2018 Volume 59 Issue 4 Pages 648-655
Hexagonal silver nanoparticles are directly formed in a solution of 6.6 mM silver citrate and 132 mM ammonia irradiated by 1.98–2.46 eV visible light. The corresponding silver ion concentration is 19.8 mM, which is several orders of magnitude higher than those employed in other silver-nanoparticle-formation experiments. In the present study, the roles of silver citrate and ammonia on nanoparticle formation are investigated through experiments in which the concentrations of silver citrate (SC) and ammonia (NH3) are altered. Silver nanoparticles are efficiently formed when [SC] is the 1.65–6.6 mM range and the [NH3]/[SC] ratio is ∼8–16. Further, hexagonal nanoplates are dominantly formed when [SC] = 1.65–6.6 mM and [NH3]/[SC] = ∼16. Within this range, hexagonal nanoplate formation is insensitive to solution concentration. Concentrations of SC less than 1.65 mM, or NH3 ≥ 132 mM, inhibit the formation of silver nanoparticles. These observations suggest that aggregates composed of diammine silver complexes and citrate are formed at specific concentration ranges of SC and NH3, and they assist in the photoreduction of silver ions by 1.98–2.46 eV visible light. Furthermore, the lateral growth of platelet seeds is proposed to be the dominant mechanism for the formation of hexagonal nanoplates at [SC] values of 1.65–6.6 mM and [NH3]/[SC] = ∼16.
Fig. 11 Silver citrate-ammonia concentration map divided into three regions by nanoparticle morphology. In Region I, the dominant products are polyhedral nanoparticles. In Region II, hexagonal nanoplates are predominantly and effectively formed. In Region III, hexagonal nanoplates dominate at fluences up to 42 J/cm2, while multiply twinned decahedrons are formed at higher fluences. The nanoparticle formation rate in Region III was much slower than those in Regions I and II. The dashed line indicates the solubility limit of silver citrate.
Metal nanoparticles dispersed in a dielectric medium absorb light strongly due to localized surface plasmon resonance (LSPR). The characteristic light-absorption properties of metal nanoparticles by LSPR is significant to a variety of plasmonics fields, such as solar cells,1,2) photocatalysts,3,4) surface-enhanced Raman scattering (SERS),5–8) and biological diagnosis.9–12) The wavelength (or energy) of light used to excite LSPR is dependent not only on the metallic elements that comprise the nanoparticles, but also on the sizes and shapes of the nanoparticles. Silver nanoparticles are the best candidates for plasmonics applications because their lower electrical resistivities lead to stronger LSPR excitations, and because the wavelength for LSPR excitation of silver nanoparticles covers almost the entire visible region of the electromagnetic spectrum through the control of nanoparticle size and shape. In addition, the higher light-absorption saturation points and long-term durabilities of silver nanoparticles compared to those of other inorganic and organic pigments are also significant advantages. Apart from spherical silver nanoparticles, various types of non-spherical silver nanoparticles, including plates,13–31) rods,13,14,32) decahedrons,5,8,14,33–40) wires,14,41) cubes,6,14,42) bars,43) and bipyramids,44,45) have also been previously prepared. The synthesis of platelet silver nanoparticles was originally reported by Mirkin’s group in 2001.30,46) In their synthesis, a colloidal solution of spherical silver nanoparticles (seeds) was prepared by UV light irradiation or by initially adding a reducing agent. These nanoparticles were used as plasmonically active seeds (plasmonic seeds). The colloidal solution with added citrate was then irradiated with visible light using a fluorescent lamp (plasmon-mediated synthesis (PMS)).
In contrast, we found that hexagonal-platelet-shaped silver nanoparticles are directly formed in a 6.6 mM silver citrate (SC) solution containing 132 mM ammonia (NH3) by monochromatic visible-light irradiation without the prior preparation of spherical nanoparticle seeds.47) Citrate is not normally strong enough to reduce metal ions without the assistance of thermal agitation or UV radiation. Our silver-nanoparticle-formation result suggests that the reduction of silver ions can even be photo-assisted by light in the visible region. The silver ion concentration of 19.8 mM (one SC molecule has three silver counterions) used in our previous studies47,48) is considerably higher than those employed in other experiments (typically below 1 mM,8,27,30,31) see Table 1 for comparison). Citrate was often used as a stabilizer in the formation of platelet silver nanoparticels. It was reported that the zeta potential of triangle nanoplates in 0.3 mM citrate solution decreased with the increase in pH and it reached −35 mV at pH 11.15) The high stability of the hexagonal silver nanoplates was attained in our experiments47,48) because of the high citrate and ammonia concentrations used. On the other hand, it has been reported that the rate of platelet-silver-nanoparticle formation is suppressed,49) or that the shapes of the silver nanoplates are altered, by changing the concentration ratio of silver ions to citrate.50,51) Furthermore, it is known that soluble diammine silver complexes are formed by the addition of excess of NH3. The formation of soluble diammine silver complexes can contribute to reduce the particle size distribution. A monodisperse silver nanoparticle formation was reported by addition of NH3 just after starting the nucleation and growth stage.52) In the experiment, the excess silver ions were removed by forming soluble diammine silver complexes and the further formation of nuclei and the growth of already formed nanoparticles were prevented.
In the present study, we examine the effect of SC and NH3 concentrations on the formation of silver nanoparticles by visible-light irradiation in order to assess the roles of SC and NH3 on the photo-assisted reduction as well as the morphologies of the nanoparticles formed.
Silver nitrate (>99.8%), trisodium citrate dehydrate (>99%), ammonia solution (25% w/w), and ethanol (>99.5%) used for the purification and dehydration of the silver citrate, were purchased from Wako Pure Chemical Industries, Ltd., Japan, and used as received. Silver citrate (SC), synthesized from silver nitrate and trisodium citrate dehydrate, was dissolved in ultrapure water (resistivity > 18 MΩ cm) containing ammonia in a dark room following a previously reported method.47)
The NH3-containing SC solution was poured into a Petri dish with a glass lid and irradiated by cyan light (2.36 mW/cm2) from the top of the solution, as previously described.47) Irradiation by cyan light (Eir = 2.46 eV) was adopted in the present study because previous studies showed that the rate of silver-nanoparticle formation per irradiation fluence is highest, and the threshold fluence for hexagonal nanoplate formation is lowest, using this light. In order to maintain a constant solution temperature, irradiation was conducted in a dark incubator maintained at a temperature of 309 K.
The ultraviolet-visible (UV-vis) absorption spectra of the solutions were acquired using a V-660iRm spectrophotometer (Jasco, Japan) over the 300–900 nm wavelength range, and a UV-3100PC (Shimadzu, Japan) instrument was used for the 300–1200 nm range. The sizes and shapes of the silver nanoparticles formed in the SC solutions were investigated by transmission electron microscopy (TEM, JEM-2010, JEOL, Japan; operating voltage = 200 keV).
As already mentioned, the silver ion concentration of 19.8 mM used in our previous studies47,48) was considerably higher than those employed in other experiments (typically below 1 mM,8,27,30,31) see Table 1). Further, 132 mM NH3 was added for the dissolution of 6.6 mM silver citrate in water. Since three silver ions are associated with one SC molecule, an ammonia concentration in excess of 39.6 mM is required in order that all silver ions form water-soluble diammine silver complexes at an SC concentration of 6.6 mM. Experimentally, it was found that [NH3] ≥ 53 mM was required for the complete dissolution of 6.6 mM SC. To survey the effect of the NH3 concentration on nanoparticle formation, [NH3] was varied between 53 and 158 mM while [SC] was held constant at 6.6 mM. Figure 1 displays UV-vis spectra of solutions irradiated with various light fluences (Φ), up to 42 J/cm2. At [NH3] of 132 mM (Fig. 1(b), [SC] and [NH3] used in our previous studies47,48)), a strong absorption peak appeared at around 2.4 eV when the solution was irradiated above the threshold fluence (Φth ∼ 14 J/cm2, see Fig. 6), which was attributed to LSPR of the hexagonal silver nanoplates (Abshex) in our previous studies.47,48) When [NH3] was 106 or 79 mM (Figs. 1(c) and (d), respectively), Abshex was similar to that of the [NH3] = 132 mM solution. When [NH3] was 158 mM (Fig. 1(a)), the Abshex intensity was observed to be about half that of the [NH3] = 132 mM solution.
UV-vis absorption spectra of SC solutions prepared by irradiation with cyan light (Eir = 2.46 eV) at different fluences; the SC concentration was maintained at 6.6 mM and [NH3] was varied to: (a) 158 mM, (b) 132 mM, (c) 106 mM, (d) 79 mM, and (e) 53 mM.
On the other hand, the absorption spectra of the [NH3] = 53 mM solution (Fig. 1(e)) are significantly different to those of the [NH3] ≥ 79 mM solutions (Figs. 1(a)–(d)). Figure 1(e) reveals a major broad absorption peak at approximately 3 eV, with small shoulders at approximately 2.3 and 3.6 eV. Figure 2 shows TEM images of the [SC]/[NH3] = 6.6 mM/53 mM solution after irradiation at Φ = 42 J/cm2. Various polyhedral nanoparticles with sizes of approximately 30 nm were observed in the images.
TEM images of the [SC]/[NH3] = 6.6/53 mM solution after irradiation with cyan light (Eir = 2.46 eV) at Φ = 42 J/cm2.
The effect of NH3 on nanoparticle formation was further surveyed by varying [SC] from 2.2 to 6.6 mM at [NH3] = 132 mM. Figure 3 shows the UV-vis spectra upon irradiation at fluences up to Φ = 42 J/cm2. At [SC] = 3.3 mM (Fig. 3(b)), a similar Abshex was observed to that of the [SC] = 6.6 mM solution (Fig. 3(a)) but the intensity was lower and the peak position was slightly shifted to higher energy when spectra at the same fluence are compared. When [SC] was decreased to 2.2 mM (Fig. 3(c)), Abshex was negligible compared to that of the [SC] = 6.6 mM solution (Fig. 3(a)). The formation of hexagonal nanoparticles is considerably inhibited at [SC] = 2.2 mM and [NH3] = 132 mM.
UV-vis absorption spectra of SC solutions prepared by irradiation with cyan light (Eir = 2.46 eV) at different fluences; only the SC concentration was varied and the NH3 concentration was kept constant at 132 mM. (a) [SC] = 6.6 mM ([NH3]/[SC] = 20), (b) [SC] = 3.3 mM ([NH3]/[SC] = 40), and (c) [SC] = 2.2 mM ([NH3]/[SC] = 60).
The results in Figs. 1 and 3 indicate that the SC-to-NH3 concentration ratio is an important factor for the formation of hexagonal nanoparticles. The effect of silver-ion concentration on the formation of the hexagonal silver particles was investigated by varying [SC] between 0.33 and 6.6 mM while maintaining a constant concentration ratio of 16 between SC and NH3 (i.e., [NH3]/[SC] = 16); Fig. 4 shows the UV-vis spectra of these SC solutions after irradiation with fluences up to Φ = 42 J/cm2. Figures 4(a)–(c) display very similar UV-vis spectra, even as [SC] was decreased from 6.6 to 2.2 mM. Figure 5 shows the TEM images of the particles formed at [SC] = 2.2 mM after irradiation at Φ = 42 J/cm2; hexagonal nanoplates with mean edge lengths of 34 nm were observed. Hexagonal nanoplates with edge lengths of 25–36 nm were observed for the [SC]/[NH3] = 6.6/132 mM solution after irradiation at Φ = 42 J/cm2 in our previous study.48) The dependence of Abshex on Φ when [SC] = 1.65 mM and [NH3] = 26 mM (Fig. 4(d)) was also similar to those of solutions with higher SC and NH3 concentrations; however, the peak position was slightly higher-energy-shifted compared to those observed at higher SC and NH3 concentrations (Figs. 5(a)–(c)). These observations indicate that hexagonal nanoplates are similarly formed at lower SC concentrations (down to 1.65 mM) when [NH3]/[SC] = 16.
UV-vis absorption spectra of SC solutions prepared by irradiation with cyan light (Eir = 2.46 eV) at different fluences. SC and NH3 concentrations were varied while maintaining a constant [NH3]/[SC] ratio of 16. [SC] and [NH3] are: (a) 6.6 and 106 mM, (b) 3.3 and 53 mM, (c) 2.2 and 35 mM, (d) 1.65 and 26 mM, (e) 0.83 and 13 mM, and (f) 0.33 and 5.3 mM, respectively.
TEM images of nanoparticles formed at [SC] = 2.2 mM and [NH3] = 35 mM ([NH3]/[SC] = 16) after irradiation at Φ = 42 J/cm2 (the scale bars are 50 nm), and the edge-length distribution.
In order to clarify the effects of [SC] and [NH3] on hexagonal silver-nanoplate formation, the intensity and position of Abshex in Figs. 1 and 4 were determined by Gaussian peak fitting; the fluence dependences are summarized in Fig. 6. Figures 1, 4 and 6(b) reveal that the hexagonal silver nanoplates are most effectively formed at [NH3] = 79–106 mM when [SC] = 6.6 mM. The Φth values corresponding to Abshex were almost the same (approximately 14 J/cm2) for all solutions when [SC] = 1.65–6.6 mM and [NH3]/[SC] = 16. On the other hand, the absorption spectra of the [SC] ≤ 0.83 mM and [NH3] ≤ 13 mM solutions (Figs. 4(e)–(f)) were significantly different from those at the higher [SC] and [NH3] concentrations; instead of an Abshex at approximately 2.4 eV, broader absorption peaks with maxima at approximately 2.2 and 3.1 eV were observed. We also noted that when SC is 0.33 mM the intensities of these absorption peaks are much lower than those of Abshex at [SC] = 1.65–6.6 mM after irradiation at Φ = 42 J/cm2, indicating that silver nanoparticle formation becomes inhibited at [SC] lower than 1.65 mM at [NH3]/[SC] = 16. When [SC] was varied while maintaining a constant [NH3]/[SC] ratio of 20, absorption spectra similar to those in Fig. 4 were observed. The intensity and position of Abshex on fluence observed at [NH3]/[SC] = 20 are also summarized in Fig. 6. Similar Φth values of approximately 12 J/cm2 were found for solutions with [SC] = 2.2–6.6 mM and [NH3]/[SC] = 20. Furthermore, the increases in Abshex for solutions in which [SC] = 2.2, 3.3, and 6.6 mM were nearly identical.
(a) Peak energy (Ehex) and (b) intensity of the predominant LSPR absorption peak of hexagonal silver nanoplates (Abshex) as functions of light-irradiation fluence (Φ). The right-hand axis of the upper panel corresponds to the mean side length of the hexagons (Dhex) estimated from Ehex using our previous result.48) The lines are linear fits to each data set. The [NH3]/[SC] ratio is 20 for (black) triangles, 16 for (red) circles, and 12 for (blue) diamonds.
After irradiation with light at Φ ≤ 42 J/cm2, as mentioned above, the dominant products formed at [SC] = 1.65–6.6 mM and [NH3]/[SC] = 16–40 were hexagonal nanoplates with side lengths of 25–36 nm, while those at [SC] = 6.6 mM and [NH3] = 53 mM, near the solubility limit, were polyhedral with the sizes of ∼30 nm. On the other hand, the formation of silver nanoparticles was considerably inhibited at [SC] = 6.6 mM and [NH3]/[SC] = 40–60. We explored changes the nanoparticle morphology by further irradiation at Φ ≥ 42 J/cm2. Figures 7(a)–(e) display the UV-vis spectra when [SC]/[NH3] = 2.2/132 mM, 3.3/132 mM, 6.6/106 mM, 6.6/53 mM, and 0.33/5.3 mM solutions, respectively, were irradiated with light. When [SC] = 6.6 mM and [NH3] = 106 mM (Fig. 7(c)), in which hexagonal nanoplates were dominantly formed in solution at Φ ≤ 42 J/cm2, Abshex exhibited continuous growth in intensity and shifts to lower energy when Φ was increased to 165 J/cm2. Figure 8 shows TEM images of the particles formed at [SC] = 6.6 and [NH3] = 106 mM after irradiation at Φ = 165 J/cm2, in which hexagonal nanoplates with a mean edge length of 45 nm are seen. In contrast, the UV-vis absorption spectra displayed in Figs. 7(a), (d), and (e) are significantly different to that in Fig. 7(c). In Fig. 7(a), the Abshex peak was observed at around 2 eV at Φ = 42 J/cm2 exhibited a low-energy shift to the infrared region when Φ > 64 J/cm2. Two broad peaks appeared at approximately 2.3 and 2.9 eV after irradiation at Φ ≥ 123 J/cm2; these peaks grew with increasing Φ and their peak energies shifted to approximately 2.0 and 2.8 eV at Φ = 234 J/cm2, respectively, indicating that the nanoparticles responsible for these absorption peaks are formed at higher fluences. Figure 9 displays a TEM image of the [SC]/[NH3] = 2.2/132 mM solution after irradiation at Φ = 234 J/cm2, in which multiply twined decahedrons and large triangular plates are observed. Absorptions of multiply twined decahedrons with similar sizes were reported to occur at approximately 2.1 and 2.9 eV.34,53,54) For the [SC]/[NH3] = 3.3/132 mM solution (Fig. 7(b)), Abshex was observed to be similar to that of the [SC]/[NH3] = 6.6/106 mM solution (Fig. 7(c)), but its intensity was lower. In addition, low-absorbing peaks at approximately 2.3 and 2.8 eV appeared after irradiation at Φ = 165 J/cm2. Hexagonal plates, triangular plates, and multiply twined decahedrons are observed in the TEM image of the [SC]/[NH3] = 3.3/132 mM solution after irradiation at Φ = 165 J/cm2 (Fig. 10(a)).
UV-vis absorption spectra of SC solutions prepared by irradiation with cyan light (Eir = 2.46 eV) at different fluences and different SC and NH3 concentrations at Φ values above 42 J/cm2. (a) [SC]/[NH3] = 2.2/132 mM, (b) 3.3/132 mM, (c) 6.6/106 mM, (d) 6.6/53 mM, and (e) 0.33/5.3 mM.
TEM image of the [SC]/[NH3] = 6.6/106 mM solution ([NH3]/[SC] = 16, Fig. 7(c)) after irradiation at Φ = 165 J/cm2.
TEM images of the [SC]/[NH3] = 2.2/132 mM solution after irradiation at Φ = 234 J/cm2.
TEM images of (a) the [SC]/[NH3] = 3.3/132 mM solution after irradiation at Φ = 165 J/cm2, (b) the [SC]/[NH3] = 6.6/53 mM solution after irradiation at Φ = 168 J/cm2, and (c) and (d) the [SC]/[NH3] = 0.33/5.3 mM solution after irradiation at Φ = 174 J/cm2. The scale bars in (a)–(d) are 100 nm.
When the [SC]/[NH3] = 6.6/53 mM solution was irradiated above 42 J/cm2, the absorption was observed to increase in intensity while the overall spectral profile was maintained, as shown in Fig. 7(d); the corresponding TEM images reveal the predominance of polyhedral nanoparticles with diameters of approximately 30 nm, in addition to hexagonal and triangular nanoplates (Fig. 10(b)). The absorption spectrum of the [SC]/[NH3] = 0.33/5.3 mM solution (Fig. 7(e)) is very similar to that of [SC]/[NH3] = 6.6/53 mM solution (Fig. 7(d)) but with less-intense absorptions. As was observed in the [SC]/[NH3] = 6.6/53 mM solution, polyhedral nanoparticles with the diameters of approximately 30 nm were predominantly observed in addition to hexagonal nanoplates (Figs. 10(c) and (d)).
From the morphologies of the nanoparticles formed, the SC-NH3 concentration map can be divided into three regions, as shown in Fig. 11. In Region I, which is near the solubility limit and where [NH3]/[SC] = ∼8 and [SC] = 1.65–6.6 mM, the dominant products are polyhedral nanoparticles, with small amounts of hexagonal and triangular plates formed at higher fluences. In Region II, in which [NH3]/[SC] = ∼16 and [SC] = 1.65–6.6 mM, hexagonal nanoplates are predominantly formed at fluences up to 165 J/cm2. At [NH3] = ∼132 and [SC] = 2.2–3.3 mM (Region III) hexagonal nanoplates dominate at fluences up to 42 J/cm2, while multiply twinned decahedrons are formed at higher fluences. However, the nanoparticle formation rate in Region III was much slower than those in Regions I and II.
Silver citrate-ammonia concentration map divided into three regions by nanoparticle morphology. In Region I, the dominant products are polyhedral nanoparticles. In Region II, hexagonal nanoplates are predominantly and effectively formed. In Region III, hexagonal nanoplates dominate at fluences up to 42 J/cm2, while multiply twinned decahedrons are formed at higher fluences. The nanoparticle formation rate in Region III was much slower than those in Regions I and II. The dashed line indicates the solubility limit of silver citrate.
Silver nanoparticles are most efficiently formed in Regions I and II, which indicates that the reduction of silver ions is most effectively assisted by visible light at the corresponding concentrations. Silver citrate dissolves in water upon addition of citric acid instead of ammonia. Figure 12 displays the UV-vis spectra of a 6.6 mM SC solution with 60 mM citric acid after irradiation with cyan light of varying fluences up to 195 J/cm2. No absorptions due to silver-nanoparticle formation are observed in these spectra. These observations indicate that the NH3 concentration, as well as the SC-to-NH3 concentration ratio, is important factors that control the reduction of silver ions when irradiated by visible light. We propose that some aggregates of diammine silver complexes and citrate are formed and play key roles in the light-assisted reduction of silver ions by visible light. Citrate is known to act as a chelation ligand towards metal ions and is used for the preparation of ceramics from metal-chelate solutions using sol-gel processes.55) The aggregation of chemical moieties containing metal ions in solution has been reported during the precursor stage of nanoparticle formation by X-ray small angle scattering56,57) and X-ray absorption fine structure analyses.58)
UV-vis absorption spectra of 6.6 mM SC and 60 mM citrate solutions prepared by irradiation with cyan light (Eir = 2.46 eV) at different fluences.
From the intensities of the light absorbed in Figs. 1, 3, 4 and 7, it is clear that silver-nanoparticle formation is inhibited when [NH3] exceeds 132 mM or [SC] is below 1.65 mM. While the morphologies of the nanoparticles formed are different in Regions I and II, silver nanoparticles were effectively formed. Polyhedral nanoparticles are predominantly formed in Region I whereas hexagonal nanoplates were almost exclusively formed in Region II when irradiated at fluences below 165 J/cm2. Close examination of the polyhedral nanoparticles formed in Region I at Φ = 42 J/cm2 (Fig. 2) reveal that they are composed of smaller particles; some of them appear to be small coalesced triangular plates. It is known that citrate acts as a surface-protecting group that is preferentially absorbed on the (111) planes of silver.46,59) A triangular plate surrounded by (111) planes can be prepared by the truncation of one corner of a (111)-facetted tetrahedron. In Region II, the formation of such platelet seeds begins at Φth = ∼14 J/cm2 and lateral growth leads to hexagonal plates above Φth. In Region I, on the other hand, the platelet seeds are similarly formed, but lateral growth hardly occurs. Instead their aggregation leads to the formation of polyhedral nanoparticles. In other words, the SC-NH3 concentration ratio is the critical factor that controls the lateral growth of the platelet seeds into hexagonal plates. In Region I, near the solubility limit, the dissolution of silver citrate, through the formation of the diammine silver complex, is not complete and the concentration of free citrate may be insufficient for surface protection.
In Region III, the nanoparticle formation rate was very slow, with small amounts of hexagonal and triangle nanoplates formed by irradiation at Φ ≤ 42 J/cm2. On the other hand, multiply twinned decahedrons become formed at higher fluences. The surface energy of a multiply twinned decahedron is rather low because its surfaces are the energetically most stable (111) planes. The fusion of five regular tetrahedrons with common edges cannot completely fill the available space, leaving a wedge-shaped void with a gap angle of 7.35°. In multiply twinned decahedrons or multiply twinned seed crystals, considerable internal strain or disorder exists at the interfacial boundaries in order to accommodate the geometric mismatch. The energy barrier for the formation of a multiply twinned seed is considerably high, resulting in a rate of formation that is likely to be much lower than that for the formation a hexagonal nanoplate. It is known that the nuclei of hexagonal or triangular nanoplates are formed under kinetically favorable conditions (or at faster formation rates) whereas those of nanodecahedrons are formed under thermodynamically favorable conditions (or at slower formation rates).14,53,54)
Hexagonal silver nanoplates are formed in a 6.6 mM SC solution with 132 mM ammonia by irradiation with 1.98–2.46 eV visible light. The 19.8 mM silver ion concentration used (one SC molecule has three silver counterions) is considerably higher than those employed in other experiments, in which concentrations are typically below 1 mM. In the present study, the effect of SC and NH3 concentration on the photo-assisted silver ion reduction and the formation of silver nanoparticles by visible-light irradiation were investigated.
The lower NH3-concentration limit for the complete dissolution of 6.6 mM silver citrate in water was found to be 53 mM. This value is slightly larger than 39.6 mM numerically calculated from the molar ratio required to form water-soluble diammine silver complexes. From the morphologies of the nanoparticles formed, the SC-NH3 concentration map can be divided into three regions: Region I in which [NH3]/[SC] = ∼8 and [SC] = 1.65–6.6 mM and where polyhedral nanoparticles dominate; Region II in which [NH3]/[SC] = ∼16 and [SC] = 1.65–6.6 mM and where hexagonal nanoplates dominate; and Region III in which [NH3] = ∼132 and [SC] = 2.2–3.3 mM and where hexagonal nanoplates dominate at fluences up to 42 J/cm2 and multiple-twinned decahedrons are formed at the higher fluences. The formation rate and the threshold fluence for hexagonal-nanoplate formation are noted to be independent of concentration in Region II. Silver nanoparticle formation is inhibited at [SC] < 1.65 mM or [NH3] ≥ 132 mM. No silver nanoparticles were formed in the absence of ammonia in a 6.2 mM SC solution with 60 mM citric acid by irradiation with fluences up to 195 J/cm2. Since water-soluble diammine silver complexes are formed in the presence of excess NH3, aggregates composed of diammine silver complexes and citrate with fixed ratios are efficiently formed in Regions I and II and play important roles in the photoreduction of silver ions by visible light. Furthermore, we suggest that the lateral growth of platelet seeds is promoted in Region II and hexagonal nanoplates are predominantly formed in the corresponding concentration range.
The authors thank Prof. Hiroshi Mizubayashi (University of Tsukuba) for valuable discussions and Prof. Kikuo Yamabe (University of Tsukuba) for supplying the ultrapure water. The TEM observations and some of the UV-vis measurements were carried out at the Chemical Analysis Division, Faculty of Pure and Applied Sciences and the OPEN FACILITY, Research Facility Center for Science and Technology, University of Tsukuba.
