Semic-Lab Semiconductors Lab Dep. of Eng. Sciences Univ. of Patras Univ. of Patras Modulated photocurrent as a powerful method to reveal transport by the majority carriers of disordered semiconductors and to resolve all the kinds of probed states Maura Pomoni, Athina Giannopoulou and Panagiotis Kounavis Department of Engineering Sciences, University of Patras, 26504 Patra, Greece Semic-Lab Semiconductors Lab Dep. of Eng. Sciences Univ. of Patras Univ. of Patras Possible contribution from both carriers complicates interpretation of photoconductivity measurements This limitation is overcome in the modulated photocurrent (MPC) Specific features in the MPC spectra can be used to reveal whether the transport of the majority carriers dominates In this case, a DOS spectroscopy based on a general formula can be used to evaluate the DOS parameters of the various species of states with which the majority carriers interact Semic-Lab The essential parameter of the MPC is the Univ. of Patras Semiconductors Lab Dep. of Eng. Sciences Univ. of Patras out of phase Y signal Out of phase MPC Y signal Y eGac Experimental setup Phase shift sin ac Mobility of the majority carriers Modulated light Generation rate N ( E ) Phase shift LED Modulated + bias light Φ amplitude of MPC Osciloscope Measured iac reference signal Φ iac Lock-In DOS Amplifier electrodes Modulated photoconductivity Vdc a-Si:H The analysis have shown that Semic-Lab Semiconductors Lab Dep. of Eng. Sciences Univ. of Patras Univ. of Patras nac nin jnout pac pin jpout Gac X n jYn Gac X p jYp iac A ( nnout p pout ) ( nnin p pin ) 2 n nout p pout tan n nin p pin 2 1/ 2 Y eGac Y Yn or sin ac Y Yp Y signal is related to the gap states with which the electrons and holes interact and contribute to Yn , Yp Univ. of Patras Semic-Lab Semiconductors Lab Dep. of Eng. Sciences Univ. of Patras Electrons the majority carriers n>>p The states contributing Yn High Frequency (HF) regime Low Frequency (LF) regime ti ncni pcip ti ncni pcip Yn Y1n Y2 n Y3n 10 17 10 14 10 11 EF Etp Etn E ωn holes electrons 0.0 -3 20 -1 10 DOS (eV cm ) -1 -3 DOS (eV cm ) Yn Y1n Y2n Y3n Trapping & detrapping Deep trapping 0.5 1.0 1.5 Energy (eV) Y1n>>Y2n + Y3n and so Yn=Y1n 10 20 10 17 EF Etp holes 10 14 10 11 0.0 Etn Eωn electrons 0.5 1.0 1.5 Energy (eV) In some cases Y2n + Y3n=0 and so Yn=Y1n Semic-Lab Semiconductors Lab The states contributing Yp Dep. of Eng. Sciences Univ. of Patras Univ. of Patras Holes the minority carriers High Frequency (HF) regime Low Frequency (LF) regime ti ncni pcip ti ncni pcip 10 17 10 14 10 11 Etp EF Etn -3 Eωp -1 10 20 Yp Y1 p DOS (eV cm ) -1 -3 DOS (eV cm ) Yp Y1 p Y2 p Y3 p Y1 p holes 0.0 0.5 1.0 1.5 Energy (eV) 10 20 10 17 10 14 Etp EF Etn Eωp 11 10 0.0 Y1p>>Y2p+Y3p and so Yp=Y1p holes 0.5 1.0 1.5 Energy (eV) Yn Y1n H ( , )c D ( En )kT =1/τωn Effective trapping rate of electrons 2 n>>p 2 c t nc pc i t i n 10 10 1 HF regime ωt 9 10 Ep Etp Etn En v ωt c The so-called H function ti H ( , ) 1 arctan( ) 1 1/τωp i t 8 1/τωn 10 7 do not reflect the DOS 1/ 2 ti ncni pcip 10 Yn, Yp Semiconductors Lab Dep. of Eng. Sciences Univ. of Patras Eωn 0.5 1.0 1.5 Energy (eV) 1/τ (s ) i t 16 -1 H ( , ) 0 E i kT ln 2 i t Etn Eωp i p 2 EF Etp Etp Ep Etn En i t c 18 -1 LF regime c n trapping =1/τωp Effective rate of holes H ( , tv )c vp D v ( Ep )kT -3 Yp Y1 p DOS (eV cm ) Univ. of Patras Semic-Lab ω<<ωt i ω>>ωt LF -3 -1 1 3 Yn=1/τωn Reflects the DOS of the CB side Yp=1/τωp Reflects the DOS of the VB side i HF 5 10 10 10 10 10 ω (rad/s) 2 Univ. of Patras μpτωp= μnτωn 1 / p Y 1 / n -1 Y (s ) 10 8 Y μn=μp 10 -1 9 1 3 10 10 10 ω (rad/s) 1/τωn Mixed contributions From both carriers 5 -1 Y (s ) 10 10 9 10 8 10 1 3 10 10 10 ω (rad/s) 10 0 μpτωp<< μnτωn 2 4 1.0 E (eV) 10 10 10 10 ω (rad/s) Semic-Lab Semiconductors Lab Dep. of Eng. Sciences Univ. of Patras 1.5 Comparable densities below and above EF EF is abοve midgap so that electrons the majority carriers n>>p A DOS spectroscopy is impossible How can we know whether the majority carriers dominate μp<<μn 1 / p Y 1 / ωn Y μn=10μp -2 0.5 Minority carriers (holes) dominate 5 1/τωn 1/τωp 7 D (E) EF 1 / p Y 1 / n Y μp=10μn -1 0.0 μpτωp>> μnτωn -1 Y (s ) 10 10 17 c v D (E) μp>>μn 10 1/τ ωp 8 -1 μp=μn 1/τωp1/τωn 10 19 -3 9 21 DOS (cm eV ) 10 DOS model 10 6 A DOS spectroscopy can be achieved Majority carriers (electrons) dominate ? Majority carriers (electrons) dominate Y signal 10 Bias light dependence Y0 8 2 Y Two bias light levels at ωtΗ=ωtc=ncnc 0 2 Y moderate bias ω (rad/s) 1 Y H ( , tc )cnc D c ( En )kT n 2 Y0 weak bias near dark equilibrium For tc c c H ( , tc ) 1 Y0 Y signal drops by a factor of 2 μn=10μp 10 2 H ( , tc ) 1 / 2 c 7 10 1/τωn 10 n>>p H ωt ωt -1 Y (s ) 10 9 Semiconductors Lab Dep. of Eng. Sciences Univ. of Patras cn D ( En )kT 10 4 This can be used to determine the capture coefficient Y follows 1/τωn c c n etH n p Normalized Y/Y0 spectrum 1.0 Y H ( , )Y0 c t Y H ( , tH ) Y0 If the majority carriers dominate and Y signal follows 1/τωn Y/Y0 (r.u.) Univ. of Patras Semic-Lab H 0.5 0.0 Y/Y0 μnτωn>>μpτωp 0.1 1 H 10 100 ω/ωt (r.u.) H the normalized Y/Y0 ratio follows the universal H function H ( , H ) 1 2 arctan( t ) t v …providing that the capture coefficient cnv of the states below EF for the cn Ce c 1 majority carriers is much lower than that cnc of the states above EF cn Y Y1n Y2 n Y3n Y1n cnv Ce c 1 cn because 7 10 -1 10 -3 1 -1 Ce 0.1 1 2 8 DOS (eV cm ) -1 Y (s ) 1/τωn 10 n Y2n Y3n 0 Y1n Y2 n Y3n 9 10 1 μn=10μp 3 10 10 10 ω (rad/s) Y Y1n 5 1 n 10 20 10 17 10 14 10 11 17 0.0 Semiconductors Lab Dep. of Eng. Sciences Univ. of Patras D (E) EF 0.5 1.0 E (eV) 1.5 Recombination through the states below EF increases EF Etp 1.0 Etn Eωn holes 0.0 electrons 0.5 1.0 1.5 Energy (eV) Y H ( , tH ) Y0 The decay of Y signal in the LF regime is steeper than 1/τωn In general, 10 Semic-Lab c v D (E) Y/Y0 (r.u.) cnv cnc -1 μpτωp<< μnτωn 10 19 -3 Univ. of Patras 10 21 DOS (cm eV ) Majority carriers (electrons) dominate Y signal, but Y<1/τωn cnv cnc For H 0.8 Ce 0.6 0.4 1 μn=10μp 0.1 1 2 10 100 H ω/ωt (r.u.) The normalized Y/Y0 spectrum is below the universal spectrum of H function for Ce=cnv/cnc≥1 if Y differs from the 1/τωn the normalized Y/Y0 ratio does not follow the universal H function 10 Y0 For H t H Y 8 Y H ( , tH ) Y0 μp=10μn 10 0 2 10 10 ω (rad/s) 4 -1 Y (s ) μpτωp>> μnτωn 10 Y Y0 H ωt Y Y H ( , tH ) Y0 μn=μp 10 0 1.0 0.5 2 H 0.1 1 H 10 100 ω/ωt (r.u.) 2 8 μpτωp= μnτωn 0.0 Y0 Minority carriers (holes) dominate μpτωp>> μnτωn Y Y0 ωt -1 10 Semiconductors Lab Dep. of Eng. Sciences Univ. of Patras Y does not follow 1/τωn 2 Y (s ) μpτωp= μnτωn Bias light dependence 9 Y/Y0 (r.u.) Univ. of Patras Mixed contributions from electrons and holes Semic-Lab Majority carriers (electrons) do not dominate Y signal 4 10 10 ω (rad/s) 10 6 the normalized Y/Y0 ratio is above the universal H function for tH DOS spectroscopy Y/Y0 (r.u.) Univ. of Patras Semic-Lab 1.0 If the normalized Y/Y0 ratio follows H function Semiconductors Lab Dep. of Eng. Sciences Univ. of Patras H 0.5 0.0 Y/Y0 μnτωn>>μpτωp 0.1 1 H ω/ωt (r.u.) 10 9 10 8 Y0 2 Y 2Y c This formula can be used for a DOS spectroscopy LF regime 4 ωt -3 -1 DOS (cm eV ) -3 -1 DOS (cm eV ) 10 10 10 15 introduced DOS 1.0 1.1 1.2 1.3 E (eV) 1.4 10 -1 1 10 ω (rad/s) 4 etH n p Alternatively ωtc can be obtained from the DOS in the frequency regime L 10 10 ω (rad/s) 10 and Y signal drops by a factor of 2 c (b) introduced DOS 10 μn=10μp 2 c n c ωt 15 0 H ( , tc ) 1 / 2 at ωtΗ=ωtc=ncnc 16 16 7 10 1/τωn 10 the capture coefficient is obtained coefficient from H ωt ωt -1 Y (s ) The majority carriers (electrons) dominate 1 Y H ( , tc )cnc D c ( En )kT & Y follows 1/τωn n 2 p D c ( En ) etH H ( , tH )kT 10 100 at ωtL/4 =ωtc/4 3 5 10 is the onset of LF regime (plateau) Semic-Lab Experimental spectra of a-As2Se3 The majority carriers (holes) dominate and Y signal follows 1/τωp Univ. of Patras DOS spectroscopy D v ( Ep ) 10 ωt 8 10 1 and Y signal drops by a factor of 2 H Y 2 10 10 10 3 10 4 10 5 c ω (rad/s) v p p e H t p 10 18 10 17 10 1.0 ωt 1 10 -1 10 0 ω/ωt 10 H Neutral centers (b) 1 10 2 (r.u.) the normalized Y/Y0 ratio follows the universal H function (a) 3 10 5 (b) -3 H -1 0.5 H 10 ω (rad/s) Capture radius 2.8 Ǻ DOS (eV cm ) Y/Y 0 (r.u.) 19 -3 10 2 -1 10 p etH H ( , tH )kT 2Y at ωtΗ=ωtc=ncnc (a) Y0 9 DOS (eV cm ) -1 Y (s ) 10 Semiconductors Lab Dep. of Eng. Sciences Univ. of Patras 10 18 10 17 0.5 0.6 Eω-EV (eV) 0.7 Exponential dependence (valence band-tail) Various species of states Semic-Lab Semiconductors Lab Y Univ. of Patras 1 n 2 H ( , tc )cnc D c ( E cn )kT -1 -3 0 v D (E) 18 10 10 c EF D (E) 16 10 Y (r.u.) c n DOS (cm eV ) c 100c hc n 20 10 hc D (E) 14 10 0.0 0.5 1.0 1.5 Energy (eV) nc H t hc t c hc n hc n etH n p 9 10 -1 Y (s ) From the decay of Y signal by the factor of 2 ωtH is determined ωt hc ωt H 8 10 7 10 2 H ( , thc )cnhc D hc ( E hcn )kT Experimental spectra of a-Si:H DOS model Additional states having a 100 times higher capture coefficient Dep. of Eng. Sciences Univ. of Patras hc D (E) ccc D D(E) (E) D (E) 10 -1 10 -2 Y0 2 ωt H the experimental Y signal follows 1/τωn (a) 10 0 6 10 -1 1 3 5 7 10 10 10 10 10 ω (rad/s) c hc n 2 4 6 10 10 10 ω (rad/s) H hc hc From t t ncn etH n p 10 6 cm 3 / s The highest capture coefficient Various species of states DOS spectroscopy Semiconductors Lab p D ( En ) etH H ( , tH )kT 2Y c Univ. of Patras -3 250K 300K hc D (E) 14 1.0 1.2 1.4 Energy (eV) 1.6 300K 15 10 hc D (E) 14 10 tL tc ncnc c c n 10 -1 etL n p 10 1 3 10 10 ω (rad/s) -3 -1 t 10 D (E) cnc 2 10 9 cm3 / s Normal db’s 10 HF hc D (E) 15 10 0.7 0.6 0.5 0.4 10 ωt 17 cnhc 10 6 cm3 / s db’s with a Si-H back bond 0.3 EC-E (eV) DOS (eV cm ) ωt c 4 ω lc D (E) 16 EFn lc 16 L -1 is the onset of LF (plateau) 10 -1 170K -3 at ωtL/4 =ωtc/4 200K 16 HF DOS (cm eV ) Dhc(Eωn) c 10 DOS (eV cm ) -3 Dc(Eωn) LF LF 17 D (E) 10 a-Si:H 18 -1 DOS (cm eV ) Vertical line the signature of various species of states Dep. of Eng. Sciences Univ. of Patras Provides the DOS of both species of states model 10 Semic-Lab or a three center Si-H-Si bond L From tL tc ncnc 10 10 5 The states with the lowest capture coefficient 16 ωt c c n hc D (E) 15 10 H 0 2 10 10 4 ω (rad/s) 10 6 etL n p 2 10 9 cm 3 / s Experimental spectra from the literature where the majority carriers do not dominate Semic-Lab Semiconductors Lab Dep. of Eng. Sciences Univ. of Patras U. of Patras μc-Si:H a-Si:H lightly p-type doped from MPC measurements of Bruggemann J Mat. Sc.14, 629 (2003) from the MPC measurements of Kleider & Longeaud Sol. St. Phen.44&46 596 (1995) Y0 2 Y (r.u.) ωt Y Y does not follow H -1 10 1 3 4 10 10 ω (rad/s) 10 5 Y 2 H ( , )c D ( E n )kT c t c n c c 10 3 10 2 10 1 10 0 Bias light dependence 10 3 10 4 10 5 10 6 ω (rad/s) H (b) Y/Y 0 (r.u.) The normalized Y/Y0 ratio does not follow the universal H function 2 10 1/τωn at lowest ω 0 153 K 183 K 0.1 10 -1 0 1 2 10 10 10 H ω/ωt (r.u.) 10 Mixed contributions from electrons and holes 1 / p Y 1 / p Mixed contributions from electrons and holes reasonable for the the lightly p-type doped material A DOS spectroscopy is impossible Y/Y 0 (r. u.) Y does not follow 1/τωn Y signal exponential dependence 0 Y (r. u.) 10 H -1 10 -2 10 -2 10 -1 10 0 1 10 10 H ω/ωt The normalized Y/Y0 ratio at lowest ω does not follow the universal H function Conclusions Semic-Lab Semiconductors Lab Dep. of Eng. Sciences Univ. of Patras U. of Patras If Y signal follows the universal H function around each ωti. the transport of the majority carriers dominates giving the highest mobility effective trapping time Y signal follows the effective trapping rate of the majority carriers into the probed states. HF A DOS spectroscopy using a general formula gives The states with the highest capture coefficient The states with the lowest LF capture coefficient If the Y signal deviates from the universal frequency dependence of H function, then there are possible contributions from both carriers. The applicability of our analysis was demonstrated in a-As2Se3, undoped and lightly p-doped a-Si:H samples and μc-Si:H.
© Copyright 2026 Paperzz