined by (1), is expressible, by virtue of (2), as Y = F [1 – exp (–αW)] (1 – R)(3)

which describes the detection-yield dependence upon chosen illumination-wavelength λ (through the λ-related quantities F, α, and R) and applied reverse bias V (through the phodetector diodic depletion-region width W).

A noteworthy prediction, then, regarding the radiation-hardness optoelectronic reliability of the photodetector nanointerface, as exemplified by its detection yield behaviour and sustainability, is that it would be essentially effected upon by the radioactivity dose δ influencing chiefly its (internal) quantum efficiency F (potentially codified, thus, as F (λ; δ) ), with the non-excluded possibility that, for each cumulative radioactivity-particle intaking, transcending a critical photonic flux of illumination could through some “photonic-congestion state” adequately liberate the detection yield from radioactive-radiation impediments.

The experimentally measured, now, conductivity current through the illuminated reverse-biased diodicphosensors comprises the optoelectronic response photocurrent I and the dark saturation-recombination current (saturation of the dark reverse current having been reached at the sufficiently high reverse bias regularly applied), both flowing in the reverse sense with respect to a forward- biased diodic photodetector’s dark conduction situation.

As, then, is well known [5-7], the dark saturation— recombination diodic current is at a given absolute sensorambient temperature determined by the dark depletion-zone built-in voltage value, indicatively measuring for commercial Si p-i-n photodiodes below 200 - 300 nA against registered total conductivity-current values of units or tens of a μA through wide-range and high-level photonic fluxes. These facts render the directly measured overall conductivity-current permeating the photodetector nanointerface adequately approaching the net optoelectronic detection photocurrent I.

3. Experiment

The experimental configuration comprises the infrared (IR) LASER-beam emitter unit, an optical fiber waveguide, and the photodetector nanodevice part: Dou bleheterojunction AlGaAs IR-LASER diodes emitting in the 0.78 μm band are launched into the front end of a 1 m long. 3 mm in diameter, single-mode, Eska high performance plastic optical fiber connected at its rear end to (preferentially up to this stage but not exclusively) commercially conventional, narrow receiving-angle, linear response, fast switching-time, Si p-i-n photodiodes exhibiting peak responsivity for incoming wavelengths between 0.75 and 1.00 μm.

Each double-heterostructure AlGaAs IR-LASER diode is controlled through a special, high-fidelity, current source sustaining up to 54 mA of injection current. The optical power at the output port of the Eska fiber waveguide is exactly measured at each LASER diode injection-current level utilised by a Melles Griot, 4-digit, universal optical-power metre bearing a 10-mm-aperture Si-detector head. These optical power P measurements along with each illuminated Si p-i-n photodiode’s area A lead to the respective energetic intensity Θ = P/A (μW/cm²) values, which through the straightforward relation Θ = h (c/λ) Φ, with h being Planck’s action constant and c being the universal constant of the speed of light, furnish the photonic-flux Φ (photons/( cm²·s)) data as witnessed by the photodetector nanodevice investigated each time, for the successive LASER-diode injection-current levels employed.

The optoelectronic response of the diodic photodetector studied per experiment whilst sensing the incident IR signal is on the other hand monitored at each experimental step in terms of the photocurrent I flowing through its nanointerface. For the series of optoelectronic-reliability experiments (performed at room temperature) concerning Si p-i-n photodetectors in particular and reported here, the order of magnitude of the impingent IR photonic flux Φ ranges from 4.6 × 10¹⁴ to 8.0 × 10¹⁵ photons/( cm²·s), whereas the resulting detection-photocurrent I permeating the Si photodiodes varies from, around, 2 to 70 μA.

For each Si p-i-n IR photodetector nanointerface studied, the optoelectronic-response I-Φ characteristic curve is experimentally traced both prior to and after exposure to some decided cumulative radioactive α-particle dose δ (α-particles/cm²) materialising through the bombarding   of the photodetector at a constant α-particle flux (α-particules/(cm²·s)) for a prederemined time-interval.

The radioactive α-particle source utilised for the experiments described here is of ²⁴¹Am nuclide with a 2.87 mm face-diameter and a mean emitted α-particle energy- value of approximately 5 MeV. Care is taken that the ²⁴¹Am source is properly situated in almost direct contact to each exposed photodetector nanodevice for the time period desired. The exact cumulative α-particle dose to which the photodetector has been, thus, exposed is evaluated by consideration of this time interval along with the accurate α-particle flux at the site of the exposed photodiode, measured by a Leybold Heraeus Geiger-Muller counter.

The basic processing of the experimental (plausibly singly or piece-wise linear) I-Φ characteristics for any photosensor considered provides the absolute value(s) (in photogenerated flowing charge-carriers per incident visible or IR illumination-photon) of the nanodevice’s detection yield Y, which in accordance with its notion and definition (1) may be obtained through the photodetector optoelectronic-response-curve slope (ΔI/ΔΦ) calculated by a least-squares fitting as：

Y = [1/(eA)] (ΔI/ΔΦ)(4)

4. Radioactivity-Exposed Nanophotodetector Optoreliability-Assessment 

Detailed investigations during the series of experiments have, with respect to the set of commercial encapsulated Si p-i-n photodetectors and α-particle ²⁴¹Am source, traced a cumulative-dose threshold δ₀ of around 4.2 × 10⁹ α-particles/cm² necessary to surpass for clearly measurable radioactivity-induced performance-effects, interestingly comparably to relevant findings referring to observable particle-irradiation-produced operation-modifications in AlGAs/GaAs quantum-well IR-photodetectors [5].

Fundamental features of our systematic experimental correlation between magnitude of radioactivity-induced performance alteration of photodetector nanointerfaces and cumulative particle-irradiation-dose intaken include the following (traceable also in representative Figure 1. deriving from IR at λ = 780 nm illumination), registered for Si Photodiodes exposed as above: 

1) Overall attenuation of the detection photocurrent I for each photonic flux Φ impinging upon the photodetector’s illuminated area by a factor dropping to, about one fifth its low-flux-regime value upon proceeding (in almost the middle of the employed flux-range [4.6 × 10¹⁴ to 8.0 × 10¹⁵ ] photons/(cm²·s)) to the high-flux region. The relative magnitude of the detection photocurrent after-exposure attenuation factors appears uniform for the different α-particle cumulative exposure-dose-levels employed (up to about Δ = 3.2 × 10¹² α-particles/cm² , wherefrom extinction-like effects seem to be setting on), though expectedly their absolute values are connected with the respective total radioactivity-dose for each experimentally traced I-Φ characteristic curve.

2) Splitting of the single pre-exposure optoelectronicresponse linearity into two distinct consecutive after exposure linearity-regimes of different slope (embodying the differing detection yield), stemming away from a clearly defined kink observed, uniformly for any tested cumulative particle-irradiation-dose) at an incomingphotonic-flux level of Φ* = 5 × 10¹⁵ photons/(cm²·s), in about the middle of the employed photonic-flux range. It is noteworthy that the high photonic-flux after exposure linearity-regime extends with a slope (detection yield) essentially equal to the one marking the single pre-exposure I-Φ linearity, for any α-particle cumulative exposure-dose δ surpassing the relevant threshold δ₀ and up to the “extinction demarcation” Δ.

3) Optoelectronic response dynamic stability, as manifested by the inverse of the relaxation-time preceding the fully reliable reading of the successive detection photocurrent I values, degraded by almost a factor of 2.

The respective optoelectronic-yield Y value is straight forwardly deduced to be approximately 21 photogenerated flowing charge carriers per incident infrared photon On the other hand, Figure 2. traces the evolution of the detection yield Y, concerning the optoelectronic response of the previous photodetector class to the applied IR illumination photonic fluxes Φ transcending the “photonic-congestion state” ignitionpoint Φ*= 5 × 10¹⁵ photons/(cm²·s), against cumulative particle-irradiation dose δ increasing from radioactivity-consequence threshold δ₀ of, around, 4.2 × 10⁹ α-particles/cm²  to beyond the “extinction demarcation” Δ = 3.2 × 10¹² α-particles/cm² .

The above prime characteristics of the after-exposure optoelectronic performance of tested photodetector nan- ointerfaces appear understandably consistent with the adequately established observation of other research- ers that particle-bombardment-invoked lattice displacement damage results in degradation of carrier mobility and lifetime, causing a decrease in the device’s optoelectronic dark-current and responsivity [4,5]. And yet, novel trends and invariants are herewith plausibly revealed.

5. Conclusions

In conclusion, evidence of the character of the optoelectronic reliability of representative radioactivity-exposed

nanophotodetectors and the degree of functionally tolerable radioactivity-induced responsivity de-emphasis, against increasing cumulative radioactivity-dose, is notionally considered and modelled. Experimental results regarding the sample class of commercial encapsulated Si p-i-n IR photodetectors exposed to α-particle bombarding, are obtained both consistent with the modelling scheme and conducive to reliable predictions for relevant-devices functioning either in a Space-application environment or as intended radioactivity-sensors and counters.

Notably, the optoelectronic reliability (as codified by the detection yield) of photodetector nanodevices appears rather immune to radiation hardship, provided that the cumulative particle-irradiation-dose intaken does not exceed a pertinent “extinction demarcation” and that the illuminating photonic-flux transcends the “photoniccongestion state” ignition-point.

REFERENCES