Krüger, Berger, Marso, Reetz, Eickhoff, Loo, Vescan, Thönissen, Lüth, Arens-Fischer, Hilbrich and Theiss (Jpn. J. Appl. Phys. 36, L24–L26, 1997) demonstrated color-sensitive silicon photodiodes in which porous-silicon multilayer interference filters, formed by electrochemical anodization inside the p+-type part of a p+-n junction, set the diode's spectral response while the porous silicon had no significant influence on the electrical characteristics. The integrated filters had only moderate optical quality. The letter's outlook identified ~2–3 µm p+ epitaxial layers, higher anodization current densities for the low-index layers and correspondingly stronger illumination during anodization as the route to standalone-class filter quality inside the diode, leaving the influence of that stronger illumination on filter formation uninvestigated, and the group's full-length companion paper (Thin Solid Films 297, 241–244, 1997) reported the illumination trade-off qualitatively: the high power densities needed to enable high anodization current densities degrade filter performance. The surviving problem is: determine how far, and by which process means, porous-silicon interference filters can be integrated into a functioning junction photodiode with optical quality approaching that of the best standalone porous-silicon filters demonstrated on plain substrates, such that the diode's spectral response is set by the high-quality filter while the porous silicon has no significant influence on its electrical characteristics relative to a reference diode fabricated identically without the porous-silicon stack. The problem remains open.

Background

Silicon photodiodes are broadband detectors: their responsivity spans the whole visible and near-infrared range without wavelength discrimination. Color-sensitive or multispectral silicon detection therefore requires spectral shaping, achieved today mainly by dye/pigment color-filter arrays deposited on top of the photodiodes of complementary metal–oxide–semiconductor (CMOS) image sensors, or by conventional dielectric interference filters — stacks of alternating transparent thin films such as SiO2 and TiO2 whose optical thicknesses are tuned to give wavelength-dependent transmission. Such interference filters require vacuum deposition of many layers, which is slow and costly relative to standard silicon processing.

Porous silicon (PS) offers a monolithic alternative. It is produced by electrochemical anodic etching ('anodization') of silicon in hydrofluoric-acid-based electrolyte; the porosity, and hence the effective refractive index of the layer, is set mainly by the anodization current density, while the material stays largely transparent in the visible. Alternating the current density during etching produces multilayer stacks that act as one-dimensional photonic structures: Bragg reflectors (quarter-wave pairs of high- and low-index layers, giving a stop band with very high reflectivity — values up to about 99% were demonstrated) and Fabry–Pérot filters (a cavity layer between two Bragg mirrors, giving a narrow transmission peak — linewidths around 11 nm were demonstrated). These benchmarks, established by the Jülich group on plain, uniformly doped p-type silicon substrates (Berger et al., 1994, and later work), motivated the attempt to integrate such filters directly into photodiodes.

In the 1997 demonstration of Krüger et al., the PS multilayer filter was formed inside the upper, highly doped p-type (p+) part of a p+-n junction photodiode. Because the anodic current must flow through the device, the wafer was anodized under reverse bias while being illuminated (a halogen lamp filtered to wavelengths above about 630 nm, at roughly 0.5 W/cm²) to supply the carrier flow through the junction; the porous stack was confined to the p+ epitaxial layer — a doped crystalline silicon layer grown on the substrate, here about 1.1 µm thick — leaving a crystalline conducting channel between the porous region and the junction. The diode's external quantum efficiency (photogenerated electrons per incident photon, measured as spectral responsivity) showed minima near 500, 590 and 660 nm for the three integrated filters, tracking the reflectance maxima of the corresponding filters, and the dark current–voltage characteristics were nearly identical to reference diodes made without porous silicon — the diode ideality factor (the slope parameter of a diode's exponential current–voltage characteristic, 1.0 for ideal diffusion-limited behavior) rose only from about 1.2 to about 1.3–1.4. The process was fast (a single filter formed in under 20 s) and used laboratory equipment far cheaper than deposition systems, and filter regions could be defined lithographically on a scale of a few micrometers.

What the letter left open is the gap between the integrated and standalone filter quality. The filters inside the diodes had only moderate optical contrast and selectivity, well below the best plain-substrate results. According to the letter's outlook, reproducing the improved standalone characteristics within the photodiode geometry would require thicker p-type epitaxial layers — on the order of 2–3 µm — and higher anodization current densities for the low-refractive-index layers, which in turn would necessitate stronger illumination during anodization; at the intensities used, illumination caused only a small red shift of the filter spectra, but the effect of the stronger illumination required for better filters was not investigated in the letter itself. The group's full-length companion paper (Krüger et al., Thin Solid Films 297, 241 (1997)) then established the illumination trade-off qualitatively, as its abstract states: 'The optimal illumination power density turned out to be a compromise: high power densities are necessary to enable high anodization current densities, but this results in a degraded filter performance.' This documents the trade-off but neither demonstrates standalone-class integrated filters nor quantifies the ceiling. The unresolved question thus couples three interlocked constraints: optical quality of the multilayer, the electrochemical process window (current density, illumination, epitaxial thickness), and preservation of the diode's electrical behavior — the series resistance of the remaining crystalline channel, the ideality factor, and the dark current.

Later literature developed porous-silicon filters further on plain substrates (ageing suppression, laterally displaced filter arrays, commercial optical components) and built junction-free filtered detectors, but never demonstrated a junction photodiode carrying a standalone-class integrated porous-silicon filter. Mainstream color imaging adopted dye-based filter arrays, so the question survives primarily as a problem of monolithic all-silicon spectral shaping within the porous-silicon integration niche.

Problem Statement

How far can the optical performance of porous-silicon interference filters integrated into a silicon p+-n junction photodiode be improved toward the quality of the best porous-silicon filters demonstrated on plain substrates (the Bragg-reflector and Fabry–Pérot benchmarks described in Background), while the diode's spectral response remains determined by the integrated filter and the porous silicon has no significant influence on the electrical characteristics relative to a reference diode fabricated identically without the porous-silicon stack? Concretely, the question couples three interlocked aspects: (i) how close the best integrated-filter optical quality actually achievable in a color-sensitive silicon photodiode comes to the standalone plain-substrate filter quality described in Background (the ≈99% Bragg-reflector reflectivity and ≈10-nm-class Fabry–Pérot linewidth benchmarks), with the diode's spectral response determined by that filter; (ii) the anodization and illumination process conditions and the p+ epitaxial layer design that govern that best achievable quality, taking into account the trade-off documented in the group's companion full-length paper (Thin Solid Films 297, 241 (1997)), whose abstract reports that the high illumination power densities needed to enable high anodization current densities degrade filter performance; and (iii) the residual effect of the integrated stack on diode electrical behavior (dark current–voltage characteristics, ideality factor, series resistance) relative to the reference diode fabricated identically without the porous-silicon stack, where the 'no significant influence' criterion is that the ideality factor stays within about 0.1–0.2 of the reference value (the margin the 1997 device reported, about 1.2 rising to 1.3–1.4), the dark current–voltage characteristics remain essentially unchanged, and the series resistance is compared on the same reference basis.

References

  1. M. Krüger, M. G. Berger, M. Marso, W. Reetz, T. Eickhoff, R. Loo, L. Vescan, M. Thönissen, H. Lüth, R. Arens-Fischer, S. Hilbrich, W. Theiss, 'Color-Sensitive Si-Photodiode Using Porous Silicon Interference Filters', Japanese Journal of Applied Physics 36, L24–L26 (1997). DOI: 10.1143/jjap.36.l24, https://doi.org/10.1143/jjap.36.l24
  2. M. Krüger, S. Hilbrich, M. Thönissen, D. Scheyen, W. Theiß, H. Lüth, 'Suppression of ageing effects in porous silicon interference filters', Optics Communications 146, 309–315 (1998). DOI: 10.1016/S0030-4018(97)00513-0, https://doi.org/10.1016/S0030-4018(97)00513-0
  3. M. Krüger, M. Marso, M. G. Berger, M. Thönissen, S. Billat, R. Loo, W. Reetz, H. Lüth, S. Hilbrich, R. Arens-Fischer, P. Grosse, 'Color-sensitive photodetector based on porous silicon superlattices', Thin Solid Films 297, 241–244 (1997). DOI: 10.1016/S0040-6090(96)09414-X, https://doi.org/10.1016/S0040-6090(96)09414-X
  4. M. Marso, M. Wolter, R. Arens-Fischer, P. Kordoš, H. Lüth, 'Formation of laterally displaced porous silicon filters using different fabrication methods', ASDAM 2000 — Proceedings of the Third International EuroConference on Advanced Semiconductor Devices and Microsystems, Smolenice, Slovakia, pp. 95–98 (2000). DOI: 10.1109/ASDAM.2000.889458, https://doi.org/10.1109/ASDAM.2000.889458
  5. V. Torres-Costa, R. J. Martín-Palma, J. M. Martínez-Duart, 'All-silicon color-sensitive photodetectors in the visible', Materials Science and Engineering: C 27, 954–956 (2007). DOI: 10.1016/j.msec.2006.06.026, https://doi.org/10.1016/j.msec.2006.06.026
  6. V. Torres-Costa, R. J. Martín-Palma, 'Application of nanostructured porous silicon in the field of optics. A review', Journal of Materials Science 45, 2823–2838 (2010). DOI: 10.1007/s10853-010-4251-8, https://doi.org/10.1007/s10853-010-4251-8
  7. V. Kochergin, H. Föll, 'Novel optical elements made from porous Si', Materials Science and Engineering: R: Reports 52, 93–140 (2006). DOI: 10.1016/j.mser.2006.05.002, https://doi.org/10.1016/j.mser.2006.05.002
  8. V. Kochergin, H. Föll, 'Commercial applications of porous Si: optical filters and components', physica status solidi (c) 4, 1933–1940 (2007). DOI: 10.1002/pssc.200674331, https://doi.org/10.1002/pssc.200674331
  9. S. Zangooie, R. Jansson, H. Arwin, 'Reversible and irreversible control of optical properties of porous silicon superlattices by thermal oxidation, vapor adsorption, and liquid penetration', Journal of Vacuum Science & Technology A 16, 2901–2912 (1998). DOI: 10.1116/1.581438, https://doi.org/10.1116/1.581438
  10. Michael G. Berger, S. Frohnhoff, Ruediger Arens-Fischer, M. Thoenissen, C. Dieker, H. Muender, Hans Luth, Wolfgang Theiss, M. Arntzen, 'Fabrication and properties of multilayer porous silicon filters', Optical Interference Coatings, SPIE Proceedings Vol. 2253, 865 (1994). DOI: 10.1117/12.192162, https://doi.org/10.1117/12.192162
  11. L. A. Balagurov, D. G. Yarkin, G. A. Petrovicheva, E. A. Petrova, A. F. Orlov, S. Ya. Andryushin, 'Highly sensitive porous silicon based photodiode structures', Journal of Applied Physics 82, 4647–4650 (1997). DOI: 10.1063/1.366203, https://doi.org/10.1063/1.366203
  12. Seon-Hwa Park, Dongwoo Seo, Young-You Kim, Ki-Won Lee, 'Organic vapor detection using a color-difference image technique for distributed Bragg reflector structured porous silicon', Sensors and Actuators B: Chemical 147, 775–779 (2010). DOI: 10.1016/j.snb.2010.03.075, https://doi.org/10.1016/j.snb.2010.03.075
  13. Ghenadii Korotcenkov, Nima Naderi, 'Photodetectors Based on Porous Silicon', in Porous Silicon: From Formation to Applications, Volume 3: Optoelectronics, Microelectronics, and Energy Technology Applications, CRC Press / Taylor & Francis, pp. 35–50 (2016). DOI: 10.1201/b19042-4, https://doi.org/10.1201/b19042-4
  14. M. Krüger, M. G. Berger, M. Marso, M. Thönissen, S. Hilbrich, W. Theiß, R. Loo, T. Eickhoff, W. Reetz, P. Grosse, H. Lüth, 'Integration of Porous Silicon Interference Filters in Si-Photodiodes', ESSDERC '96: Proceedings of the 26th European Solid State Device Research Conference, pp. 891–894 (1996); no DOI registered — IEEE Xplore, https://ieeexplore.ieee.org/document/5435938