Seed-Layer Free Zinc Tin Oxide Tailored Nanostructures for Nanoelectronic Applications: Effect of Chemical Parameters.

Semiconductor nanowires are mostly processed by complex, expensive, and high temperature methods. In this work, with the intent of developing zinc tin oxide nanowires (ZTO NWs) by low-cost and low-complexity processes, we show a detailed study on the influence of chemical parameters in the hydrothermal synthesis of ZTO nanostructures at temperatures of only 200 °C. Two different zinc precursors, the ratio between zinc and tin precursors, and the concentration of the surfactant agent and of the mineralizer were studied. The type and the crystallinity of the nanostructures were found to be highly dependent on the used precursors and on the concentration of each reagent. Conditions for obtaining different ZTO nanostructures were achieved, namely, Zn2SnO4 nanoparticles and ZnSnO3 nanowires with length ∼600 nm, with the latter being reported for the first time ever by hydrothermal methods without the use of seed layers. Optical and electrical properties were analyzed, obtaining band gaps of 3.60 and 3.46 eV for ZnSnO3 and Zn2SnO4, respectively, and a resistivity of 1.42 kΩ·cm for single ZnSnO3 nanowires, measured using nanomanipulators after localized deposition of Pt electrodes by e-beam assisted gas decomposition. The low-temperature hydrothermal methods explored here proved to be a low-cost, reproducible, and highly flexible route to obtain multicomponent oxide nanostructures, particularly ZTO NWs. The diversity of the synthesized ZTO structures has potential application in next-generation nanoscale devices such as field effect nanotransistors, nanogenerators, resistive switching memories, gas sensors, and photocatalysis.

Introduction

The increasing demand to have smart and multifunctional surfaces on all sorts of objects and shapes is pushing flexible and transparent electronics to unprecedented performance and integration levels.[1] For this end, it is highly desirable a material system offering sustainability in terms of raw materials and processes to synthesize its low-dimensional structures, combined with a wide range of properties to enable its use on transistors, sensing, or even energy-harvesting components. Metal oxides are one of the material classes with the highest potential to fulfill all these needs. In fact, ZnO-based nanostructures have been widely explored over the past decade.[2,3] ZnO nanowires are a good example of the multifunctionality of oxides, enabling for instance nanogenerators to act as gas sensors and biosensors.[4,5] Moving from single to multicomponent oxides, e.g., from ZnO to zinc tin oxide (ZTO), has been one of the current trends, enabling one to obtain different properties by adjusting the cationic ratio, resulting in a wider range of applications for a given material system.[6] Concerning sustainability, ZTO also has a great advantage over other multicomponent oxides as the well-established indium–gallium–zinc oxide (IGZO) in thin-film technologies, since it avoids the use of critical raw materials as In and Ga.[7]

ZTO can crystallize through solid-state reaction in the metastable perovskite (orthorhombic or face centered, fcc)[8] or rhombohedral[9] forms (ZnSnO3) and the more stable inverse spinel orthostannate (Zn2SnO4).[10,11] In Figure S1 in Supporting Information the crystalline structure of both phases is presented.

Zn2SnO4 is an n-type semiconductor with mobilities higher than 10–15 cm2 V–1 s–1 and a wide band gap of 3.6 eV being reported in nanostructures.[12] On the other hand, ZnSnO3 has been reported as an excellent piezoelectric material, with a piezoelectric potential along the c-axis of ∼59 μC/cm2, more than 1 order of magnitude higher than that of ZnO (∼5 μC/cm2),[13−15] and also as a ferroelectric material.[16] Its band gap was reported as being 3.9 eV, higher than for Zn2SnO4.[17,18] These ZTO nanostructures can be synthesized by vapor phase processes as chemical vapor deposition (CVD)[19] and thermal evaporation,[6] which present high efficiency. However, these are cumbersome and expensive techniques, which demand high temperatures (>700 °C). Thereby, solution-based methods are imperative to decrease complexity, cost, and temperature while still enabling good performance of the synthesized nanostructures. Solution-based hydrothermal methods were already used to obtain ZTO nanostructures such as nanoparticles (NPs),[20] nanowires (NWs),[21] nanorods,[22] octahedrons,[23] nanocubes (NCs),[24,25] and nanoflowers.[26] These nanostructures have demonstrated interesting properties for numerous applications as photocatalysis,[27] sensors,[28−30] nanogenerators,[31−33] resistive switching memories,[34,35] and solar cells,[36] reinforcing the multifunctionality of ZTO for next-generation nanoscale devices.

A proper control of the synthesis process to achieve the target structure and shape is crucial. As an example, for gas sensing it was already reported that within ZnSnO3 structures an orthorhombic phase (as the one obtained in this study) possesses a much higher sensibility than the fcc. The dimension of the obtained structures also plays an important role, with higher specific surface areas resulting in improved gas sensing performance.[37]

However, two important drawbacks need to be solved: first, it is well-known that obtaining a single phase (ZnSnO3 or Zn2SnO4) and a single nanostructure shape (e.g., NP or NW) by solution processes is quite challenging.[38,39] This can limit the usefulness of ZTO for different applications, as the properties are heavily dependent on phase and shape; also, low-cost hydrothermal methods, highly desirable from an upscaling perspective, always require seed layers to achieve ZnSnO3 NWs.[38,40] While the use of a seed layer can enable easier fabrication of vertical structures such as gate-all-around transistors,[41,42] synthesizing ZTO NWs without a seed layer also brings multiple advantages: imposes fewer constraints to the synthesis conditions to be studied, which is crucial for investigating in detail the role of each synthesis parameter in controlling phase, shape, and size of the nanostructures;[10,43] allows for fewer processing steps to obtain the nanostructures; provides higher degree of freedom to integration by relying on a wide variety of available transfer methods to obtain random and aligned networks of NWs on any substrate;[44] finally, the nanostructures do not incorporate on their final shape any undesired residuals from the seed layers.[45]

In this paper, we present different multicomponent ZTO nanostructures produced by a seed-layer-free, one-step hydrothermal method, at only 200 °C. The chemical and structural influence on the solution-based synthesis of the zinc salt, the ratio between zinc and tin precursors, the concentration of the surfactant agent (H2O:EDA ratio), and the mineralizer (NaOH) concentration were studied with the aim of obtaining ZTO NWs. We are particularly interested in 1D structures given their efficient charge transport, crucial for conceiving nanoelectronic devices.[46]

Herein we show a simple hydrothermal method where we can control the phase and shape of the nanostructures by tuning the chemical parameters of the synthesis. ZnSnO3 NWs were successfully achieved without the support of seed layers and using two different zinc precursors.

Results and Discussion

Introduction: Governing Equations To Obtain ZTO NWs

In a typical hydrothermal method to achieve ZTO nanostructures, the synthesis product is seldom composed by a single crystalline phase. In fact, ZnSnO3 NWs, Zn2SnO4 NPs, NCs and octahedrons with nanoplates, ZnO NWs, SnO2 NPs and mixtures of them are usually obtained (Figure S2).[10] It is thus imperative to revise the governing equations representing the chemical processes to achieve each of these phases when Zn and Sn precursors are present.

The chemical reaction processes for the formation of ZnSnO3 nanostructures can be represented by the following equations:[47]

Concerning Zn2SnO4, its formation can be represented as follows:[12,43]

While these equations provide an ideal scenario to obtain ZnSnO3 and Zn2SnO4 nanostructures, it has to be taken into account that the modification of the concentration of the precursors and the mineralizer during the reaction can promote the formation of other species/structures. Moreover, both ZTO phases have ZnSn(OH)6 as an intermediary phase (Figure S2), which can also appear as an end product for synthesis with short durations and/or low temperatures. Given this, the detection of OH– groups by FTIR spectroscopy is quite useful for inferring about the completeness of the reaction (Figure S3).

Regarding the formation of ZnO nanostructures, it is normally associated with a high alkaline concentration,[43] and it can be represented bywhere tin species are washed away after reaction.

Finally, with respect to the SnO2 nanostructures,[43] its formation is favored by a lower alkaline concentration and can be represented by the following equation:Similar to tin species in the case of a highly alkaline environment, the zinc species are washed away after synthesis.

While the cationic ratios and chemical parameters mentioned above dictate which nanostructures within the Zn–Sn–O system are obtained, understanding the growth mechanism for each nanostructure would require a detailed analysis of the effect of physical parameters such as time, temperature, and pressure,[43,48,49] which is currently underway.

Influence of the Zn:Sn Molar Ratio

The type of precursor and the ratio between the metallic elements in the synthesis are crucial to define the nanostructures’ shape, size, and crystallinity. We first studied the different ratios between zinc and tin precursors (2:1, 1:1, and 1:2) using two different zinc sources, zinc acetate (ZnAc) and zinc chloride (ZnCl2). For these studies a NaOH molar concentration of 0.240 M and a H2O:EDA volume ratio of 7.5:7.5 mL:mL were used, based on the synthesis reported by Li et al.,[12] where Zn2SnO4 NWs were grown on a stainless steel mesh for dye-sensitized solar cells application.

Synthesis using ZnAc as the Zn precursor, with a Zn:Sn molar ratio of 1:2, results in inconclusive XRD analysis, showing the possible presence of different phases whose diffraction peaks overlap (Figure a): tetragonal phase of SnO2 (ICDD card 01-077-0452), ZnSnO3 orthorhombic perovskite phase, and Zn2SnO4 inverse spinel phase (ICDD card 00-024-1470). For the ZnSnO3 orthorhombic perovskite phase, peaks can be identified by the ICDD card 00-028-1486. It should be noted that although this card was removed from the ICDD database due to the similarities with a mixture of Zn2SnO4 and SnO2 phases, several reports in the literature[48,50−52] still refer to it. In fact, when we performed peak indexing in different ZnSnO3 nanowires samples, an orthorhombic phase was always determined. These findings are also supported by Raman analysis as will be discussed later.

Figure 1

XRD patterns for three different Zn:Sn molar ratios (1:2, 1:1, and 2:1) using (a) ZnAc precursor and (b) ZnCl2 precursor. Identification is following ICDD cards 00-028-1486 (deleted), 00-024-1470, and 01-077-0452 (Figure S2).

SEM analysis (Figure a) supports the argument that some ZTO (ZnSnO3 and/or Zn2SnO4) NWs are starting to form. This is reinforced by Raman spectroscopy analysis (Figure S4a), which shows the predominance of the vibrational band at 631 cm–1 associated with the expansion and contraction of the Sn–O bond peak[53] but also that peaks at 538 and 676 cm–1 start to appear, corresponding to internal vibrations of the oxygen tetrahedron in Zn2SnO4 and to the characteristic Raman M–O bonds stretching vibration mode in the MO6 octahedron of ZnSnO3 and/or Zn2SnO4, respectively.[43]

Figure 2

SEM micrographs of nanostructures obtained with ZnAc precursor and Zn:Sn molar ratios of (a) 1:2, (b) 1:1, and (c) 2:1.

For synthesis with a 1:1 molar ratio of Zn:Sn, ZnSnO3 NWs are predominantly obtained, as shown by XRD and SEM analysis (Figure b and Figure b). The identification of ortho-ZnSnO3 is not immediately clear: it can be mistaken not only with SnO2 (as seen before for 1:2 Zn:Sn molar ratio) but also with Zn2SnO4 inverse spinel-cubic phase (ICDD card 00-024-1470). EDS analysis on isolated wires shows the ratio Zn:Sn of 1:1 (Figure S5), supporting the ZnSnO3 phase, which was also identified by Kovacheva et al.[54] on a similar XRD spectra. Raman spectroscopy shows that the intensity of the 676 cm–1 peak (Zn2SnO4 or ZnSnO3) increases while the intensity for the 538 cm–1 peak (Zn2SnO4) decreases when compared with the sample using 1:2 Zn:Sn ratio. This confirms the predominance of the ZnSnO3 phase for the ratio 1:1. (Figure S4a). While SnO2 NPs could not be confirmed by SEM analysis, deeper investigation using SEM and EDS revealed a plausible explanation for the SnO2 peak: Figure S5 shows that within the same sample some structures comprising agglomerated NWs could be found. Such structures were reported by Mao et al.[55] and have been described as ZnO-doped SnO2. This can be explained by the higher solubility of chlorides in solvents based on ethylenediamine when compared to acetates: the prior dissolution of tin chloride would lead initially to the formation of SnO2 nanostructures, which could then be doped by the Zn present in the solution, which falls in line with the Zn and Sn distribution measured by EDS as seen in Figure S6 (see also Figure S7).

XRD data obtained for the synthesis with Zn:Sn molar ratio of 2:1 are similar to the data of 1:1 condition (Figure b), suggesting the presence of ZnSnO3 perovskite phase. However, SEM in Figure c readily shows that besides the ZnSnO3 NWs some other structures are present. A more detailed analysis reveals a mixture of ZnSnO3 and Zn2SnO4 octahedrons and NWs, microtubes comprising agglomerates of ZnSnO3 NWs and ZnO nanoplatelets (Figure S8 and Figure S9). Raman spectroscopy data support these results, showing the predominance of the 676 cm–1 peak associated with ZnSnO3 and/or Zn2SnO4 over the 631 cm–1 peak from SnO2 but also the presence of the 538 cm–1 peak corresponding to Zn2SnO4. Furthermore, a small peak at 574 cm–1 is detected for this synthesis condition, attributed to a vibrational mode of ZnO,[56] concomitant with the SEM analysis.

It can be concluded that for the ZnAc precursor a 1:1 Zn:Sn molar ratio is the one allowing us to obtain ZnSnO3 NWs without a large fraction of other Zn- and/or Sn-based nanostructures. As such, this ratio was chosen for the following studies. It should be noted that Li et al. reported Zn2SnO4 NWs following a similar synthesis but using a stainless steel mesh as seed, favoring the growth of ZTO nanostructures similar to the cubic structure of the seed.[12]

The same study was performed using ZnCl2 as zinc precursor, with the different Zn:Sn molar ratios of 1:1, 1:2, 2:1. Using a Zn:Sn molar ratio of 1:2, similar XRD data are obtained when compared to the ZnAc precursor, suggesting that a tetragonal phase of SnO2 and/or a ZnSnO3 perovskite phase exists (Figure b). SEM analysis reveals that both SnO2 NPs and ZTO (ZnSnO3 and/or Zn2SnO4) NWs are obtained but now with more relevance to the latter (Figure a). Raman spectroscopy data confirm this, exhibiting a more intense peak for ZnSnO3 and/or Zn2SnO4 (676 cm–1) than for SnO2 (636 cm–1), as seen in Figure S4b.

Figure 3

SEM micrographs of nanostructures obtained with ZnCl2 precursor and Zn:Sn molar ratios of (a) 1:2, (b) 1:1, and (c) 2:1.

For a solution with Zn:Sn molar ratio of 1:1, the XRD spectra show mainly the phase Zn2SnO4, as depicted in Figure b, with some small peaks attributed to ZnSnO3 and SnO2. SEM images show several types of nanostructures: while ZnSnO3 NWs are clearly observed (Figure b), a more detailed inspection also reveals a large amount of Zn2SnO4 NCs (Figure S10). For the Zn:Sn molar ratio of 2:1 mostly ZnSnO3 NWs are obtained (Figure c). Still, a few ZnO columnar nanoplatelets with hexagonal phase (ICDD card 00-036-1451) are also shown in Figure S11c. These trends are confirmed by the Raman analysis (Figure S4b), analyzing the evolution of peaks at 538 cm–1 (Zn2SnO4), 676 cm–1 (Zn2SnO4 and/or ZnSnO3), and 631 cm–1 (SnO2). Similar structures were already reported by Tian et al.[57] for the synthesis of ZnO nanostructures. Figure S11a shows an example of ZnSnO3 NW agglomerates obtained in this condition, which is shown by EDS to have a 1:1 Zn:Sn ratio (Figure S12). Curiously, by looking at the hexagonal form in the middle, we can suggest that the ZTO NWs are grown from an initial hexagonal ZnO microtube/wire. The initial formation of the ZnO NWs can be attributed to the higher solubility of ZnCl2 when compared to SnCl4·5H2O,[55] contributing to the faster formation of these structures relative to the ZnSnO3 NWs, which is an issue when trying to obtain single phase ZTO nanostructures. However, according to Miyauchi et al., these ZnO NWs can be removed with an acid solution of HNO3,[17] allowing us to achieve only ZnSnO3 NWs in the sample. As such, the Zn:Sn ratio of 2:1 was the selected condition for the following studies with the ZnCl2 precursor, allowing us to obtain ZnSnO3 NWs with an average length of 605 nm and a diameter of around 65 nm.

For both zinc precursors, for the Zn:Sn molar ratio of 1:2, SnO2 NPs are predominantly obtained, which can be attributed to the higher concentration of tin precursor in the solution. Even so, the initial growth of ZnSnO3 NWs is already observed. When the molar ratio is 1:1, the results differ for the two precursors. For ZnAc this was shown to be the best ratio in terms of promoting the growth of a single phase of ZnSnO3 NWs. On the other hand, for ZnCl2, this ratio leads to a mixture of Zn2SnO4 NCs (predominant) and ZnSnO3 NWs. As for the 2:1 molar ratio of Zn:Sn, it is the condition that promotes better results when using ZnCl2, resulting in ZnSnO3 NWs (although mixed with ZnO NWs), while for ZnAc a higher mixture of phases, with predominance of Zn2SnO4 NCs, is achieved.

In summary, whichever the ratios, using ZnAc promotes the presence of SnO2 while using ZnCl2 results in higher amounts of ZnO. This can be explained by the different precursors’ solubility: while SnCl4·5H2O is more soluble than ZnAc, promoting a faster growth of tin-based structures, ZnCl2 is more soluble than tin chloride, promoting a preferential growth of zinc-based structures. Table S1 presents the nanostructures sizes for the different conditions of both zinc precursors, obtained through the SEM images and using the software ImageJ. In general, it is possible to observe that the nanostructures produced using ZnCl2 as zinc precursor have longer sizes than the nanostructures produced using ZnAc.

Influence of the Surfactant Concentration

Oriented growth and morphological control of nanostructures are highly dependent on surfactant use.[43] This section presents the study of the influence of the H2O:EDA volume ratio for the two selected conditions from the previous study: Zn:Sn = 1:1 molar ratio using the ZnAc precursor and Zn:Sn = 2:1 molar ratio using the ZnCl2 precursor. The H2O:EDA volume ratios used were 15:0, 9:6, 8:7, 7.5:7.5, 8:7, 9:6, and 15:0 mL:mL. For all conditions the mineralizer’s (NaOH) concentration was kept as 0.240 M.

For the ZnAc precursor and using only H2O as a solvent a mixture of ZnSnO3 NWs and Zn2SnO4 nanoplates and octahedrons comprising nanoplates is obtained (Figure a and Figure a), with Zn2SnO4 being the predominant phase. These types of octahedron structures were already reported by Ji et al.[23] With increasing of EDA up to 7.5:7.5, there is a trend for Zn2SnO4 NPs to disappear while ZnSnO3 NWs dimensions get larger. SnO2 NPs initially appear as isolated structures moving to SnO2-filled ZnSnO3 NWs as EDA volume is increased (Figure a, Figure , and Figure S13a). For 7:8 volume ratio, there is an increase in the presence of ZnO NWs and ZnSnO3 NPs with face centered cubic structure (Figure a and Figure i,j), due to the higher amount of EDA. Note that ZnSnO3 face centered cubic structure and ZnSn(OH)6 exhibit coincident Raman peak (603 cm–1)[49] and XRD peaks (Figures S13 and S2, respectively). Still, for this synthesis condition it is verified by FTIR analysis that no OH– groups are present (Figure S14), confirming the ZnSnO3 identification. The ratio 6:9 produces mainly ZnO NWs and only a few ZnSnO3 NPs as shown by SEM and XRD (Figure m and Figure a, respectively). When no H2O is used as solvent (0:15), mostly SnO2 NPs and ZnO NWs are obtained, with some ZnSnO3 NPs being also present (Figure a and Figure o). These results are confirmed by Raman spectroscopy (Figure S13a) and can be explained by the significantly higher solubility of SnCl4·5H2O in EDA compared to ZnAc, inducing a faster and preferential growth of SnO2 NPs and the later growth of ZnO NWs. Still, probably due to the long duration of the synthesis, some ZnSnO3 NPs are grown, since as previously discussed ZTO nanostructures can originate from SnO2 nanostructures.

Figure 4

XRD patterns when using (a) ZnAc precursor (with 1:1 Zn:Sn ratio) and (b) ZnCl2 precursor (with 2:1 Zn:Sn ratio) for different H2O:EDA volume ratios. Identification is following ICDD cards 00-028-1486 (deleted), 00-011-0274, 00-024-1470, 01-077-0452, and 00-06-1451 (Figure S2).

Figure 5

SEM micrographs of the nanostructures obtained by synthesis using ZnAc/ZnCl2 as precursors, respectively, with the different H2O:EDA volume ratios of (a, b) 15:0, (c, d) 9:6, (e, f) 8:7, (g, h) 7.5:7.5, (i–l) 7:8, (m, n) 6:9, and (o, p) 0:15.

For ZnCl2 as precursor and when using only water as a solvent (15:0 H2O:EDA), Zn2SnO4 NPs with octahedral shape are obtained (Figure b). The XRD spectra (Figure b) show a pure cubic-spinel-type phase for these nanostructures. This is the most stable phase and shape for zinc tin oxide in the absence of EDA, in line with the literature even for other conditions of hydrothermal synthesis.[23,58,59] With the addition of EDA, for a H2O:EDA ratio 9:6, Zn2SnO4 nanostructures with a different shape than the octahedral and some ZnSnO3 NWs are obtained, as seen by SEM (Figure d). By XRD (Figure b), it can be verified the mixture of ZnSnO3 and Zn2SnO4 phases, with Zn2SnO4 still being predominant. By increase of EDA’s concentrations (8:7, 7.5:7.5, and 7:8), the size of ZnSnO3 NWs is increased and this phase becomes predominant. As Figure f shows, for the 8:7 ratio, only ZnSnO3 NWs are grown, but they present a large size distribution. For the 7.5:7.5 ratio, ZnSnO3 and large wires of ZnO are obtained, as it was already discussed in section Introduction: Governing Equations To Obtain ZTO NWs. For the 7:8 ratio the ZnSnO3 NWs are highly agglomerated, as shown in Figure l. These structures are similar to the obtained for ZnAc (7.5:7.5, Zn:Sn = 1:1), as described by Mao et al.[55] The H2O:EDA ratio of 6:9 gives miscellaneous results, as Figure n and Figure S15 show, with a mixture of several types of structures being obtained. It is possible to observe Zn2SnO4 NCs and octahedrons comprising nanoplates, ZnSnO3 NWs, SnO2 NPs, and also ZnO columnar nanoplatelets agglomerates, with all of these phases being identified by XRD (Figure b). By use of only EDA (0:15), ZnSnO3 NPs and ZnO NWs are formed (Figure p). The strong ZnO peaks in XRD (Figure b) appear due to the large size of ZnO NWs when comparing with the ZnSnO3 NPs. Overall, the XRD (Figure b) shows that for higher EDA’s concentration, the formation of ZnO becomes preferential over ZTO, which is in agreement with the SEM/EDS and Raman (Figure S13b) analysis. This preferential formation of ZnO nanostructures for the higher EDA conditions can be explained by the strong coordination ability between the ZnCl2 and EDA molecules.[60,61]

FTIR analysis helps us to understand and to explain the results for the synthesis where EDA has a higher concentration than H2O. Figure S14a shows that when using ZnAc as zinc precursor, for the conditions with H2O:EDA ratios of 7:8, 6:9, and 0:15, precursor or solvent residuals are still present in the final product, due to the low solubility of ZnAc in EDA. In comparison, for the synthesis using ZnCl2, residuals are observed when only EDA is used as a solvent (Figure S14.b), which can be attributed to the higher solubility of ZnCl2 in EDA, compared to that of ZnAc. For all the other conditions no precursor peaks can be traced by FTIR analysis.

Table S2 summarizes the type of nanostructures obtained for the different conditions and their respective dimensions. The trend verified in the previous study regarding the longer sizes of nanostructures synthesized using ZnCl2 as compared to ZnAc is again observed here. It is also reinforced the higher presence of Sn-based structures in the synthesis using ZnAc, and a predominance of Zn-based structures in the synthesis using ZnCl2. The addition of EDA favors the formation of ZnSnO3 NWs over the more energetically stable Zn2SnO4 NPs, which we attribute to the pH increase, as reported by Miyauchi et al.[17] Up to a certain EDA concentration the length and diameter of these NWs are increased. For both precursors, an optimal H2O:EDA volume ratio of 7.5:7.5 was determined for the formation of ZnSnO3 NWs. As the EDA concentration is increased beyond this point, ZnO nanostructures start to be dominant, even for the ZnAc precursor. This suggests that the predominant factor defining the type of nanostructure obtained in an environment with a high EDA concentration starts to be the pH value, as will be discussed in more detail in the next section. Thus, despite the relevance of EDA to achieve the desired ZTO NWs, the presence of the water is imperative to the formation of these nanostructures, not only to ensure the complete dissolution of the precursors but also to balance the pH in solution.

Influence of NaOH Concentration

NaOH plays an important role in the growth of the nanostructures, acting as a mineralizer agent, having a direct influence on the definition of the crystalline phase that is produced.

On the basis of the previous studies presented in this manuscript, the synthesis conditions for this study were set as Zn:Sn = 1:1 ratio when using the ZnAc precursor, Zn:Sn = 2:1 ratio with the ZnCl2 precursor, keeping a H2O:EDA ratio of 7.5:7.5 in both conditions. Concentrations of NaOH of 0.100 M, 0.175 M, 0.240 M, and 0.350 M were used to understand the influence of the mineralizer on the synthesis. The results for both zinc precursors presented the same trend, being discussed here simultaneously. The poorer concentration of NaOH (0.100 M) synthesis results in SnO2 NPs, whichever precursor is used, as seen in Figures and 7. Lehnen et al.[22] explained this behavior as an effect of the fast hydrolysis of Sn4+ cations, leading to the preferential formation of SnO2. With the increasing of the NaOH concentration to 0.175 M, SnO2 NPs are still obtained but Zn2SnO4 NPs are now predominant. Still, SnO2 is more evident for the ZnAc precursor, as shown in the XRD spectra (Figure a), which would be expected based on the higher solubility of SnCl4·5H2O compared to ZnAc. Increasing the NaOH concentration to 0.240 M, the Zn2SnO4 phase is no longer present and ZnSnO3 NWs are now produced, both in dispersed and in agglomerate shapes, as well as some ZnO NWs as already discussed in the section Influence of the Zn:Sn Molar Ratio. Finally, for the NaOH concentration of 0.350 M, only ZnO NWs are obtained when using ZnCl2, while when using ZnAc both ZnO NWs and ZnSnO3 NPs (ICDD card 00-011-0274) are observed. This trend of preferential growth of ZnO in alkaline solutions is well-known, since divalent metal ions do not hydrolyze in acidic environments.[62] Even with higher concentrations of NaOH (0.500 M) the resulting structures were verified to be the same as for 0.350 M but with a lower reaction yield. These results are in agreement with the literature, as Lehnen et al.[22] showed the same tendency for specific pH values: for pH ≈ 1 SnO2 NPs are obtained, pH ≈ 8.5 yields Zn2SnO4 NPs, and higher pH values yields ZnO mixed with ZnSn(OH)6.

Figure 6

XRD pattern of the nanostructures obtained for different NaOH concentrations, using (a) ZnAc and (b) ZnCl2 as zinc precursors. Identification is following ICDD cards 00-028-1486 (deleted), 00-011-0274, 00-024-1470, 01-077-0452, and 00-036-1451 (Figure S2).

Figure 7

SEM micrographs of the nanostructures obtained by synthesis using different NaOH concentrations.

Our results show a similar trend; i.e., lower pH leads to SnO2 structures and higher pH favors ZnO ones, even if the starting pH (without adding NaOH) is already 12. As such, this trend is not dependent on the pH value itself but in the variation of the NaOH concentration for a specific synthesis.

Table S3 summarizes all these findings, showing that low mineralizer concentrations favor the growth of tin oxide structures over zinc oxide ones, with the trend being reversed as the NaOH concentration increases. Having in mind the specific goal of obtaining ZTO NWs, the optimum mineralizer concentration is around 0.240 M.

As seen in the previous sections, data in Figure and Table S3 reinforce the trend of obtaining nanostructures with larger dimensions using ZnCl2 precursor instead of ZnAc, given the higher solubility of ZnCl2.

Reproducibility

For a comparison of the precursors in terms of reproducibility, synthesis with the selected conditions (Zn:Sn molar ratio of 1:1 for ZnAc and 2:1 for ZnCl2, H2O:EDA volume ratio of 7.5:7.5 and 0.240 M of NaOH) was repeated at least three times. The results showed a better reproducibility for ZnCl2 than for the ZnAc precursor as discussed next. Figure S16a shows that for three syntheses in the same conditions using ZnCl2 the results are very similar, showing only some differences in the presence of some residual ZnO NWs (confirmed by XRD spectra, in Figure S17b) and some variation in the average size of the ZnSnO3 NWs (Figure b). On the other hand, for the synthesis with ZnAc, three runs already show a quite significant variation on the XRD spectra (Figure S17a), even if all show the predominance of the ZnSnO3 phase (see also SEM images in Figure S16b). Furthermore, the size of the obtained NWs differs substantially for the multiple runs with ZnAc (Figure a). As discussed previously, ZnAc has a poor solubility in EDA when compared with ZnCl2 and SnCl4·5H2O. This can be one of the factors for the poor reproducibility when using ZnAc as zinc precursor. Moreover, when using ZnCl2, since tin precursor is also a chloride, the reaction will be less complex: not only is the number of chemical species lower, but also as Cl– reacts with Na+, the number of possible reactions reduces. Also, in the case of ZnAc, the presence of the ionic species H+, O2–, and C+ can lead to an imbalance in the reaction precluding the formation of zinc tin oxide nanostructures.

Figure 8

Comparison of the obtained nanostructures dimensions in different repetitions of synthesis using (a) ZnAc and (b) ZnCl2 as zinc precursor, under similar conditions. Average lengths and diameters are given in parentheses.

ZnSnO3 Nanowires: Optical and Electrical Characterization

Optical Characterization

To study the optical properties of the produced ZnSnO3 NWs, the absorption under UV and visible radiation was measured. Figure a shows the well-known Tauc relation that follows the equationwhere α is the absorption coefficient, h is the Planck constant, ν is the frequency, A is an energy-independent parameter, Eg is the optical band gap, and x is a coefficient related to the electronic transition (x = 2 for allowed direct transition[63]). The Eg values, inferred by applying the equation to the linear region of the plots, are 3.53 and 3.60 eV for the ZnAc and ZnCl2 precursors, respectively. For comparison, optical properties of pure Zn2SnO4 NPs, produced when using only H2O as a solvent (section Influence of the Surfactant Concentration), were also evaluated (Figure b). A band gap of 3.46 eV was achieved, lower than for the ZnSnO3 phase, confirming the trends verified by other authors.[12,17,18] These results also confirm the trends typically observed in multicomponent oxides, where the Eg of the resulting structure tends to be closer to the Eg of the predominant binary compound (ZnO or SnO2).[64]

Figure 9

Tauc’s plots (a) and photoluminescence spectra (b) for ZnSnO3 NWs using two different zinc precursors ZnAc and ZnCl2 and for Zn2SnO4 octahedrons, obtained using ZnCl2 precursor with only H2O as solvent.

Still, the Eg values obtained in this work (for both ZnSnO3 and Zn2SnO4) are slightly lower than the ones reported in literature (∼0.15–0.30 eV). This difference can be explained by a higher defect density in these nanostructures, resulting in absorption close to band edges.[65] This is a consequence of the low temperature of the reported solution-based process (200 °C) as compared to the >600 °C typically used for physical processes. These band gap values suggest a potential applicability of these nanostructures for photocatalysis and photosensors.[66,67]

Photoluminescence (PL) was also evaluated for these samples (Figure b). A more evident peak at around 366 nm is observed, which can be attributed to ZnO-based nanostructures.[68] By observing the PL spectra, it is clear that this peak is more evident in the samples where ZnO-based structures were identified (see for instance Table S2), corroborating our previous analysis. Small peaks at 427 and 488 nm can also be identified. The presence of these UV peaks is associated with oxygen vacancies, which are major defects in this type of material,[69,70] emissions in this region being common for ZTO nanostructures.[71] Other authors also attribute a 428 nm PL peak to SnO contaminations or nanocrystals formed during ZTO synthesis,[67] which is also a plausible justification for our samples.

Electrical Characterization

Electrical measurements on single NWs, although highly desirable for determining its electrical properties, presents itself as a real challenge. In the literature the reported characterizations are related to NWs with lengths of >10 μm. In this work we present electrical characterization of a single NW with length below 1 μm. The data discussed below are for a single NW with length and diameter of 769 and 63 nm, respectively, produced using ZnCl2 precursor, with a Zn:Sn molar ratio of 2:1, H2O:EDA volume ratio of 7.5:7.5 mL, and 0.240 M of NaOH. This condition was selected based on the fact that it is the one enabling higher reproducibility and a larger fraction of dispersed ZnSnO3 NW as the synthesis product. The electrical measurement was performed inside SEM using nanomanipulators after in situ deposition of Pt electrodes (Figure ). The obtained data (Figure ) show linear I–V characteristics, following an ohmic behavior. The background current between two electrodes deposited in the same configuration and distance (but without a NW connecting them) was also measured, as shown in the inset of Figure . This background current, resulting from a residual deposition of the electrodes material over the sample, was taken into account on the calculation of the NW resistivity. The significantly lower background current level compared to the actual measurement performed in the NW reinforces the validity of the nanostructure measurement. Considering Ohm’s law and the physical dimensions of the NW, a resistivity of 1.42 kΩ·cm was obtained, which is significantly higher than the one reported by Xue et al. (∼73 Ω·cm)[72] for ZnSnO3 NWs and by Karthik et al. (6 Ω·cm in vacuum) for Zn2SnO4 NWs.[67]

Figure 10

Images of (a) SEM nanomanipulators and (b) the W tips of the nanomanipulators contacting the Pt electrodes during electrical characterization of a single ZTO NW.

Figure 11

I–V curve for a single ZnSnO3 NW contacted by two Pt electrodes, measured inside SEM using nanomanipulators. The inset shows an I–V curve used for background current extraction, taken from a similar sized Pt electrode structure but without any NW connecting.

As explained for the optical properties analysis, the higher defect density associated with the low-temperature solution-based process is the most plausible explanation for the higher electrical resistivity reported here (Xue et al.[72] synthesized NWs by thermal evaporation at 990 °C and Karthik et al.[67] synthesized by vapor phase methods at 900 °C).

We consider that concluding whether the resistivity obtained here is too high for our targeted electronic applications would require device fabrication, e.g., field-effect transistors, where many other challenges need to be addressed, such as contact properties or electrostatic coupling between dielectric and semiconductor NW. Even if the synthesized ZTO NWs are too resistive for a certain application, there is still room for improvement, for instance, by passivation of surface-related defects by postannealing in different environments and/or by coating with encapsulation films.[73,74]

Conclusions

ZnSnO3 NWs produced by a solution process without the use of a seed layer and with temperatures of only 200 °C were reported for the first time. To accomplish this, the work presented a detailed study on the influence of the different chemical parameters on the hydrothermal synthesis of ZTO nanostructures. More specifically the role of Zn:Sn ratio, surfactant concentration (EDA), and mineralization agent (NaOH) concentration for two zinc precursors (ZnAc and ZnCl2) was studied. By adjustment of these parameters, the potential to achieve ZTO structures with different phases and morphologies was shown. It was found that an intricate interdependence of the different chemical parameters would enable multiple synthesis conditions to result in the final goal of obtaining ZTO NWs. Still, it was concluded that ZnCl2 allowed for a more stable (with less mixture of phases/structures) and more reproducible reaction than ZnAc, with longer ZnSnO3 NWs being obtained. Hence, the best condition proved to be using ZnCl2 as zinc precursor, with a Zn:Sn molar ration of 2:1, H2O:EDA volume ratio of 7.5:7.5 mL:mL, and a NaOH concentration of 0.240 M. These ZnSnO3 NWs presented lengths and diameters of around 600 and 65 nm, respectively. Electrical characterization of a single NW with a length of <1 μm was successfully done inside SEM, using Pt electrodes deposited by localized e-beam assisted gas decomposition. Optical and electrical properties were comparable with those reported for ZTO NWs produced by physical processes, which employ considerably higher fabrication temperatures. As such, low-temperature hydrothermal methods proved to be a low-cost, reproducible, and highly flexible route to obtain multicomponent oxide nanostructures, particularly ZTO NWs. Moreover optical and electrical properties showed a great potential for applications such as photocatalysis, nanogenerators, nanotransistors, and gas senors/photosensors.

Experimental Section

Synthesis of Nanostructures

ZTO nanostructures were synthesized via hydrothermal method, using a modified version of the synthesis reported by Li et al.,[12] without the use of a seed layer (in ref (12) a stainless steel mesh seed-layer is used). Figure S18a shows the schematic of the synthesis where the precursor concentrations used were 0.020 M SnCl4·5H2O and 0.040 M Zn(CH3COO)2·2H2O. The precursors were separately dissolved in 7.5 mL of Millipore water and were then mixed together. Afterward, 7.5 mL of the surfactant ethylenediamine (EDA) were added, and the mixture was left stirring for 30 min. Finally 0.240 M NaOH was added. The precursors were smashed in a mortar before being added to water to help dissolution. The reagents used were all commercially available: zinc acetate dihydrate 99.0% (Zn(CH3COO)2·2H2O), sodium hydroxide ≥98% (NaOH), and ethylenediamine 99% (EDA) from Sigma-Aldrich, tin(IV) chloride 5-hydrate (SnCl4·5H2O) extra pure from Riedel-de Haën and zinc chloride 98% (ZnCl2) from Merck.

To study the influence of the zinc precursor, zinc acetate was replaced by zinc chloride, maintaining the same concentration of zinc in the solution. Different Zn:Sn ratios (molar concentration) were studied, namely, 2:1, 1:1, and 1:2. The ratio between H2O and EDA was varied (H2O:EDA, 15:0, 9:6, 8:7, 7.5:7.5, 7:8, 6:9, 0:15), as well as the concentration of NaOH (0.100 M, 0.175 M, 0.240, and 0.350 M). When the solution was ready, it was transferred into a 45 mL Teflon-lined stainless-steel autoclave, filling 80% of the total autoclave volume. The mixture was kept in an electric oven (Thermo Scientific) at 200 °C for 24 h, with a heating ramp of 200 °C/h. The autoclave was cooled to ambient temperature naturally. The resultant precipitate, comprising the nanostructures, was centrifuged at 4000 rpm and washed several times with deionized water and isopropyl alcohol, alternately. The nanostructures were finally dried at 60 °C, in vacuum, for 2 h, as schematized in Figure S18b.

Characterization of Nanostructures

Structural characterization by X-ray diffraction (XRD) was performed using a PANalytical’s X’Pert PRO MRD diffractometer with Cu Ka radiation. The XRD data were acquired in the 10–90° 2θ range with a step size of 0.033°, using the nanostructures in the form of powder. Fourier transform infrared (FTIR) spectroscopy data were recorded using an attenuated total reflectance (ATR) sampling accessory (Smart iTR) equipped with a single bounce diamond crystal on a Thermo Nicolet 6700 spectrometer. The spectra were acquired with a 45° incident angle in the range of 4000–525 cm–1 and with a 4 cm–1 resolution. Raman spectroscopy measurements were carried out in a Renishaw inVia Reflex micro-Raman spectrometer equipped with an air-cooled CCD detector and a HeNe laser operating at 50 mW of 532 nm laser excitation. The spectral resolution of the spectroscopic system is 0.3 cm–1. The laser beam was focused with a 50× Leica objective lens (N Plan EPI) with a numerical aperture of 0.75. An integration time of 2 scans (10 s each) was used for all measurements to reduce the random background noise induced by the detector, without significantly increasing the acquisition time. The intensity of the incident laser was 50 μW. All spectra were obtained in triplicate for each sample at room temperature in the 100–1600 nm range. After measurements a baseline subtraction was performed in order to identify the different vibrational bands. The band gap of the ZTO nanostructures was estimated from reflectance spectra recorded in the 200–800 nm range with a PerkinElmer lambda 950 UV/vis/NIR spectrophotometer using the Tauc plot method. The photoluminescence (PL) measurements were performed at room temperature, using a PerkinElmer LS.55 instrument with a xenon lamp as excitation source with an excitation wavelength of 325 nm. The morphology and element analysis of the samples were performed using scanning electron microscopy (SEM) and energy dispersive X-ray spectroscopy (EDS) inside a Carl Zeiss AURIGA CrossBeam workstation. The electrical characterization of single-NWs using a two-point probe structure was also performed inside the Auriga system. First, the NWs were dispersed in isopropyl alcohol, yielding a low concentration solution, sonicated during 5 min, and were finally drop-casted on a Si/SiO2 substrate. After drying, isolated NWs were contacted by Pt electrodes, deposited using localized e-beam assisted decomposition of a C5H4CH3Pt(CH3)3 precursor introduced close to the sample surface using a gas injector system. Kleindiek nanomanipulators with W tips were then used to access the current–voltage characteristics, together with a semiconductor parameter analyzer (Agilent 4155 C).