Chemical transformations of nanomaterials for energy applications

Abstract

Chemical Transformations of Nanomaterials (CTNMs) is a powerful method to modify as-synthesized nanomaterials into forms with novel composition and structure. CTNMs allows access to nano-based materials that are difficult to synthesize by conventional means, including metastable materials through partial transformations. These new nanomaterial forms can have property and stability advantages when used for energy applications. This short review highlights recent research progress in chemical transformations of non-metallic nanomaterials as a means to create useful energy materials, such as metastable structures, with a focus on fuel cell catalysts and battery technologies. In addition to discussing progress with chemically transformed materials, this review briefly highlights several metastable materials that have been used for energy applications but were synthesized through conventional means. Synthetic protocols of CTNMs could be used to produce these metastable energy materials, which may result in novel properties. Finally, the review concludes with a forward looking perspective of metastable materials that have yet to be synthesized, but are good candidates for catalysis/batteries as metastable materials. These proposed systems could yield next generation catalysts/battery materials.

---

R. D. Robinson

Richard Robinson is an Assistant Professor at Cornell University. His research is centered on understanding the fundamental physics of nanomaterials, and applying chemical and structural transformations to engineer the properties of nanoparticles for energy applications. His group is targeting new materials for electrochemical energy storage, thermoelectrics, and catalysis. Richard received his BS and MS in Mechanical Engineering from Tufts University. Then, after several years at Accenture technology consulting, he returned to school and received his PhD in Applied Physics from Columbia University, under Irving Herman. During that time he worked briefly at Bell Laboratories in the Physical Research Laboratory. Richard's postdoctoral fellowship was at the University of California, Berkeley/LBNL in the research group of Paul Alivisatos. There, he worked on nanoparticle synthesis and chemical transformations of nanoparticles. In 2008 Richard began a faculty position at Cornell University in the Materials Science Department, and began work on nanomaterials for energy applications. He has won a number of awards including, the 3M Non-tenured Faculty Award, the NSF CAREER award, and the R&D 100 Award.

---

1. Introduction

In the past decade, chemical transformations of nanomaterials (NMs) has emerged as a powerful new means for nanosynthetic chemistry.^1–27 Chemical transformations of nanomaterials (CTNMs) are reactions that are performed on first-generation (as-synthesized) nanomaterials to modify their chemical composition, atomic structure, morphology, and/or size. CTNMs include redox reactions (either addition or exchange) and ionic exchange (cation and anion exchange) (Fig. 1). The popularity of CTNMs has been driven by their ability to create new NM forms^14,18,28 that are difficult or inaccessible from conventional syntheses (e.g., the “hot-injection” method) such as, Cu₂S nanorods¹² and unique core–shell/shell combinations like PbSe–CdSe/ZnS.⁷ But perhaps the most exciting and untapped feature of CTNMs is its capacity to form metastable phases, such as striped heterostructured nanorods,^3,6 crystalline–amorphous core–shell and interdiffused NMs,^29,30 and doped NMs^21,23,31 (Fig. 2). Thus CTNMs is emerging as a powerful tool to tailor the composition and phase of nanomaterials, as well as produce a series of compositionally interesting metastable structures.

Fig. 1 Possible routes for chemical transformations of nanomaterials to produce metastable structures for energy applications. Control of composition and phase through chemical transformations can be used to produce metastable materials, such as core–shell heterostructures with either a graded or sharp interface. These nanomaterials can be tailored for energy applications.

Fig. 2 Metastable Intermediate states during chemical transformation of nanomaterials. (left) TEM images of spontaneous nanorod superlattice formation formed through partial cation exchange. (A) Original 4.8 × 64 nm CdS nanorods. (B and C) Transformed CdS–Ag₂S superlattices. Inset to (C) is a histogram of Ag₂S segment spacing (center-to-center, nm). (right) An amorphous Co–P phase precedes the nanoscale Kirkendall effect. (top right) Schematic of Co nanoparticles with amorphous Co–P shell converting into Co₂P shell and subsequent initiation of Kirkendall hollowing. (bottom right) False color image of high-resolution TEM showing amorphous Co–P shell (red) on crystalline Co core (blue). Adapted with permission from ref. 3, Copyright 2007 AAAS, and ref. 29, Copyright 2011 Royal Society of Chemistry.

A number of previous CTNMs studies have focused on changes to the electronic and/or optical properties.^7,13,21,31 Less work has been done specifically utilizing CTNMs to manipulate composition to produce useful energy materials. Of interest is the use of chemical transformations to access metastable materials, which for the balance of this article refers to heterostructures (graded composition or sharp interfaces, e.g., striped CdS–Ag₂S nanorods^3,6), doped phases (e.g., Co-rich Co₂P (supra-stoichometric)²⁹), interdiffused phases (e.g., ε-Co with CoO³⁰), and/or amorphous structures (e.g., ε-Co crystalline core with a Co–P amorphous shell²⁹). Also included in the metastable class of materials is the formation of crystalline structures of materials that are normally unstable when synthesized via conventional means (e.g., hexagonal Cu₂Se).²⁰ All of these types of metastable structures can be considered as a new class of unique materials that could potentially lead to the next generation of nano-energy materials^32,33 (e.g., for fuel cells and/or Li-ion batteries). CTNMs can be a powerful means to convert NMs into useful new metastable structures for energy applications (Fig. 1).

Metastable and mixed phase metals have notably already proven their worth in the area of catalysis, where certain alloys have been extensively studied in order to alter the electronic structure of the parent material to achieve higher activity and/or stability than their pure phase/composition analogues. This was famously demonstrated with thin films of Pt₃Ni(111) that were shown to have 10× the oxygen reduction reaction (ORR) catalytic activity of pure Pt(111) surfaces and to be over 90 times more efficient at ORR than Pt nanoparticles, most likely due to a d-band shift of the Pt lattice due to alloying with Ni.³⁴ Bimetallic core–shell metal nanoparticles have demonstrated similar exceptional catalytic properties due to synergetic effects (e.g., Au–Pd).³⁵ Pt–Pt₃Co³⁶ catalysts have been reported to have the highest mass activity of all Pt–Co catalysts for ORR, showing the applicability of bimetallic core–shell nanomaterials. Metastable ionic/covalent nanomaterials, however, are less rigorously studied in relation to energy applications; clearly there is much room for new discoveries in this area.

Herein, we present a forward-looking review article that discusses the use of chemical transformations of nanomaterials to produce useful products for energy applications, specifically, the hydrogen evolution reaction (HER), the oxygen reduction reaction (ORR), and Li-ion battery materials. This article will focus on non-metallic NMs. In Section 2 we present an overview of the concepts of CTNMs. Section 3 gives several examples of nanomaterials for energy applications that were synthesized via chemical transformation protocols. This section is designed to illustrate the utility of these materials for catalysis and electrochemical energy storage. Section 4 highlights metastable materials used in energy applications that have not been formed through CTNMs but should be accessible by CTNMs. In the final section (Section 5) we propose new metastable material systems for energy applications that can be synthesized through chemical transformations of nanomaterials, and should have certain advantages over their pure phase analogues.

2. Overview of Chemical Transformations of Nanomaterials (CTNMs)

From a review of recent literature it is clear that the new focus of nanoparticle synthesis and characterization has migrated from first-generation (as-synthesized) to nanomaterials that are derived from the first-generation nanomaterials.³⁷ The large surface to volume ratio of nanomaterials, and finite number of constituent atoms, allows chemical transformations to behave differently at the nanoscale than in bulk. For instance, in nanomaterials the high activation barriers for diffusion may be reduced, permitting rapid diffusion of atoms in a solid,¹ which can then enable the formation of new metastable structures.³ CTNMs is defined as reactions performed on as-synthesized nanomaterials to chemically adjust their composition, structure, and/or morphology. The chemical transformation may be an addition or an exchange reaction. The driving forces for these reactions are either a redox couple or the thermodynamics of the solvation of the ions in solution versus in a solid. In the simplest case of a redox exchange reaction, metal (M) particles in the presence of excess cations/species (N^+z) results in the oxidation of M along with the reduction of the cationic species, leading to N⁰ particles provided that N^+z/N has higher reduction potential than M. An example of a redox exchange chemical transformation as recently shown with Mn₃O₄ in the presence of Fe(ClO₄)₃, that transformed into hollow Mn₃O₄–Fe₂O₃ nanoboxes with the final product being hollow Fe₂O₃ nanocrystals.³⁸ This was explained through the reduction potential of the Fe⁺³/Fe⁺² (E = 0.77 V SHE) driving the reduction of Mn₃O₄/Mn⁺² (E = 1.82 V SHE).

In a redox addition reaction when the initial NM atoms become oxidized and the foreign atoms are reduced there is the possibility for a nanoscale Kirkendall effect, but if the initial atoms are reduced then the Kirkendall effect is not usually observed.³⁹ An example of a redox addition reaction with the nanoscale Kirkendall effect, metal M particles annealed in the presence of O₂ (a covalent system) result in an oxidation of the metal and a reduction of the oxygen, based upon their relative reduction potentials, producing M_xO_y. The asymmetric diffusion rates between the two atomic components in the diffusion couple can result in a hollow nanomaterial, called the nanoscale Kirkendall effect. This was originally demonstrated in the cobalt metal system, transforming the Co NMs into hollow cobalt chalcogenides^2,40 (Fig. 3). Researchers had originally believed that the Kirkendall hollowing was a straightforward mechanism in nanomaterials, directly related to the bulk Kirkendall effect. However, recent work using X-ray absorption spectroscopy and density functional theory on the cobalt oxide and cobalt phosphide systems has found that the mechanism of the nanoscale Kirkendall effect more complex than previously believed: the nanoscale Kirkendall effect is not a one-step process where hollowing simply occurs due to asymmetric diffusion rates between the anion and cation, but involves several distinct steps resulting from phase-related diffusion constants and processes.^29,30 For instance, for nanoparticles of cobalt converting into cobalt oxide, oxygen diffuses inward into the ε-Co in the beginning of the transformation, forming an amorphous Co–O layer. When the major phase of this layer becomes crystalline CoO outward diffusion of Co becomes more favorable.³⁰ The oxygen diffuses inward through an interesting indirect two-step mechanism, with the O hopping between interstitial sites and type I Co vacancies.³⁰ The cobalt phosphide system was shown to follow a similar process, with phosphorous in-diffusion initially faster than cobalt out-diffusion, but once the Co–P amorphous shell crystallizes to Co₂P then the diffusion rates invert and the characteristic Kirkendall hollowing is seen.²⁹ Characterization of this process identified a new metastable amorphous state that precedes the Kirkendall hollowing (Fig. 2, right).²⁹

Fig. 3 Redox addition reaction and accompanying nanoscale Kirkendall effect. (left) Schematic showing Co nanoparticles in presence of sulfur transforming into Co₉S₈via redox addition, and resulting in hollow Co₉S₈ due to the nanoscale Kirkendall effect. (A) TEM images of initial Co nanoparticles. (B) Co₃S₄ hollow nanoparticles after injection of sulfur into Co nanoparticle solution. (C) HRTEM images of Co₃S₄ (left) and Co₉S₈ (right). (D) TEM image of hollow Co₉S₈. (E) Formation of hollow CoSe. Adapted with permission from ref. 2 and 40, Copyright 2004 AAAS.

Ionic exchange is another method to chemically transform nanomaterials. The most prevalent of these is the cation exchange in which a compound such as CdSe in the presence of Ag⁺ ions transforms into Ag₂Se + Cd⁺² (Fig. 4). The driving force behind these exchanges is solvation, meaning that the parent cation is more soluble in solution than in the solid, resulting in a net dissolution of the parent cation in the presence of excess daughter cations.⁴¹ The process is also aided by the high surface area/volume of the nanomaterials, which lowers the activation energy for the diffusion of the cations.⁴¹

Fig. 4 Reversible cation exchange in nanocrystals. (top) Schematic showing cation exchange between CdSe and Ag₂Se nanocrystals. (below) TEM images of: (A) initial CdSe nanoparticles, (B) cation exchanged Ag₂Se particles, and (C) recovered CdSe from second cation exchange. (D–F) Cation exchanged CdTe and Ag₂Te tetrapods. Adapted with permission from ref. 1, Copyright 2004 AAAS.

The reaction time for ion exchange can be much faster in nanoparticles than would be expected by scaling a bulk characteristic time to a nano-scale material. Considering diffusion, using the values reported for large Cd particles,⁸ the diffusivity of Cd in CdS extrapolated to room temperature is ∼10⁻²¹ cm² s⁻¹.⁴² For a nanomaterial with a characteristic dimension of 10 nm the diffusion time is then ∼10⁸ s.⁴³ However, cationic diffusion-based exchanges occur on the order of milliseconds at room temperature in nanoparticles,^1,44 indicating that other effects may be at play.

Cation exchange reactions have been used to convert CdS to Cu₂S, forming interfaces as the Cu₂S forms at the ends of the precursor nanorods and then grows inwards, which was attributed to high stability of the CdS–Cu₂S interfaces.¹⁰ Cation exchange has been used to dope particles³¹ and to create highly luminescent particles.⁴⁵

The other form of ionic exchange is the anion exchange. Anionic exchange is not as common as cationic exchange, but there are several reports^11,46–49 in the literature, such as the conversion of Al₂O₃ to AlN.⁵⁰ Anionic exchanges occur when a compound such as MX is in solution (or in contact with) anion Y⁻², which replaces the precursor anion to form M₂Y + X⁻. It has been shown for the conversion of ZnS to ZnO as the exchange between S⁻² and O⁻² occurs, the Zn⁺² ions start diffusing outwards. As the Zn⁺² diffuses outwards, O diffuses faster than S, leading to the formation of hollow ZnO that has the same crystalline structure as the original ZnS.¹¹ An interesting anion exchange was reported to create heterostructured rods, in which Ag₈W₄O₁₆ nanorods were transformed into Ag₈W₄O₁₆–AgCl by reacting NaCl with the precursor nanorods.⁵¹ Recently, our group has shown that (NH₄)₂S can be used as a low temperature anion exchange reagent in solution by transforming CoO nanoparticles to Co₃S₄.²⁴

As described above, the use of chemical transformations has yielded exciting new nanomaterials with interesting structures and compositions. These new material forms give rise to an expanded set of properties that can have useful applications. Of the novel properties that result for CTNMs, doping of nanomaterials through partial cation exchange has gained attention for controlling optical properties,⁵² making them useful for photovoltaics⁵³ as well as plasmonics.^54,55 For example, it has been shown that Ag⁺ doping of CdSe shows enhanced fluorescence combined with transition from n-type to p-type depending on doping level.²³ However, other applications of CTNMs have received less attention. With such composition and structure control, CTNMs should be useful for tailoring metastable materials for alternative energy sources, particularly fuel cell catalysis as well as Li-ion battery materials.

The field of fuel cell catalysis has attracted interest due to the propensity of clean renewable energy. Generally, the hydrogen fuel cell is regarded as the most promising configuration because of the high potential differences between the anodic reaction (H₂ oxidation) and the cathodic reaction (ORR). Since a fuel cell is in essence a galvanic cell with replaceable cathodic and anodic reactants, a large potential difference between the two compartments is ideal. With a larger potential gap, the reaction becomes more energetically and thermodynamically favorable, thus more power will be produced. In the hydrogen fuel cell, the potential difference is approximately 0.7–0.9 V, depending upon the catalyst used. However, a critical component is the high cost of the precious metals used as catalysts along several stages in the cycle of fuel cells, from generating the H₂ fuel to reducing the O₂ during operation. The development of a catalyst to generate the hydrogen fuel through the hydrogen evolution reaction (HER) is a key focus of research within the past decade. The most common catalysts for this reaction are the Pt-group metals, specifically Pt⁵⁶⁻⁵⁹ and Pd.^57,60 As a result of their high cost and rarity in nature, much work has been done to limit the amount of these materials or to find promising alternatives such as MoS₂.^61–63

Within the operation of the fuel cell the majority of catalysis research is focused on the cathodic reaction, ORR. As with HER, it is widely accepted that the best metal catalyst for ORR is Pt.^58,64 To reduce the amount of Pt needed researchers are investigating the use of Pt alloys such as Pt₃Ni,^34,65,66 as well as the use of thin films^67–73 and nanomaterials.^74,75 In addition, work has been aimed at finding alternatives to Pt, mainly the use of metal chalcogenides such as Co₃O₄.^76–78 It is worth noting that metastable materials have not been extensively tested for HER/ORR, which, considering the unique electronic effects that are possible in metastable compositions, warrants further investigation. Additionally, chemical transformations could impart advantages such as manipulating composition of the NM to be a controlled amalgam of two well-known catalysts to take advantage of each of their strengths, or formation of graded core–shell structures that could impart a soft-confinement potential.⁷⁹

For electrochemical energy storage, Li-ion batteries have already been commercialized on a large scale with the main cathodic material being LiCoO₂. LiCoO₂ is well known to degrade under high applied voltages but improvements continue to be made.^80,81 New cathodic materials with low cost and low charging times are being investigated, such as LiFePO₄.^82–87 It should be noted that the low charging time of LiFePO₄ is due to low electronic conductivity^86,87 in the plateau region. However, work in this area continues as battery materials with higher capacitance coupled with longer cycling ability are needed. In particular, the use of cobalt oxides has been well studied with Co₃O₄^88–91 being among the most examined anodic materials. Most of the battery nanomaterials are synthesized directly into their stable form. To the best of our knowledge, little work has been done on the use of CTNMs in order to engineer advanced materials. The ability to access new cathodic/anodic phases could lead to materials with high stability/performance.

The next section briefly summarizes materials synthesized via chemical transformation protocols for fuel cell catalysis as well as Li-ion battery technology. Specifically, the materials presented highlight the different methodologies of CTNM discussed in the previous section as shown in Fig. 1.

3. Chemical transformations of nanomaterials used in energy applications

In this section we discuss direct examples of CTNMs for energy applications. We focus on several examples from literature where the as-synthesized nanomaterial was chemically transformed to produce a more useful material for the specific application. The reactions chosen highlight energy applications that are among the most heavily investigated: HER, ORR and Li-ion battery anode materials.

3.1. Hydrogen Evolution Reaction (HER)

The use of noble metals for catalysis has been studied in part because of their high stability. Nanomaterial catalysts replaced bulk catalysts due to the scarcity of noble metals, coupled with the high surface area/volume ratio afforded by nanomaterial-based catalysts. The scarcity of Pt-group metals has necessitated an in-depth analysis of non-Pt based catalysts for HER. Interest has also been drawn away from electrochemical catalysis and more toward renewable methods, such as photocatalysis.

CdS–PdX. The cation exchange reaction was used to fabricate photocatalysts for HER consisting of CdS nanorods with Pd₄S tips.⁹² For the synthesis, CdS nanorods were injected into Pd(acetylacetonate)₂, oleylamine, and oleic acid and heated to 200 °C. The synthesis resulted in 5.8 nm rods with Pd-based tips, which X-ray Diffraction (XRD) identified as Pd₄S (Fig. 5). Cation exchange is the most likely explanation for the growth since the final product has the same dimensions as the CdS starting material, but with Pd₄S tips. However, the expected product of a cation exchange should be PdS not Pd₄S indicating that other factors were at play. Pd₄S is not expected since it is not the thermodynamic product: the Pd₄S contains Pd⁰, and speculation for the mechanism was based on the knowledge that long chain amines can reduce Pd⁺² to Pd⁰.

Fig. 5 CdS–PdX hybrid structures for HER. (a–d) CdS nanorods with tip growth of Pd₄S formed through a modified cation exchange reaction. (bottom) Photocatalytic H₂ production by CdS with tips of Pd₄S, compared against heterostructures of CdS rods with small PdO dots, CdS with body exchanged Pd₄S, and CdS rods with large PdO dots. Adapted with permission from ref. 92, Copyright 2011 Wiley.

The photocatalytic properties were tested in sulfide/sulfite solution (Fig. 5). The nanorods with a small amount of Pd₄S are 2 times (∼9.5 μg h⁻¹) more efficient at H₂ evolution than the small particle PdO based CdS hybrids that were also synthesized. A major consideration in this reaction is the ability to produce hydrogen consistently over time. Even though the yield of the CdS–Pd₄S nanorods was low based upon the amount of photons adsorbed/mass H₂ generated, the catalytic response was essentially a linear function. It is interesting to note that the amount of Pd₄S dictates the activity (more Pd₄S leads to a worse performance). The proposed mechanistic rationale was through a raised Fermi level in the Pd₄S catalyst, in which the high Fermi level permits better electron transfer to nearby water molecules. Another possibility is that larger Pd₄S sites near the CdS–Pd₄S interface favor electron–hole recombination versus electron transfer to water.

MoO₃–MoS₂. The use of metal chalcogenides for electrocatalysis has been pursued due to their low cost in relation to Pt. MoS₂ has been of interest for the hydrogen evolution reaction because of its stability in acidic media, however, performance of this material for HER is relatively low since conductivity is limited in the bulk.⁶¹ To this end MoO₃–MoS₂ core–shell nanowires were synthesized for the HER with the metal oxide as the core.⁶¹ Nanowires of MoO₃ synthesized via chemical vapor deposition were sulfidized at 200 °C and higher temperatures in 90% H₂/10% H₂S (Fig 6). This is an example of an anion exchange with the S²⁻ replacing the O²⁻. The XRD results show that the sulfidization treatment results in some loss of crystalline order of the MoO₃ as the MoS₂ forms.

Fig. 6 MoO₃–MoS₂ core–shell nanowires obtained through anion exchange for HER. (a) Original MoO₃ nanowires. (b–d) Anion exchanged nanowires at reaction temperatures 200, 300, and 400 °C. (bottom left) Electrochemical activity of initial MoO₃ wires (blue) and MoO₃–MoS₂ core–shell nanowires anion exchanged at 150 °C (red). (bottom right) Electrochemical activity of MoO₃–MoS₂ core–shell nanowires anion exchanged at 200 and 300 °C, and FTO support substrate. Adapted with permission from ref. 61, Copyright 2011 American Chemical Society.

The Mo-based nanowires were then tested for their electrochemical performance for the HER (Fig. 6) in 0.5 M H₂SO₄. Compared to MoO₃, MoS₂ has high stability in acidic solution, while MoO₃ is unstable and readily decomposes. The MoS₂ begins to evolve H₂ at ∼−0.15–0.2 V negative of the H₂ equilibrium potential, which is approximately 50 mV more negative than on Pt(111) and Pt(100) catalysts,⁵⁶ meaning, the amount of energy to initiate HER on Pt^56,57 is less than the Mo-based catalyst. The Mo-based catalyst has an extremely high current density of 10 mA cm⁻² that is higher than Pt single crystals^56,57 (MoO₃–MoS₂ has a more negative onset potential than Pt as HER electrochemical activity is potential-based), indicating good activity for the reaction. The catalyst was exceptionally durable: the material is stable for over 10 000 cycles with no loss of activity as the stability is attributed to the metal sulfide shell protecting the metal oxide core from dissolution. It is important to note that the top performing MoO₃–MoS₂ core–shell nanowires were obtained through a partial anion exchange, with the wires retaining crystal character from both MoS₂ and MoO₃. This reinforces the benefits of metastable materials produced through CTNMs for their system.

3.2. Oxygen Reduction Reaction (ORR)

Co–CoO. Nanoparticle catalysts of cobalt oxide have been recently extended beyond the popular Co₃O₄ with the use of Co–CoO core–shell nanoparticles for ORR.⁹³ The nanoparticle synthesis of the cobalt core consisted of the thermal decomposition of the [Co₂(CO)₈] precursor, in a solution of oleic acid, dioctylamine, and tetrahydronaphthalene. The CoO shell was grown by oxidizing cobalt core (redox addition reaction). Butylamine was used to remove the native long-chain surfactants. Transmission Electron Microscopy (TEM) images of the initial Co nanoparticles show their high monodispersity with an average size of 10 nm (Fig. 7). Exposed to air in ambient conditions resulted in 8 nm Co cores with 1 nm CoO shells. These 8 + 1 nm core–shell structures were the best performing structure in ORR. Further heating of the core–shell resulted in the formation of hollow CoO nanoparticles through the nanoscale Kirkendall effect.

Fig. 7 Co–CoO core–shell nanoparticles formed through redox addition reaction for ORR. (A) TEM images of initial Co nanoparticles. (B) Final Co–CoO. (C and D) Co–CoO nanoparticles after different air exposure times (17 h, 96 h). (E) Co–CoO nanoparticles deposited on graphene (G–Co–CoO). (bottom left) ORR polarization curves of the G–Co–CoO and Pt nanoparticles on glassy carbon (C–Pt). Solution: O₂ saturated 0.1 M KOH. Scan rate: 10 mV s⁻¹. Rotation rate: 400 rpm. (bottom right) Chronoamperometry at −0.3 V for the ORR on the G–Co–CoO nanoparticles and commercial C–Pt catalysts. Solution: O₂ saturated 0.1 M KOH. Rotation rate: 200 rpm. Adapted with permission from ref. 93, Copyright 2012 Wiley.

The Co–CoO catalyst was tested for ORR and showed favorable results when dispersed on graphene (Fig. 7). Compared to Pt nanoparticles, the Co–CoO catalyst has comparable activity: the E_1/2 potential difference between the Pt and Co–CoO is 0.025 V, but the Co–CoO has a higher current density and steeper polarization curve. (The E_1/2, which is the half width potential in the polarization curve, is a catalyst comparison characterization tool that can be used as an alternative to comparing the onset potential for ORR). Longer nanoparticle annealing times shift the ORR kinetics to more negative potentials, due to the structural changes in the nanoparticles (NPs) as the oxide shell thickens. Under chronoamperometric conditions, the Co–CoO catalyst has higher stability than pure Pt for 60 000 s (Fig. 7). Rotating Disk Electrode (RDE) results show that the Co–CoO catalyst reduces oxygen via a 4e-pathway based upon information resolved from the Levich plot. (RDE rotates the working electrode whereby laminar flow of solution towards and across the electrode results in current signal if redox species is present). The Co–CoO catalyst has higher activity and durability for ORR than Pt even though the E_1/2 is less than Pt. Overall, these results indicate that this material would be an encouraging replacement for Pt in alkaline fuel cells. It is noteworthy that peak performance was achieved through a partial chemical transformation, resulting in a 1 nm CoO shell. Previous work on this system by our group using X-ray absorption spectroscopy points to an amorphous CoO and potentially interdiffused interface between structures of CoO and Co.³⁰

3.3 Li-ion batteries

Cu–Cu_xS. Active searches are underway to find next generation materials for anodes in Li-ion batteries. These materials should have high power density as well as high charge–discharge rates coupled with high cycling ability. Copper sulfide has gained widespread interest as a Li battery anode, and metal sulfide use for Li-ion batteries has been investigated for many years.⁹⁴ A main problem from the Cu_xS group is the ethylene carbon (electrolyte) losing capacitance due to Li polysulfide formation during charge–discharge cycling.⁹⁵ The stability can be improved by capping of sulfide by polyethylene glycol (PEG) inside porous carbon. To avoid the use of carbon, core–shell Cu–Cu_xS nanotubes were synthesized. Briefly, Cu nanowires were solvated in EtOH/PEG followed by addition of thiourea in EtOH.⁹⁵ After purging with Ar for 10 min, the reaction was heated to 90–95 °C for 30 min, and then degassed for 10 min. Finally the reaction was heated to 180–190 °C for 30, 60, and 120 min. This synthesis is the result of a galvanic (redox) addition reaction and nanoscale Kirkendall hollowing as evidenced by Field-Emission Scanning Electron Microscopy (FESEM), which shows the initial Cu nanowires transform into nanotubes (Fig. 8). Scanning Area Electron Diffraction (SAED), XRD, and X-ray Photoelectron Spectroscopy (XPS) showed that the nanomaterials were polycrystalline with a blend of Cu and Cu_xS, depending on the reaction time. All cases showed either core–shell characteristics from the partial exchange (Cu core with Cu_xS shell for the 30 and 60 minute reactions) or a mixture of phases (Cu₂S and CuS for the 120 minute reaction).

Fig. 8 Cu–Cu_xS core–shell nanotubes obtained through addition reaction for Li-ion battery electrodes. (a) Low-magnification and (inset) high-magnification FESEM image of Cu–Cu_xS core–shell nanotubes. (b) Cycling performance of the Cu–Cu_xS nanotube samples with different synthesis reaction times (30, 60, and 120 min). Current density: 200 mA g⁻¹. Potential Range: 0.001–3.0 V. (c) Cycling performance (varying current densities) of the Cu–Cu_xS nanotubes with different synthesis times (30, 60, and 120 min). (d) Charge–discharge curves (1,3, and 5 cycles) for Cu–Cu_xS. Current density: 200 mA g⁻¹. Adapted with permission from ref. 95, Copyright 2012 American Chemical Society.

The Cu–Cu_xS nanotubes were tested as both a cathode and anode for a Li-ion battery (Fig. 8). Although there is not an extensive plateau in the charging–discharging regime for this material, the redox transformations during charge–discharge are reversible, indicating stable battery performance. Maximum specific capacities occurred for the 120 minute sample with a fifth-cycle discharge capacity of 535 mA h g⁻¹. However, the 60 minute sample had a better specific capacity at high current densities (≥2 A g⁻¹). All samples showed stability up to 50 cycles during cycling performance at a current density of 200 mA g⁻¹, and had stable cycling performance under a set of different current densities. As a general measure, long term battery anodes should feature both high charge–discharge capacities coupled with high stability. The Cu–Cu_xS nanomaterial exhibits high stability under cycling conditions, making it a good candidate for next generation Li-ion batteries.

4. Metastable materials with energy applications

In this section we discuss examples of metastable materials with energy applications that were not formed through chemical transformations, but are candidates for the CTNM process. We focus on several exciting examples from literature where the nanomaterial was synthesized directly using common practices. A synthetic route for these materials could someday be achieved using CTNMs, which may lead to enhanced properties.

4.1. Fuel cell catalysis

Zn_xCd_1−xS for HER. Nano-based metal sulfides have been widely researched as photocatalysts because of their advantageous properties, with CdS rising to the top due to its ability for visible light absorption and well-aligned conduction band edge, which is negative to the H₂O/H₂ redox couple.^96–98 However, CdS is susceptible to photocorrosion and degrades in photocatalytic reactions.⁹⁹ To overcome this issue Zn_xCd_1−xS has been explored and is now one of the most efficient and extensively used photocatalysts, mainly for dye degradation.^98,100 Zn_xCd_1−xS has also been tested in the HER field as a co-catalyst with MoS₂ to evolve hydrogen.⁶² In this case, the photocatalytic properties of the Zn_xCd_1−xS nanomaterial were tested in 0.35 M Na₂S + 0.25 M Na₂SO₃ solution. The catalyst works by creating electron–hole pairs in the presence of light. The electrons migrate to the MoS₂ which evolves H₂, with maximum output of ∼9.5 mmol g_cat⁻¹ over a time span of 24 hours.⁶² The Zn_0.2Cd_0.8S stoichiometry was found to be the most active of all compositions as hydrogen generation increased linearly with time (Fig. 9), implying that the hydrogen generation is constant with respect to time.⁶² Constant hydrogen production with time is the ideal situation for a HER catalyst. In order to evaluate the durability of the catalyst, the particles were immersed in fresh solution for 30 minutes and used for HER, followed by solution change. The catalyst showed no degradation after catalyst turnover.

Fig. 9 Zn_xCd_1−xS nanoparticles for HER. Zn_xCd_1−xS hydrogen photogeneration under visible light irradiation in 0.35 M Na₂S + 0.25 M Na₂SO₃ solution. (left) Zn_xCd_1−xS–MoS₂ 3% hybrid photocatalysts: (a) control sample of Zn_0.2Cd_0.8S without MoS₂, (b) x = 0.8, (c) x = 0.6, (d) x = 0.4 and (e) x = 0.2. (right) Zn_0.2Cd_0.8S with different MoS₂ loadings. Reprinted with permission from ref. 62, Copyright 2013 Royal Society of Chemistry.

In 2013, Zn_xCd_1−xS nanoparticles were used for HER without a co-catalyst.⁹⁷ The Zn_xCd_1−xS particles were mixed with carbon nanotubes (CNTs) and allowed to react in an autoclave at 140 °C for 12 h followed by Ar purging for 10 minutes.⁹⁷ H₂ generation was tested in aqueous Na₂S + Na₂SO₃ (Fig. 10). Overall, the Zn_xCd_1−xS particles with CNT has higher rates of production (1.5×) compared to the non-CNT material. It was determined that the rate of production was heavily influenced by the Cd amount, with the Zn_0.83Cd_0.17S being the best solid solution (6.03 mmol h⁻¹ g⁻¹). (Note: this result cannot be quantitatively compared with noble metal catalysts such as Pt, as the Pt electrocatalyzed reaction is heavily dependent on the applied potential, which affects the rate of evolution of hydrogen). A possible reaction mechanism was proposed: electrons excited from the S 3p valence band to the metal C-band (hybrid of the Cd 5s5p and Zn 4s4p), creating a hole in the S valence band. The smaller bandgap of the CNT allows the electron transfer to the CNT, followed by the reduction of H⁺ ions to H₂. Lastly, the sulfur ions consume the electron holes in the Zn_xCd_1−xS.⁹⁷ It is interesting that hydrogen generation is maximum at a high Zn content (x > 0.75), both with and without CNTs.

Fig. 10 Zn_xCd_1−xS nanomaterials for HER. The rate of H₂ evolution on Zn_xCd_1−xS and Zn_xCd_1−xS/CNTs photocatalysts in sulfide/sulfite solution. Reprinted with permission from ref. 97, Copyright 2013 Royal Society of Chemistry.

It's important to note that the XRD results showed that all of these nanoparticles were not a mixture of CdS and ZnS, but a solid solution.^97,98,100 This implies that additional structuring of the cations through CTNMs could lead to modified properties (e.g., tunable bandgap).

MnCo₂O₄ for ORR. The use of metal oxide nanoparticles for ORR has been increasingly investigated. The Dai group has found useful performance and properties of cobalt oxides such as Co₃O₄⁷⁶ and CoO¹⁰¹ on conductive supports such as graphene and CNTs. Generally, ORR metal oxides have a maximum catalytic loading, owing to lack of conductivity when this threshold is reached. If too much oxide catalyst is loaded, then the catalyst is non-conductive and results in little if any current flow even though only the topmost layers of material are in contact with redox species. However, by depositing the oxide nanoparticles on highly conductive supports such as graphene or nitrogen reduced graphene oxide (N-rmGO) a synergistic effect between the particles and the support is formed, enhancing the conductivity and activity.¹⁰²

This synergistic relationship between particles and substrate was demonstrated using MnCo₂O₄ nanoparticles supported on nitrogen-reduced graphene oxide, which was shown to have good ORR activity and high stability in alkaline solution (Fig. 11).¹⁰² Unlike Pt which is known to strongly adsorb anions (e.g. OH⁻)¹⁰³ in the same potential region that ORR occurs, metal oxides are not as prone to adsorbates. XRD results showed that Mn was substituted for the Co in the crystal lattice. Energy Dispersive Spectroscopy (EDS) and XPS results confirmed the Mn : Co ratio of 1 : 2 coupled with ∼4% N in the graphene structure.

The catalyst loading was kept to ∼20% wt: higher loading will result in insulating behavior and no electrochemical reaction. The Mn doped catalyst was compared with other Co oxides such as Co₃O₄ and CoO. It was found that the Mn-based catalyst has higher activity for ORR evidenced by the oxide peak being the most positively shifted (Fig. 11a). When comparing the ORR curves of MnCo₂O₄ and Pt, the behavior is almost identical, with the E_1/2 of each catalyst being only 0.02 V apart. However, Pt does have the slightly higher half-width potential of the two catalysts (Fig. 11b,c). Rotating Ring Disk Electrode (RRDE) results show that the MnCo₂O₄ particles reduce oxygen via a 4e-pathway, which is the preferred pathway since the 2e-pathway produces the HO₂⁻ intermediate in alkaline solution. Chronoamperometric results indicated a higher tolerance for ORR compared to Pt nanoparticles, with the MnCo₂O₄ retaining 97% activity compared to 60% activity for Pt after 20 000 seconds (Fig. 11d). The Mn-based catalyst shows the same catalytic activity towards ORR as Pt nanoparticles, but with much higher stability, making this a good candidate for an alkaline fuel cell cathode.

Fig. 11 MnCo₂O₄ nanoparticles obtained for ORR. (a) CV curves of MnCo₂O₄ hybrid, MnCo₂O₄ mixture, Co₃O₄ hybrid, and N-rmGO (nitrogen reduced graphene oxide) catalysts on glassy carbon (GC) electrodes (Pt ORR peak represented by dashed line). Solution: O₂ saturated (solid line) or N₂-saturated (dash line) 1 M KOH. (b) Rotating-disk electrode curves of MnCo₂O₄ hybrid, MnCo₂O₄ mixture, Co₃O₄ hybrid, N-rmGO, and Pt nanoparticles (Pt/C) on carbon fiber paper electrodes. Scan rate: 5 mV s⁻¹ Rotation rate: 1600 rpm. (c) ORR polarization curves of MnCo₂O₄ hybrid, Co₃O₄ hybrid, and Pt/C on carbon fiber papers solution: 1 M KOH. (d) Chronoamperometric curves at 0.70 V vs. RHE for MnCo₂O₄ hybrid, MnCo₂O₄ mixture, and Pt nanoparticles (Pt/C) on carbon fiber paper electrodes. Solution: O₂ saturated 1 M KOH. Adapted with permission from ref. 102, Copyright 2012 American Chemical Society.

4.2. Li-ion battery

TiOF₂. TiO₂ has been the subject of intense study owing to high capacity retention at high charge–discharge rates along with good cycling stability.¹⁰⁴ However, the theoretical specific capacity of TiO₂ is lower than graphite (167.5 vs. 372 mA h g⁻¹), while the Li intercalation voltage is higher (1.7 V vs. <0.5 V).¹⁰⁵ Compared to TiO₂, metal oxyfluorides have become attractive anode materials as a result of the theoretically lower Li intercalation voltages than TiO₂ (ca. 1.3 V vs. ca. 1.7 V) as well as higher specific capacity (∼400 mA h g⁻¹vs. 167.5 mA h g⁻¹). TiOF₂ was recently synthesized by injecting a mixture of TiF₄ salt and oleic acid into octadecene at 160 °C for 10 minutes.¹⁰⁵ FTIR data aided in explanation of the synthetic route as well as elucidating the role of the oleic acid in the synthesis as the oleic acid provided water molecules and, in turn, oxygen. TEM images show the length of the nanotube walls is in the range of 40–60 nm, and the SAED results show that the tubes are single crystalline along the long axis, with (001) orientation.

Fig. 12 shows the charge–discharge curves for the nanotubes. As with most battery anode materials the initial charge–discharge rates can be discounted due to solvent degradation. The second set of curves show a discharge of 513.7 mA h g⁻¹ and a charge of 451 mA h g⁻¹, which is a columbic efficiency of 87.8%. While the initial charging–discharging rate is not exceptionally high, the retention of charge with the number of cycles indicates good stability. The durability was assessed with cycling results show that nanotubes are stable for 60 cycles with only a minimal loss of capacitance.

Fig. 12 TiOF₂ nanotubes in Li-ion battery electrodes. The first two charge–discharge profiles of TiOF₂ nanotubes. Current density: 400 mA g⁻¹. The measurements were conducted with a potential window of 0.005–3.0 V. Reprinted with permission from ref. 105, Copyright 2012 Wiley.

CuInS₂. CuInS₂, which normally is studied for photovoltaic applications, has been tested as a possible anode material for Li batteries.¹⁰⁶ Two sets of nanospheres were synthesized: one set hollow and the other solid. XRD results confirmed the formation of the tetragonal chalcopyrite phase, in which the In and Cu occupy the same sites.^106,107 SAED result show that the spheres are polycrystalline with the (112), (220), and (116) diffraction rings being present. The EDS result showed that 1.1 : 0.9 : 1.9 was the Cu : In : S ratio.

Fig. 13 (top) shows charge–discharge curves for the CuInS₂ nanospheres. The second and third cycles are within 85% of each other based upon examination of the charge–discharge curves as the discharging capacities for the second and third cycles are 1004 and 903 mA h g⁻¹. Fig. 13 (bottom) shows the cycling performance of the nanospheres. Of interest is that the hollow spheres have improved stability in their specific capacity (charge–discharge) after the first 10 cycles and this is maintained over the course of 20 cycles with an initial loss of over 50% specific activity.

Fig. 13 CuInS₂ nanospheres obtained in Li-ion battery electrodes. (top) Charge–discharge curves for the first three cycles for the CuInS₂ hollow nanospheres. Current density: 30 mA g⁻¹ (bottom) Specific capacity versus cycle number for charge (filled) and discharge (open) on the CuInS₂ hollow nanospheres (■,□) and nanoparticles (●,○). Current density: 30 mA g⁻¹ between 3 and 0.005 V. Adapted with permission from ref. 106, Copyright 2012 Elsevier.

Li_1.2Mn_0.6−0.5xNi_0.2−0.5xCo_xO₂. Li-rich oxides are being heavily researched since they should have higher capacity than LiCoO₂, the commercial battery cathode. Li_1.2Mn_0.6−0.5xNi_0.2−0.5xCo_xO₂ (x = 0.06, 0.13, and 0.2) was studied as a potential cathode material because of the variable length of the plateau in the discharge region.¹⁰⁸ This plateau involves the irreversible oxidation of the O²⁻ ions and a release of oxygen from the layered oxide lattice. Another possibility for the variable plateau length is the oxidation of the transition-metal ions past +4. These reactions in the plateau result in a lower oxidation state for the metal ions at the end of the first discharge, which permits a higher reversible capacity in the ensuing cycles.

Fig. 14 shows the charge–discharge data for the Li_1.2Mn_0.6−0.5xNi_0.2−0.5xCo_xO₂ cathode material. It is interesting that increasing the Co concentration first increases then decreases the discharge capacity, as well as the charge capacity. The maximum is reached at x = 0.06, with the discharge being ∼230 mA h g⁻¹. The plateau length of the discharge region decreases with increasing Co concentration due to the overlap of the Co^+3/+4 t_2g and O 2p electron overlap (Fig. 14). Finally, the cycling performance of the material showed ∼10% capacity drop for 40 cycles.

Fig. 14 Li_1.2Mn_0.6−0.5xNi_0.2−0.5xCo_xO₂ particles in Li-ion battery electrodes. (top) First charge–discharge curves for the Li_1.2Mn_0.6−0.5xNi_0.2−0.5xCo_xO₂ nanoparticles at C/25 rate with a potential window of 2 to 4.5 V. (bottom) Qualitative energy diagram depicting the positions of metal 3d bands with respect to the top of the O₂ 2p band. Adapted with permission from ref. 108, Copyright 2012 Royal Society of Chemistry.

5. Future work – proposed new systems

As has been seen in the previous sections, non-metallic metastable nanomaterials and chemically transformed nanomaterials have a proven worth in energy applications. The shortfall in this field, however, is the paucity of work specifically examining metastable materials that result from partial chemical transformation reactions (e.g., partial ionic exchanges). Partial reactions during CTNMs can yield interesting metastable intermediates that could impart additional functionality upon the materials. In this section we propose new systems that could be formed by CTNMs, with potential battery and catalysis benefits as metastable phases.

In the ORR, both CoO^93,101 and Co₃S₄^109,110 are useful catalysts. Cobalt sulfide is used in acidic media which is too harsh for metal oxides, but the cobalt oxide is an overall better catalyst. To this end, a metastable Co_xO_yS_4−y structure (either graded core–shell or a single mixed phase) should be durable enough for acidic solution but have the higher activity of cobalt oxide.

Combinations of metal oxides could be useful as catalysts for the ORR reaction. For example, Co₃O₄,^76–78 Fe₃O₄,¹¹¹ and Mn₃O₄¹¹² are all catalysts and have been significantly tested in alkaline solutions. The Fe based nanoparticles have similar activity but shifted half width potential to more negative potentials compared to Co₃O₄, and have higher durability than Pt. The Mn nanomaterials however, have slightly higher current densities than the Co-based catalysts. Metastable mixtures of Fe_xMn_3−xO₄ and Co_xMn_3−xO₄ (where x does not equal 1or 2)^102,113 would combine the durability of Fe, the activity of Mn with the high activities/durability of Co and Mn catalysts. This would also allow for characterization in alkaline media.

It is well documented that PtRu^114,115 alloys have high catalytic capabilities. Ru chalcogenide compounds have also garnered attention for ORR reactions, specifically RuSe¹¹⁶ and RuSe₂,¹¹⁷ because they are stable in acidic solution and also because RuSe₂ has a comparable half width potential to Pt (within 0.1 V). A proposed metastable material that could be accessed through chemical transformations of nanomaterials that would be an attractive catalyst is Ru_xCo_3−xSe_yO_4−y, due to the durability in acidic solution from the selenide content coupled with activity from both Ru and Co.

For Li-ion battery anodes materials Co–CoO,¹¹⁸ CoO,^119,120 Co₃S₄,¹¹⁰ and Co₃O₄^{89–91,121–125} have been researched. The Co–CoO structure has a slightly higher stability, but Co₃O₄ has higher theoretical capacity.¹¹⁸ As an illustration of the activity/stability of cobalt chalcogenides, it has been reported that Co₃O₄ coupled with reduced graphene oxide has a 71% capacity retention during cycling.¹²⁴ A metastable material such as Co_xO_y or graded core–shell CoO–Co₃O₄ could possess both the stability and durability of both parent phases.

Fe₃O₄¹²⁶ and FeO¹²⁷ have also been tested as anode materials for Li-ion batteries, with the FeO having high cycling performance (higher columbic efficiency) compared to Fe₃O₄ nanoparticles and overall better performance than CoO or Co₃O₄. The mixed phases of Fe_xCo_1−xO and Fe_xCo_3−xO₄ should be of interest since the Co materials have higher charging capacities than the Fe oxides, but the Fe oxides are more stable. Also, because of the high efficiency of metal sulfides, materials such as Fe_xCo_3−xS₄ and Co₃S_yO_4−y would also be worthy of attention, combining the stability of the sulfides and the oxide activities.

Cathodic materials for Li batteries continue to be the focus of research, with LiFePO₄^82–87,128 being among the most well-studied materials. In addition, variations of LiFePO₄ such as Li₂FeSiO₄¹²⁹ and LiMn_2−xMo_xO₄¹³⁰ have been recently tested with the silicon containing material having high stability under cycling conditions and comparable capacity to LiFePO₄. The Mo nanomaterial has lower stability than the Si material and comparable capacity. While most work has been on Li metal oxides, metal sulfides have long been known to be excellent cathodes with TiS₂^94,131 being well-studied. Other metal sulfides such as MoS₂^132,133 have been accruing interest as Li battery anodes. MoS₂ have shown comparable in capacity to the previously listed oxides but does not easily oxidize at the high potentials of battery applications,¹³² although LiMn_2−xMo_xO₄ and MoS₂ are cathodic in nature. To this end, a focus on materials such as Mn_xMo_1−xS₂, LiMnMoS₄, and Ti_xMo_1−xS₂ would combine the conductivity of Mo, the stability of Mn, and the activity of Ti.

6. Summary

Chemical transformation of nanomaterials is a young field of research that has provided access to new and exotic nanomaterials. Previous work has shown that the CTNMs method can create metastable materials that have unique electronic, optical, and plasmonic properties. In this highlight review we have shown prominent examples of materials synthesized via CTNMs that are useful for energy applications. We have also pointed to some candidate metastable materials which have high activities/durability for HER, ORR and Li-ion battery technology and could be engineered through CTNMs. Lastly, we have proposed new materials based upon CTNMs that could yield next generation HER and ORR catalysts, and Li-ion batteries. CTNMs could open a new frontier in nanosynthetic chemistry for energy applications by allowing access of nanomaterials that have both unique electronic and stability properties.

Acknowledgements

This work was supported in part by the National Science Foundation under award number CHE-1152922, and the Cornell Center for Materials Research (CCMR) with funding from the Materials Research Science and Engineering Center program of the National Science Foundation (cooperative agreement DMR-1120296). R.D.R. was supported as part of the Energy Materials Center at Cornell (EMC²), an Energy Frontier Research Center funded by the U.S. Department of Energy, Office of Science, Office of Basic Energy Science under Award number DE-SC0001086.

References

1. D. H. Son, S. M. Hughes, Y. D. Yin and A. P. Alivisatos, Science, 2004, 306, 1009–1012.
2. Y. D. Yin, R. M. Rioux, C. K. Erdonmez, S. Hughes, G. A. Somorjai and A. P. Alivisatos, Science, 2004, 304, 711–714.
3. R. D. Robinson, B. Sadtler, D. O. Demchenko, C. K. Erdonmez, L. W. Wang and A. P. Alivisatos, Science, 2007, 317, 355–358.
4. H. J. Fan, U. Gosele and M. Zacharias, Small, 2007, 3, 1660–1671.
5. A. Cabot, V. F. Puntes, E. Shevchenko, Y. Yin, L. Balcells, M. A. Marcus, S. M. Hughes and A. P. Alivisatos, J. Am. Chem. Soc., 2007, 129, 10358–10630.
6. D. O. Demchenko, R. D. Robinson, B. Sadtler, C. K. Erdonmez, A. P. Alivisatos and L. W. Wang, ACS Nano, 2008, 2, 627–636.
7. J. M. Pietryga, D. J. Werder, D. J. Williams, J. L. Casson, R. D. Schaller, V. I. Klimov and J. A. Hollingsworth, J. Am. Chem. Soc., 2008, 130, 4879–4885.
8. A. Cabot, R. K. Smith, Y. D. Yin, H. M. Zheng, B. M. Reinhard, H. T. Liu and A. P. Alivisatos, ACS Nano, 2008, 2, 1452–1458.
9. A. Cabot, M. Ibanez, P. Guardia and A. P. Alivisatos, J. Am. Chem. Soc., 2009, 131, 11326–11328.
10. B. Sadtler, D. O. Demchenko, H. Zheng, S. M. Hughes, M. G. Merkle, U. Dahmen, L. W. Wang and A. P. Alivisatos, J. Am. Chem. Soc., 2009, 131, 5285–5293.
11. J. Park, H. Zheng, Y.-w. Jun and A. P. Alivisatos, J. Am. Chem. Soc., 2009, 131, 13943–13945.
12. J. M. Luther, H. M. Zheng, B. Sadtler and A. P. Alivisatos, J. Am. Chem. Soc., 2009, 131, 16851–16857.
13. P. K. Jain, L. Amirav, S. Aloni and A. P. Alivisatos, J. Am. Chem. Soc., 2010, 132, 9997–9999.
14. J. T. Zhang, Y. Tang, K. Lee and O. Y. Min, Science, 2010, 327, 1634–1638.
15. D. V. Talapin and Y. D. Yin, J. Mater. Chem., 2011, 21, 11454–11456.
16. G. D. Moon, S. Ko, Y. Min, J. Zeng, Y. N. Xia and U. Jeong, Nano Today, 2011, 6, 186–203.
17. M. Ibanez, J. D. Fan, W. H. Li, D. Cadavid, R. Nafria, A. Carrete and A. Cabot, Chem. Mater., 2011, 23, 3095–3104.
18. E. R. Essinger-Hileman, D. DeCicco, J. F. Bondi and R. E. Schaak, J. Mater. Chem., 2011, 21, 11599–11604.
19. K. Miszta, D. Dorfs, A. Genovese, M. R. Kim and L. Manna, ACS Nano, 2011, 5, 7176–7183.
20. H. B. Li, M. Zanella, A. Genovese, M. Povia, A. Falqui, C. Giannini and L. Manna, Nano Lett., 2011, 11, 4964–4970.
21. D. Mocatta, G. Cohen, J. Schattner, O. Millo, E. Rabani and U. Banin, Science, 2011, 332, 77–81.
22. M. Casavola, M. A. van Huis, S. Bals, K. Lambert, Z. Hens and D. Vanmaekelbergh, Chem. Mater., 2012, 24, 294–302.
23. A. Sahu, M. S. Kang, A. Kompch, C. Notthoff, A. W. Wills, D. Deng, M. Winterer, C. D. Frisbie and D. J. Norris, Nano Lett., 2012, 12, 2587–2594.
24. H. Zhang, L. V. Solomon, D.-H. Ha, S. Honrao, R. G. Hennig and R. D. Robinson, Dalton Trans., 2013, 42, 12596–12599.
25. J. A. Medford, A. C. Johnston-Peck and J. B. Tracy, Nanoscale, 2013, 5, 155–159.
26. J. W. Wang, A. C. Johnston-Peck and J. B. Tracy, Chem. Mater., 2009, 21, 4462–4467.
27. J. G. Railsback, A. C. Johnston-Peck, J. W. Wang and J. B. Tracy, ACS Nano, 2010, 4, 1913–1920.
28. C. D. Donega, Chem. Soc. Rev., 2011, 40, 1512–1546.
29. D.-H. Ha, L. M. Moreau, C. R. Bealing, H. Zhang, R. G. Hennig and R. D. Robinson, J. Mater. Chem., 2011, 21, 11498–11510.
30. D.-H. Ha, L. M. Moreau, S. Honrao, R. G. Hennig and R. D. Robinson, J. Phys. Chem. C, 2013, 117, 14303.
31. V. A. Vlaskin, C. J. Barrows, C. S. Erickson and D. R. Gamelin, J. Am. Chem. Soc., 2013, 135, 14380.
32. J. Maier, Phys. Chem. Chem. Phys., 2009, 11, 3011–3022.
33. J. Maier, Adv. Mater., 2009, 21, 2571–2585.
34. V. R. Stamenkovic, B. Fowler, B. S. Mun, G. Wang, P. N. Ross, C. A. Lucas and N. M. Markovic, Science, 2007, 315, 493–497.
35. R. Ghosh Chaudhuri and S. Paria, Chem. Rev., 2011, 112, 2373–2433.
36. D. Wang, H. L. Xin, R. Hovden, H. Wang, Y. Yu, D. A. Muller, F. J. DiSalvo and H. D. Abruna, Nat. Mater., 2013, 12, 81–87.
37. F. Stellacci, Nat. Mater., 2005, 4, 113–114.
38. M. H. Oh, T. Yu, S.-H. Yu, B. Lim, K.-T. Ko, M.-G. Willinger, D.-H. Seo, B. H. Kim, M. G. Cho, J.-H. Park, K. Kang, Y.-E. Sung, N. Pinna and T. Hyeon, Science, 2013, 340, 964–968.
39. U. Jeong and Y. Xia, Angew. Chem., Int. Ed., 2005, 44, 3099–3103.
40. J. M. Buriak, Science, 2004, 304, 692–693.
41. J. B. Rivest and P. K. Jain, Chem. Soc. Rev., 2013, 42, 89–96.
42. Diffusivities were calculated from the Arrhenius equation (D = D₀e^(−E/RT)) using the activation energy and diffusion coefficients from ref. 8. The temperature was set at room temperature (293.15 K).
43. Time was calculated using the equation 〈x〉² = q_iDt, where the mean free path is the square of nanometer dimensions (x = 10⁻⁷ cm) or bulk dimensions (x = 0.1 cm). q_i was set at 6, in order to allow for a comparison in 3 − D, while D was resultant from ref. 42.
44. E. M. Chan, M. A. Marcus, S. Fakra, M. ElNaggar, R. A. Mathies and A. P. Alivisatos, J. Phys. Chem. A, 2007, 111, 12210–12215.
45. P. K. Jain, B. J. Beberwyck, L.-K. Fong, M. J. Polking and A. P. Alivisatos, Angew. Chem., Int. Ed. Engl., 2012, 124, 2437–2440.
46. H. L. Cao, X. F. Qian, C. Wang, X. D. Ma, J. Yin and Z. K. Zhu, J. Am. Chem. Soc., 2005, 127, 16024–16025.
47. L. Dloczik and R. Könenkamp, Nano Lett., 2003, 3, 651–653.
48. J. Geng, B. Liu, L. Xu, F.-N. Hu and J.-J. Zhu, Langmuir, 2007, 23, 10286–10293.
49. F. Dawood and R. E. Schaak, J. Am. Chem. Soc., 2008, 131, 424–425.
50. Q. H. Zhang and L. Gao, J. Am. Ceram. Soc., 2006, 89, 415–421.
51. X. Wang, S. Li, H. Yu and J. Yu, J. Mol. Catal. A: Chem., 2011, 334, 52–59.
52. A. W. Cohn, K. R. Kittilstved and D. R. Gamelin, J. Am. Chem. Soc., 2012, 134, 7937–7943.
53. B. D. Mao, C. H. Chuang, F. Lu, L. X. Sang, J. J. Zhu and C. Burda, J. Phys. Chem. C, 2013, 117, 648–656.
54. A. Grubisic, E. Ringe, C. M. Cobley, Y. Xia, L. D. Marks, R. P. Van Duyne and D. J. Nesbitt, Nano Lett., 2012, 12, 4823–4829.
55. I. Kriegel, J. Rodriguez-Fernandez, A. Wisnet, H. Zhang, C. Waurisch, A. Eychmueller, A. Dubavik, A. O. Govorov and J. Feldmann, ACS Nano, 2013, 7, 4367–4377.
56. N. M. Markovic, B. N. Grgur and P. N. Ross, J. Phys. Chem. B, 1997, 101, 5405–5413.
57. N. M. Markovic, C. A. Lucas, V. Climent, V. Stamenkovic and P. N. Ross, Surf. Sci., 2000, 465, 103–114.
58. V. Stamenkovic, N. M. Markovic and P. N. Ross Jr, J. Electroanal. Chem., 2001, 500, 44–51.
59. T. J. Schmidt, P. N. Ross and N. M. Markovic, J. Electroanal. Chem., 2002, 524, 252–260.
60. L. Kibler, Electrochim. Acta, 2008, 53, 6824–6828.
61. Z. Chen, D. Cummins, B. N. Reinecke, E. Clark, M. K. Sunkara and T. F. Jaramillo, Nano Lett., 2011, 11, 4168–4175.
62. M. Nguyen, P. D. Tran, S. S. Pramana, R. L. Lee, S. K. Batabyal, N. Mathews, L. H. Wong and M. Graetzel, Nanoscale, 2013, 5, 1479–1482.
63. M. A. Lukowski, A. S. Daniel, F. Meng, A. Forticaux, L. Li and S. Jin, J. Am. Chem. Soc., 2013, 135, 10274–10277.
64. E. Antolini, Mater. Chem. Phys., 2003, 78, 563–573.
65. V. Stamenković, T. J. Schmidt, P. N. Ross and N. M. Marković, J. Electroanal. Chem., 2003, 554–555, 191–199.
66. Y. Liu, C. M. Hangarter, U. Bertocci and T. P. Moffat, J. Phys. Chem. C, 2012, 116, 7848–7862.
67. K. Uosaki, S. Ye, H. Naohara, Y. Oda, T. Haba and T. Kondo, J. Phys. Chem. B, 1997, 101, 7566–7572.
68. S. R. Brankovic, J. X. Wang and R. R. Adzic, Surf. Sci., 2001, 474, L173–L179.
69. M. F. Mrozek, Y. Xie and M. J. Weaver, Anal. Chem., 2001, 73, 5953–5960.
70. J. Zhang, Y. Mo, M. B. Vukmirovic, R. Klie, K. Sasaki and R. R. Adzic, J. Phys. Chem. B, 2004, 108, 10955–10964.
71. Y. G. Kim, J. Y. Kim, D. Vairavapandian and J. L. Stickney, J. Phys. Chem. B, 2006, 110, 17998–18006.
72. M. Fayette, Y. Liu, D. Bertrand, J. Nutariya, N. Vasiljevic and N. Dimitrov, Langmuir, 2011, 27, 5650–5658.
73. J. Nutariya, M. Fayette, N. Dimitrov and N. Vasiljevic, Electrochim. Acta, 2013 DOI:10.1016/j.electacta.2013.01.052.
74. K. Kinoshita, J. Electrochem. Soc., 1990, 137, 845–848.
75. S. Guo and S. Sun, J. Am. Chem. Soc., 2012, 134, 2492–2495.
76. Y. Y. Liang, Y. G. Li, H. L. Wang, J. G. Zhou, J. Wang, T. Regier and H. J. Dai, Nat. Mater., 2011, 10, 780–786.
77. J. Xu, P. Gao and T. S. Zhao, Energy Environ. Sci., 2012, 5, 5333–5339.
78. D. Wang, X. Chen, D. G. Evans and W. Yang, Nanoscale, 2013, 5, 5312–5315.
79. X. Y. Wang, X. F. Ren, K. Kahen, M. A. Hahn, M. Rajeswaran, S. Maccagnano-Zacher, J. Silcox, G. E. Cragg, A. L. Efros and T. D. Krauss, Nature, 2009, 459, 686–689.
80. Y. S. Jung, P. Lu, A. S. Cavanagh, C. Ban, G.-H. Kim, S.-H. Lee, S. M. George, S. J. Harris and A. C. Dillon, Adv. Energy Mater., 2013, 3, 213–219.
81. Y. Ou, J. Wen, H. Xu, S. Xie and J. Li, J. Phys. Chem. Solids, 2013, 74, 322–327.
82. J. Hong, C. S. Wang, X. Chen, S. Upreti and M. S. Whittingham, Electrochem. Solid-State Lett., 2009, 12, A33–A38.
83. F. Omenya, N. A. Chernova, S. Upreti, P. Y. Zavalij, K. W. Nam, X. Q. Yang and M. S. Whittingham, Chem. Mater., 2011, 23, 4733–4740.
84. J. Hong, X. L. Wang, Q. Wang, F. Omenya, N. A. Chernoya, M. S. Whittingham and J. Graetz, J. Phys. Chem. C, 2012, 116, 20787–20793.
85. S. Upreti, N. A. Chernova, J. Xiao, J. K. Miller, O. V. Yakubovich, J. Cabana, C. P. Grey, V. L. Chevrier, G. Ceder, J. L. Musfeldt and M. S. Whittingham, Chem. Mater., 2012, 24, 166–173.
86. F. Omenya, N. A. Chernova, Q. Wang, R. B. Zhang and M. S. Whittingham, Chem. Mater., 2013, 25, 2691–2699.
87. F. Omenya, N. A. Chernova, R. B. Zhang, J. Fang, Y. Q. Huang, F. Cohen, N. Dobrzynski, S. Senanayake, W. Q. Xu and M. S. Whittingham, Chem. Mater., 2013, 25, 85–89.
88. C.-T. Hsieh, J.-S. Lin, Y.-F. Chen and H. Teng, J. Phys. Chem. C, 2012, 116, 15251–15258.
89. M. Ren, S. Yuan, L. Su and Z. Zhou, Solid State Sci., 2012, 14, 451–455.
90. L. Pan, H. Zhao, W. Shen, X. Dong and J. Xu, J. Mater. Chem. A, 2013, 1, 7159–7166.
91. A. K. Rai, J. Gim, L. T. Anh and J. Kim, Electrochim. Acta, 2013, 100, 63–71.
92. Y. Shemesh, J. E. Macdonald, G. Menagen and U. Banin, Angew. Chem., Int. Ed., 2011, 50, 1185–1189.
93. S. Guo, S. Zhang, L. Wu and S. Sun, Angew. Chem., Int. Ed., 2012, 51, 11770–11773.
94. M. S. Whittingham, Prog. Solid State Chem., 1978, 12, 41–99.
95. R. Cai, J. Chen, J. Zhu, C. Xu, W. Zhang, C. Zhang, W. Shi, H. Tan, D. Yang, H. H. Hng, T. M. Lim and Q. Yan, J. Phys. Chem. C, 2012, 116, 12468–12474.
96. R. Banerjee, R. Jayakrishnan and P. Ayyub, J. Phys.: Condens. Matter, 2000, 12, 10647.
97. L. Wang, Z. Yao, F. Jia, B. Chen and Z. Jiang, Dalton Trans., 2013, 42, 9976–9981.
98. W. Wang, W. Zhu and H. Xu, J. Phys. Chem. C, 2008, 112, 16754–16758.
99. C. Xing, Y. Zhang, W. Yan and L. Guo, Int. J. Hydrogen Energy, 2006, 31, 2018–2024.
100. X. Xu, R. Lu, X. Zhao, Y. Zhu, S. Xu and F. Zhang, Appl. Catal., B, 2012, 125, 11–20.
101. Y. Liang, H. Wang, P. Diao, W. Chang, G. Hong, Y. Li, M. Gong, L. Xie, J. Zhou, J. Wang, T. Z. Regier, F. Wei and H. Dai, J. Am. Chem. Soc., 2012, 134, 15849–15857.
102. Y. Liang, H. Wang, J. Zhou, Y. Li, J. Wang, T. Regier and H. Dai, J. Am. Chem. Soc., 2012, 134, 3517–3523.
103. N. M. Marković, T. J. Schmidt, V. Stamenković and P. N. Ross, Fuel Cells, 2001, 1, 105–116.
104. N. Li, G. Liu, C. Zhen, F. Li, L. Zhang and H.-M. Cheng, Adv. Funct. Mater., 2011, 21, 1717–1722.
105. Y. Zeng, W. Zhang, C. Xu, N. Xiao, Y. Huang, D. Y. W. Yu, H. H. Hng and Q. Yan, Chem.–Eur. J., 2012, 18, 4026–4030.
106. W. Zhang, H. Zeng, Z. Yang and Q. Wang, J. Solid State Chem., 2012, 186, 58–63.
107. X. T. Lu, Z. B. Zhuang, Q. Peng and Y. D. Li, CrystEngComm, 2011, 13, 4039–4045.
108. C.-C. Wang and A. Manthiram, J. Mater. Chem. A, 2013, 1, 10209–10217.
109. Y. Feng, T. He and N. Alonso-Vante, Chem. Mater., 2007, 20, 26–28.
110. N. Mahmood, C. Zhang, J. Jiang, F. Liu and Y. Hou, Chem.–Eur. J., 2013, 19, 5183–5190.
111. Z.-S. Wu, S. Yang, Y. Sun, K. Parvez, X. Feng and K. Müllen, J. Am. Chem. Soc., 2012, 134, 9082–9085.
112. Y. Gorlin, C.-J. Chung, D. Nordlund, B. M. Clemens and T. F. Jaramillo, ACS Catal., 2012, 2, 2687–2694.
113. H. Zhu, S. Zhang, Y.-X. Huang, L. Wu and S. Sun, Nano Lett., 2013, 13, 2947–2951.
114. N. M. Markovic, H. A. Gasteiger, P. N. Ross Jr, X. Jiange, I. Villegas and M. J. Weaver, Electrochim. Acta, 1995, 40, 91–98.
115. J.-H. Choi, K.-J. Jeong, Y. Dong, J. Han, T.-H. Lim, J.-S. Lee and Y.-E. Sung, J. Power Sources, 2006, 163, 71–75.
116. A. Ezeta-Mejía, O. Solorza-Feria, H. J. Dorantes-Rosales, J. M. H. López and E. M. Arce-Estrada, Int. J. Electrochem. Sci., 2012, 7, 8940–8957.
117. P. H. C. Camargo, Z. Peng, X. Lu, H. Yang and Y. Xia, J. Mater. Chem., 2009, 19, 1024–1030.
118. L. Zhang, P. Hu, X. Zhao, R. Tian, R. Zou and D. Xia, J. Mater. Chem., 2011, 21, 18279–18283.
119. X. Zheng, G. Shen, Y. Li, H. Duan, X. Yang, S. Huang, H. Wang, C. Wang, Z. Deng and B.-L. Su, J. Mater. Chem. A, 2013, 1, 1394–1400.
120. W.-H. Ryu, J. Shin, J.-W. Jung and I.-D. Kim, J. Mater. Chem. A, 2013, 1, 3239–3243.
121. R. Yang, Z. Wang, J. Liu and L. Chen, Electrochem. Solid-State Lett., 2004, 7, A496–A499.
122. G. Wang, J. Liu, S. Tang, H. Li and D. Cao, J. Solid State Electrochem., 2011, 15, 2587–2592.
123. H. Duliang, W. Jun, T. Fei, C. Mindong, W. Yong, D. D. Meng and L. Guiqing, Micro Nano Lett., 2012, 7, 773–777.
124. B. G. Choi, S.-J. Chang, Y. B. Lee, J. S. Bae, H. J. Kim and Y. S. Huh, Nanoscale, 2012, 4, 5924–5930.
125. X. W. Lou, D. Deng, J. Y. Lee, J. Feng and L. A. Archer, Adv. Mater., 2008, 20, 258–262.
126. B. Das, M. V. Reddy and B. V. R. Chowdari, J. Alloys Compd., 2013, 565, 90–96.
127. M. Gao, P. Zhou, P. Wang, J. Wang, C. Liang, J. Zhang and Y. Liu, J. Alloys Compd., 2013, 565, 97–103.
128. Y. Lin, Y. Lin, T. Zhou, G. Zhao, Y. Huang, Y. Yang and Z. Huang, Mater. Chem. Phys., 2013, 138, 313–318.
129. X. Wu, X. Wang and Y. Zhang, ACS Appl. Mater. Interfaces, 2013, 5, 2510–2516.
130. S. Jayapal, R. Mariappan and S. Piraman, J. Solid State Electrochem., 2013, 17, 2157–2165.
131. G. Che, K. B. Jirage, E. R. Fisher, C. R. Martin and H. Yoneyama, J. Electrochem. Soc., 1997, 144, 4296–4302.
132. X. Fang, C. Hua, X. Guo, Y. Hu, Z. Wang, X. Gao, F. Wu, J. Wang and L. Chen, Electrochim. Acta, 2012, 81, 155–160.
133. J.-Z. Wang, L. Lu, M. Lotya, J. N. Coleman, S.-L. Chou, H.-K. Liu, A. I. Minett and J. Chen, Adv. Energy Mater., 2013, 3, 798–805.

---