Nanoshell quantum dots: Quantum confinement beyond the exciton Bohr radius

The ability to tune the bandgap of quantum confined semiconductors offers new opportunities for the development of functional nanomaterials. In their nanocrystalline form, semiconductors can exhibit unique optical, electronic, and magnetic properties that are usually not observed in the bulk size. Some examples of confinement-enabled features include a tunable absorption spectrum, bandgap photoluminescence (PL), singlet-to-triplet exciton conversion, and multiple exciton generation, all of which become available when the size of the semiconductor nanoparticle is reduced to or below its corresponding exciton Bohr radius.

Presently, the light-emitting properties of colloidal nanocrystals (NCs) represent the most developed functionality offered by quantum confined semiconductors. These are currently utilized in areas of biosensing and bioimaging,^1–4 solar concentrators,^5–9 and light emitting diodes.^10–14 Particularly notable are the recent advances in the development of nanocrystal phosphors for light-emitting displays,^15–17 which already outperform rival materials based on heteroepitaxial semiconductors and organic LEDs, in several important categories.^15,18,19

In contrast to emission-based applications of semiconductor nanocrystals, light-harvesting technologies utilizing quantum confined semiconductors are currently less developed and require further innovative research to ensure future competitiveness. One of the key problems lies in the high surface-to-volume ratio of most quantum confined semiconductors, which results in the enhanced density of “midgap” surface states (e.g., in the case of cadmium or lead chalcogenides). These states tend to have a significant contribution to the non-radiative carrier decay. In addition to a large surface-to-volume ratio, a small particle volume naturally accelerates the non-radiative Auger decay of multiple excitons (MX),^20–22 greatly diminishing the feasibility of the MX phenomenon in nanocrystal-based photovoltaic,^23,24 laser,^25,26 and photoelectrochemical devices.^27,28 In this regard, increasing the effective volume of quantum-confined semiconductors while maintaining quantum confinement effects represents an important step toward improving charge-transport and MX characteristics of colloidal QDs,²⁹ as was previously demonstrated in the case of 1- and 2-dimensional nanostructures [e.g., nanotubes,³⁰ nanoplatelets (NPLs),^31–33 nanorods,^34–36 and atomically coherent superlattices³⁷].

Nanoshell quantum dots^29,38,39 represent another viable geometry that enables band tuning in larger-size semiconductor colloids [Figs. 1(a) and 1(b)]. According to the schematic illustration in Fig. 2(a), the potential energy gradient across a semiconductor nanoshell is designed to promote the surface localization of photoinduced charges. In this arrangement, all three particle dimensions can be increased in size without causing the loss of the quantum confinement in the shell layer. As a result, the nanoshell architecture can enable a controllable increase in the quantum confinement volume while maintaining the lowest possible surface area for a given nanoparticle size [Fig. 1(c)]. Meanwhile, a nearly spherical shape of nanoshells is expected to facilitate their assembly into close-packed solids, a task, which is notoriously challenging in the case of one- or two-dimensional semiconductor colloids.

The benefits of employing nanoshell quantum dots for energy conversion applications have already been demonstrated using CdSe semiconductors as a model system of a quantum confined shell.³⁸ In this case, the quantum confinement of photoinduced charges was achieved in a thin layer of CdSe (2–6 nm) grown over the ∼10-nm, bulk-size CdS core. In the first part of this perspective, we will discuss how an enhanced exciton volume associated with the CdS_bulk/CdSe nanoshell morphology can lead to the suppressed Auger recombination and an improved conductivity of solution-processed films. We will also demonstrate a possibility of combining two-dimensional excitons in nanoshell QDs with zero-dimensional excitons in spherical QDs within the same nano-object, which enables a dual photoluminescence signal at two distinguishable spectral bands. In the second part of this perspective, we will discuss the possibility of constructing nanoshell colloids from other semiconductor combinations for which electronic properties are better suited for various applications of QDs. The synthetic challenges associated with the fabrication of nanoshell QDs comprising several core/shell semiconductor combinations will be highlighted.

The morphology of nanoshell QDs was inspired by a popular class of Quantum-Dot–Quantum-Well (QDQW) nanostructures,^40–51 whose synthesis was perfected over the last two decades. The quantum confined layer in QDQW colloids (e.g., CdSe and HgS) is usually sandwiched between the core and shell domains comprising a wider-gap semiconductor (e.g., CdS and ZnS). The resulting energy gradient leads to the formation of two-dimensional excitons that reside primarily in the intermediate shell and therefore exhibit an increased confinement volume. In the case of nanoshell QDs discussed here, the core domain is increased in size to beyond the material’s exciton Bohr radius.^38,39 An important advantage of such a bulk-seeded nanoshell geometry is associated with the ability to preserve the radial confinement of photoinduced charges, regardless of the particle size. This phenomenon is evidenced by the onset of the bandgap emission in ≈18-nm CdS_bulk/CdSe nanoshell QDs [Figs. 2(b) and 2(c)], which stems from the quantum confined CdSe shell layer. Notably, the emission of these nanoparticles exhibited a characteristic size-dependent behavior tunable via the shell thickness [Fig. 2(f)].

Electrical conductivity of nanoparticle assembles represents one of the performance characteristics that are expected to improve with the increased particle volume. It is well known that the charge transport across nanocrystal films plays an important role in quantum dot-based solar cells and field effect transistors (FETs).⁵² In these applications, small-size semiconductor nanoparticles suffer from a large density of interfacial defects, which leads to charge carrier losses. It is, therefore, reasonable to assume that increasing an average particle size in nanocrystal solids will be beneficial to charge transport characteristics. This strategy had been previously implemented through the use of 1- and 2-dimensional QD architectures (e.g., nanoplatelets and nanorods), where one or two dimensions of the incorporated nanoparticles exceeded the exciton Bohr radius. Integrating these materials into close-packed films, however, was complicated by the challenging assembly protocols, which required non-trivial processing steps. For instance, the alignment of nanorods or nanotubes perpendicular to the substrate improves the charge transport characteristics of a device but is increasingly difficult to achieve for large aspect-ratio structures. Likewise, nanosheets and tetrapods can be challenging to assemble into ordered patterns that assimilate the advantages of their geometries. As a result, one- or two-dimensional colloidal nanocrystals (NCs) still rarely appear in record-efficiency devices.⁵³ 

A nearly spherical morphology of CdS_bulk/CdSe nanoshell QDs allows using a simple, spin coating-based strategy for the deposition of close-packed solids [Figs. 2(d) and 3(b)]. In our earlier work,³⁸ we estimated the photoconductivity of 19.6 ± 1.0 nm CdS_bulk/CdSe nanoshell assembles by comparing it to that of similarly processed assemblies of 3.9-nm, zero-dimensional CdSe NCs. Both solids were developed on fluorine-doped tin oxide (FTO) substrates and capped with small-area 30-nm-thick Au electrodes. The original bulky ligands on nanoparticle surfaces were displaced using a dimethylformamide (DMF) ligand stripping strategy⁵⁴ followed by mild annealing at 120 °C. According to Fig. 3(a), the average photocurrent of nanoshell solids surpassed that of similarly processed CdSe NC films by a factor of 7, supporting the premise that larger-size quantum-confined nanoshell QDs exhibit improved charge transport characteristics compared to zero-dimensional QDs.

Multiple exciton (MX) generation is beneficial to many applications of quantum-confined semiconductors, including photoinduced energy conversion,^27,28 stimulated emission in QD lasers,^25,26,55,56 and carrier multiplication.^57–59 Since the MX generation is enabled by the quantum confinement regime, semiconductor colloids have to be sufficiently small to support this effect. Unfortunately, a small particle volume is also known to promote the non-radiative Auger decay of multiple excitations, greatly diminishing the MX feasibility. Even in the case of longer-lived biexciton populations (n = 2), the Auger decay time constant could be as short as just a few picoseconds (e.g., CdSe or PbSe NCs),^60,61 representing the predominant mechanism of MX losses in laser and photovoltaic applications of these materials.^23,24

Within the framework of interacting formalism,⁶² the Auger recombination rate decreases linearly with the nanoparticle volume (Γ⁻¹ ∼ V^0.9−1.1).^{21–22,63,65} It is, therefore, reasonable to expect that the general solution for enhancing the biexciton lifetime of semiconductor NCs should be sought through nanoscale geometries that offer a reduced confinement in one or two spatial dimensions (e.g., alloyed core/shell nanoparticles,^65,66 nanorod-shaped heterostructures,^67–71 and nanoplatelets^72–76). Among these, nanoshell QDs exhibit one of the largest quantum-confinement volumes [Fig. 1(d)], which makes this geometry particularly attractive for MX applications. Our recent work has confirmed these predictions demonstrated that a quantum-well (QW) nanoshell architecture effectively suppresses Auger recombination processes leading to long-lived biexciton populations (Fig. 4).²⁹ In particular, we observed that the biexciton lifetime of CdSe-based QW nanoshells (CdS_bulk/CdSe/CdS) was increased more than thirty times relative to zero-dimensional CdSe NCs [Figs. 4(a)–4(c)]. The slower biexciton decay in QW nanoshells was attributed to a large confinement volume, which compared favorably to other existing MX architectures [Fig. 1(d)].

The possibility of achieving the quantum confinement regime in the surface layer of a semiconductor nanoparticle has inspired the development of dual-emitting, double-well nanostructures featuring the PbS/CdS/CdSe core/barrier/shell morphology [Figs. 5(a), 5(c), and 5(d)].³⁹ These QDs offered a rather unique regime of the carrier localization, characterized by the mixture of zero- and two-dimensional excitons residing in the PbS core and CdSe surface domains of the composite nano-object, respectively. An interstitial barrier (CdS) separating the two confinement regions allowed regulating the rate of energy and charge transfer between the two excitation types enabling a dual emission from the core and the shell semiconductors.

Dual-color emitting fluorophores have attracted a considerable amount of attention as ratiometric phosphors for sensing applications. The intensity ratio of the two emission bands originating from different recombination processes within the same nano-object has been demonstrated to depend on environmental factors such as the pH balance in the surrounding media,^77–81 gas pressure,⁸² and the solvent temperature.^83,84 The employment of semiconductor QDs as ratiometric sensors comes with its own unique advantages associated with the size-dependent tuning of the emission spectrum across visible and infrared ranges and the photostability of inorganic colloids to photodegrading environments.^84–101 

The feasibility of PbS/CdS/CdSe-based ratiometric sensors was explored through concentration measurements of cationic and anionic species in solution. As a representative example of cations, methyl viologen (MV²⁺) was introduced at various concentrations to nanocrystal samples in chloroform. Upon the adsorption of MV²⁺ by nanoparticle surfaces, the emission of the PbS core became strongly attenuated, while the emission of the CdSe shell subsided only by about 50% [Fig. 5(b)]. The corresponding PbS-to-CdSe emission intensity ratio was, therefore, reduced proportionally to the concentration of MV²⁺. PbS/CdS/CdSe nanocrystals were subsequently exposed to a hole scavenging agent, 3-mercaptopropionic acid (MPA),^102,103 representing anionic molecules. Similarly to MV²⁺, MPA was introduced to a chloroform solution of nanocrystals at various concentrations and the mixture was stirred until the changes in the emission profile were no longer observed. In this case, binding of the MPA thiol group to nanocrystals promoted the CdSe (1S_h) → MPA hole transfer causing PL quenching. The trend was clearly seen in the photoluminescence spectra of the MPA/nanocrystal mixture [Fig. 5(b)] showing a noticeable reduction in the intensity of the CdSe emission band. Consequently, the observed ratiometric changes quantitatively reflected the concentrations of MV²⁺ and MPA analytes in chloroform.

Nanoshell QDs represent a novel class of colloidal nanocrystals where tunable optoelectronic properties are possible over the extended range of particle sizes. The heterostructured morphology of nanoshells employs the radial energy gradient to funnel the photoinduced charges into the quantum-confined surface layer, where two-dimensional excitons can then be induced, regardless of the total particle size. So far, the exploratory work focusing on evaluating the general feasibility of the nanoshell QD architecture was performed exclusively on the CdS_bulk/CdSe proof-of-concept system. These studies have reported improvements in the charge transport and MX characteristics of nanoshell QDs as compared to zero-dimensional CdSe QDs. One potential downside of the CdS_bulk/CdSe nanoshell geometry, however, concerns a relatively weak confinement of CdSe excitons caused by the delocalization of electron wave functions into the CdS_bulk core. Consequently, the photoinduced energy in CdS_bulk/CdSe nanoshells is not strictly contained within the surface layer, which could cause issues in applications where photoinduced charges are expected to interact with the external environment. For instance, in the case of catalytic processes, both the primary reaction rate and the sacrificial carrier regeneration processes are intrinsically more efficient when electron and hole wave functions are surface-localized. Shell localization of both charges is likewise beneficial for applications requiring charge transport across nanoparticle assemblies (e.g., FETs and solar cells), where surface confinement enhances the wave function overlap between neighboring dots. Consequently, the realization of a stronger confining nanoshell system represents the next logical step in the development of this type of nanomaterial.

Among the possible material candidates to replace CdS in the core domain of nanoshells, wide-gap ZnS and ZnO semiconductors appear to be particularly promising. ZnS has a relatively wide bandgap, which is needed to induce a stronger localization of both carriers in the shell. This material is also lattice matched to a number of functional semiconductors, including CuZnSnS₄, CdS, CuInS₂, and AgInS₂, which should allow their deposition in the form of a high-quality crystalline shell. Among possible nanoshell combinations utilizing the ZnS core domain, a particularly interesting set of properties is expected from semiconductors featuring a relatively small exciton Bohr radius, such as CuZnSnS₄ or CuInS₂. Quantum-confined nanoparticles of these materials exhibit a particularly large surface to volume ratio, which prevents achieving a high photoconductivity in solids. Conversely, assemblies of larger-size nanoparticles of these semiconductors could benefit from an improved charge transport but would not exhibit quantum confinement characteristics. In this regard, the nanoshell geometry of these semiconductors could offer a unique advantage of preserving the quantum confinement regime without compromising the charge transport characteristics. Even in the case of indirect gap semiconductors, the ability to preserve the quantum confinement via the nanoshell geometry could be important. While this group of semiconductors exhibits comparatively low radiative rates and is not likely to be considered for MX applications, the increased electronic density in the shell domain could lead to benefits associated with surface localization of charges. For instance, the nanoshell geometry of indirect band materials could result in the improved film conductivity since the spatial overlap of wave functions across neighboring nanoparticles would be enhanced.

The fabrication of nanoshell QDs beyond the investigated CdS_bulk/CdSe morphology poses non-trivial synthetic challenges. First, since the electronegativity of anions (e.g., chalcogenides) decreases with the increasing atomic number, narrow gap semiconductors tend to exhibit a lower reactivity in growth reactions as compared to wider-gap materials of the same cation group [see Fig. 6(a)]. This condition renders hot injection techniques ineffective for the deposition of sulfides on oxides, selenides on sulfides, and tellurides on selenides. Even in the case of reported CdS_bulk/CdSe nanoshell colloids, the shell deposition was believed to proceed via the initial nucleation of small-size CdSe clusters, which were subsequently attached to the surface of the CdS core domain [see Fig. 6(b)]. Such an attachment process was shown to proceed only in the presence of a free ligand in the reaction mixture (oleylamine), which was critical for stimulating the digestive ripening of CdSe clusters. Second, in the case of semiconductor combinations, where the two cations of the core and shell domains exhibit different reactivities, the introduction of shell precursors into the reaction mixture often results in the cation exchange, which causes either a full displacement of the weaker “core” cation (e.g., Zn → Cd in ZnS/CdS core/shell QDs), or the formation of a partly displaced dimer structure (e.g., CdS/PbS heterodimers¹⁰⁴). The limited ability of hot-injection techniques to produce a nanoshell morphology can be attested by the scarcity of literature reports on the synthesis of inverted (wide-gap core/narrow-gap shell) semiconductor combinations featuring 2D excitons.

To address the aforementioned limitations of the nanoshell synthesis, one can employ an alternative growth strategy, which relies on the crystallographic fusion of pre-fabricated inorganic nanoparticles. This approach was inspired by an earlier work on digestive ripening of semiconductor NCs,¹⁰⁵ reporting a pronounced interparticle coalescence in ligand-saturated solutions (ligand/solvent >60%). Typically, the particle coalescence process was observed when the reaction mixture exceeded a certain thermal threshold, T_c. We have recently attempted adapting this strategy to the nanoshell growth by using the CdS_bulk/CdSe model system. In this case, the shell layer was synthesized through the coalescence of bulk-size CdS NCs and small-size CdSe colloids [Fig. 6(b)]. Bulk-size core nanoparticles were first introduced into the oleylamine ligand solution and heated to just below T_c. In this temperature regime, large-size colloids were not yet prone to crystallographic fusion and remained isolated. The subsequent injection of smaller-size nanoparticles, exhibiting a lower T_c, resulted in their rapid attachment to larger-size cores. The product was then heated to above T_c to achieve nearly uniform nanoshell QDs [see Fig. 6(b)]. We were subsequently able to confirm that even in the case of previously reported CdS_bulk/CdSe nanoshell QD samples,³⁸ the growth proceeded via the attachment of small-size CdSe crystallites to large-size CdS core colloids. Based on these observations, it is reasonable to expect that such a coalescence-based strategy could be utilized for the synthesis of various semiconductor nanoshell QDs that usually defy hot-injection routes.

The utility of the nanoshell geometry has already been confirmed for several applications of QD colloids, including those relying on charge-transport and MX phenomena. We expect that the unique benefits of nanoshell QDs can also prove to be useful for other nanocrystal applications. For instance, photocatalytic systems utilizing semiconductor nanocrystals are likely to benefit from larger-volume nanoshell QDs. This is because most photochemical reactions require multiple excited electrons (e.g., 2-electron H₂ production or 4-hole water oxidation),^106,107 making it necessary for a single semiconductor nanoparticle to capture several photons in order to complete the photocatalytic cycle. Since multi-photon absorption grows proportionally to the particle volume,^25,26,55,56 larger-size colloids should result in enhanced photocatalytic rates. These expectations are strongly supported by the reports of enhanced H₂ production yield in the case of 1D CdS nanorods¹⁰⁴ and 2D CdS nanoplatelets (NPLs),^108–110 which outperform 0D nanocrystals of the same semiconductor. Therefore, we speculate that the nanoshell architecture could lead to better-preforming multi-electron catalysts.

Semiconductor quantum dots have been recently demonstrated to support a transfer of triplet excitons to a surrounding solid matrix or acceptor dyes.^111,112 The translation of triplet states from semiconductor QDs to molecular acceptors offers a general paradigm for sensitizing chemical transformations in diverse fields, including optoelectronics, solar energy conversion, and photobiology. Generally, the extraction of triplet excitons proceeds via the Dexter energy transfer process, which exhibits a short-range character. Unlike dipole–dipole FRET interaction, Dexter ET relies on the overlap of donor and acceptor electronic wave functions and, therefore, is only efficient when the charge density of the excited state extents to the nanoparticle surface. In this regard, nanoshell QDs represent a preferred geometry to be used as a triplet exciton donor since both excited charges are located on the surface. While systematic studies of triplet ET from nanoshells or QDQWs have not been performed, the feasibility of this process can be readily evidenced by inducing the CdS_bulk/CdSe nanoparticle → PCA triplet energy transfer, resulting in the 2(³PCA) → 2(³DPA) → ¹DPA (blue) upconversion process (Fig. 7).

The employment of semiconductor nanoshell QDs in lieu of zero-dimensional nanocrystals could potentially improve many aspects of nanocrystal-based applications. The ability to achieve molecular-like, tunable characteristics in bulk-size colloidal semiconductors is expected to enhance the photoconductivity and field-effect mobility of nanoparticles assemblies, increase the biexciton lifetime, and enable new paradigms in light-energy concentration. From a broader prospective, semiconductor nanoshells could emerge as an alternative to other low-dimensional colloids (nanosheets, nanotubes, and nanorods). Ultimately, the concept of charge-confining nanoshells can be applied to non-toxic and abundant material systems to be deployed in “printable” nanostructured solids. This methodology could be extended to a wide range of devices, including photocathodes, lasers, solar cells, field-effect transistors, and photodetectors.

This work was supported by the U.S. Department of Energy, Office of Science [Award No. DE-SC0016872 (M.Z.)]. J.C. was supported, in part, by NSF (Award No. DMR-1710063).