(DLTS) on Colloidal-Synthesized Nanocrystal Solids AWS
Supporting Information - Deep Level Transient Spectroscopy (DLTS) on Colloidal-Synthesized Nanocrystal Solids Deniz Bozyigit, Michael Jakob, Olesya Yarema, and Vanessa Wood∗ Laboratory for Nanoelectronics, Department of Information Technology and Electrical Engineering, Eidgenoessische Technische Hochschule Zurich E-mail: [email protected]
NC Synthesis and Characterization The PbS NCs were made following Hines et al. 1 Octadecene (75 ml), lead (II) oxide (1.8 g), and oleic acid (5 ml) were mixed in the 250 ml four-neck flask. The mixture was dried under vacuum at 150 ◦ C for 2 hours. During this time, the reaction solution slowly became colorless and transparent, indicating in situ formation of Lead (II) oleate. Then the Lead (II) oleate solution was backfilled with the nitrogen. At a temperature of 150 ◦ C, 40 ml of 0.1 M solution of hexamethyldisilathiane in pre-purified octadecene was swiftly added. The reaction mixture was allowed to cool down to 100◦ C, after which this temperature was kept for another 5 min and then quenched with water bath. The PbS NCs were purified by a standard solvent/nonsolvent procedure, using hexane and acetone. The washing cycle was repeated 3 times. Lead (II) oxide (99.999 %) is purchased from Strem Chemicals; hexamethyldisilathiane (purum grade), octadecene (90 %, techn.), oleic acid (90 %), hexane, acetone, ethanol, toluene are from Sigma Aldrich. All chemicals were used as received. We characterize the NCs by transmission electron microscopy (TEM) (Fig. (1a)) and find a diameter of 3 nm. Optical absorption and photoluminescence of the NCs in hexane are shown in Fig. (1b). ∗ To
whom correspondence should be addressed
aL
Abs. & PL int. @a.u.D
2 1.0 0.8 0.6 0.4 0.2 0.0
bL
600
Λ @nmD
800
1000
1200
Figure 1: a) TEM image of 3 nm PbS nanocrystals. b) Absorption and photoluminescence spectra of PbS nanocrystals in solution.
Device Fabrication All fabrication and characterization steps were performed in air unless otherwise noted. Devices were fabricated on cleaned, pre-patterned indium tin oxide (ITO) glass substrates from Thin Film technologies. The NC active layers were deposited using a multi-step, dip-coating process. A home-built dip-coater sequentially immersed the patterned ITO substrate in a bath of NCs with concentration of 5 mg/ml in hexane, a bath of 2 mM ethanedithiol (EDT) in acetonitrile, and a rinse bath of acetonitrile. Wait times between each step were selected to enable sufficient drying of the films between each immersion step. Repeating the processes 30 times resulted in a NC film thickness of d = 67 nm, determined by AFM. The top electrode consists of 1.5 nm of lithium fluoride (LiF), 100 nm aluminium, and 300 nm of silver thermally evaporated through a shadow mask. The additional layer of silver provides a robust and conductive electrode, which facilitates measurements in the cryostat. An active device active area of 0.02 cm2 is defined by the overlap between the ITO and Al electrodes.
3
DLTS Measurement Setup We utilized a current-based DLTS method 2 (often referred to as Q-DLTS) based on the measurement of current transients. The current transient measurements were performed in a Janis ST-500 cryostat. The devices were kept in the dark and biased using an arbitrary waveform generator (Agilent 33522A). A broadband low current amplifier (FEMTO DHPCA-100) connected to an oscilloscope (R&S RTM1054) measured the current density through the device. We applied a reverse bias (VD = −0.7 V) to the device and a voltage pulse every 10 ms for 1 ms, while simultaneously measuring the current density through the device (JD (t)). During the forward section of the pulse (VD = 0 V), traps in the SCR can be populated. At time t = 0, when the reverse bias is again applied, the populated traps can emit the captured charge carriers so as to thermalize to the new bias condition. This carrier emission process is observed as a current transient in the device current signal. In addition to the current due to trap emission (JE ), in which we are interested, the capacitive displacement current (Jcap ) and the reverse leakage current (Jleak ), also contribute to the total measured device current (JD ):
JD (t) = JE (t) + Jcap (t) + Jleak .
(1)
To eliminate the contribution of Jleak , we subtracted the baseline of the measured current (at D t = 8 ms). Further, we estimated the displacement current by Jcap,est (t) = C1 MHz dV dt . We inde-
pendently measured the high frequency capacitance C1 MHz = 225 nF/cm2 using an impedance analyzer (SOLARTRON MODULAB MTS). Based on this estimation, we chose a starting time t0 for the trap emission transients such that Jcap,est (t0 ) JE (t0 ).
Capacitance DLTS Measurements Before turning to Q-DLTS measurements, we attempted to perform conventional capacitancebased DLTS (C-DLTS) measurements on our devices. In this section, we want to briefly explain why such measurements cannot be applied in a useful fashion to the type of NC devices investigated here.
4
8
8 CHf,VL @nFD
Vd = 8-0.4V,0V,+0.4V
NC Synthesis and Characterization The PbS NCs were made following Hines et al. 1 Octadecene (75 ml), lead (II) oxide (1.8 g), and oleic acid (5 ml) were mixed in the 250 ml four-neck flask. The mixture was dried under vacuum at 150 ◦ C for 2 hours. During this time, the reaction solution slowly became colorless and transparent, indicating in situ formation of Lead (II) oleate. Then the Lead (II) oleate solution was backfilled with the nitrogen. At a temperature of 150 ◦ C, 40 ml of 0.1 M solution of hexamethyldisilathiane in pre-purified octadecene was swiftly added. The reaction mixture was allowed to cool down to 100◦ C, after which this temperature was kept for another 5 min and then quenched with water bath. The PbS NCs were purified by a standard solvent/nonsolvent procedure, using hexane and acetone. The washing cycle was repeated 3 times. Lead (II) oxide (99.999 %) is purchased from Strem Chemicals; hexamethyldisilathiane (purum grade), octadecene (90 %, techn.), oleic acid (90 %), hexane, acetone, ethanol, toluene are from Sigma Aldrich. All chemicals were used as received. We characterize the NCs by transmission electron microscopy (TEM) (Fig. (1a)) and find a diameter of 3 nm. Optical absorption and photoluminescence of the NCs in hexane are shown in Fig. (1b). ∗ To
whom correspondence should be addressed
aL
Abs. & PL int. @a.u.D
2 1.0 0.8 0.6 0.4 0.2 0.0
bL
600
Λ @nmD
800
1000
1200
Figure 1: a) TEM image of 3 nm PbS nanocrystals. b) Absorption and photoluminescence spectra of PbS nanocrystals in solution.
Device Fabrication All fabrication and characterization steps were performed in air unless otherwise noted. Devices were fabricated on cleaned, pre-patterned indium tin oxide (ITO) glass substrates from Thin Film technologies. The NC active layers were deposited using a multi-step, dip-coating process. A home-built dip-coater sequentially immersed the patterned ITO substrate in a bath of NCs with concentration of 5 mg/ml in hexane, a bath of 2 mM ethanedithiol (EDT) in acetonitrile, and a rinse bath of acetonitrile. Wait times between each step were selected to enable sufficient drying of the films between each immersion step. Repeating the processes 30 times resulted in a NC film thickness of d = 67 nm, determined by AFM. The top electrode consists of 1.5 nm of lithium fluoride (LiF), 100 nm aluminium, and 300 nm of silver thermally evaporated through a shadow mask. The additional layer of silver provides a robust and conductive electrode, which facilitates measurements in the cryostat. An active device active area of 0.02 cm2 is defined by the overlap between the ITO and Al electrodes.
3
DLTS Measurement Setup We utilized a current-based DLTS method 2 (often referred to as Q-DLTS) based on the measurement of current transients. The current transient measurements were performed in a Janis ST-500 cryostat. The devices were kept in the dark and biased using an arbitrary waveform generator (Agilent 33522A). A broadband low current amplifier (FEMTO DHPCA-100) connected to an oscilloscope (R&S RTM1054) measured the current density through the device. We applied a reverse bias (VD = −0.7 V) to the device and a voltage pulse every 10 ms for 1 ms, while simultaneously measuring the current density through the device (JD (t)). During the forward section of the pulse (VD = 0 V), traps in the SCR can be populated. At time t = 0, when the reverse bias is again applied, the populated traps can emit the captured charge carriers so as to thermalize to the new bias condition. This carrier emission process is observed as a current transient in the device current signal. In addition to the current due to trap emission (JE ), in which we are interested, the capacitive displacement current (Jcap ) and the reverse leakage current (Jleak ), also contribute to the total measured device current (JD ):
JD (t) = JE (t) + Jcap (t) + Jleak .
(1)
To eliminate the contribution of Jleak , we subtracted the baseline of the measured current (at D t = 8 ms). Further, we estimated the displacement current by Jcap,est (t) = C1 MHz dV dt . We inde-
pendently measured the high frequency capacitance C1 MHz = 225 nF/cm2 using an impedance analyzer (SOLARTRON MODULAB MTS). Based on this estimation, we chose a starting time t0 for the trap emission transients such that Jcap,est (t0 ) JE (t0 ).
Capacitance DLTS Measurements Before turning to Q-DLTS measurements, we attempted to perform conventional capacitancebased DLTS (C-DLTS) measurements on our devices. In this section, we want to briefly explain why such measurements cannot be applied in a useful fashion to the type of NC devices investigated here.
4
8
8 CHf,VL @nFD
Vd = 8-0.4V,0V,+0.4V
