Growth mechanisms and origin of localized surface plasmon resonance coupled exciton effects in Cu_2−xS thin films

Abstract

We have demonstrated a robust protocol to prepare Cu_2−xS thin films with a controlled crystal phase and size which exhibit localized surface plasmon resonance (LSPR) coupled exciton effects by a simple template free single step wet chemical method without any surfactant. The LSPR frequency can be tuned in the Cu_2−xS thin films by the growth temperature and time which controls the free carrier density. These selectively grown Cu_2−xS thin films possess a tunable band gap (2.6–1.4 eV) due to the quantum size effect. The origin of the LSPR coupled exciton effects are discussed.

---

Introduction

Localized surface plasmon resonance (LSPR) in semiconductor nanocrystals (NCs) has attracted much attention since it has been recognized as ubiquitous and originating from the resonance of free carriers.^1–4 It was believed until recently that LSPR is exhibited only by noble metals' free electrons due to resonance oscillation with incident electromagnetic waves causing strong absorption at the metal surface.⁵ However, frequency tuned LSPR has been observed in cation deficient CuX (X = S, Se) chalcogenide materials, nanowires, and tungsten oxide NCs.^4,6–8 LSPR spectra can be tuned in metal nanoparticles within the limited change in their size, shape and dielectric environment whereas in semiconducting NCs, doping alone can change the free carrier concentration with appreciable change in LSPR frequency in addition to size effects.^1–11 Among the semiconducting materials, degenerately doped Cu_2−xS is unique, because of its increased carrier concentration upon the creation of Cu vacancies, varied surface morphologies, sizes, phases, non-toxic nature and compositions from non-stoichiometric Cu rich Cu₂S to stoichiometric CuS.¹² These self-doped Cu_2−xS also serve as templates in cation exchange reactions which proceed via diffusion mechanism to prepare core/shell nanostructures with LSPR properties.^4,13 It is noteworthy to mention that Cu_2−xS is a copper deficient and heavily self-doped p-type semiconducting material, which has five stable phases at room temperature namely covellite (CuS), anilite (Cu_1.75S), digenite (Cu_1.8S), djurleite, (Cu_1.94S), and chalcocite (Cu₂S)¹⁴ with different crystal structures from rhombohedral to hexagonal.¹⁰ It has been found that LSPR wavelength is blue shifted from Cu_1.9S to Cu_1.8S to CuS¹² due to increased Cu vacancies. Hence, LSPR can be tuned from terahertz to NIR frequencies in Cu_2−xS as long as controlled size, phase and doping concentrations are achievable. However, synthesis of Cu_2−xS NCs usually involves sophisticated techniques which incorporate template, catalyst and surfactant or capping agent. They are not only expensive but also involve time consuming complex procedures like modification of template, removal of the core and ensuring the attachment of the precursor molecule. The difficulty to obtain uniform nanostructure of Cu_2−xS arises because of its very low solubility in aqueous solutions (Cu₂S, pK_Sp = 47.60; CuS, K_Sp = 35.20) and instability in the atmosphere.¹⁵ So, the synthesis of Cu_2−xS NCs with uniform nanostructures demands the assistance of organic chelating agents like amines to form the coordination compounds or ions in solution as a transition state before attaining the stable state.^16–18

Here, we report three dimensional Cu_2−xS thin films which exhibit both quantum confined exciton and LSPR effects for the samples synthesized through facile template free single step chemical bath deposition method on fluorine doped tin oxide (FTO). To the best of our knowledge, no studies related to LSPR coupled exciton effects have been reported so far, for Cu_2−xS thin films. Moreover, all the reported LSPR has been observed only for Cu_2−xS NCs dispersed in liquid medium. In this respect, we strongly believe that our method would bring an easy on chip fabrication of Cu_2−xS thin films based LSPR sensors and optoelectronic devices. If LSPR is in resonance with excitonic transition, we can enhance the absorption cross section of the excitonic transitions by means of antenna effect and/or spontaneous emission rate and the yield by means of the Purcell effect. Moreover, LSPR coupled with excitonic transition can enhance two photon absorption,¹⁹ photonic up-conversion,²⁰ sub-wavelength lasing²¹ and the energy pumped into the excitonic modes can channel into the LSPR mode making possible single quantized plasmon generation.²² The formation of the above said Cu_2−xS thin films were carried out without the assistance of any capping agent or surfactant with consistent reproducibility. The Cu_2−xS thin films surface morphology, formation mechanism, origin of LSPR and excitonic effects have been discussed in this work.

Experimental section

Chemicals and materials

All analytical grade chemicals used for the synthesis were purchased from Sigma-Aldrich and used without further purification. The three dimensional Cu_2−xS thin films were deposited on cleaned FTO glass substrates with a sheet resistance of 7 Ω cm⁻² (Hartford Glass). The FTO substrates were cleaned using acetone, ethanol and deionized (DI) water subjecting to ultra-sonication.

Synthesis

In a typical synthesis, 0.05 M each of copper chloride dihydrate (CuCl₂·2H₂O) and copper sulphate pentahydrate (CuSO₄·5H₂O) were dissolved in 50 ml of DI water and stirred for 30 minutes to get a uniform solution. To this, 0.7 M acetic acid (CH₃COOH) was added in drops followed by 1 M of thioacetamide (CH₃CSNH₂) and stirred continuously for 45 minutes to get a homogeneous solution. The well cleaned FTO substrates were placed horizontally (FTO facing down using Teflon substrate holder) in bottles containing the growth solution and was kept in a hot air oven for 2, 3, 4 and 5 hours at a temperature of 60 °C are labelled as A2, A3, A4 and A5 respectively while the films deposited at 70 °C for a time period of 1 and 2 hours were labelled as B1 and B2 respectively for ease of identification.

Experimental techniques

The crystal structure and phase purity of the synthesized Cu_2−xS thin films were analyzed using X-ray diffraction (XRD; Bruker D8-Advance) with Cu Kα radiation (λ = 1.54056) source operated at 40 kV and 30 mA in the range of 10–90°. The surface morphology of the samples were analyzed using FE-SEM (Hitachi, model S-4200) operated at 15 kV. X-ray photon spectroscopy (XPS) was performed using a VG scientific ESCALAB250 with monochromatic Al-Kα radiation of 1486.6 eV and with an electron take off angle of 90°. The pressure of the chamber was kept at 10⁻¹⁰ Torr during measurement. The survey spectrum was scanned in the binding energy (BE) range of 0.0–1400 eV in steps of 1 eV. The binding energy values reported here is relative to the carbon C 1s core level at 284.6 eV. UV-Vis spectroscopic analysis was carried out using Optizen 3220 UV spectrophotometer. The PL spectra were obtained by a Princeton Instrument – CCD (P1-MAX3). All the Cu_2−xS thin film samples were excited by He–Cd laser (Kimon, 1K, Japan) wavelength of 325 nm and power of 50 mW.

Results and discussion

Reaction mechanism

The growth mechanism of Cu_2−xS thin film is both time and temperature dependent. The film growth starts at 1 hour and ends at 5 hours in case of deposition carried out at 60 °C. After 5 hours, the film gets peeled off due to internal strain, while the deposition temperature was 70 °C, the film growth starts after 30 minutes and ends in 2 hours since the supply of S²⁻ ions is more catering to rapid sulfidation. The chemical reaction involves two steps such as (i) homogeneous nucleation and (ii) heterogeneous crystal growth. In an aqueous medium, most metal cations with suitable sulfur sources result in giving their corresponding metal sulfides as follows:

M²⁺ + CH₃CSNH₂ + H₂O → MS + H₃CONH₂ + 2H⁺ (1)

(M = Cu, Ni, Pb, Cd, Hg)

The formation of copper and sulfur ions occurs by the dissociation of copper sulfate pentahydrate, copper chloride dihydrate and thioacetamide respectively.

CuSO₄ → Cu²⁺ + SO₄²⁻ (2)

CuCl₂ → Cu²⁺ + 2Cl⁻ (3)

Hydrolysis of thioacetamide yields hydrogen sulfide (H₂S) and acetamide (CH₃CONH₂)

CH₃CSNH₂ + H₂O → CH₃CONH₂ + H₂S (4)

H₂S + H₂O → SH⁻ + H₃O⁺ (5)

SH⁻ + H₂O → S²⁻ + H₃O⁺ (6)

The role of acetic acid is to function as complexing agent and to enrich S²⁻ concentration over S atoms and is pH dependent. As the deposition is carried out in hot air oven, the temperature causes the copper salts and thioacetamide to release Cu²⁺ and S²⁻ ions respectively, while acetic acid strikes thioacetamide to release S²⁻ ions through the formation of H₂S, which reduces the copper salts to yield Cu_2−xS. Thioacetamide is a base which reduces Cu(II) to Cu(I) at lower temperature and readily forms coordination complex.

Cu²⁺ + S²⁻ ⇄ ½Cu₂S + S^o + H⁺ ⇄ CuS + H⁺ (7)

Crystal growth mechanism

The S²⁻ ions to be attached with Cu²⁺ undergo a temporary transition state of a micro level ring like Cu₃S₃ six membered coordination complex which is an intermediate before forming the final product.²³ This intermediate is highly unstable due to the activated energy and thus in the process of attaining stability, it is oscillating between a pseudo-chair and pseudo-boat like structure. Fig. 1 illustrates the formation mechanism of time and temperature dependent Cu_2−xS thin film on FTO substrate. The resultant product's morphology and the preferential growth strongly depend on the ability of the central metal ion to ligate with the surfactant, solvent and the reaction conditions like temperature, concentration and duration as well as the supramolecular non covalent interactions.^24,25 Though, the concentration of the reactants is same in both cases, when the temperature was 60 °C, the sulfidation rate is lower whereas at 70 °C it is faster and hence, the deposition starts to occur within 30 minutes and ends in 2 hours, as the rapidly formed nuclei drastically reduces the supply of nutrients in the solution, hinders the growth of Cu_2−xS thin films. When the deposition was carried out at 60 °C, homogeneous nucleation and self-aggregation of the nano flakes are formed after 2 hours. These nano aggregates function as seeds for the further selective growth.²⁶ This initiates atom by atom growth, which is due to Ostwald ripening mechanism,^27–29 where the precipitation reaction occurs between Cu²⁺ and S²⁻ ions to yield nanostructures as depicted in the Fig. 1. As the deposition time increased to 3 hours, there is almost equal amount of nano flakes with denser packing and hexagonal platelet structures are formed due to oriented attachment aggregation mechanism because of particle by particle growth.^30,31 The nano flakes grow into three dimensional crystals, which are evident at 4 hours of deposition; the flakes grow still denser due to further self-aggregation and the hexagonal crystals become stacked and less in number. After 5 hours, the nano flakes become dominant and they curtail the growth of the three dimensional crystals. This is due to Ostwald ripening mechanism being predominant over oriented attachment aggregation mechanism where the particles get adsorbed on the surface of FTO substrate, aggregate, ripen and finally the surface gets reconstructed to a particular morphology arising from heterogeneous nucleation due to acetic acid's role as stabilizing agent. On the other hand, the growth mechanism follows oriented attachment instead of self-aggregation at 70 °C. The fast release of S²⁻ ions directs the formation uniform sized three dimensional stacked nanocrystals with Cu²⁺ ions due to high temperature at 1 hour deposition time. As the deposition is prolonged for 2 hours, the stacked hexagonal crystals merge to form three dimensional Cu_2−xS thin film. The formation of three dimensional Cu_2−xS thin films results from a collective contribution from hydrogen bonds and weak van der Waals forces whereas hydrophobic, electrostatic attractions and dipolar fields are responsible for aggregation.^32–34 The formation of well-defined grain boundary is due to prolonged weak interaction between the central metal ion whereas the ligand at higher temperature of 70 °C, involves dissolution–recrystallization mechanism.³⁵

Fig. 1 Schematic representation of the formation mechanism of Cu_2−xS thin films.

Characterization of surface morphology

Fig. 2 shows typical FE-SEM images of Cu_2−xS deposited on FTO with different deposition time and temperatures. The panoramic view of the A2 film shows self-aggregated Cu_2−xS thin films from homogeneous nucleation, which in turn functions as seed for further growth.²⁶ A3 and A4 show a mixture of stacked nano flakes and two dimensional hexagonal nanocrystals. Compared to A3, A4 shows increased array of nano flakes indicative of the Oswald ripening and self-aggregation mechanisms to be predominant over oriented attachment aggregation mechanism.

Fig. 2 The SEM images of Cu_2−xS thin films deposited on FTO substrates.

The nano flakes are about 250 nm in length and few tens of nm in width, while the two dimensional nanocrystals are 200–300 nm in length and about 100 nm thickness due to stacking.

When the deposition time was prolonged up to 5 hours, Oswald ripening and self-aggregation mechanisms with heterogeneous nucleation are present and therefore, there are only dense array of nano flakes with knit coir mat like structure is formed as a stable product and it is uniform throughout as seen in A5. The SEM image B1 shows two dimensional stacked nanocrystals and increase in deposition time results in three dimensional crystals with sharp edges as shown in image B2 due to oriented attachment aggregation.

XRD analysis

The XRD patterns of the Cu_2−xS thin films deposited on FTO are shown in Fig. 3. All the Cu_2−xS thin films deposited at 60 °C (A2–A5) show cubic phase of digenite Cu_1.8S (ICDD file no. 075-2241) with a space group of Fm3̄m. On the other hand, both B1 and B2 films deposited at 70 °C have a mixed cubic phases of Cu_2−xS (ICDD file no. 02-1292) and Cu_1.8S (ICDD file no. 075-2241); however B2 shows predominant orientation (maximum intensity) peak located at 46.6 degree with cubic phase of Cu_2−xS (ICDD file no. 02-1292).

Fig. 3 X-ray diffraction spectrum of Cu_2−xS thin films deposited on FTO substrates.

The peaks at 2θ of 27.86, 46.54 and 67.31 values corresponding to (111), (220) and (400) diffracted crystal planes respectively indicate Cu_2−xS phase. The peaks at 26.6, 44.13 and 64.18 of 2θ value corresponding to (111), (220) and (400) planes respectively give an evidence for the presence of cubic Cu_1.8S phase. The origin of the peak at 2θ of 38.2 degree (indicated with black colour arrow) for the samples A4, A5, B1 and B2 could not be indexed as it does not match with the above crystal structures having maximum peak intensity at 44.5 degree. The identification of copper chalcogenide crystallographic phases is particularly difficult since they exist in a wide variety of compositions and crystal structures. For example, several phases exist for bulk copper sulfide with composition close to the ratio Cu/S = 2: low and high chalcocite, djurleite, digenite, anilite with maximum intensity diffraction peak occurs at 46 degree.¹ By this, a slight change in temperature or composition can lead to the formation of one or the other crystal structure. However, in our case the diffraction plane at 44.5 degree occurs only for cubic phase of Cu_1.8S for A2 to A5 and B1 while B2 have maximum intensity at 46.6 degree and is complex to identify all the crystal planes. However, after carefully reviewing the previous database, the diffraction peak at 2θ of 38.2 can be indexed to digenite Cu_1.8S of rhombohedral (ICDD file no. 47-1748) phase. This clearly shows that temperature and time play major role in determining the phase and surface morphology of Cu_2−xS thin films.

UV-Vis spectrum

Fig. 4 shows the UV-VIS-NIR absorbance spectrum of the Cu_2−xS thin films. All the films show absorption at UV/visible region and is red shifted as the deposition time and temperature are increased due to quantum confined exciton effect and the absorption in NIR regions corresponds to LSPR originates from free carrier density.²

Fig. 4 UV-VIS-NIR absorption spectra of Cu_2−xS thin films deposited on FTO substrates.

It is well documented that the Cu rich Cu₂S does not show plasmon resonances while the non-stoichiometric Cu_2−xS (x > 0) NCs develops an LSPR in the NIR region and shifts to blue region with strong absorption due to creation of higher free carrier (with increased copper vacancies) in the material if the Cu_2−xS NCs is exposed to air (oxidation).^1–9,36 Moreover, reduction of Cu_2−xS NCs leads to total recovery of the initial crystal structure and LSPR absorption band.^36,37 In our sample, A4 and B1 shows well pronounced LSPR peak intensity while samples A2, A3 does not show plasmon resonance within the wavelength range of 1100 nm. In our case, the increase in intensity might also come from the thickness of the samples because of increased deposition time. All the samples were prepared and characterized in ambient atmosphere without any pre and/or post heat treatment.

If we presume that oxidation had caused the creation of copper vacancies, all the samples should have shown plasmon resonance; otherwise, the oxidation rate would have been different for different samples and will be discussed later. The absorption maximum in the UV/VIS are at 366, 402, 456, 456, 473 and 521 nm respectively for A2, A3, A4, A5, B1 and B2 samples with increased absorption from A2 to B2 (indicated with an arrow). These high energy absorption peaks arise from the 1S_h–1S_e excitonic transition found in semiconductor nanoparticles.³⁸ This is an interesting result because, LSPR coupled with excitonic transition enhance two photon absorptions, photonic up-conversion and hence Cu_2−xS could find wide applications in optoelectronic devices and sensors.^8–22 Moreover, the onset of absorption shifts to longer wavelength region due to size effects. The band gaps of the Cu_2−xS thin films were calculated from the absorption spectra using Tauc method.³⁹ The films A2, A3, A4 and A5 have the band gap of 2.6, 2.3, 2.0 and 2.0 eV respectively, while B1 and B2 have 1.7 and 1.4 eV respectively. The reduction in the band gap is due to the formation of uniform crystals of increased size showing quantum confinement effect which all make Cu_2−xS with tunable properties including that of LSPR as reported elsewhere.^4,13 The band gap reduction in the case of B1 and B2 may also occur due to the presence of mixed Cu_1.8S and Cu_2−xS phases along with increased uniform crystalline nature.⁴⁰ Only the metal nanoparticles exhibit the plasmonic effect in the visible region.^8,9 Hence, we believe that our samples exhibit a hybrid nature of metal and semiconductor with increased optical and electrical properties arising from near field enhancement, plasmon-based electron transfer⁴¹ and exciton–plasmon coupling.⁴² The increased free carrier concentration or the copper vacancy due to oxidation, makes it a self-doped p-type semiconductor with much pronounced LSPR and the quantum size confinement effects.^5,39,43 A slight change in composition might have also occurred due to the changes in the surface morphology and phases caused by Gibbs free energy.¹³

X-ray photoelectron spectroscopy

The XPS analysis was used to ensure the purity of the Cu_2−xS films and help to infer the composition and ionization state of the compounds. In order to elucidate the nature of bonding of the elements, the individual Cu 2p, S 2p peaks were scanned at a higher rate of resolution. XPS data were fitted with Gaussian–Lorentzian (30% Gaussian) functions and Shirley type background using Casa XPS software. Three constraints were applied to fit S peak components such as the spin orbit splitting (1.18 eV between 2p_3/2 and 2p_1/2), the peak area ratio (2p_3/2 : 2p_1/2 = 2 : 1) and equal full width at half maximum. Fig. 5 shows the core level Cu spectra where two sets of distinct double peaks at 932.3 eV and 952.2 eV representing the Cu(I) binding energies of Cu 2p_3/2 and 2p_1/2 states respectively while another set of peaks at 933.8 eV and 953.7 eV corresponding to the Cu(II) binding energies of Cu 2p_3/2 and 2p_1/2 states respectively. The satellite peak at 947 eV indicated with downward red colour arrow proves the existence of Cu²⁺ oxidation state due to Cu vacancy.^44–46

Fig. 5 Core level XPS spectra for Cu 2p of Cu_2−xS thin films.

In the solution, the intercalation of the central metal cation into the transition metal dichalcogenide (TMDC) layers proceeds via predominant reduction of Cu(I) in the sub lattice of the sulfur ions. This is electronically balanced by the inflow of equal number of electrons to TMDC and the host electrons are accommodated in the partially filled d orbital of the central metal ion.⁴⁷ Similarly, all the S 2p peaks from Cu_2−xS samples (Fig. 6) ascribed to the doublets peaks of S 2p_3/2 and 2p_1/2 arising from spin–orbit splitting (1.18 eV). The three different oxidation states located at 160.8, 162 and 163 eV are assigned to S²⁻, disulfide and Cu deficient non-stoichiometric sulfide (X in the Fig. 6) respectively.^47,48 The binding energy values are given only for the S 2p_3/2 peaks and S 2p_1/2 peak is located at 1.18 eV at higher binding energy end. Even though, all the samples exhibited satellite peak due to Cu vacancy, only A4, A5 and B1 show pronounced LSPR in the near NIR region. The atomic percentage of Cu is in the order of A4 < B1 < A5 than the rest of the samples as calculated from the survey spectra. Formations of CuO/Cu₂O at the surface and/or grain boundary occur by breaking the Cu–S bond and liberate the sulfur to form disulfide. The disulfide concentration [ A4 (23.1%), A5 (20.2%), B1 (21.1%)] ensures that oxidation has occurred at the surface at different rates. Fig. 7 shows the oxygen peak arising from the samples without any pre/post heat treatment. Samples, A2 and A3 show respectively nil and less than 1% of oxygen; while A4, A5, B1 and B2 respectively show 4.9, 2.24, 5.29 and 2.18%.

Fig. 6 Core level XPS spectra for S 2p of Cu_2−xS thin films.

Fig. 7 XPS spectra of oxygen for Cu_2−xS thin films.

It clearly shows that oxidation rate is much higher for samples A4 and B1 compared to A5 and B2 and is a clear evidence for the evaluation of LSPR band observed in optical studies related to copper vacancies, which act as free holes and hence free carrier absorption from excess holes in the valence band arise from the non-stoichiometry between copper and sulfur. Wavelength tuning of the LSPR mode has been previously achieved by actively controlling carrier density concentrations by varying the size, shape, and stoichiometry of the Cu_2−xS. In addition to this, the surrounding dielectric medium also has an influence on the plasmonic behavior of the nanocrystals. However, in our sample, the surrounding medium is air and hence, refractive index has no role to affect the LSPR mode, but only the size and the carrier density due to non-stoichiometric nature and oxidation of the Cu_2−xS films have caused the origin of LSPR. The different rate of oxidation may be due to (i) different surface morphology as is evidenced from the SEM images (ii) concentration of Cu vacancies and formation of disulfides at the surface of the films due to surface active species.

Photoluminescence spectroscopy

Photoluminescence (PL) is the spontaneous emission of light when optical excitation takes place in the material where absorption of photons gives excitons. The photon generated excitons, after a period of relaxation, return to the ground states as electrons either via radiative or nonradiative recombination: the radiative recombination proceeds via (a) band-to-band; (b) donor to valance band; (c) conduction band to acceptor; and the nonradiative recombination via a defect level.

Since PL originates from the surfaces, the surface and interface properties of the material such as discrete energy states can be analyzed from it. The interfaces possessing impurities and defect states mostly behave as long-lived traps for the non-radiative decay.⁴⁹ Fig. 8 shows PL spectra of the Cu_2−xS thin films. The valance band electrons are excited to the conduction band with an excitation wavelength of 325 nm whose excitation energy corresponds to 3.8 eV. The PL shows enhanced emission intensity across wide wavelength ranges from 400 to 600 nm and low intensity until 900 nm. It clearly shows that, in addition to the band to band recombination corresponding to the band gap of the respective materials, trap assisted radiative recombination might have caused the low energy PL spectra. The ordered grain boundaries and increased Cu vacancies are considered as defects or impurities which can function as trap states. These trap states do serve as meta stable state for the excitons to relax before recombination.^50,51 Therefore, variation of both the reaction time and temperature have caused Cu_2−xS to possess well-defined grain boundaries, tunable band gap and enhanced surface states due to increased grain boundaries.

Fig. 8 PL spectra of the Cu_2−xS thin films deposited on FTO substrate.

Conclusions

In conclusion, three dimensional Cu_2−xS thin films have been prepared on the FTO substrate by simple and one step chemical bath deposition technique without any surfactants or capping agents using water as medium. The effects of temperature and deposition time on the surface morphology were studied and plausible mechanisms for the reaction and the crystal growth are proposed. It is found that, prolonged deposition time and elevated temperature play combined role in producing three dimensional thin films with the reduction of band gap because of quantum confinement effect. Even though, all the samples have Cu vacancies which have caused free carrier concentration in the films, only the films with higher oxygen concentration show the LSPR in the near NIR region. These selectively synthesized three dimensional Cu_2−xS thin films possess trap states arising from well-defined grain boundaries and Cu vacancies act as radiative recombination centers.

Acknowledgements

This work was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) grant funded by the Korea government (No. 2014005051).

Notes and references

1. Y. Zhao, H. Pan, Y. Lou, X. Qiu, J. Zhu and C. Burda, J. Am. Chem. Soc., 2009, 131, 4253–4261.
2. J. M. Luther, P. K. Jain, T. Ewers and A. P. Alivisatos, Nat. Mater., 2011, 10, 361–366.
3. D. Segets, J. M. Lucas, R. N. Klupp Taylor, M. Scheele, H. Zheng, A. P. Alivisatos and W. Peukert, ACS Nano, 2012, 6, 9021–9032.
4. X. Liu, X. L. Wang, B. Zhou, W. C. Law, A. N. Cartwright and M. T. Swihart, Adv. Funct. Mater., 2013, 23, 1256–1264.
5. J. A. Faucheaux, A. L. D. Stanton and P. K. Jain, J. Phys. Chem. Lett., 2014, 5, 976–985.
6. K. Manthiram and A. P. Alivisatos, J. Am. Chem. Soc., 2012, 134, 3995–3998.
7. J. Zhang and C. Noguez, Plasmonics, 2008, 3, 127–150.
8. X. Liu and M. T. Swihart, Chem. Soc. Rev., 2014, 43, 3908–3920.
9. S. W. Hsu, C. Ngo and A. R. Tao, Nano Lett., 2014, 14, 2372–2380.
10. L. Liu, H. Zhong, Z. Bai, T. Zhang, W. Fu, L. Shi, H. Xie, L. Deng and B. Zou, Chem. Mater., 2013, 25, 4828–4834.
11. G. Garcia, R. Buonsanti, E. L. Runnerstrom, R. J. Mendelsberg, A. Llordes, A. Anders, T. J. Richardson and D. J. Milliron, Nano Lett., 2011, 11, 4415–4420.
12. Y. Xie, A. Riedinger, M. Prato, A. Casu, A. Genovese, P. Guardia, S. Sottini, C. Sangregorio, K. Miszta, S. Ghosh, T. Pellegrino and L. Manna, J. Am. Chem. Soc., 2013, 135, 17630–17637.
13. X. Wang, X. Liu, D. Zhu and M. T. Swihart, Nanoscale, 2014, 6, 8852–8857.
14. N. W. Buerger, J. Chem. Phys., 1939, 7, 1067.
15. Z. Fang, X. Wang, J. Shen, X. Lin, Y. Ni and X. Wei, Cryst. Growth Des., 2010, 10, 469–474.
16. C. Wu, S. Yu and M. Antonietti, Chem. Mater., 2006, 18, 3599–3601.
17. W. Du, X. Qian, X. Ma, Q. Gong, H. Cao and J. Yin, Chem.–Eur. J., 2007, 13, 3241–3247.
18. W. P. Lim, H. Y. Low and W. S. Chin, Cryst. Growth Des., 2007, 7, 2429–2435.
19. J. Lee, P. Hernandez, J. Lee, A. O. Govorov and N. A. Kotov, Nat. Mater., 2007, 6, 291–295.
20. H. A. Atwater and A. Polman, Nat. Mater., 2010, 9, 205–213.
21. R. F. Oulton, V. J. Sorger, T. Zentgraf, R. Ma, C. Gladden, L. Dai, G. Bartal and X. Zhang, Nature, 2009, 461, 629–632.
22. A. V. Akimov, A. Mukherjee, C. L. Yu, D. E. Chang, A. S. Zibrov, P. R. Hemmer, H. Park and M. D. Lukin, Nature, 2007, 450, 402–406.
23. S. Zhan, R. Peng, S. Lin, S. W. Ng and D. Li, CrystEngComm, 2010, 12, 1385–1387.
24. M. Nagarathinam, J. Chen and J. J. Vittal, Cryst. Growth Des., 2009, 9, 2457–2463.
25. J. Zou, J. Zhang, B. Zhang, P. Zhao, X. Xu, J. Chen and K. Huang, J. Mater. Sci., 2007, 42, 9181–9186.
26. C. Wang, Z. Fang, F. Fan, X. Dong, Y. Peng, S. Hao and L. Long, CrystEngComm, 2013, 15, 5792–5798.
27. M. R. Truter, J. Chem. Soc., 1960, 997–1007.
28. H. Flaisher, R. Tenne and G. Hodes, J. Phys. D: Appl. Phys., 1984, 17, 1055.
29. W. He, H. Jia, X. Li, Y. Lei, J. Li, H. Zhao, L. Mi, L. Zhang and Z. Zheng, Nanoscale, 2012, 4, 3501–3506.
30. J. F. Banfield, S. A. Welch, H. Zhang, T. T. Ebert and R. L. Penn, Science, 2000, 289, 751–754.
31. H. G. Yang and H. C. Zeng, Angew. Chem., Int. Ed., 2004, 43, 5930–5933.
32. H. Cölfen and M. Antonietti, Angew. Chem., Int. Ed., 2005, 44, 5576–5591.
33. H. Cölfen and S. Mann, Angew. Chem., Int. Ed., 2003, 42, 2350–2365.
34. L. Zhong, J. Hu, H. Liang, A. Cao, W. Song and L. Wan, Adv. Biomater., 2006, 18, 2426–2431.
35. J. Xie, C. Wu, S. Hu, J. Dai, N. Zhang, J. Feng, J. Yang and Y. Xie, Phys. Chem. Chem. Phys., 2012, 14, 4810–4816.
36. I. Kriegel, C. Jiang, J. Rodriguez-Fernandez, R. D. Schaller, D. V. Talapin, E. da Como and J. Feldmann, J. Am. Chem. Soc., 2012, 134, 1583–1590.
37. D. Dorfs, T. Hartling, K. Miszta, N. C. Bigall, M. R. Kim, A. Genovese, A. Falqui, M. Povia and L. Manna, J. Am. Chem. Soc., 2011, 133, 11175–11180.
38. K. A. Ann Mary, N. V. Unnikrishnan and R. Philip, APL Mater., 2014, 2, 076104.
39. N. Ghobadi, Int. Nano Lett., 2013, 3, 2–5.
40. W. Han, L. Yi, N. Zhao, A. Tang, M. Gao and Z. Tang, J. Am. Chem. Soc., 2008, 130, 13152–13161.
41. A. O. Govorov, G. W. Bryant, W. Zhang, T. Skeini, J. Lee, N. A. Kotov, J. M. Slocik and R. R. Naik, Nano Lett., 2006, 6, 984–994.
42. W. Zhang, A. Govorov and G. Bryant, Phys. Rev. Lett., 2006, 97, 146804.
43. A. B. F. Martinson, S. C. Riha, E. Thimsen, J. W. Elam and M. J. Pellin, Energy Environ. Sci., 2013, 6, 1868–1878.
44. M. Wang, F. Xie, W. Li, M. Chen and Y. Zhao, J. Mater. Chem. A, 2013, 1, 8616–8621.
45. J. Ghijsen, L. Tjeng, J. van Elp, H. Eskes, J. Westerink, G. Sawatzky and M. Czyzyk, Phys. Rev. B: Condens. Matter Mater. Phys., 1988, 38, 11322–11330.
46. S. Lee, M. Park, J. Kim, H. Kim, C. Choi, D. Lee and K. Ahn, J. Electrochem. Soc., 2013, 160, H847–H851.
47. J. C. W. Folmer and F. Jellinek, J. Less-Common Met., 1980, 76, 153–162.
48. K. Laajalehto, I. Kartio and P. Nowak, Appl. Surf. Sci., 1994, 81, 11–15.
49. T. H. Gfroerer, in Encyclopedia of Analytical Chemistry, John Wiley & Sons, Ltd, 2006.
50. K. Takase, M. Koyano, T. Shimizu, K. Makihara, Y. Takahashi, Y. Takano and K. Sekizawa, Solid State Commun., 2002, 123, 531–534.
51. L. G. Liu, H. Z. Zhong, Z. L. Bai, T. Zhang, W. P. Fu, L. J. Shi, H. Y. Xie, L. G. Deng and B. S. Zou, Chem. Mater., 2013, 25, 4828–4834.

---