Highly repeatable kinetically-independent synthesis of one- and two-dimensional silver nanostructures by oriented attachment

Abstract

A repeatable and fast synthesis of one- and two-dimensional silver nanostructures with thickness of 20–25 nm, constructed from highly stable hexagonal and triangular nanoplates has been achieved. Various useful morphologies can be constructed on the gram-scale from the reduction of aqueous solutions of silver nitrate by L-ascorbic (L-ASB) acid in the presence of poly(methacrylic)acid sodium salt, within a few minutes by simple room temperature mixing. Contrary to the literature, the assembly mechanisms governing the final morphology are not kinetically controlled, and are instead selective based on concentration of the reaction mixture. These diverse silver nanostructures meet the criteria desired for feasible industrial production and incorporation into nanoscale devices.

---

Introduction

In recent years, silver nanoparticles with diverse shapes and sizes have been proposed for use in a wide range of applications such as optoelectronic sensors and devices, biological sensors, nanocomposites, and catalysts.^1–12 Due to their utility, research into the synthesis of silver nanoparticles is ongoing. However, difficulties arise when researchers attempt to progress from prototype to market. A feasible, high yield, repeatable, and environmentally friendly synthesis method is in much demand for industrial applications. In addition, electrical and thermal stability are the major weaknesses of silver nanoparticles.^13–15 Given the wide-ranging proposed applications for silver nanoparticles, and the above-mentioned issues, a feasible, high yield, and repeatable method for synthesizing high stability zero-, one- and two-dimensional silver nanoparticles would be critical to advance their wide use.

Silver has a face-centered-cubic crystal structure in which the (111) planes have the most dense packing, exhibiting the highest stability compared to other crystal planes. Among all different silver nanoparticles, the hexagonal and triangular nanoplates have maximum (111) surface texture.^16–18 Therefore, the silver nanoplates seem ideal structural blocks for fabricating high-stability supercrystals. Based on the literature a non-classical crystallization mechanism is proposed for supercrystals, in which the initial structural blocks are synthesized by classical crystallization, then undergo assembly, followed by joining to produce larger structures. The joining mechanism of the structural blocks could be diffusional sintering,^19,20 or a non-thermal diffusion-less mechanism, referred to as oriented attachment.^21–24 The product of joining through oriented attachment is usually a defect-free perfect crystal structure at the interface.²⁵

We previously examined the joining and fusion of the nanoplates through molecular dynamics simulation, which demonstrated how the nanoplates recrystallize after joining.²⁶ That research revealed that joining of the nanoparticles by oriented attachment leads to reductions of the total energy of the system because of decreasing energy of the surface atoms. Later work by other researchers confirmed our results by different simulation approaches.^27,28 However, during initial experiments the assembled morphology was found to be sensitive to experimental procedure, making repeatable synthesis by this method a challenge.

The orientation and assembly of nanoparticles during non-classical crystallization is a challenging subject which is not well understood, and is a rapidly developing field.^29–31 However, more investigation seems to be required to clarify all affecting parameters.

Nanoplates have been previously synthesized through chemical reduction of silver ions by several different recipes, which are usually slow and require high-temperatures.^18,32–35 There are also a few previous reports of supercrystal synthesis, most likely using nanoplates;^36–43 two of them have successfully synthesized one-dimensional supercrystals.^37,38 Following these promising research results, we synthesized one-dimensional belts-shape and two-dimensional porous supercrystals through joining of the silver nanoplates by a fast, high yield, and room temperature chemical reaction.²⁶ We demonstrated that our supercrystals have excellent thermal stability and good electromigration resistance thanks to their (111) surface texture that is provided by nanoplates structural blocks.^41,44 Moreover, the supercrystals were successfully employed as nanofiller in hybrid conductive adhesive and as sensing material for an airflow sensor.^41,45

The purpose of this work is to investigate the parameters affecting the synthesis of silver supercrystals by the reduction of silver nitride with ascorbic acid in presence of PMAA via the following chemical reaction:

2AgNO₃ + C₆H₈O₆ → 2Ag + 2HNO₃ + C₆H₆O₆ (1)

Based on this chemical reaction, the main factors that are able to play role in the synthesis are: (1) the ratio of silver nitrate to ascorbic acid (2) the pH of the mixture, (3) the reagent concentrations in the reaction mixture including that of the capping agent and, (4) nucleation of the crystalline silver. The pH in particular is expected to have a significant effect as it influences the redox reaction rate and the affinity of the polymer capping agent.⁴⁶ In this paper, we have applied experimental observation and mathematical modeling to investigate the influence of the aforementioned factors on the resultant nanostructure morphologies. First, we report the morphology of the silver supercrystals that were produced at different concentrations of silver nitrate. Next, to investigate the change of pH during the progress of the reaction, we developed a model, which was verified by experimental observation. This model is employed to predict the pH of the solution, to compare to experimental measurements taken during synthesis, and to discuss the effect of pH on the shape of the generated supercrystals. Finally, we monitored the progress of the reaction at different conditions through measurement of the optical transmittance of the solution by a simple photo detector, to determine the fastest possible reaction rate.

Experimental procedure

First, 0.68 g L-ascorbic acid (Alfa Aesar, 50-81-17) was dissolved into 200 mL H₂O. In another beaker, 1 mL of a poly(methacrylic acid), sodium salt, 40 wt% solution in water (PMAA) (Aldrich Chemistry) and 100 mL deionized water were mixed by gently shaking (solution B). Then, 1.6 mL of this solution was added to ascorbic acid solution and mixed to prepare reducing solution (solution B). Then, 2.1 g of AgNO₃ (Sigma-Aldrich) was poured in a 500 mL beaker and 60 mL of H₂O (except for recipe 9) added to it and gently sway the container to dissolve AgNO₃ and prepare the silver nitrate solution (solution A). When the silver nitrate crystals disappeared, the solution B was poured into this 500 mL beaker. After two minutes (10 minutes for recipe 9 and 10), the synthesized silver nanoparticles was collected by Büchner funnel vacuum filtration.

For SEM observations, two minutes after synthesis (exactly before vacuum filtration), 500 μL of nanoparticles suspension was collected by a micropipette and added to a 3 mL of H₂O. This suspension was shake for around one minute. 150 μL of this dilute suspension was poured on a clean silicon wafer and dried at 70 °C for SEM observation.

To measure transmittance of the solution during reaction, a photoconductive cell (Parallax Inc., 350-00009) and a LED (Lumex Opto/component Inc., SML-XL1110SOC-BTR) were used. They installed in two standard cuvettes (LIGHTLAB, C-6001) and the cuvettes are fix in a way that the gap between the photodetector and LED was 5 mm. The setup was connected to a data logger to record the data. This setup was placed in the reactor and transmittance of the solution was measured every 5 ms.

Modelling

A predictive model of the pH developed throughout the silver reduction was constructed. The reacting mixture is assumed to be well mixed, and to be at pH equilibrium for every point of the reaction, where the equilibrium point will shift as the relative concentrations and activity of each contributing species changes throughout the reaction. It is also assumed that every species that is in the mixture throughout the progression of the reaction is contributing to the pH equilibrium without sequestration; which is to say that it is assumed that the adsorption of the PMAA by the silver surface has a negligible effect on its buffering capacity, and that the developed dehydroascorbate is stable on the timescale of the synthesis. Ionic strength, μ, is calculated as shown in eqn (2), where c_i represents the molar concentration of an ionic species, and z_i represents the absolute value of its charge. Counter ions related to the PMAA are considered to be localized near the polymer chain and do not contribute to the overall ionic strength, and the self-ionization of water was included in the model. Strong acids and bases are treated as fully dissociated. Ascorbic acid and PMAA·Na are weak acids expected to act as buffers, and both are expected to have their dissociation pK_a values influenced by the solution ionic strength μ. For ascorbic acid, Ball demonstrated that across the applicable range of total ionic strengths the Henderson–Hasselbalch equation is acceptable when the apparent dissociation constant is varied with ionic strength according to the equation proposed by Ball (eqn (3)), with a pK_a constant of 4.21 at zero ionic strength.⁴⁷ PMAA·Na is a polyelectrolyte, and was treated according to the Katchalsky model (eqn (4)).⁴⁸

pK′ = pk − 0.5μ^0.5/(1 + 0.5μ^0.5) (3)

pH = pK_a′ + n log((1 − a)/a) (4)

For PMAA, Katchalsky and Spitnik reported analytical titration curves of across the range of interest, linearized independently for five ionic strengths.⁴⁶ The data presented in a figure by Katchalsky and Spitnik was digitized by image analysis, where it was assumed that distortions by rotation, stretch and shear were applied uniformly to the figure, and the axes were used for scale calibration. The resulting dataset was fit by a linear regression of the form shown in eqn (5), to be performed on the full set of data.⁴⁶ This obtained the model constants presented in Table 1.

Table 1 Linear model constants for dissociation of poly(methacrylic)acid, obtained from linear regression of data derived from ref. 46

	A	B	C	D
Values	6.33324	−1.32015	−1.80890	0.37048
Linearized using	X1	X2	X3
		log((1 − a)/a)	X1X2

---

For accurate modelling of the PMAA·Na, the molar concentration and salt substitution of the polyacid needed to be determined. A specimen of the stock PMAA solution was carefully weighed into a cleaned and pre-weighed 20 mL glass scintillation vial, and dried under 32 in-Hg vacuum at 80 °C for 24 hours. Once dried, the vial was sealed and re-weighed, to determine that the stock solution contained 40.01 wt% dissolved PMAA/PMAA·Na. A 19 mg specimen of this dried material was then placed in a TA Instruments Q500 TGA, degraded at 10 °C min⁻¹ to 700 °C under nitrogen, cooled to 390 °C, then reheated at 10 °C min⁻¹ under air flow to 900 °C to drive off the carbon content, leaving only white Na₂CO₃ ash. A molar analysis of the masses determined that the sodium salt substitution of the PMAA carboxylic acid groups was 86.86 mol%. This was treated in the model as 0.8686 mol NaOH strong base being added alongside each mol of PMAA.

To solve the model, the user provided the ingredient amounts intended in the two reactants solutions, the ratio of the two solutions, and a list of desired degrees of silver conversion that the model is to be solved. From these, the program calculates the total concentration of each species present at the degree of conversion being solved. The estimates of pH and the degree of dissociation of ascorbic and PMAA species were initialized at 12 and 0.5, respectively, and the initial ionic strength estimate is calculated. The program then solves for equilibrium pH at this degree of conversion by repeatedly alternating between recalculating the ionization strength and dissociation constants of each buffer species under the current species concentration estimates, and iteratively solving for the current pH estimate based on recalculations of the buffer dissociation from the last estimates of pH, μ, and dissociation constants. The PH was estimated based on the current estimated concentration of hydronium ions, obtained by summation of estimated concentrations for all hydrogen-dissociated species and subtracting the concentration of strong base. The convergence criteria used was that an acceptable pH estimate had been found for that degree of conversion when the difference between two consecutive pH estimates was <0.001 and the difference between two consecutive μ estimates was <0.0001, concurrently. Once converged, the model would recalculate concentrations based on the next degree of conversion, and reinitialize the estimates as before.

Using data from the above pH model, relative reaction kinetics were calculated for recipes 1–6, based on data reported by Kimura et al., and Moya et al., for the reduction of tris(oxalato)-cobaltate(III) by L-ascorbic acid.^49,50 This makes the assumption that the pH and ionic strength scaling for the two systems is similar, and allows us to compare relative rates on an arbitrary timescale. In accordance with Kimura et al., the rate model eqn (6) was used, where K₀ was dependent on pH and ionic strength, and calculated by eqn (7) in the range of pH < 4.5. In accordance with the findings by Moya et al., K₀ was considered constant at K₀ = −2.69 for 4.5 < pH < 7.5. This relationship was found by comparing the two published datasets, and determining that ionic strength had an insignificant effect in the studied range.

[small alpha, Greek, dot above] = K₀[AgNO₃][L-ASB] (6)

K₀ = 0.42(pH) − 4.6 (7)

Results

To investigate the effect of AgNO₃/L-ASB molar ratio on the shape of silver supercrystals, six different recipes were selected (Table 2). We performed five replicates of each recipe, both with and without the addition of balance water used to maintain constant final volume (4^th column of Table 2). The addition of water had minimal effect of the resulting nanoparticles, which otherwise exhibited repeatable morphologies specific to each recipe, providing ten replicates per recipe. Fig. 1 demonstrates the synthesized nanostructures. Recipe 1 used a very low AgNO₃/L-ASB molar ratio (0.05) which leads to ball-like nanoparticles, which are actually disordered clusters of very small structural blocks (inset of Fig. 1a). Increasing the AgNO₃/L-ASB ratio in recipe 2 changes the supercrystals to flower-shaped particles composed of short wavy ribbon-like structural blocks (inset of Fig. 1b). Recipes 3 and 4 produced small porous nanosheets of silver (Fig. 1c and d). Inset image of Fig. 1d reveals the thickness of the sheets is about 25 nm. Finally, high AgNO₃/L-ASB molar ratios in recipes 5 and 6 produced one-dimensional silver nanobelts (Fig. 1e and f).

Table 2 The recipes of the synthesis with different concentration of silver ion

Recipe number	Solution A 220 mM AgNO3	Solution B 22 mM L-ASB + PMMA	H2O (mL)
1	1	200	59	0.05
2	5	200	55	0.27
3	10	200	50	0.53
4	20	200	40	1
5	30	200	30	1.6
6	60	200	0	3.2

---

Fig. 1 SEM images of the silver supercrystals synthesized by recipe 1 to 6 (a to f, respectively).

Careful examination of the SEM images of Fig. 1 reveals a clear trend in the synthesized nanoparticles. Recipes 2 through 6 produced nanostructures composed of short nanobelt segments that have undergone a second stage of assembly with a progression in preferred joining site. Recipe 2 displays extensive joining along all segment faces, producing flower-like clusters. Recipes 3 through 6 display joining that is highly selective to the narrow edges of the short segments. This progresses from long-edge-to-long-edge “lateral” joining dominating (recipe 3), to tip-to-long-edge “branching” joining dominating (recipes 4 and 5), with tip-to-tip “linear” joining becoming more frequent in the progression from recipe 5 to 6. This is either due to an increasing preference towards linear joining, or an increasing aversion to long-edge joining. Recipe 1 appears to be consistent with this progression, representing clusters of particles that have joined indiscriminately.

In our experiments, all recipes have the same initial ascorbic acid and PMAA concentration. What varies across recipes is the silver content of the final mixture. This is accomplished by adding differing volumes of 220 mM silver nitrate solution (solution A). As the reduction reaction progresses ascorbic acid (weak) is converted into nitric acid (strong), lowering mixture pH. Therefore, pH and molar ratio are correlated in these first six recipes. For this reason, modeling work was carried out in an attempt to delineate the two effects, but failing that to help plan further experiments.

The primary effect of molar ratio is expected to be on reaction kinetics, which are also affected by pH 43, 44. For each recipe, pH evolution versus reduction reaction progress was calculated using the proposed model (Fig. 2a). To examine the model accuracy, the pH of the reactor was measured for all recipes after reaction completion (Fig. 2b). As can be seen, very good agreement was obtained. Precise measurement of the pH during the reaction was not practical due to the response time of conventional pH meters. Using the pH model results and data from Kimura et al., and Moya et al. it is possible to estimate the relative reaction rates for each recipe (Fig. 2c).^49,50 This allows us to assume that the pH and ionic strength scaling for the reduction of Ag⁺ is similar to that tris(oxalato)-cobaltate(III), and thus compare relative rates on an arbitrary timescale. In examining Fig. 2, it is apparent that both pH and reaction rate change monotonically across recipes 1 to 6. Had pH dominated, leading to a reversal in predicted rate, it would have been possible to rule out a rate effect and by extension direct molar ratio effects. As it stands, more experiments are necessary. Two observations of significance can be made when the rates are integrated numerically to estimate relative reaction times as in Fig. 2d. First, due to a crossover in stoichiometric ratio though rates increase monotonically, predicted times do not. Second, though recipes 4 and 6 have very different rates they have similar times to completion as they are equidistant from the stoichiometry crossover point.

Fig. 2 (a) pH change during progression of synthesis redox reaction for recipe 1 to 6; (b) comparison between predicted final pH of the recipe 1 to 6 and experimental final pH; (c) the reaction rate of recipe 1 to 9 versus conversion; (d) calculated molarity of silver converted for recipe 1 to 9 versus time.

To distinguish between the effect of pH and AgNO₃/L-ASB molar ratio on the morphology of the products, we manipulated the pH by adding NaOH or HNO₃ to solution B (Table 3). Recipes 7 and 8 are predicted to have a pH above that of recipe 1 for the bulk of the reaction (Fig. 3a) and silver content identical to recipe 6. Recipe 9 has a pH closer to recipe 6 (Fig. 3a), but its silver content is identical to recipe 3. Fig. 3b–d show the SEM images of the nanoparticles synthesized. In recipe 7 (high initial pH) the product is wavy silver nanobelts and individual overgrown hexagonal and triangular silver particles (Fig. 3b). In recipe 8 (the highest initial pH investigated), assembly of the structural blocks was eliminated, allowing the silver grow as individual particles (Fig. 3c). Low pH in recipe 9 resulted in a combination of wavy and straight one-dimensional silver nanostructures (Fig. 3d). Although the silver content of this recipe is almost the same as recipe 3, and the conversion time is most similar to recipes 4 and 6, the resultant silver nanostructure is most similar to recipes 5 and 6.

Table 3 Recipes of the syntheses, adjusting pH using NaOH or HNO₃

Recipe number	Solution A (mL)	Modified reducing solution
		Solution B (mL)	Adjusted pH
7	60	200	5.70	3.2
8	60	200	7.40	3.2
9	10	200	1.85	0.53

---

Fig. 3 (a) pH changes during synthesis of supercrystals by recipe 6 to 9; SEM image of the synthesized particles by pH manipulated in: (b) recipe 7; (c) recipe 8; (d) recipe 9.

While performing the above studies, it was observed that the size of the silver nanobelts synthesized by recipe 6 is very sensitive to preparation method of the silver nitrate solution. To investigate this phenomenon, the results from two methods of precursor solution preparation were compared. In the first, silver nitrate was dissolved at the same concentration investigated thus far using 5 minutes of severe shaking, and the resulting solution used for synthesis. Fig. 4a shows the synthesized sliver nanobelts after severe shaking of silver nitrate solution. Comparison between Fig. 4a and 1d clearly demonstrates that shaking of silver nitrate solution is able to reduce the length of the synthesized silver nanobelts. On the other hand, it is reported in the literature that dissolution of the silver nitrate is not a simple one-step process 45, but that the silver nitrate dissolves through some intermediate stages such as silver trimers (Ag₃³⁺ and Ag₃⁺) and these silver trimers can be considered as nucleation site for nanoplates.⁴⁵ Therefore, it can be hypothesized that silver trimers play the role of precursor for silver hexagonal and triangular nanoplates, and that aggressive mixing decreases the number of silver precursors available in the reactor. To examine this idea, we designed the second precursor solution experiment, recipe 10, in which 2.1 g AgNO₃ was slowly dissolved in 10 mL water by very gentle movement of the container to minimize mixing. The concentration of this solution is 6 times higher than recipe 6, and displayed two liquid phases, the denser of which was believed to be near the solubility limit. Under those conditions, it is believed that the silver trimers would be more stable, and a greater fraction would survive until mixed with the reducing agent. To keep total concentration of the system same as recipe 6, 50 mL water was added to the reactor simultaneously with reducing solution (Table 4). Fig. 4b demonstrates the synthesized silver nanobelts and the inset of this image demonstrates the thickness of the nanobelts. The nanobelts synthesized by recipe 10 are much longer than recipe 6. Fig. 4c shows one of the synthesized nanobelts, which is around 95 μm in length. Insets of Fig. 4 contain high-resolution views of two assembly locations, and confirm continuity of the nanobelt.

Fig. 4 (a) Short nanobelts, synthesized by recipe 6 by using aggressively shaken silver nitrate solution; (b) long nanobelts, synthesized by recipe 10; (c) high magnification SEM image of a long belt-shape supercrystal synthesized by recipe 10.

Table 4 Recipe of the synthesis with high silver nitrate concentration

Recipe number	Silver nitrate solution	Solution B (mL)	H2O (mL)
10	2.1 g AgNO3 + 10 mL H2O	200	50	3.2

---

It is noted that the differing reaction rates among the recipes were not negligible and can be visibly differentiated by changing color of the solution from a clear solution to black. Therefore, we measured the transmittance of the solutions during synthesis. The results of these measurements for recipe 6, 8, and 10 are presented in Fig. 5. In addition, it was possible to record the changing pH of the solution during synthesis for recipe 10, due to its exceptionally slow reaction rate, and the results have been superimposed. Fig. 5 demonstrates that the synthesis with recipe 8, in which the addition of NaOH raised the pH of the solution above the pK_a value of the ascorbic acid and PMAA, displayed a reaction rate higher than recipe 6. More interestingly, recipe 10 has a very low reaction rate, even though the fully-mixed concentration, pH, and AgNO₃/L-ASB molar ratio of these two recipes were similar. The only difference between these two recipes was higher concentration of the precursor silver nitrate solution, and the preparation method of this solution.

Fig. 5 Kinetic of synthesis reaction measured by light-transmission for recipe 6, 8, and 10. Evolution of pH during synthesis by recipe 10 is superimposed on the graph. The presented pH data is the average of 6 replicates.

Discussion

Literature would argue that the type of nanoparticle growth and assembly observed here achieves varying morphologies through kinetically controlled processes.^21,35 However, in this system the kinetics of the process was determined not to be the controlling factor. The reduction reaction appears to be the rate-limiting step, as evidenced by the agreement between light-transmission and pH data for recipe 10, and the general agreement between light-transmission and kinetic predictions for recipes 6 through 8. Based on morphology, recipes 6 and 10 were the most similar, and yet their reduction kinetics differ by orders of magnitude. Meanwhile, recipe 4 had reduction degree of conversion kinetics most similar to recipe 6, while the two display significantly different morphology. Similarly, recipes 9 and 10 were both observed to significantly deviate from their predicted kinetics due to long incubation times. Although they displayed similar incubation times and reaction durations, recipe 9 and 10 produced significantly different final morphologies. This insensitivity to kinetics appears to be why the repeatable synthesis of these nanostructures can be completed so rapidly. A working hypothesis on why the morphologies display such insensitivity to kinetics is that they each represent a local equilibrium under their experimental conditions.

Comparing the morphologies of recipes 1 through 6, 9, and 10, it can be seen that the progression in preference for lateral, branching, or linear joining correlates either with the final pH or with the concentration of nitrate ions. Based on our model, correlation with final pH may be due to the changing ionization of PMAA from 3% to 0.4% as the pH dropped from approximately 4 to 1.5. This is a relatively small change, and so this having such a significant effect would be surprising, though possible. This would be in agreement with published papers, in which the alignment and assembly of their particles was influenced by coulombic interaction between the particles and effect of solvation forces in a collision–recrystallization growth process.^29–31 These mechanisms were influenced by the geometry of their particles, and the local charges associated with them, generally agreeing with the pH sensitivity reported here. However, those studies were modelled with highly-idealized radially symmetric particle geometries representing simplified cases, and so more work would be necessary to verify this is applicable to the nano-structures studied here.

Another potential source of the pH sensitivity is the unreacted L-ascorbic acid, which changes ionization from 19% to 0.3% across this pH regime as predicted by model. However, for most recipes investigated the lowering pH throughout the reaction is caused by the consumption of the ascorbic acid, in some cases to completion (recipes 6, 10). Therefore, another reasonable hypothesis is that it may be either the uncharged L-ascorbic acid or the dehydroascorbic acid product of the reduction that is involved in this process. From this data set we cannot rule out the correlation of the nitrate ion with the morphology; however, we do not have a hypothesis of a nitrate-controlled joining selectivity mechanism at this time.

It should be noted that recipe 7 and 8 represent a different regime in the PMAA ionization. In recipe 7, where the high pH value produced initial PMAA ionization of 38% as predicted by our model, the majority of the silver is in the form of prismatic crystals that we hypothesize are overgrown primary structural blocks, accompanied by limited quantities of assembled nanostructures. In recipe 8, where the PMAA initial ionization was 85% (calculated by model), the assembled nanostructures are absent. Therefore, highly ionized PMAA may be inhibiting assembly, due to charge repulsion between the coated silver nanoparticles.

There is an indication that the silver ion concentration may affect the length of the short secondary nanobelt segments prior to their final assembly, based on comparison within Fig. 1. If this is the case, we hypothesize the trend would be related to the increasing concentration of primary silver nanoplates. However, silver content is partially correlated with the trend in linear joining, complicating analysis of the length of secondary segments. Recipe 10 displays exceptionally long and linear nanobelts, at the same silver concentration as recipe 6, and differs only in the AgNO₃ dissolving procedure. The long reaction time of recipe 10 may indicate a change in the nucleation characteristics caused by this gentle mixing, believed to be an increase in the number-concentration of the silver trimer seeds and correspondingly a decrease in the availability of the silver ions for reduction by the L-ascorbic acid. It is believed that this has impacted the manner in which the primary nanoplates assemble into the secondary nanobelt segments, leading to a smaller number of longer secondary segments which, in turn, are able assemble into the achieved very long and linear tertiary nanobelts.

Conclusions

Repeatable and fast synthesis of one- and two-dimensional silver nanostructures, constructed from highly stable hexagonal and triangular nanoplates, has been achieved by controlling the reactant concentrations and handling. We found that the primary nanoplate structural blocks will assemble into secondary one-dimensional belt-shaped segments, followed by secondary assembly and joining into various larger scale tertiary structures. The selectivity of preferred joint locations during this secondary assembly directed the development of the final morphology. We also demonstrated that these are not kinetically controlled processes, and that the selectivity of joining location during secondary assembly correlated with pH and the concentration of nitrate ions in solution. There were some indications that the concentration of silver ions correlated with the developed length of the secondary belt segments, however this was correlated with the factors that also selected for more linear secondary assembly modes, complicating analysis.

References

1. N. C. Bigall, M. Reitzig, W. Naumann, P. Simon, K.-H. V. Pee and A. Eychmuller, Angew. Chem., Int. Ed., 2008, 47, 7876–7879.
2. X. Chen and H. J. Schluesener, Toxicol. Lett., 2007, 176, 1–12.
3. J. L. Elechiguerra, J. L. Burt, J. R. Morones, A. C. Bragado, X. Gao, H. H. Lara and M. J. Yacaman, J. Nanobiotechnol., 2005, 3, 6.
4. X. Luo, A. Morrin, A. J. Killard and M. R. Smyth, Electroanalysis, 2006, 18, 319–326.
5. S. Panigrahi, S. Praharaj, S. Basu, S. K. Ghosh, S. Jana, S. Pande, T. Vo-Dinh, H. Jiang and T. Pal, J. Phys. Chem. B, 2006, 110, 13436–13444.
6. P. Peng, H. Huang, A. Hu, A. P. Gerlicha and Y. N. Zhou, J. Mater. Chem., 2012, 22, 15495–15499.
7. N. Pradhan, A. Pal and T. Pal, Colloids Surf., A, 2002, 196, 247–257.
8. M. H. Rashid and T. K. Mandal, J. Phys. Chem. C, 2007, 111, 16750–16760.
9. X. Rena, X. Menga, D. Chena, F. Tanga and J. Jiaob, Biosens. Bioelectron., 2005, 21, 433–437.
10. J. P. Wilcoxon and B. L. Abrams, Chem. Soc. Rev., 2006, 35, 1162–1194.
11. L. Zhao, K. L. Kelly and G. C. Schatz, J. Phys. Chem. B, 2003, 107, 7343–7350.
12. P. Peng, A. Hu, B. Zhao, A. P. Gerlich and Y. N. Zhou, J. Mater. Sci., 2012, 47, 6801–6811.
13. H. H. Khaligh and I. A. Goldthorpe, Nanoscale Res. Lett., 2013, 8, 235–241.
14. T.-B. Song, Y. Chen, C.-H. Chung, Y. M. Yang, B. Bob, H.-S. Duan, G. Li, K.-N. Tu, Y. Huang and Y. Yang, ACS Nano, 2014, 8, 2804–2811.
15. B. Stahlmecke, F.-J. M. Z. Heringdorf, L. I. Chelaru, M. H.-V. Hoegen and G. Dumpich, Appl. Phys. Lett., 2006, 88, 053122.
16. J. L. Elechiguerra, J. Reyes-Gasga and M. J. Yacaman, J. Mater. Chem., 2006, 16, 3906–3919.
17. M. Tsuji, Y. Maeda, S. Hikino, H. Kumagae, M. Matsunaga, X.-L. Tang, R. Matsuo, M. Ogino and P. Jiang, Cryst. Growth Des., 2009, 9, 4700–4705.
18. D. Aherne, D. M. Ledwith, M. Gara and J. M. Kelly, Adv. Funct. Mater., 2008, 18, 2005–2016.
19. S. Iwama and T. Safashi, Jpn. J. Appl. Phys., 1980, 19, 1039–1044.
20. S. Iwama and K. Hayakawa, Jpn. J. Appl. Phys., 1981, 20, 335–340.
21. M. Niederberger and H. Colfen, Phys. Chem. Chem. Phys., 2006, 8, 3271–3287.
22. P. Peng, L. Liu, A. P. Gerlich, A. Hu and Y. N. Zhou, Part. Part. Syst. Charact., 2013, 30, 420–426.
23. N. Gehrke, H. Colfen, N. Pinna, M. Antonietti and N. Nassif, Cryst. Growth Des., 2005, 5, 1317–1319.
24. S. E. H. Murph, C. J. Murphy, A. Leach and K. Gall, Cryst. Growth Des., 2015, 15, 1968–1974.
25. L. Lu, A. Kobayashi, Y. Kikkawa, K. Tawa and Y. Ozaki, J. Phys. Chem. B, 2006, 110, 23234–23241.
26. E. Marzbanrad, A. Hu, B. Zhao and Y. Zhou, J. Phys. Chem. C, 2013, 117, 16665–16676.
27. W. Lv, X. Yang, W. Wang, Y. Niu, Z. Liu and W. He, ChemPhysChem, 2014, 15, 2688–2691.
28. W. Lv, W. He, X. Wang, Y. Niu, H. Cao, J. H. Dickerson and Z. Wang, Nanoscale, 2014, 6, 2531.
29. Y. Pan, W. Lv, Y. Niu, K. Wen, X. Hou, J. Gu, M. Zou, L. Ye, W. Wang, K. H. L. Zhang and W. He, RSC Adv., 2015, 5, 54605.
30. W. He, CrystEngComm, 2014, 16, 1439–1442.
31. D. A. Welch, T. J. Woehl, C. Park, R. Faller, J. E. Evans and N. D. Browning, ACS Nano, 2016, 10, 181–187.
32. Y. Xiong and Y. Xia, Adv. Mater., 2007, 19, 3385–3391.
33. C. Lofton and W. Sigmund, Adv. Funct. Mater., 2005, 15, 1197–1208.
34. M. Maillard, S. Giorgio and M.-P. Pileni, Adv. Mater., 2002, 14, 1084–1086.
35. M. H. Kim, D. K. Yoon and S. H. Im, RSC Adv., 2015, 5, 14266–14272.
36. Y. Li and C. P. Wong, Appl. Phys. Lett., 2006, 89, 112112 (112111–112113).
37. D. Zhang, L. Qi, J. Ma and H. Cheng, Chem. Mater., 2001, 13, 2753–2755.
38. D. Zhang, L. Qi, J. Yang, J. Ma, H. Cheng and L. Huang, Chem. Mater., 2003, 16, 872–876.
39. R. Riahifar, B. Raissi, C. Zamani and E. Marzbanrad, Comput. Mater. Sci., 2015, 108, 183–191.
40. R. Dang, L. Song, W. Dong, C. Li, X. Zhang, G. Wang and X. Chen, ACS Appl. Mater. Interfaces, 2014, 6, 622–629.
41. E. Marzbanrad, B. Zhao and N. Zhou, Nanotechnology, 2015, 26, 445501–445508.
42. A. P. Thompson, S. J. Plimpton and W. Mattson, J. Chem. Phys., 2009, 131, 154107.
43. T. W. Sirk, S. Moore and E. F. Brown, J. Chem. Phys., 2013, 138, 064505.
44. E. Marzbanrad, G. Rivers, P. Peng, B. Zhao and N. Y. Zhou, Phys. Chem. Chem. Phys., 2015, 17, 315–324.
45. B. M. Amoli, E. Marzbanrad, A. Hu, Y. N. Zhou and B. Zhao, Macromol. Mater. Eng., 2014, 229, 739–747.
46. A. Katchalsky and P. Spitnik, J. Polym. Sci., 1947, 2, 432–446.
47. E. G. Ball, J. Biol. Chem., 1937, 118, 219–239.
48. S. E. Kudaibergenov, Adv. Polym. Sci., 1999, 144, 115–197.
49. M. Kimura, M. Yamamoto and S. Yamabe, J. Chem. Soc., Dalton Trans., 1982, 2, 423–427.
50. H. D. Moya and N. Coichev, J. Braz. Chem. Soc., 2006, 17, 364–368.

---