Reversible Electrochemistry of Mercury Chalcogenide Colloidal Quantum Dot Films Menglu Chen and Philippe Guyot-Sionnest* James Franck Institute, The University of Chicago, 929 East 57th Street, Chicago, Illinois 60637, United States S Supporting Information *
ABSTRACT: The absolute positions of the energy levels of colloidal quantum dots of Hg(S, Se, Te), which are of interest as mid-infrared materials, are determined by electrochemistry. The bulk valence bands are at −5.85, −5.50, and −4.77 eV (±0.05 eV) for zinc-blend HgS, HgSe, HgTe, respectively, in the same order as the anions porbital energies. The conduction bands are conversely at −5.20, −5.50, and −4.77 eV. The stable ambient n-doping of Hg(S, Se) quantum dots compared to HgTe arises because the conduction band is sufficiently lower than the measured environment Fermi level of ∼ −4.7 eV to allow for n-doping for HgS and HgSe quantum dots even with significant electron confinement. The position of the Fermi level and the quantum dots states are reported for a specific surface treatment with ethanedithiol and electrolyte environment. The positions are however sensitive to different surface treatments, providing an avenue to control doping. Electrochemical gating is further used to determine the carrier mobility in the films of the three different systems as a function of CQD size. HgSe and HgS show increasing mobility with increasing particle sizes while HgTe shows a nonmonotonous behavior, which is attributed to some degree of aggregation of HgTe QDs. KEYWORDS: colloidal quantum dots, electrochemistry, electronic structure, doping, mobility, infrared
C
properties. Therefore, it is important to assess the absolute energy of the states of the CQDs, their sensitivity to the surface condition and the origin of the Fermi level, in order to ultimately control doping in these CQD films. Electrochemistry is a powerful technique. It provides the measurement of the Fermi level and reversible electrochemistry allows the absolute measurements of the filled and empty state energies with the application of a voltage. 13 With a bipotentiostat, it gives the conduction as a function of the density of states of the CQD films.14 Therefore, a single cyclic voltammetry experiment can readily provide information that could otherwise be obtained by the combination of photoemission,15 inverse photoemission, and field effect transistor (FET) gating. While reversible electrochemistry is limited by the chemical stability of the materials, irreversible electrochemistry can also be used to learn about chemical processes and decomposition at CQD surfaces.16−18 Another distinction is that the environment must be an electrolyte, allowing for ionic conduction. However, the electrolyte environment may be arguably more natural than a high vacuum for colloidal materials.
olloidal quantum dots (CQD) are attractive alternatives to epitaxial materials for optoelectronic applications due to low cost and solution processing.1 In particular, CQDs of zinc-blend mercury chalcogenides, Hg(S, Se, Te),2 are of interest as solution-based materials in the mid-IR spectral range. HgTe CQDs are typically undoped with size-dependent bandgaps and have shown promise as mid-IR detectors.3 Recent advances include background limited photovoltaic (PV) devices,4 the simplified fabrication of highresolution mid-infrared cameras,5 and multispectral detectors.6 In contrast to HgTe, HgS and HgSe CQDs are stably n-doped in ambient conditions,7,8 which allowed the realizations of CQDs intraband photodetectors.7−9 A suggested explanation for the markedly different doping of the mercury chalcogenides has been the relative positions of the environmental Fermi level and the energy levels of the dots,7,8,10 but there have been no measurements to confirm this hypothesis. It has also been observed that exposure of the CQDs to various conditions could strongly change the doping level as determined by the strength of the intraband transitions.7,8,10−13 This has been proposed to arise through a displacement of the CQD states with respect to the environment Fermi level,7 but this relative energy shift has also not been tested. In general, small infrared gaps place strong requirements on the control of the Fermi level around the CQDs since this determines the doping which strongly affects the photodetector © 2017 American Chemical Society
Received: February 13, 2017 Accepted: March 17, 2017 Published: March 17, 2017 4165
DOI: 10.1021/acsnano.7b01014 ACS Nano 2017, 11, 4165−4173
Article
www.acsnano.org
Article
ACS Nano
Figure 1. (a, b, c) Cyclic voltammetry of films of HgS (5 nm), HgSe (6.8 nm), and HgTe (7.1 nm) CQDs, respectively. The blue lines are the charging currents, the black lines are the conductance, and the red arrow indicates the rest potential. Small arrows indicate the scan directions, with reduction being toward negative potential and oxidation toward positive potentials. (d, e, f) Infrared absorption of films of the same HgS, HgSe, and HgTe CQDs films, respectively. (g, h, i) TEM images of the same colloid dispersions. The scale bars are 20 and 50 nm. HgS has a size distribution of 13%, HgSe is ∼9%, and HgTe is ∼10% (partially aggregated).
mentioned, we use the EdT/HCl/IPA treatment in air as it affords repeatable redox positions (within 0.05 V) and rest potentials, as well as good film conductance. The reversibility of the electrochemistry is evident in Figure 1a−c. This is consistent with simple charge transfer to an electroactive film.13 Extending the scan range in the oxidation and reduction directions leads to increasing Faradaic currents as electrons or holes are consumed in irreversible chemical reactions. For reversible electrochemistry involving only charge flowing back and forth into the CQD state, the potential must stay in the range of the electrochemical stability, with no reactions. To estimate stability limits, we use (HgX)n + 2e− → Hg(HgX)n−1 + X2− for the reduction decomposition potential, Er,dec, and (HgX)n → (HgX)n−1X + Hg2+ + 2e− for the oxidation decomposition potential Eo,dec.21,7 These can be determined using standard aqueous redox potential and standard energy of formation ΔGof,HgX, such that Er,dec = E0X/X2− + ΔGof,HgX/2F and Eo,dec = E0Hg2+/Hg − ΔGof,HgX/2F where F is the Faraday constant. The region of stability is [−0.99 V, + 0.85 V] for HgS, [−1.17 V, + 0.76 V] for HgSe, and [−1.3 V, + 0.75 V] for HgTe, with potentials referenced to SCE. The oxidation limit in our experiments can therefore be explained by the oxidation decomposition, around +0.8 V/SCE for all three systems. However, the reduction limit is experimentally much more positive than −1 V/SCE, and this may be due to other reactions not considered above or impurities. Figure 1a,b shows that both HgS and HgSe show n-type conductivity with reversible charging current waves but no ptype conductivity. At positive potentials, the conductivity remains in the noise and there is no reversible charging current when the potential scan is reversed. Therefore, if holes are injected, the CQDs undergo some irreversible process on the
This article presents results from an extensive study of the reversible electrochemistry of films of the three mercury chalcogenides CQDs. The redox potentials of the CQD states, the absolute position of the bulk bands, the doping, the origin of the Fermi level, and the mobility as a function of CQD size are determined.
RESULTS AND DISCUSSION Typical Electrochemistry Results. Figure 1a−c show the typical results from a cyclic voltammetry experiment for CQD films of HgS (average diameter 5 nm), HgSe (average diameter 6.8 nm), and HgTe (average diameter 7.1 nm). In this work, HgS,12 HgSe,7 and HgTe19 CQDs are prepared following reported methods. For electrochemical measurements, we use thin films of dodecanethiol capped CQDs (
