Surface Depletion Layers in Plasmonic Metal Oxide Nanocrystals

Article Cite This: Acc. Chem. Res. XXXX, XXX, XXX−XXX

pubs.acs.org/accounts

Surface Depletion Layers in Plasmonic Metal Oxide Nanocrystals Published as part of the Accounts of Chemical Research special issue “Nanochemistry for Plasmonics and Plasmonics for Nanochemistry”. Stephen L. Gibbs, Corey M. Staller, and Delia J. Milliron*

Downloaded via EAST CAROLINA UNIV on August 20, 2019 at 16:11:44 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

McKetta Department of Chemical Engineering, University of Texas at Austin, Austin, Texas 78712-1589, United States CONSPECTUS: Strong infrared (IR) light−matter interaction and spectral tunability combine to make plasmonic metal oxide nanocrystals (NCs) a compelling choice for IR applications. In particular, visible transparency paired with strong, dynamically tunable IR absorption has motivated their implementation in electrochromic smart windows, but these NCs hold promise for a far broader range of plasmonically driven processes such as surface-enhanced infrared sensing, photothermal therapy, and enhanced photocatalysis. These unique properties result from localized surface plasmon resonance (LSPR) sustained by a relatively low free charge carrier concentration, which in turn requires consideration of distinct materials physics relative to traditional plasmonic materials (i.e., metals). Particularly important is the formation of insulating shells devoid of charge carriers (depletion layers) near the NC surface. Surface states as well as applied surface potentials can give rise to a potential difference between the NC surface and its core that depletes free charge carriers from the surface, forming an insulating shell that reduces the conductivity in NC films, lowers the dielectric sensitivity of the LSPR, and diminishes the incident electric field enhancement. In this Account, we report recent investigations of depletion layers in plasmonic metal oxide NCs that have advanced understanding of the semiconductor physics underlying the optoelectronic properties of these NCs and the electrochemical modulation of their LSPR, establishing a conceptual framework with which to broaden their applicability and optimize their performance. As a result of surface depletion, larger, highly doped NCs have improved dielectric sensitivity compared with their smaller, lightly doped counterparts. Concentrating dopants near the NC surface compresses the depletion layer, resulting in improved conductivity of NC films. Moreover, atomic layer deposition of alumina to infill NC films enhances the film conductivity by more than 2 orders of magnitude, ascribed to the elimination of depletion effects by reactive removal of surface water species. At the conclusion, we reflect on how our newfound understanding of surface depletion in plasmonic metal oxide NCs is quickly leading to rational material design. This insight is already resulting in significant performance improvements, and the same principles can be applied to new, exciting opportunities in hot carrier extraction and resonant IR energy transduction.

1. INTRODUCTION Metal oxide nanocrystals (NCs) are large-band-gap semiconductors that, when doped with aliovalent substitutional impurities or other charged defects, gain a significant concentration of free charge carriers, ne, enabling a wide range of tunability of their electronic and optical properties.1,2 Colloidally synthesized metal oxide NCs can easily be processed into transparent conductive oxide (TCO) films that are visibly transparent, making them functional materials for electrochromic windows and optoelectronic devices.3−8 Furthermore, the free charge carriers within each NC collectively oscillate in response to incident light, thereby sustaining a localized surface plasmon resonance (LSPR). Because of their ability to efficiently collect and confine the energy from incident light near the LSPR frequency, ωLSPR, nanoparticles (NPs) of traditional plasmonic materials such as Au and Ag have shown promise for applications ranging from enhanced catalysis9−12 to photothermal therapy13,14 and sensing down to the single-molecule level.15−17 LSPR is most effective at enhancing light−matter interactions when ωLSPR is on- or near-resonant to the © XXXX American Chemical Society

photoinduced process of interest, so it is desirable to develop a library of plasmonic materials with ωLSPR ranging widely across the electromagnetic spectrum. The ωLSPR of doped metal oxide NCs is easily tunable throughout the infrared (IR) wavelengths because of a relatively low and variable ne. While their optical properties show promise for diverse applications, the underlying plasmon physics of doped metal oxide NCs is not as well established as that of metal NPs. Though comparable in many ways, the distinguishing properties of doped metal oxide NCs, namely, their semiconductor band structure, IR-range ωLSPR, and relatively low ne, demand unique consideration. Of principal importance, and the focus of this Account, is the presence of depletion regions near the surface of doped metal oxide NCs that have a greatly reduced ne, even below the limit required for metallic electronic and optical properties (Figure 1). Received: May 31, 2019

A

DOI: 10.1021/acs.accounts.9b00287 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research

developments. Fundamental understanding of plasmonic metal oxide NCs has advanced markedly on the basis of systematic data sets leveraging well-controlled synthetic routes, creating a foundation for their viability in developing impactful plasmonpowered processes.

2. THEORY OF SURFACE DEPLETION IN PLASMONIC METAL OXIDE NANOCRYSTALS Space-charge layers arise at the surface of homogeneous semiconductors because changes in electronic structure arise when the lattice periodicity is interrupted and surface defects are introduced. The interaction between the uncoordinated surface atoms and nearby molecules determines the surface-state energy. Space-charge layers (including depletion layers) induced by surface states or an applied surface potential are often assumed to have a spatial extent that far exceeds the radius of the NC, so much so, in fact, that ne is taken to be independent of the distance from the surface. This is a good approximation if the NC is very small and sparsely doped; however, it becomes invalid for heavily doped or larger, moderately doped NCs. To appreciate this, Poisson’s equation yields the voltage (ϕ) profile near the surface of a planar electrode (eq 1):

Figure 1. When charge carriers are depleted from the NC surface, NCs of high ne (blue) form insulating shells of low ne (gray). This diminishes the extinction cross section, weakens the dipole−dipole interactions between nearby NCs, and hinders electron transport in NC films in comparison with the nondepleted (flat-band) case.

Recent work not only confirms the presence of such depletion layers but also establishes that they form a barrier that insulates the charge carriers from their surroundings, detrimentally affecting the performance. For example, the enhanced electric fields at the NC surface, which are responsible for sensitized spectroscopic detection of molecular vibrations in surfaceenhanced infrared spectroscopy (SEIRS), decay exponentially with depletion layer thickness, reducing the sensitivity.18,19 Moreover, as the depletion layer thickness increases, the conductivity in NC films diminishes because cores of high ne are physically separated by a potential barrier that mobile carriers must overcome.20,21 Neglecting the presence of depletion layers in metal oxide NCs has led to inaccuracies in evaluating their intrinsic material properties and clouded strategies for maximizing their performance. To advance the development of doped metal oxide NCs for plasmonic-powered technologies, it is essential to deepen the understanding of their fundamental physical and chemical properties. This Account presents recent findings on plasmonic optical and electron transport properties of doped metal oxide NC dispersions and NC thin films. Experiment and simulation indicate that understanding depletion layers is imperative for extracting intrinsic material properties and rationalizing and predicting material performance. We discuss methods to chemically and electrochemically control depletion effects on the optical and electronic properties with an eye toward further

ρ (x ) d2ϕ =− εSε0 dx 2

(1)

where ρ(x) is the position-dependent charge density and εS is the relative, static dielectric constant. The built-in potential, EBI, is the difference between the potential far from the surface (i.e., the bulk potential) and the surface potential. For a given EBI, the depletion width, W, is defined by eq 2 at ionized defect density Nd: W=

2εSε0E BI eNd

(2)

Consider a prototypical doped metal oxide: Sn-doped indium oxide (ITO). For a dopant concentration of 1 atom % Sn (Nd ≈ 3 × 1020 cm−3) and an applied potential difference of 1.0 V, the depletion width is 1.8 nma length scale on the order of the radius for smaller NCs. The depleted region constitutes a tiny volume fraction in, for example, an ∼100 nm thin-film electrode, so there is a nearly negligible effect on the total number of free

Figure 2. Simulated surface depletion in an ITO NC. (A) Simulated intra-NC radial potential profiles for the conduction-band minimum (CBM) (solid lines) and the Fermi level (EF) (dashed lines) for a 20.6 nm diameter ITO NC containing 3 atom % Sn when the surface potential dictates flatband conditions (blue) and depletion conditions (red). (B) ne profiles under flat-band (blue) and depletion (red) conditions. The dashed gray line denotes the Mott criterion for metallic conductivity in ITO. (C) Two-dimensional cross sections of the ne profiles in (B) under (top) flat-band and (bottom) depletion conditions. Adapted from ref 20. Copyright 2018 American Chemical Society. B


Article


where ωp is the bulk plasma frequency, e and m* are the charge and effective mass of an electron, ε∞ is the constant background polarizability, εm is the dielectric constant of the surrounding medium, and γ is the electron damping value. For metals, ne is a fixed, intrinsic value, and tuning of ωLSPR into the IR region requires large particle sizes and anisotropic shapes, which are difficult to achieve colloidally. Even with these geometric attributes, ωLSPR in colloidal metal NPs is still limited to the nearIR region.27−30 For doped semiconductor NCs, however, ne is dependent on the number of defects incorporated into the NC lattice. Often the defect is an extrinsic donor atom, in which case an increase in the dopant concentration Nd will increase ne. Thus, while metals have intrinsic ne and researchers rely on shape and size modification to tune ωLSPR, the resonance of doped metal oxide NCs can be shifted simply by incorporating variable concentrations of dopants during synthesis. This tunability allows for independent variation of ne and the NC shape and size to fundamentally probe and optimally tune these materials. Applying Mie theory for particle sizes much smaller than the wavelength of incident light shows that the scattering cross section is negligible and the extinction cross section, σext, can be derived by solving Maxwell’s equations for light interacting with a sphere: ÅÄÅ ÑÉ Å ε(ω) − εm ÑÑÑ ÑÑ σext(ω) = 3Vk εm ImÅÅÅÅ ÅÅÇ ε(ω) + 2εm ÑÑÑÖ (5)

charge carriers (electrons in this case). By contrast, in an electrode composed of NCs, the depleted regions constitute a large volume fraction of the overall film, which means that the total number of free electrons can be altered dramatically by modulating the surface potential. While the space-charge width in eq 2 is a reasonable approximation for NC depletion layers, accurate calculation of the number of electrons depleted and the spatial profile of the potential requires solving Poisson’s equation in spherical coordinates, which is nontrivial and must be done numerically.22 Staller et al. numerically solved Poisson’s equation to simulate band profiles and ne profiles for a range of ITO NCs (Figure 2).20 The built-in potential, EBI, determines whether the surface will accumulate charge, deplete charge, or remain neutral (i.e., the flat-band condition, when EBI is zero and therefore W is zero). Under an oxidizing surface potential, the CBM bends upward near the surface as a result of the potential gradient in the depletion layer. This band bending impacts the radial profile of ne, which is constant at flat-band but drops off significantly near the surface with an oxidizing potential. Even at a moderate EBI of a few electron volts, the NC in this example loses 17% of its electrons. As mentioned in the Introduction, this loss creates an insulating barrier between the plasmonic core and its surroundings, reducing the strength of their optical and electronic interaction. The rest of this Account describes efforts to quantify and control depletion by three methods: electrochemical modulation, chemical surface treatment, and intra-NC dopant distribution. These techniques improve the doped metal oxide NC performance for electrochemical, optical, and electronic applications.

in which ε(ω) = ε∞ −

3. ELECTROCHEMICAL MODULATION OF SPACE-CHARGE LAYERS IN PLASMONIC METAL OXIDE NCS The substantive impact of space-charge layers is clear when understanding the operation of nanostructured electrochromic thin films, which are promising for energy-saving smart windows because they can have improved switching speed and spectral tunability compared with conventional thin films. The improvements derive in part from capacitive electrochromisman appreciable change in optical transmittance with applied voltage in response to interfacial charge-compensating ions. There are a few reports discussing space-charge layers in electrochromic films of doped NCs,23−25 yet quantitative analysis and theoretical simulation of these and their effect on material performance are just emerging. Through careful analysis of extinction spectra, we can probe the extent of surface spacecharge layers.

ωp =

ε∞ + 2εm

nee 2 m*ε0

(6)

3.2. Electrochemical Modulation of Space-Charge Layers in NC Thin Films

An important alteration to the particle dielectric function must be considered when a surface depletion layer is present. When the radial electron concentration is nonuniform, the dielectric function also varies radially, causing a significant change in the extinction spectrum. To account for this, Zandi et al. employed a core−shell effective medium approximation (EMA).18 After the ne profile is found numerically and sliced into small step sizes, the dielectric function of each shell is calculated using eq 6, and the resulting values are collected into an effective dielectric function for the particle using a Maxwell−Garnett EMA. The singleparticle extinction spectrum is then calculated by plugging the effective dielectric function into eq 5. To simulate the optical response of an NC film, a second level of effective medium modeling represents these particles within the medium of a porous NC film. Zandi et al. tested the validity of this approach by comparing simulations to experimental observations of the transmittance of ITO NC films under a varying electrochemical potential. A series of ITO NCs with dopant concentrations ranging from 1− 8 atom % Sn were synthesized with both small (D ≈ 7 nm) and large (D ≈ 14 nm) diameters. A custom spectroelectrochemical

Whereas traditional plasmonic metals have intrinsic ne ∼ 1022 cm−3, electron concentrations of ne ∼ 1019−1021 cm−3 are achieved for a variety of doped metal oxide materials. The LSPR peak position is strongly dependent on ne according to eqs 3 and 4:26 ωp2

ω 2 + iγω

The cross section quantifies the strength of the light−matter interaction and is dependent on the host dielectric constant εm, the particle volume V, the incident-light wave vector magnitude k, and the particle dielectric function ε. The frequencydependent dielectric function (eq 6) considers contributions to the polarizability coming from (i) electrons abiding by the Drude theory of metals (free electrons) and (ii) a constant ε∞.

3.1. Theory of Extinction in Plasmonic Metal Oxide NCs

ωLSPR =

ωp2

− γ2 (3)

(4) C


Article


that falls in the depletion layer (f W) depends on the applied potential. At a given potential, f W is lower for large, highly doped NCs because the surface area-to-volume ratio drops and W decreases. Hence, the depletion layer is confined to the nearsurface regions of large, highly doped NCs, but it can extend deep into the core of small, lightly doped NCs. Accumulating (depleting) charge anywhere in the NC volume has the effect of increasing (decreasing) the LSPR peak intensity, but predicting changes in peak location is more nuanced. When W is small, the electron concentration in the plasmonic core of the NC, responsible for the optical absorption, remains nearly constant, and thus, ωLSPR shifts only slightly, mostly as a result of changes in the dielectric surroundings of that plasmonic core. Since changes to ne occur almost exclusively within the depletion layer, ωLSPR shifts more substantially only when W is larger than the NC radius. Recent work by Tandon et al. reinforced the prominent influence of depletion layers on the optical response of electrochemically modulated NC films. They observed for even larger (>20 nm), highly doped NCs that the peak shift upon charging can be eliminated or even reversed in direction, meaning that highly reduced NC films exhibit a red shift even as the LSPR gains intensity (Figure 5).31 The authors explained this phenomenon by simulating extinction spectra of NC films while taking into account dipolar coupling interactions between NCs. Coupling is known to cause a red shift of LSPR,32−34 and the authors concluded that the coupling strength increases when the depletion width is diminished under application of a reducing potential. For larger particles, the red shift due to potential-dependent coupling overcomes the typical blue shift due to increased ne.

cell for in situ FTIR spectroscopy under a varying potential was employed (Figure 3). The changes in the extinction spectra were

Figure 3. (A) Schematic and (B) photograph of the in situ FTIR spectroelectrochemical cell. Adapted with permission from ref 18. Copyright 2018 Springer Nature Publishing AG.

recorded starting from the most oxidizing potential, and then these were analyzed for trends in the peak intensity (σext) and peak location (ωLSPR). For films of small, lightly doped NCs, the peak intensity increases and shifts to higher energy upon charging (Figure 4A(i)), while films of large, heavily doped NCs show an increase in peak intensity with nearly constant peak location (Figure 4B(i)). Under the simplistic assumption of a uniform profile of electrons at all potentials, the peak for any size or doping level is expected to shift according to eqs 3 and 4. The discrepancy between this expectation and the experimental results, particularly for larger and more highly doped NCs, is explicable when the effects of the NC size and dopant concentration on modulation of the depletion layers are considered. In particular, the fraction of the total NC volume

Figure 4. In situ FTIR spectroelectrochemistry of ITO NC films. For films of (A) 1 atom % Sn, 7.4 nm diameter and (B) 10 atom % Sn, 11.5 nm diameter ITO NCs: (i) Experimental extinction spectra of NC films modulated across a voltage window of 2 to −2 V. All of the spectra are referenced to the spectrum in the most oxidized state (2 V). (ii) Simulated intra-NC electron concentration profiles. (iii) Simulated extinction spectra. Adapted with permission from ref 18. Copyright 2018 Springer Nature Publishing AG. D


Article


Figure 5. In situ FTIR spectroelectrochemistry of Cr−Sn-codoped In2O3 NC films. Cr is a neutral dopant that does not affect ne. For films of (A) 0 atom % Cr−10 atom % Sn, 4.7 nm diameter and (B) 20 atom % Cr−10 atom % Sn, 21.6 nm diameter NCs: (i) Experimental extinction spectra for a NC film modulated across a voltage window of 2 to −1.5 V. All of the spectra are referenced to the spectrum in the most oxidized state (2 V). (ii) Simulated intra-NC electron concentration profiles. (iii) Simulated extinction spectra. Adapted from ref 31. Copyright 2019 American Chemical Society.

Figure 6. Surface depletion hinders NFE and sensitivity. (A) ITO NCs electrochemically modulated in dispersion: (i) Experimental extinction of ITO NCs upon reduction. (ii) Simulated extinction spectra for isolated NCs. (iii−v) Simulated NFE maps for NCs whose extinction spectra are shown in (ii). (B) Dielectric sensitivity of ωLSPR for ITO NCs: (i) Experimental sensitivity calculated after dispersion of ITO NCs in various solvents. (ii) Modeled sensitivity for comparison. Adapted (A) from ref 19 and (B) with permission from ref 18. Copyright 2018 American Chemical Society and Springer Nature Publishing AG, respectively.

interaction in films, Agrawal et al. reported the extinction of ITO NCs undergoing electrochemical charging while freely suspended in a dilute dispersion (Figure 6).19 The spectra were fit using the same core−shell model as described above, without

3.3. Electrochemical Modulation of NC Dispersions

It is challenging to draw quantitative conclusions about dynamic intra-NC properties when coupling effects are present. To observe NC optical modulation unclouded by NC−NC E


Article


Figure 7. Identification of adsorbed water species as surface states in metal oxide NCs using in situ sheet resistance measurements. (A) Bare and alumina-capped ITO NC films. (B) ZnO NC films during exposure to inert and forming gas. Adapted from ref 21. Copyright 2016 American Chemical Society.

barrier at each hopping step.35,36 Hopping conduction is described in the most general case by the Miller−Abrahams model:37

the second level of effective medium modeling required for NC films. The modulation of ωLSPR in the NC dispersions exceeds that in similar NC films since coupling effects partially counteract the modulation in the latter case. These measurements also more clearly reveal the shift in ωLSPR due solely to the depletion layer because each NC interacts only with the surrounding static dielectric (i.e., the solvent). The depletion layer effectively surrounds the plasmonic NC core with a highdielectric shell, contributing to the red shift in the LSPR peak upon oxidation. The dielectric shell represented by the depletion layer also impacts the electric field strength near the NC surface. Photothermal applications, SEIRS, and other plasmonically enhanced processes depend on the confinement of incident light to nanoscale volumes, as measured by the near-field enhancement (NFE): E NFE = E0

ij Eij yz i 2rij y zz σ ∝ A expjjjj− zzzz expjjj− j kBT zz k a { k {

(8)

where σ is the conductivity, A is a material-dependent constant, rij is the distance between sites i and j, a is the inverse of the wave function decay rate (i.e., the electron localization length), and Eij is the energetic barrier encountered in moving from site i to site j. The depletion layer separates the electron localization volumes in neighboring NCs, increasing rij and consequentially decreasing the film conductivity. Measuring the electronic conductivity of NC films can be a sensitive tool for monitoring changes in W and determining their chemical and physical origins. Thimsen et al. observed that highly resistive films of bare 6.5 nm ZnO NCs were made >8 orders of magnitude less resistive following the deposition of alumina by atomic layer deposition (ALD).38 They proposed that this effect was the result of the reaction of trimethylaluminum (TMA) with adsorbed water species (i.e., surface hydroxyls), which form mid-band-gap surface states on metal oxide surfaces. On the basis of the above discussion, alumina deposition may therefore be expected to alleviate depletion, thereby reducing barriers to electron hopping. Another possibility, that alumina causes interfacial doping of ZnO NCs, was acknowledged by the authors. Following up, Ephraim et al. ruled out interfacial doping and further substantiated that removal of adsorbed water species is responsible for the enhanced conductivity in alumina ALD-capped NC films.21 First, the possibility of n-type doping due to the formation of • , was eliminated by depositing substitutional defects, AlZn alumina on ITO NC films, AlxIn (Figure 7A). A 1−2 order of magnitude decrease in sheet resistance was observed. The modest improvement in the sheet resistance of the ITO NC film compared with that of the ZnO NC film is attributed to the higher dopant concentration in ITO NCs, resulting in a thinner depletion layer before ALD. Second, Ephraim et al. conducted in situ measurements of bare ZnO NC films under exposure to Ar gas and 15% H2 in Ar gas (forming gas) at 200 °C (Figure 7B). Heating to 200 °C in Ar removes physisorbed water from the NC surfaces, resulting in a >2 order of magnitude decrease in sheet resistance. Forming gas is known to remove chemisorbed water species from metal oxides. Upon exposure to forming gas, the NC film sheet

2

(7)

where E0 is the incident field strength and E is the near-field strength. Simulations of the NFE at increasingly reducing potentials and thus decreasing W (Figure 6A(iii−v)) show that the maximum NFE just outside the NC is significantly increased as W decreases. For a given W, smaller NCs are impacted more acutely, leading to simulations that predict a size dependence of the sensitivity of ωLSPR to changes in the dielectric environment. Indeed, in experiments, larger NCs show greater dielectric sensitivity, though the differences become insignificant at higher doping concentrations where W is smaller (Figure 6B). The insulating shell arising from depletion is expected to diminish the strength of the coupling of the LSPR to molecular vibrations, negatively impacting SEIRS. It may also hamper hot electron extraction, posing a challenge for plasmon-induced photochemistry. Overall, depletion of electrons from the nearsurface region impacts the performance across all applications and motivates further efforts to understand and control it beyond electrochemical modulation.

4. MITIGATING DEPLETION THROUGH CHEMICAL MODIFICATION OF THE NC SURFACE As mentioned earlier, TCO films represent a potential application of doped metal oxide NCs for which conductivity is an essential design parameter. Electron transport in NC films typically proceeds through a series of inelastic tunneling events (so-called hopping conduction), and depletion layers pose a F


Article


Figure 8. Dependence of the NC film conductivity on the near-surface dopant concentration before and after alumina ALD. (A) Room-temperature conductivity of bare ITO NC films (red) and alumina-capped ITO NC films (black). (B) Extracted values of the electron localization length in bare NC films (red) and the metallic grain diameter in alumina-capped NC films (black). Adapted from ref 20. Copyright 2018 American Chemical Society.

found to conduct electrons through Efros−Shklovskii variablerange hopping with a Gaussian dispersion of energy levels (ESVRH-GD).42 In this conduction mechanism, the diameter of the nondepleted core is defined by the electron localization length. From fits to the ES-VRH-GS model, the electron localization lengths of bare NC films were determined to be longer for higher near-surface dopant concentrations, consistent with a decrease in W (Figure 8B). After capping with alumina by ALD, the film conductivity increased by 1−2 orders of magnitude, and the electron transport in NC films was found to follow a granular metal conduction mechanism (Figure 8A).43,44 In this conduction mechanism, the metric analogous to the localization length is the metallic grain size. Figure 8B shows that the metallic grain size for alumina-capped films is independent of the nearsurface dopant concentration and is approximately equal to the NC diameter (20 nm), meaning that W is nearly zero. This confirms that the dependence of the bare NC film conductivity on the near-surface dopant concentration arises from variation of the depletion layer thickness.

resistance asymptotically approaches the sheet resistance of a ZnO NC film with alumina-capped surfaces. These results confirmed that the decrease in NC film resistivity was due to the removal of hydroxyls at the metal oxide surface.

5. TUNING THE DEPLETION WIDTH BY CONTROLLING THE INTRA-NC DOPANT DISTRIBUTION Because hydroxylation of metal oxide surfaces is spontaneous under ambient conditions, strategies for controlling W are of high importance. For many applications, embedding NCs in alumina is not feasible because it renders the NCs nondispersible and prevents access to the NC surface for sensing, detection, and catalysis. Hence, the ability to control depletion without modifying the surface is valuable. We have discussed the decrease in depletion width in highly doped NCs; however, depletion layer compression is possible even for lightly doped NCs if the dopants are concentrated in the near-surface region where they can screen the surface potential. Staller et al.20 leveraged advances in metal oxide NC synthesis39−41 to control the radial dopant distribution in 20 nm ITO NCs while holding the overall NC dopant concentration at 3 atom % Sn. This control was afforded by a slow-injection synthesis in which NC cores of a given dopant concentration and size are synthesized, purified, and then used as seeds for a shelling procedure in which the shell dopant concentration and thickness are tunable. The conductivity of bare NC films was found to exponentially depend on the near-surface dopant concentration over a conductivity range spanning about an order of magnitude (Figure 8A). This trend supports the hypothesis that W is reduced as the near-surface dopant concentration increases. This explanation was verified by passivating NC surfaces by alumina ALD to ameliorate the effects of depletion on NC film conductivity, after which all of the samples showed a 1−2 order of magnitude increase in conductivity. Importantly, the aluminacapped films show no apparent dependence of the film conductivity on the near-surface dopant concentration, consistent with the conclusion that at constant overall doping, the extent of the depletion layer is the determining factor in electronic conductivity. Fits to the variable-temperature conductivity support the conclusions drawn from room-temperature measurements about how depletion impacts electron transport in NC films. On the basis of the temperature dependence, bare NC films were

6. CONCLUSION AND OUTLOOK In this Account, we have focused on surface depletion as a physical phenomenon that is crucial to understanding and overcoming performance limitations of doped metal oxide NCs. Depletion layers influence the character of optical modulation in electrochromic devices, diminish the sensitivity of LSPR to changes in the environment, pose barriers for electron transport, reduce coupling strength, and could potentially reduce hot electron transfer rates. However, understanding these effects reveals strategies to overcome these limitations. We have shown that the optical response upon electrochemical modulation of NCs in thin films and dispersions depends heavily on f W and W, which decrease with increasing NC size and doping density. As well, W can be decreased with surface treatment by ALD to remove adventitious hydroxyls and by preferentially segregating dopant atoms near the NC surface. Already, understanding the physics and chemistry influencing depletion layers has enabled orders of magnitude improvements in the conductivity of NC films, improved dielectric sensitivity, and predicted improvements in NFE surrounding NCs. Furthermore, we have observed and accurately simulated an anomalous red shift with increasing carrier concentration upon electrochemical reduction. This newfound understanding offers another parameter for G


Article

Accounts of Chemical Research tuning the dynamic optical response of NC−NC interactions in films. Looking ahead, these strategies and others will enable doped metal oxide NCs to find application in new contexts. One underexplored aspect of these NCs’ photophysics is the generation of hot electrons and their utilization for photochemistry. Employing ultrafast spectroscopy techniques, experiments in the past decade have made leaps and bounds to characterize the energetic distribution of electrons upon excitation of LSPR.10,11,45−47 Fitting such data, Johns, Blemker, and co-workers showed that in plasmonic metal oxide NCs, a significant population of electrons is excited to energies much higher than in a thermal distribution.48 In fact, the hot electrons can have energies well exceeding those in conventional plasmonic metal NPs because of the low electronic heat capacity of metal oxide NCs. These hot electrons can dissipate heat to the particlethe foundation for plasmon-assisted photothermal therapyor can in principle be transferred out of the particle to drive photochemical reactions. Numerous reports have already demonstrated the promise of resonant hot electron transfer to molecules from plasmonic NPs, mostly composed of traditional metals.9,12,49−51 Hot-electron-driven processes utilizing doped metal oxide NCs represent an exciting frontier considering that catalytic processes may be facilitated by molecular adsorption to the metal oxide surface and that these materials could extend plasmonic photocatalysis well into the IR region. Ultimately, the most exciting possibility may be that resonant coupling between IR LSPR and molecular vibrations could enable efficient channeling of incident radiation into specific modes that influence chemical reaction pathways. The foundation for such exploration lies in studies of the fundamental questions about the nature and tunability of LSPR in metal oxide NCs that are now underway. As depletion layers play a critical role in the physics and chemistry occurring at the NC surface, harnessing control over them will be a major key to unlocking the potential of these materials. In the near term, we expect that further advances in the synthesis of NCs with controlled size, shape, and dopant placement will not only enhance our current understanding but also open the door to new applications, potentially including efficient hot carrier extraction and resonant energy transduction following IR excitation.

■

Engineering from the University of Texas at Austin under Prof. Delia Milliron. He is currently a Thin Films Development Engineer at Micron Technology. His research interest is in the quantitative analysis of the electronic and optical properties of doped semiconductor nanocrystals and understanding how they are influenced by variations in NC surface chemistry. Delia J. Milliron received her A.B. and Ph.D. in Chemistry from Princeton University and the University of California, Berkeley, respectively. Following appointments at IBM Research and Lawrence Berkeley National Laboratory, she joined the faculty at the University of Texas at Austin, where she is currently the T. Brockett Hudson Professor of Chemical Engineering. Her research explores the optoelectronic and electrochemical properties of inorganic colloidal nanocrystals and their processing into functional materials. Advancing understanding and control of doped metal oxide nanocrystals is a primary focus of research in the Milliron group.

■

ACKNOWLEDGMENTS This research was supported by the National Science Foundation (NSF), including NASCENT, an NSF ERC (EEC-1160494, C.M.S.), CHE-1609656, a Graduate Research Fellowship (Award DGE-1610403, S.L.G.), and the Welch Foundation (F-1848).

■

REFERENCES

(1) Agrawal, A.; Johns, R. W.; Milliron, D. J. Control of Localized Surface Plasmon Resonances in Metal Oxide Nanocrystals. Annu. Rev. Mater. Res. 2017, 47, 1−31. (2) Agrawal, A.; Cho, S. H.; Zandi, O.; Ghosh, S.; Johns, R. W.; Milliron, D. J. Localized Surface Plasmon Resonance in Semiconductor Nanocrystals. Chem. Rev. 2018, 118, 3121−3207. (3) Buonsanti, R.; Llordes, A.; Aloni, S.; Helms, B. A.; Milliron, D. J. Tunable Infrared Absorption and Visible Transparency of Colloidal Aluminum-Doped Zinc Oxide Nanocrystals. Nano Lett. 2011, 11, 4706−4710. (4) Garcia, G.; Buonsanti, R.; Runnerstrom, E. L.; Mendelsberg, R. J.; Llordes, A.; Anders, A.; Richardson, T. J.; Milliron, D. J. Dynamically Modulating the Surface Plasmon Resonance of Doped Semiconductor Nanocrystals. Nano Lett. 2011, 11, 4415−4420. (5) Runnerstrom, E. L.; Llordés, A.; Lounis, S. D.; Milliron, D. J. Nanostructured Electrochromic Smart Windows: Traditional Materials and NIR-Selective Plasmonic Nanocrystals. Chem. Commun. 2014, 50, 10555−10572. (6) Runnerstrom, E. L.; Kelley, K. P.; Sachet, E.; Shelton, C. T.; Maria, J.-P. Epsilon-near-Zero Modes and Surface Plasmon Resonance in Fluorine-Doped Cadmium Oxide Thin Films. ACS Photonics 2017, 4, 1885−1892. (7) Stanley, R. Plasmonics in the Mid-Infrared. Nat. Photonics 2012, 6, 409−411. (8) Noginov, M. A.; Gu, L.; Livenere, J.; Zhu, G.; Pradhan, A. K.; Mundle, R.; Bahoura, M.; Barnakov, Yu. A.; Podolskiy, V. A. Transparent Conductive Oxides: Plasmonic Materials for Telecom Wavelengths. Appl. Phys. Lett. 2011, 99, 021101. (9) Christopher, P.; Xin, H.; Linic, S. Visible-Light-Enhanced Catalytic Oxidation Reactions on Plasmonic Silver Nanostructures. Nat. Chem. 2011, 3, 467−472. (10) Christopher, P.; Moskovits, M. Hot Charge Carrier Transmission from Plasmonic Nanostructures. Annu. Rev. Phys. Chem. 2017, 68, 379− 398. (11) Brongersma, M. L.; Halas, N. J.; Nordlander, P. PlasmonInduced Hot Carrier Science and Technology. Nat. Nanotechnol. 2015, 10, 25−34. (12) Li, K.; Hogan, N. J.; Kale, M. J.; Halas, N. J.; Nordlander, P.; Christopher, P. Balancing Near-Field Enhancement, Absorption, and Scattering for Effective Antenna−Reactor Plasmonic Photocatalysis. Nano Lett. 2017, 17, 3710−3717.

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Telephone: (512)232-5702. ORCID

Stephen L. Gibbs: 0000-0003-2533-0957 Corey M. Staller: 0000-0001-8665-2840 Delia J. Milliron: 0000-0002-8737-451X Notes

The authors declare no competing financial interest. Biographies Stephen L. Gibbs received his B.S. in Chemical Engineering from the University of Florida. He is now a Ph.D. candidate working in Prof. Delia Milliron’s research group. His research interest is in the synthesis and quantitative analysis of the electronic and optical properties of doped semiconductor nanocrystals. Corey M. Staller received his B.S. in Chemical Engineering and B.S. in Economics from the University of Missouri and his Ph.D. in Chemical H


Article

Accounts of Chemical Research (13) Cole, J. R.; Mirin, N. A.; Knight, M. W.; Goodrich, G. P.; Halas, N. J. Photothermal Efficiencies of Nanoshells and Nanorods for Clinical Therapeutic Applications. J. Phys. Chem. C 2009, 113, 12090−12094. (14) Hirsch, L. R.; Stafford, R. J.; Bankson, J. A.; Sershen, S. R.; Rivera, B.; Price, R. E.; Hazle, J. D.; Halas, N. J.; West, J. L. Nanoshell-Mediated near-Infrared Thermal Therapy of Tumors under Magnetic Resonance Guidance. Proc. Natl. Acad. Sci. U. S. A. 2003, 100, 13549−13554. (15) Willets, K. A.; Van Duyne, R. P. Localized Surface Plasmon Resonance Spectroscopy and Sensing. Annu. Rev. Phys. Chem. 2007, 58, 267−297. (16) Neubrech, F.; Huck, C.; Weber, K.; Pucci, A.; Giessen, H. Surface-Enhanced Infrared Spectroscopy Using Resonant Nanoantennas. Chem. Rev. 2017, 117, 5110−5145. (17) Kneipp, K.; Wang, Y.; Kneipp, H.; Perelman, L. T.; Itzkan, I.; Dasari, R. R.; Feld, M. S. Single Molecule Detection Using SurfaceEnhanced Raman Scattering (SERS). Phys. Rev. Lett. 1997, 78, 1667− 1670. (18) Zandi, O.; Agrawal, A.; Shearer, A. B.; Reimnitz, L. C.; Dahlman, C. J.; Staller, C. M.; Milliron, D. J. Impacts of Surface Depletion on the Plasmonic Properties of Doped Semiconductor Nanocrystals. Nat. Mater. 2018, 17, 710−717. (19) Agrawal, A.; Kriegel, I.; Runnerstrom, E. L.; Scotognella, F.; Llordes, A.; Milliron, D. J. Rationalizing the Impact of Surface Depletion on Electrochemical Modulation of Plasmon Resonance Absorption in Metal Oxide Nanocrystals. ACS Photonics 2018, 5, 2044−2050. (20) Staller, C. M.; Robinson, Z. L.; Agrawal, A.; Gibbs, S. L.; Greenberg, B. L.; Lounis, S. D.; Kortshagen, U. R.; Milliron, D. J. Tuning Nanocrystal Surface Depletion by Controlling Dopant Distribution as a Route toward Enhanced Film Conductivity. Nano Lett. 2018, 18, 2870−2878. (21) Ephraim, J.; Lanigan, D.; Staller, C.; Milliron, D. J.; Thimsen, E. Transparent Conductive Oxide Nanocrystals Coated with Insulators by Atomic Layer Deposition. Chem. Mater. 2016, 28, 5549−5553. (22) Seiwatz, R.; Green, M. Space Charge Calculations for Semiconductors. J. Appl. Phys. 1958, 29, 1034−1040. (23) Boschloo, G.; Fitzmaurice, D. Spectroelectrochemistry of Highly Doped Nanostructured Tin Dioxide Electrodes. J. Phys. Chem. B 1999, 103, 3093−3098. (24) Zum Felde, U.; Haase, M.; Weller, H. Electrochromism of Highly Doped Nanocrystalline SnO2:Sb. J. Phys. Chem. B 2000, 104, 9388− 9395. (25) Pflughoefft, M.; Weller, H. Spectroelectrochemical Analysis of the Electrochromism of Antimony-Doped Nanoparticulate TinDioxide Electrodes. J. Phys. Chem. B 2002, 106, 10530−10534. (26) Mayer, K. M.; Hafner, J. H. Localized Surface Plasmon Resonance Sensors. Chem. Rev. 2011, 111, 3828−3857. (27) Jain, P. K.; Lee, K. S.; El-Sayed, I. H.; El-Sayed, M. A. Calculated Absorption and Scattering Properties of Gold Nanoparticles of Different Size, Shape, and Composition: Applications in Biological Imaging and Biomedicine. J. Phys. Chem. B 2006, 110, 7238−7248. (28) Lee, K.-S.; El-Sayed, M. A. Gold and Silver Nanoparticles in Sensing and Imaging: Sensitivity of Plasmon Response to Size, Shape, and Metal Composition. J. Phys. Chem. B 2006, 110, 19220−19225. (29) Chen, H.; Kou, X.; Yang, Z.; Ni, W.; Wang, J. Shape- and SizeDependent Refractive Index Sensitivity of Gold Nanoparticles. Langmuir 2008, 24, 5233−5237. (30) Nehl, C. L.; Hafner, J. H. Shape-Dependent Plasmon Resonances of Gold Nanoparticles. J. Mater. Chem. 2008, 18, 2415−2419. (31) Tandon, B.; Agrawal, A.; Heo, S.; Milliron, D. J. Competition between Depletion Effects and Coupling in the Plasmon Modulation of Doped Metal Oxide Nanocrystals. Nano Lett. 2019, 19, 2012−2019. (32) Nordlander, P.; Oubre, C.; Prodan, E.; Li, K.; Stockman, M. I. Plasmon Hybridization in Nanoparticle Dimers. Nano Lett. 2004, 4, 899−903. (33) Funston, A. M.; Novo, C.; Davis, T. J.; Mulvaney, P. Plasmon Coupling of Gold Nanorods at Short Distances and in Different Geometries. Nano Lett. 2009, 9, 1651−1658.

(34) Rechberger, W.; Hohenau, A.; Leitner, A.; Krenn, J. R.; Lamprecht, B.; Aussenegg, F. R. Optical Properties of Two Interacting Gold Nanoparticles. Opt. Commun. 2003, 220, 137−141. (35) Talapin, D. V.; Lee, J.-S.; Kovalenko, M. V.; Shevchenko, E. V. Prospects of Colloidal Nanocrystals for Electronic and Optoelectronic Applications. Chem. Rev. 2010, 110, 389−458. (36) Zabet-Khosousi, A.; Dhirani, A.-A. Charge Transport in Nanoparticle Assemblies. Chem. Rev. 2008, 108, 4072−4124. (37) Miller, A.; Abrahams, E. Impurity Conduction at Low Concentrations. Phys. Rev. 1960, 120, 745−755. (38) Thimsen, E.; Johnson, M.; Zhang, X.; Wagner, A. J.; Mkhoyan, K. A.; Kortshagen, U. R.; Aydil, E. S. High Electron Mobility in Thin Films Formed via Supersonic Impact Deposition of Nanocrystals Synthesized in Nonthermal Plasmas. Nat. Commun. 2014, 5, 5822. (39) Ito, D.; Yokoyama, S.; Zaikova, T.; Masuko, K.; Hutchison, J. E. Synthesis of Ligand-Stabilized Metal Oxide Nanocrystals and Epitaxial Core/Shell Nanocrystals via a Lower-Temperature Esterification Process. ACS Nano 2014, 8, 64−75. (40) Jansons, A. W.; Hutchison, J. E. Continuous Growth of Metal Oxide Nanocrystals: Enhanced Control of Nanocrystal Size and Radial Dopant Distribution. ACS Nano 2016, 10, 6942−6951. (41) Crockett, B. M.; Jansons, A. W.; Koskela, K. M.; Johnson, D. W.; Hutchison, J. E. Radial Dopant Placement for Tuning Plasmonic Properties in Metal Oxide Nanocrystals. ACS Nano 2017, 11, 7719− 7728. (42) Efros, A.; Shklovskii, B. Electronic Properties of Doped Semiconductors; Springer: Berlin, 1984. (43) Beloborodov, I. S.; Lopatin, A. V.; Vinokur, V. M. Universal Description of Granular Metals at Low Temperatures: Granular Fermi Liquid. Phys. Rev. B: Condens. Matter Mater. Phys. 2004, 70, 205120. (44) Beloborodov, I. S.; Lopatin, A. V.; Vinokur, V. M.; Efetov, K. B. Granular Electronic Systems. Rev. Mod. Phys. 2007, 79, 469−518. (45) Linic, S.; Christopher, P.; Ingram, D. B. Plasmonic-Metal Nanostructures for Efficient Conversion of Solar to Chemical Energy. Nat. Mater. 2011, 10, 911−921. (46) Zhou, L.; Swearer, D. F.; Zhang, C.; Robatjazi, H.; Zhao, H.; Henderson, L.; Dong, L.; Christopher, P.; Carter, E. A.; Nordlander, P.; et al. Quantifying Hot Carrier and Thermal Contributions in Plasmonic Photocatalysis. Science 2018, 362, 69−72. (47) Link, S.; El-Sayed, M. A. Spectral Properties and Relaxation Dynamics of Surface Plasmon Electronic Oscillations in Gold and Silver Nanodots and Nanorods. J. Phys. Chem. B 1999, 103, 8410−8426. (48) Johns, R. W.; Blemker, M. A.; Azzaro, M. S.; Heo, S.; Runnerstrom, E. L.; Milliron, D. J.; Roberts, S. T. Charge Carrier Concentration Dependence of Ultrafast Plasmonic Relaxation in Conducting Metal Oxide Nanocrystals. J. Mater. Chem. C 2017, 5, 5757−5763. (49) Aslam, U.; Rao, V. G.; Chavez, S.; Linic, S. Catalytic Conversion of Solar to Chemical Energy on Plasmonic Metal Nanostructures. Nat. Catal. 2018, 1, 656. (50) Boerigter, C.; Campana, R.; Morabito, M.; Linic, S. Evidence and Implications of Direct Charge Excitation as the Dominant Mechanism in Plasmon-Mediated Photocatalysis. Nat. Commun. 2016, 7, 10545. (51) Swearer, D. F.; Zhao, H.; Zhou, L.; Zhang, C.; Robatjazi, H.; Martirez, J. M. P.; Krauter, C. M.; Yazdi, S.; McClain, M. J.; Ringe, E.; et al. Heterometallic Antenna-reactor Complexes for Photocatalysis. Proc. Natl. Acad. Sci. U. S. A. 2016, 113, 8916−8920.

I


Surface Depletion Layers in Plasmonic Metal Oxide Nanocrystals

Recommend Documents