Noise spectra and dark current investigations in n⁺-p-type Hg₁₋xCdxTe (x ≈ 0.22) photodiodes

The dark current and noise spectra were investigated in Hg₁₋xCdxTe (x≅0.22) photodiodes at zero and low reverse bias voltages. The photodiodes were prepared by boron implantation into LPE films. The 1/f noise is proved to be correlated with tunneling current via the deep defect states in the gap at...

Повний опис

Збережено в:
Бібліографічні деталі
Дата:1999
Автори: Ivasiv, Z.F., Sizov, F.F., Tetyorkin, V.V.
Формат: Стаття
Мова:English
Опубліковано: Інститут фізики напівпровідників імені В.Є. Лашкарьова НАН України 1999
Назва видання:Semiconductor Physics Quantum Electronics & Optoelectronics
Онлайн доступ:http://dspace.nbuv.gov.ua/handle/123456789/119879
Теги: Додати тег
Немає тегів, Будьте першим, хто поставить тег для цього запису!
Назва журналу:Digital Library of Periodicals of National Academy of Sciences of Ukraine
Цитувати:Noise spectra and dark current investigations in n⁺-p-type Hg₁₋xCdxTe (x ≈ 0.22) photodiodes / Z.F. Ivasiv, F.F. Sizov, V.V. Tetyorkin // Semiconductor Physics Quantum Electronics & Optoelectronics. — 1999. — Т. 2, № 3. — С. 21-25. — Бібліогр.: 16 назв. — англ.

Репозитарії

Digital Library of Periodicals of National Academy of Sciences of Ukraine
id irk-123456789-119879
record_format dspace
spelling irk-123456789-1198792017-06-11T03:02:32Z Noise spectra and dark current investigations in n⁺-p-type Hg₁₋xCdxTe (x ≈ 0.22) photodiodes Ivasiv, Z.F. Sizov, F.F. Tetyorkin, V.V. The dark current and noise spectra were investigated in Hg₁₋xCdxTe (x≅0.22) photodiodes at zero and low reverse bias voltages. The photodiodes were prepared by boron implantation into LPE films. The 1/f noise is proved to be correlated with tunneling current via the deep defect states in the gap at low reverse biases U≤0.1 V. In the photodiodes, where the tunneling current is found to be dominating, the 1/f noise is observed up to frequencies 10⁴ Hz. The decrease of tunneling current results in the decrease of 1/f noise. 1999 Article Noise spectra and dark current investigations in n⁺-p-type Hg₁₋xCdxTe (x ≈ 0.22) photodiodes / Z.F. Ivasiv, F.F. Sizov, V.V. Tetyorkin // Semiconductor Physics Quantum Electronics & Optoelectronics. — 1999. — Т. 2, № 3. — С. 21-25. — Бібліогр.: 16 назв. — англ. 1560-8034 PACS: 85.60.D, 07.57.K, 85.60.G, 73.40 http://dspace.nbuv.gov.ua/handle/123456789/119879 en Semiconductor Physics Quantum Electronics & Optoelectronics Інститут фізики напівпровідників імені В.Є. Лашкарьова НАН України
institution Digital Library of Periodicals of National Academy of Sciences of Ukraine
collection DSpace DC
language English
description The dark current and noise spectra were investigated in Hg₁₋xCdxTe (x≅0.22) photodiodes at zero and low reverse bias voltages. The photodiodes were prepared by boron implantation into LPE films. The 1/f noise is proved to be correlated with tunneling current via the deep defect states in the gap at low reverse biases U≤0.1 V. In the photodiodes, where the tunneling current is found to be dominating, the 1/f noise is observed up to frequencies 10⁴ Hz. The decrease of tunneling current results in the decrease of 1/f noise.
format Article
author Ivasiv, Z.F.
Sizov, F.F.
Tetyorkin, V.V.
spellingShingle Ivasiv, Z.F.
Sizov, F.F.
Tetyorkin, V.V.
Noise spectra and dark current investigations in n⁺-p-type Hg₁₋xCdxTe (x ≈ 0.22) photodiodes
Semiconductor Physics Quantum Electronics & Optoelectronics
author_facet Ivasiv, Z.F.
Sizov, F.F.
Tetyorkin, V.V.
author_sort Ivasiv, Z.F.
title Noise spectra and dark current investigations in n⁺-p-type Hg₁₋xCdxTe (x ≈ 0.22) photodiodes
title_short Noise spectra and dark current investigations in n⁺-p-type Hg₁₋xCdxTe (x ≈ 0.22) photodiodes
title_full Noise spectra and dark current investigations in n⁺-p-type Hg₁₋xCdxTe (x ≈ 0.22) photodiodes
title_fullStr Noise spectra and dark current investigations in n⁺-p-type Hg₁₋xCdxTe (x ≈ 0.22) photodiodes
title_full_unstemmed Noise spectra and dark current investigations in n⁺-p-type Hg₁₋xCdxTe (x ≈ 0.22) photodiodes
title_sort noise spectra and dark current investigations in n⁺-p-type hg₁₋xcdxte (x ≈ 0.22) photodiodes
publisher Інститут фізики напівпровідників імені В.Є. Лашкарьова НАН України
publishDate 1999
url http://dspace.nbuv.gov.ua/handle/123456789/119879
citation_txt Noise spectra and dark current investigations in n⁺-p-type Hg₁₋xCdxTe (x ≈ 0.22) photodiodes / Z.F. Ivasiv, F.F. Sizov, V.V. Tetyorkin // Semiconductor Physics Quantum Electronics & Optoelectronics. — 1999. — Т. 2, № 3. — С. 21-25. — Бібліогр.: 16 назв. — англ.
series Semiconductor Physics Quantum Electronics & Optoelectronics
work_keys_str_mv AT ivasivzf noisespectraanddarkcurrentinvestigationsinnptypehg1xcdxtex022photodiodes
AT sizovff noisespectraanddarkcurrentinvestigationsinnptypehg1xcdxtex022photodiodes
AT tetyorkinvv noisespectraanddarkcurrentinvestigationsinnptypehg1xcdxtex022photodiodes
first_indexed 2025-07-08T16:50:06Z
last_indexed 2025-07-08T16:50:06Z
_version_ 1837098275079979008
fulltext 21© 1999, Institute of Semiconductor Physics, National Academy of Sciences of Ukraine Semiconductor Physics, Quantum Electronics & Optoelectronics. 1999. V. 2, N 3. P. 21-25. PACS: 85.60.D, 07.57.K, 85.60.G, 73.40 Noise spectra and dark current investigations in n+-p-type Hg1-xCdxTe (x ≅≅≅≅≅ 0.22) photodiodes Z.F. Ivasiv, F.F. Sizov, V.V. Tetyorkin Institute of Semiconductor Physics, NAS of Ukraine 45, prospect Nauki, 252650 Kyiv, Ukraine Abstract. The dark current and noise spectra were investigated in Hg1-xCdxTe (x≅0.22) photodiodes at zero and low reverse bias voltages. The photodiodes were prepared by boron implantation into LPE films. The 1/f noise is proved to be correlated with tunneling current via the deep defect states in the gap at low reverse biases U≤0.1 V. In the photodiodes, where the tunneling current is found to be dominat- ing, the 1/f noise is observed up to frequencies 104 Hz. The decrease of tunneling current results in the decrease of 1/f noise. Keywords: long wavelength intra-red photodiode, trap assisted tunneling, 1/f noise. Paper received 07.09.99; revised manuscript received 30.09.99; accepted for publication 11.10.99. 1. Introduction Hg1-xCdxTe ternary compounds are recognized to be the most important materials for fabrication of long wavelength infrared (LWIR) photodiodes for 3-5 and 8-12 µm wave- length regions [1]. Ion implantation seems to be the most important method for manufacturing single n+-p-type photodiodes as well as multi-element focal plane arrays [1]. Carrier transport mechanisms in implanted LWIR photo- diodes has been investigated both theoretically and experi- mentally by many authors [1-9]. It has been found that in general case at temperatures close to 77 K total dark cur- rent are composed of several components: bulk diffusion, generation-recombination (GR) current in the depletion re- gion, trap assisted tunneling (TAT) current, band-to-band (BTB) tunneling current and surface leakage current. GR and TAT currents dominate the total dark current at zero and low reverse bias voltage as well as BTB current is domi- nant at large reverse bias voltage. In the photodiodes with improved characteristics dark current is dominated by the diffusion current. Due to the fact that carrier transport mechanisms listed above are influenced by deep defect states in the gap, fitting procedure can be used for estimation of several important parameters of these states such as con- centration Nt, energy Et and capture rates [4-9]. Because of some discrepancies exist in the models used for the fitting calculation the values of above mentioned parameters are changed in a rather wide range (e.g., Nt is changed from ~1012 cm-3 up to ~1017 cm-3 [6,7,8,14]). As to investigations of noise mechanisms in HgCdTe photodiodes the origin of the 1/f noise still seems to be the most important and unsolved problem. Several sources of the 1/f noise have been proposed in application to these photodiodes [10-15]. One of the possible source is lattice defects of different types unintentionally or intentionally introduced during growth of starting material. Fluctuation of deep point defects, shallow etch pits, dislocation multi- plication and clustering are recognized to be possible for 1/f noise source. As shown in [12], 1/f noise in LWIR photo- diodes is a bulk phenomenon associated with defects in the depletion region. However, the states at the interface of HgCdTe and passivating layer can also contribute to 1/f noise [11,13,14]. Finally, the dark current is suspected to be a possible source of 1/f noise, too. In several papers [4, 9] the correlation between 1/f noise and bulk BTB tunneling cur- rent was observed. Quite the contrary, the correlation be- tween 1/f noise and surface BTB tunneling current has been found in the papers published earlier [15]. The role of trap assisted tunneling current at zero and low reverse bias voltages seems to be not investigated yet. The aim of this work is to investigate dark current and 1/f noise in LWIR HgCdTe photodiodes in order to find possible correlation between these phenomena. More than thirty diodes were investigated to obtain statistical results. 2. Experimental results and discussion. The starting materials were undoped Hg1-xCdxTe (x≅0.22) films of p - type conductivity. The films were used for p-n- junction fabrication by boron ion implantation. The hole concentration in the films was close to 1⋅1016 cm-3 at tem- perature 77 K. Z.F. Ivasiv et al.: Noise spectra and dark current investigations... 22 SQO, 2(3), 1999 The dark current in the photodiodes has been measured at 77 K. By differentiation of I-U curves the dynamic re- sistance R characteristics versus voltage bias U were ob- tained. To estimate the cut-off wavelength λc values the photoresponse spectra were measured at 82 K. The noise spectra were investigated in the frequency range 0.07÷ 104 Hz at 77 K using CK4-73 spectrum analyzer. Typical normalized photoresponse spectra of the photodiodes are shown in Fig.1. The cutoff wavelength λc for the photodiodes investigated are shown in Table 1 to- gether with intrinsic carrier concentration, ni, and energy band gap, Eg, values. The last two parameters were calcu- lated using appropriate empirical expressions from [2]. Table 1.Fitting parameters used for modeling calculations 1. Band gap, eV, T=77 K 0.130 2. Cutoff wavelength, λñ, µm, T=82K 10.3±0.2 3. Intrinsic carrier concentration, ni, cm-3, T =77 K 4.87⋅1012 4. Trap density, Nt, cm-3 1013-1016 5. Capture rate, Cp, cm3/s 10-7-10-6 6. Trap energy, Et, eV 0.72 Eg Measured and calculated current-voltage characteristics are shown in Fig.2 (a,b). In calculations several components were taken into account: bulk diffusion current, generation current in the depletion region and TAT current. The for- mulae used for calculations of each current component are briefly discussed in Appendix. In calculation of TAT cur- rent thermal and tunnel transitions from the valence band to deep defect states in the gap followed by tunnel transi- tions to the conduction band were taken into account. The dark current caused by band-to-band tunneling was not taken into account because of investigations were mainly per- formed at zero and low reverse bias voltage (U ≤ 0.1 V). Surface leakage current seems to be negligibly small in these photodiodes (see Appendix). It is seen from Fig.2 that the photodiodes with different type of current-voltage I-U characteristics were investigated. In the photodiode with the lowest values of the reverse cur- rent (curve 1in Fig.2(a)) the tendency to its saturation at reverse bias is clearly seen. With increasing the reverse cur- rent this tendency diminishes (curve 2). Further increasing of the current is accompanied by appearance of the soft breakdown characteristic (curve 3). Sharp soft breakdown characteristics were observed in several photodiodes, Fig.2(b). Typical dependencies of the R0A vs. voltage bias U (R0 is the differential resistance at zero bias, A is the junction area) for the photodiodes investigated are shown in 2 4 6 8 10 12 1.0 0.8 0.6 0.4 0.2 0.0 W ave leng th , m icro n s S en si ti vi ty , a rb .u n . I, x 10 A-6 Fig.1. Normalized photoresponse spectra of implanted n+-p photodiodes Hg1-xCdxTe (x ≅ 0.22) at 82 K. I, x 10 A-6 Fig.2. Measured and calculated current-voltage characteristics of n+-p photodiodes at 77 K. The dots represent measured data and the solid lines represent theory. The fit was obtained for Nt = 1.2⋅1013; 2⋅1014; 4⋅1014 for curves 1-3 (a), respectively. The values Et = 0.72Eg, Cp = 8⋅10-7 and Cn = 0.01Cp were the same for all three curves. To fit experimental and calculated curves in (b) trap density was taken to be Nt =3.2⋅1015; 4.8⋅1015 cm-3, for curve 1 and 2, respectively. Also, the values Et = 0.72Eg, Cp = 1⋅10-7 and Cn = 0.01Cp were the same for these curves. b a Z.F. Ivasiv et al.: Noise spectra and dark current investigations... 23SQO, 2(3), 1999 Fig. 3 (a,b). At zero bias the R0A value less than 1.0 Ω×cm2 has been found to be typical for the photodiodes in which soft breakdown was observed. Note that maximum in R0A vs. U curves is shifted towards zero bias in these photodiodes. The parameters used in calculations I-U and R0A-U characteristics are listed in Table 1. Noise power spectral density measurements were per- formed in the frequency range of 0.7÷104 Hz for the same photodiodes in which the dark current was measured. Typi- cal noise spectra measured at 77 K are shown in Fig. 4 to- gether with the calculated curves. As one can see, each noise curve consist of frequency-dependent as well as frequency- independent parts which can be attributed to 1/f noise and to generation-recombination noise, respectively. The transition region between them is rather narrow and it shifts towards the higher frequency as the dark current increases. From the measured and calculated data shown in Fig.2(a,b) one can conclude the following. The best fit was obtained for energy of deep defect states Et ≅ 0.72Eg (Et is measured from the top of the valence band). This value correlate well with previously observed one Et ≅ 0.75Eg [7] in photodiodes prepared by boron implantation to bulk material. Also this value is in agreement with the results of DLTS measurements in undoped HgCdTe crystals [16]. It seems that these centers are introduced into the bulk of HgCdTe materials during their preparation and can be at- tributed to different states of the Hg vacancies [7]. Some contradiction exists in the concentration of deep defect centers with previously published values Nt= = 0.1-10 NA [7], where NA is the shallow acceptor concentra- tion (NA is approximately equal to the hole concentration). The values of Nt obtained in present work range from ~1013 cm-3 to ~1016 cm-3 in photodiodes with NA ≅ 1016 cm-3. This contradiction may be explained as follows. The con- centration of shallow acceptors as well as deep defects in HgCdTe implanted photodiodes is not uniform in the re- gion adjacent to n+-p-junction. Nemirovsky et al. [4] have found that n+-p-junctions can be formed with low concen- tration of shallow acceptors near the junction. Since the measurements were performed at zero and low reverse bias voltage the concentration Nt derived from the fitting proce- dure can be smaller than in the bulk (e.g., in implanted Hg1-xCdxTe (x≅0.22) photodiodes investigated previously [7] Nt reached the bulk value at reverse bias U≅0.6V). The capture rates found at this study Cp ≈10-7÷10-6 cm3/s and Cn = (0.1÷0.01)Cp differ much from the result previ- ously obtained for Cp= (10-10÷10-9) and Cn = (10÷100) Cp [8,14]. It should be pointed out that in these studies deep defect centers in p-HgCdTe material were assumed to be donor-like type. However, in present investigation the best fit was obtained for acceptor-like centers in the gap. This result are in accordance with the results of comprehensive study performed by Rosenfeld and Bahir [7]. The correlation between 1/f noise and tunneling dark current was previously observed by Nemirovsky et. al. [4] at rather high reverse bias voltages U ≥ 200 mV. In present investigation the correlation is revealed at low reverse bias U ≤ 100 mV. The result obtained seems to be important because of HgCdTe photodiodes are mainly used at zero and low reverse bias voltage operating conditions [1]. The calculation of the 1/f noise was performed by the formula (A6) [4]. To fit experimental and calculated data the con- stant α was assumed to lie in the range from 10-8 to 10-7 . These values agree with the value α = 10-7 found in [4]. Because of the increase of the dark current in the investi- gated photodiodes is related with the tunneling through deep defect states in the gap, one can assume the origin of both phenomena to be the same. Also it should be pointed out that no correlation has been observed between GR current in the depletion region and 1/f noise. The mechanism of 1/f noise in diodes with dominant trap-assisted tunneling current is unknown yet. One possi- bility of its origin has been pointed out in [14]. The funda- mental fluctuations in the process rates can exist due to the infrared divergent coupling of free carriers to low-frequency photons and other infraquanta. R A , cm Ω 0 2 Fig.3. Typical resistance-area versus bias voltage characteristics for several photodiodes at 77 K. The designations are the same as in Fig.2a (curve 1) and Fig.2b (curves 1 and 2). a b R A , cm Ω 0 2 Z.F. Ivasiv et al.: Noise spectra and dark current investigations... 24 SQO, 2(3), 1999 Conclusion Dark current have been measured and calculated in n+-p- type Hg1-xCdxTe (x ≅ 0.22) photodiodes at 77 K. It has been found that generation in the depletion region and tunneling through deep defect states in the gap are dominant carrier transport mechanisms at low reverse bias voltage. From the fitting procedure the density of the tunneling centers Nt = 1013 –1015 cm-3 and their energy E t ≅ 0.72Eg has been estimated. The 1/f noise measured at low reverse bias can be attributed to the tunneling dark current through deep defect states. References 1. C.T.Elliot and N.T.Gordon, Infrared Detectors, in Handbook on Semi- conductors, Vol.4, pp.841-936, Ed. by C.Hilsum, North-Holland, Amsterdam (1993). 2. J.R.Lowney, D.G.Seiler, C.L.Littler and I.T.Yoon, Intrinsic carrier concentration of narrow gap mercury cadmium telluride// J.Appl.Phys. 71, pp.1253-1258 (1992). 3. J.Y.Wang, Effect of trap tunneling on the performance of long-wave- length Hg1-xCdxTe photodiodes // IEEE Trans. on Electron Devices ED - 27 (1), pp.48-57 (1980). 4. Y.Nemirovsky, D.Rosenfeld, R.Adar, and A.Kornfeld, Tunneling and dark currents in HgCdTe photodiodes // J.Vac.Sci.Technology A7 (2), pp.528-535 (1989). 5. N.L.Bazhenov, S.I.Gasanov and V.I.Ivanov-Omskii, Excess currents in narrow-gap Hg1-xCdxTe p-n junctions // Infrared Phys. 34 (1), pp.37-41 (1993). 6. Akira Ajisawa and Naoki Oda, Improvement in HgCdTe diode char- acteristics by low temperature post-implantation annealing // J.Electron. Materials 24 (9), pp.1105-1111 (1995). 7. D.Rosenfeld and G.Bahir, A model for the trap-assisted tunneling mechanism in diffused n-p and implanted n+-p HgCdTe photodiodes // IEEE Trans. on Electron Dev. 39 (7), pp.1638-1645 (1992). 8. Y.Nemirovsky, R.Fastow, M.Meyassed, and A.Unikovsky, Trapping effects in HgCdTe //J.Vac.Sci.Technol. B.9 (3), pp.1829-1839 (1991). 9. R.E.Dewames, J.G.Pasko, E.S.Yao, A.H.B.Vanderwyck, and G.M.Williams, Dark current generation mechanisms and spectral noise carrent in long-wavelength infrared photodiodes // J.Vac.Sci.Technol. A6, p.2655 (1988). 10. Tian, Ji Zhongyun, Lei Ruoxi, Gu Shengqiong, Boqi, Measurements of noise of photovoltaic HgCdTe and InSb infrared detector // Proc.SPIE 2894, pp.174-179 (1996). 11. Wenmu He and Zeynep Celik-Butler, Temperature dependence of 1/f noise in Hg1-xCdxTe MIS infrared detectors // Trans. Electron Dev. 42 (1), pp.160-165 (1995). 12. R.Schiebel, D.Bartolomew, M.Bevan, R.S.List, and M.Ohlson, Noise and material defects in HgCdTe diodes // Journal of Elecrtonic Ma- terials 24 (9), pp.1299-1203 (1995). 13. S.R.Babu, K.Hu, Manthripragada et al., Improved HgCdTe detectors with novel anti-reflection coating //Proc. SPIE 2816, pp.84-89 (1995). 14. Wenmu He and Zeynep Celik-Butler, 1/f noise and dark current com- ponents in HgCdTe MIS infrared detectors // Solid - State Electron- ics 1 (1), pp.127-132 (1996). 15. W.A.Radford, and C.E.Jones, 1/f noise in ion-implanted and double- layer epitaxial HgCdTe photodiodes // J.Vac.Sci.Technol. A3 (1), pp.183-188 (1985). 16. C.E.Jones, V.Nair, and D.L.Polla, Generation-recombination centers in p-type HgCdTe // Appl. Phys. Lett. 39 (3), pp. 248-250 (1981). Appendix The dark current in the photodiodes investigated can be given by the sum of several components, namely surfBTBTATGRD JJJJJJ ++++= , (A1) Fig. 4. Measured (dots) and calculated (solid lines) noise spectra for the photodiodes with R0A = 1.0 Ω⋅cm2 (1) and R0A = 0.3 Ω⋅cm2 (2) at 77 K. The reference shot noise level (2qJ)1/2 at reverse bias 10 mV is 1.7⋅10-13 and 7.92⋅10-13 A/Hz-1/2, respectively. U , V /H z 0 1000 10000100 f , H z 10 U , V /H z c 1000100 f , H z 101 1 2 Z.F. Ivasiv et al.: Noise spectra and dark current investigations... 25SQO, 2(3), 1999 were JD is diffusion current, JGR is the depletion genera- tion-recombination current, JTAT is the trap-assisted cur- rent, JBTB is the band-to-band tunneling current and Jsurf is the surface leakage current. For asymmetric junctions the first two components are given by expressions: 2 1 0 2       ⋅⋅= n ni D q Tk p nq J τ µ , (A2) 1010 2 np Wqn J pn i RG ττ + = , (A3) where p0 is the hole concentration, µn and τn are the elec- tron mobility and the minority carrier lifetime in the bulk, where ( ) 1 0 −= tnn Ncτ , ( ) 1 0 −= tpp Ncτ , cn and cp are the capture coefficients for electrons and holes, respectively, Nt is the trap density, p1 = Nv⋅exp(-Et/kT), n1 = Nc⋅exp(-Eg+ +Et/kT), Nv and Nc are the effective density of states in the valence and conduction bands, W is the depletion region width, Et is the trap energy measured from the valence band edge. The second and third terms in (A1) were intensively discussed earlier [1]. In this work, the appropriate equa- tions for JTAT current were taken from [7]. The trap-assisted tunneling current is given by         + + = ccvvp tTAT nnpc WqNJ ωω 11 1 , (A4) were ωcNc and ωvNv are the tunneling rates . The surface leakage current caused by the fast surface states is given by [14] 2 Snq J i surf = , (A5) where S is the surface recombination velocity. In properly pre- pared photodiodes the value of S does not exceed 100 cm/s [14]. As one can see from the formula (5), the contribution of the surface leakage current to the dark current can be neglected in our case. In order this current to be dominant, the value S should be higher than 103 cm/s. However, so high values of S were not observed in the diodes investi- gated. The calculation of the 1/f noise current was carried out with Jn =     = f JTATα , (A6) where JTAT current was calculated using (A4).