Aggregation, dissolution and cyclic regeneration of Ag nanoclusters based on pH-induced conformational changes of polyethyleneimine template in aqueous solutions

Abstract

This paper reports a dramatic cyclic regeneration of polyethyleneimine-templated silver nanoclusters (PEI-AgNCs) based on the pH-induced conformational changes of polyethyleneimine (PEI) in aqueous solution. The PEI-AgNCs have been synthesized and found to be highly sensitive to the pH of the solution in air. The studies show that small AgNCs would gather to form larger silver nanoparticles (AgNPs) by adjusting the pH to 1.5 with a nitric acid solution. The AgNPs in solution was then transformed gradually to Ag(I) ions with stirring in air. Subsequently, the above Ag(I) ions were reduced again to AgNCs by changing the pH to about 9 with a NaOH solution and adding a certain amount of formaldehyde as a reductant to solution. The fluorescence and UV-visible absorption spectra recorded this process in detail. The transmission electron microscopy images, X-ray powder diffraction patterns, and Fourier transform infrared spectra further demonstrated that the cyclic transformation existed among AgNCs, AgNPs, and Ag(I) ions. The amino-rich PEI plays a crucial role in the regeneration of PEI-AgNCs. A large number of amino groups on PEI could be reversibly protonated by adjusting the pH of solution, leading to a change of the interaction between Ag and PEI, which has laid foundation for this work.

---

Introduction

The small-scale noble metal nanoclusters (NCs), typically consisting of a small number of metal atoms with sizes smaller than 2 nm, have attracted wide attention not only due to their unique physical and chemical properties between metal atoms and large nanoparticles^1–4 but for their broad range of applications in many fields such as labeling,⁵ catalysis,⁶ antibacterial agents,⁷ bioimaging,⁸ and sensing^9–12 etc. Silver nanoclusters (AgNCs), as the most frequently studied cheapest noble metal NCs,¹³ have been one of the focal studies in the past years owing to the different characteristics of silver metal, including photography,^14–16 catalysis,¹⁷ intense plasmon absorption band in visible-light region,¹⁸ surface enhanced Raman spectroscopy,^19–21 and the antibacterial properties known to all,²² etc. By far, a large number of research works from many laboratories have been reported on the synthesis and application of AgNCs.^23–27 Despite this, we have to admit that the research focusing on fundamental properties and behavior of this new nanomaterial in fluorescent family still does not go deeply.

Transformation of Ag nanoparticles (AgNPs) to Ag(I) ions, meanwhile, has already attracted much attention. For example, Tarasankar et al.¹³ reported the reversible formation and dissolution of AgNPs in aqueous surfactant media. Liu et al.²⁸ demonstrated that AgNPs can be oxidized to soluble Ag(I) ions, especially in the presence of elevated soluble oxygen concentrations at low pH. Li et al.²⁹ indicated the early stage aggregation kinetics of AgNPs in the presence of common electrolytes at neutral pH. Wojtysiak et al.³⁰ explored the influence of oxygen in the forming process of AgNPs. However, even though the separate studies on transformation of AgNPs to Ag(I) ions and reduction of Ag(I) ions to AgNCs by reductant are already well known, transformation from AgNCs to AgNPs and to Ag(I) ions and further the cyclic-conversion among AgNCs, AgNPs, and Ag(I) ions in a single system are not yet mentioned in previous reports. Especially for the cyclic-conversion among AgNCs, AgNPs, and Ag(I) ions in aqueous solution, the key challenge is the properties of templates to synthesize the AgNCs, which should be rapid and reversible change in a particular and controlled way and make the silver in an appropriate state correspondingly at the same time. On the other hand, the cyclic-conversion among AgNCs, AgNPs, and Ag(I) ions in this paper is substantially different from the reported cyclic reduction–decomposition synthesis,²⁴ in which Xie and co-workers used a reduction–decomposition–reduction cycle to modify the non-fluorescent AgNCs intermediates and further obtained highly luminescent AgNCs with a well-defined size and structure. The “reduction–decomposition–reduction cycle” was only to play an important role in modification, not to explore itself properties of as-prepared luminescent AgNCs. Meanwhile, our study on the transformation among AgNCs, AgNPs, and Ag(I) ions also differs from the published studies about aggregation-induced emission of Au–thiolate NCs,^31,32 in which the aggregation of Au–thiolate NCs changed its size and a facile synthesis procedure for highly luminescent Au–thiolate NCs was developed, whereas our study in this paper aimed at exploring the aggregation and dissolution of PEI-AgNCs and we were surprised to find the interesting cyclic-conversion among AgNCs to AgNPs and to Ag(I) ions.

Our previous study³³ has shown that pH response of PEI-AgNCs is due to the charge distribution of different amine groups of PEI as a function of pH and the change of conformation on the local structure of PEI. Based on the above phenomena, a highly sensitive fluorescent and colorimetric pH sensor was developed. In addition, PEI-AgNCs will aggregate to form larger AgNPs in acidic media because of the change of chain conformation of PEI. However, the aggregation and dissolution of PEI-AgNCs have not been explored in detail, and the deprotonation of PEI and regeneration of fluorescent AgNCs have not yet been investigated. Thus, we further explored the transformation from AgNCs to AgNPs and to Ag(I) ions in detail in this work. We first added a certain amount of nitric acid to the PEI-AgNCs solution to keep its pH at about 1.5, the colorless solution was observed to turn brown instantly and the fluorescence of PEI-AgNCs disappeared, while the brown solution was breached gradually in air with stirring. The experimental results show that the change of color is due to the aggregation and dissolution of AgNCs and this process was recorded by fluorescence spectroscopy and UV-vis absorption spectroscopy. At the same time, we further investigated the influences of oxygen and ionic strength on the aggregation and dissolution process.

Specially, a more interesting thing was that Ag(I) ions contained in colorless solution and obtained in the above process were reduced again and the fluorescent AgNCs were regenerated by making the solution alkaline and obtaining deprotonated PEI with a sodium hydroxide (NaOH) solution and adding a certain amount of formaldehyde to the solution. The fundamental behavior of regenerated AgNCs was observed at least 6 days by fluorescence spectroscopy and UV-vis absorption spectroscopy. The results showed the regenerated AgNCs are almost identical to the original. That is, cyclic regeneration of PEI-AgNCs can be realized in aqueous solutions by subtly adjusting pH value of solution and adding corresponding reductant.

Hence, our work based on the small AgNCs capped by PEI not only tried to explore the transformation from AgNCs to AgNPs and to Ag(I) ions, but also showed an interesting cyclic-conversion among PEI-AgNCs, AgNPs, and Ag(I) ions in aqueous solution. Furthermore, the mechanism of transformation has been primarily discussed. What's more, the finding provided a fresh perspective to understand more fundamental properties and behavior of metal nanoclusters and we believe that this study will also help researchers deeply explore more wide application of AgNCs in various fields.

Experimental section

Materials

Silver nitrate (AgNO₃, 99%), hyperbranched polyethyleneimine (PEI, M_r 10 000, 99%), 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), and formaldehyde (35 wt%) were used for synthesizing the PEI-AgNCs. Nitric acid (HNO₃) and sodium hydroxide (NaOH) were used for adjusting the pH of solution. Potassium nitrate (KNO₃) was used for adjusting the ionic strength. All other nitrate solutions used in the experiment were 0.1 M, including lithium nitrate (LiNO₃), sodium nitrate (NaNO₃), magnesium nitrate (Mg(NO₃)₂), calcium nitrate (Ca(NO₃)₂), cupric nitrate (Cu(NO₃)₂), zinc nitrate (Zn(NO₃)₂), cobalt nitrate (Co(NO₃)₂), nickel nitrate (Ni(NO₃)₂), cadmium nitrate (Cd(NO₃)₂), yttrium nitrate (Y(NO₃)₃), and aluminium nitrate (Al(NO₃)₃). All reagents used are of analytical reagent grade from Aladdin Chemical Reagent Company (China). Double-distilled water (18.2 MΩ cm) was used throughout the experiments.

Experimental instrumentations

Fluorescence spectra were recorded on an F-2700 spectrofluorophotometer (Japan) with a photomultiplier tube (PMT) voltage of 400 V and slits of 10 nm for both excitation and emission. UV-visible measurements were carried out with a Shimadzu UV-2450 UV-vis spectrophotometer (Japan). TEM images were collected using a Hitachi 7500 transmission electron microscope (TEM) with an accelerating voltage of 80 kV and the samples were not further processed under experimental conditions. X-ray powder diffraction (XRD) patterns were obtained on a Shimadzu XRD-7000 diffractometer using the Cu Kα radiation with the scan speed of 2° min⁻¹ and the scanning angle range of 2θ from 20 to 80°, the samples were prepared as follows: original PEI-AgNCs, AgNPs, and the dissolved Ag(I) ionic samples were treated by adding the solution to a glass substrate dropwise and making them dried thoroughly in air. The regenerated PEI-AgNCs sample was dialysed for 7 h using a cellulose ester dialysis membrane with the molecular weight cut-off of 14 000 before the above operation. The Fourier transform infrared (FT-IR) spectra were obtained using a Bruker IFS (Germany) 113v spectrometer after pelleting the fine powder with KBr. The samples of original PEI-AgNCs, AgNPs, and the Ag(I) ions for the FT-IR analysis were prepared by a complete drying process in air and the regenerated PEI-AgNCs solution sample was dialysed for 7 h using a cellulose ester dialysis membrane with the molecular weight cut-off of 14 000 before drying. The X-ray photoelectron spectroscopy (XPS) measurements were made on a Thermo ESCALAB 250 X-ray photoelectron spectrometer (USA), and the AgNPs sample was obtained by adding a certain amount of HNO₃ to original PEI-AgNCs in air and further vacuum dried fully. A PHS-3C meter (China) was used to adjust solution pH values. The solution was stirred using a 79-1 magnetic stirrer hot plate (China).

Preparation of polyethyleneimine-templated silver nanoclusters

Typically, according to our previous report,³⁴ 100 μL of 0.094 g mL⁻¹ PEI and 50 μL of 1 mM HEPES were first dissolved in 95 μL water by stirring for 2 min, and then 250 μL of 100 mM AgNO₃ was added and the solution was homogenized by stirring for 2 min. Then 5 μL of formaldehyde solution (35 wt%) was added with vigorous stirring, and the mixture was heated at 70 °C for 10 min. The final solution was stored at room temperature for at least 72 h before further use. Without any centrifugation or purification, the PEI-AgNCs were diluted 100-fold with distilled water and the diluted silver nanoclusters were used for further studies of its aggregation, dissolution, and cyclic regeneration in aqueous solution, except that the regenerated PEI-AgNCs solution samples were treated by dialysis process before drying to avoid the interference from nitrate and obtain more accurate XRD patterns and FT-IR spectra.

Aggregation, dissolution and regeneration of polyethyleneimine-templated silver nanoclusters

Briefly, a certain amount of HNO₃ (450 μL, 0.65 M) was added to the PEI-AgNCs solution (10 mL) which was diluted 100-fold with water to keep its pH at about 1.5 under vigorous stirring. AgNCs quickly aggregated to form large silver nanoparticles and a brown solution was observed. The solution was then stirred at room temperature (∼25 °C) for 4 h, and the silver nanoparticles gradually dissolved to form Ag(I) ions in the highly acidic media with the action of oxygen and the color of solution changed from brown to colorless. Subsequently, 360 μL of 1 M NaOH was added to the above solution to make the pH value ∼9 and the solution was stirred for another 3 h so that the reaction can be carried out completely. Next, a certain amount of formaldehyde (7 μL, 3.5 wt%) was added to the solution and stirring was kept until the solution was mixed thoroughly. Finally, the solution was placed away from light to form PEI-AgNCs again without stirring. The process of regeneration was continuously recorded by fluorescence and UV-visible spectra for 6 days. Furthermore, we explored the effects of oxygen and ionic strength on the dissolution of AgNPs by bubbling nitrogen into the solution of PEI-AgNCs to eliminate oxygen and adding 10 and 100 mM KNO₃ to the solution to change its ionic strength, respectively.

Results and discussion

Synthesis of the polyethyleneimine-templated silver nanoclusters

The PEI-templated AgNCs in aqueous solution were obtained based on our previous work.³⁴ The diluted solution of Ag nanoclusters is nearly colorless under visible light and shows two absorption bands centered at 268 and 354 nm. An intense blue emission is emitted by the PEI-AgNCs under a UV light and the fluorescence excitation and emission wavelengths are located at 375 and 455 nm, respectively (Fig. S1a in the ESI†). A TEM image confirmed that the average diameter of PEI-AgNCs is about 1.8 nm (Fig. S1b in the ESI†). Moreover, many control experiments have been performed to verify the origin of fluorescence of PEI-AgNCs in our previous work and the results indicated that the emission indeed generates from PEI-AgNCs rather than pure PEI, PEI oxidized by formaldehyde, or PEI–Ag(I) ions complexes.³⁵ In addition, the possible fundamentals of blue emission of the PEI-AgNCs have also been discussed. The emission of metal nanoclusters is dependent on the number of atoms per cluster and the interactions between metal clusters and scaffolds used in the synthesis of nanoclusters. The scaffolds may change the electron density of metal nanoclusters by charge transfer from the ligands to the metal core or direct donation of delocalized electrons of electron-rich atoms or groups of the ligands to the metal core.³⁶ In this experiment, the ligands of PEI would directly donate the delocalized electrons to the Ag core owing to the electron-rich amino groups of PEI branches and their efficient combination with Ag atoms, which would enhance the electron density of the Ag nanoclusters, but how the interaction between ligands and the metal core influences the fluorescence is still unclear.³⁵

Aggregation and dissolution of polyethyleneimine-templated silver nanoclusters

The interconversion among PEI-AgNCs, AgNPs, and Ag(I) ions was firstly studied under experimental conditions. Just as we have reported previously, the value of pK_a of the PEI used for our study is approximately 9.92 without salt and is about 10.35 in 100 mM KNO₃ solution and the chain conformation of PEI is strongly dependent on the medium pH because it possesses lots of protonated nitrogen atoms.³³ It is modeled at three protonation levels: high pH (>10) with PEI fully deprotonated, neutral pH (∼7) with all primary amines protonated, and low pH (<4) with most of amines protonated.^37,38 Therefore, when we adjusted the pH value to 1.5 by adding 0.65 M HNO₃ to 10 mL diluted PEI-AgNCs solution with stirring, almost all of amines on PEI were protonated at such a low pH, which made PEI lose its protective effect on AgNCs and the AgNCs quickly aggregated to form larger size of AgNPs, resulted in the fluorescence quenching and an obvious color change from colorless to brown of the solution.³³ Then the AgNPs gradually dissolved in solution and then produced Ag(I) ions in the highly acidic media with the action of oxygen^13,28 and the color of solution changed from brown to colorless over time (Fig. 1c), and this bleaching process needed probably 4 h with stirring in air. Here, we can observe that the aggregation of PEI-AgNCs occurs immediately in the solution at pH 1.5 and the dissolution is also relatively fast at such a low pH. Therefore, we obtained protonated PEI and studied the aggregation and dissolution of PEI-AgNCs in aqueous solutions at pH 1.5.

Fig. 1 (a) Fluorescence and (b) UV-vis absorption spectra before (no. 0) and after adjusting pH to 1.5 of the PEI-AgNCs in aqueous solution for different times (10 min, 1 h, 2 h, 3 h, 4 h). (c) The corresponding photographs under the daylight and ultraviolet lamp, respectively. The insets in (a) and (b) show the fluorescence and UV-vis absorption spectra (from 1 to 10) of the solution obtained once a minute after adjusting pH to 1.5.

The process described above was monitored by fluorescence (Fig. 1a) and UV-vis absorption spectra (Fig. 1b). The fluorescence of the system was almost completely quenched just when the pH value of the solution became 1.5. At the same time the characteristic absorption bands of PEI-AgNCs at 268 and 354 nm disappeared, whereas a surface plasma resonance absorption peak of AgNPs at about 400 nm^39–42 appeared in the UV-vis spectrum correspondingly. This phenomenon was obviously attributed to aggregation of AgNCs in acidic media to produce AgNPs. And the fluorescence had little change while the characteristic absorption peak of AgNPs weakened and eventually vanished in the next hours. At the same time, we observed a redshift of the absorption peak at about 400 nm as the dissolution of AgNPs, illustrating that the particles size increased and the aggregation was still going on although AgNPs was dissolving.⁴³ This phenomenon was consistent with the classic Derjaguin–Landau–Verwey–Overbeek (DLVO) theory of colloidal stability.²⁹

The effects of oxygen and ionic strength on the aggregation and dissolution of PEI-AgNCs have been studied in detail. The results indicated that eliminating oxygen had very little impact on the aggregation of PEI-AgNCs and formation of AgNPs, while higher ionic strength made the aggregation of PEI-AgNCs more obvious in a slight degree (Fig. S2 in the ESI†). On the other hand, the two factors both hindered the transformation of AgNPs to Ag(I) ions in different degrees (Fig. S3 in the ESI†). It can be seen from Fig. S3† that higher ionic strength has relatively little effect on dissolution of AgNPs, whereas eliminating oxygen markedly decreased the dissolution rate of AgNPs. Despite of this, AgNCs would still transform to Ag(I) ions in the presence of nitric acid even without oxygen.

The possible reasons for this result are as follows: the aggregation of AgNCs was mainly attributed to instantaneous protonation of amino groups on the PEI under highly acidic condition, making the donating electron ability of N atoms weakened dramatically and the amino groups lost the good coordination ability with Ag atoms, thereby PEI could no longer efficiently keep AgNCs stable³³ in aqueous solution. Meanwhile, the smaller AgNCs without template would spontaneously aggregate to form larger AgNPs to increase their stability according to the classic DLVO theory of colloidal stability,⁴⁴ and this process was so quick that the effects of oxygen could be negligible, whereas the results including the decrease in fluorescence intensity, the increase in absorbance, as well as the red shift of absorption peaks, demonstrated that higher ionic strength made the aggregation of PEI-AgNCs more obvious to some extent. On the other hand, the transformation of AgNPs to Ag(I) ions was slower, and was mainly due to the existence of oxygen in air, which resulted in the formation of an Ag₂O film on bare silver particles and then the Ag₂O would gradually dissolve in acidic solution over time.^13,33 Meanwhile, it has been reported that XPS measurements could be used to determine the valence state of Ag in AgNCs^45,46 and the oxidation state of the Ag on the surface of AgNPs was further determined by XPS in this experiment (Fig. S4 in the ESI†). The two binding energy peaks of Ag 3d were centered at 368 and 374 eV, which were consistent with the standard spectrum reported before.^45,46 Therefore, the oxygen played an important role in this process and eliminating oxygen would certainly hinder the transformation. In addition, oxygen-induced dissolution of AgNPs depended closely on electrolyte type and concentration,²⁹ so high ionic strength also influenced the dissolution of AgNPs to some extent.

Moreover, an obvious absorption peak at about 300 nm is observed in the UV-vis absorption spectrum of solution after adding HNO₃. The appearance of the absorption peak can be attributed to the amounts of NO₃⁻ in the solution. We find that nitric acid and many kinds of nitrates all show an absorption peak at about 300 nm in the UV-region, which is consistent with the published report,⁴⁷ meanwhile, the absorbance at about 300 nm depends on the concentration of nitrate (NO₃⁻) of the solution to some extent and is directly proportional to the concentration of NO₃⁻ for the same nitrate, for example the KNO₃. Fig. S5a (ESI)† provides the UV-vis absorption spectra of nitric acid, different types of nitrates, and several kinds of nitrates mixtures selected randomly. All of which contain same concentration of NO₃⁻ (0.1 M), and the absorbance values of nitric acid and those nitrates at about 300 nm are also given (Fig. S5b in the ESI†). Fig. S6 (ESI)† shows that the absorbance of the KNO₃ solution at 300 nm increases with increasing concentration of KNO₃.

Regeneration of polyethyleneimine-templated silver nanoclusters

Based on the above research and reversibility of protonation, we wondered if deprotonated PEI will have good coordination ability with Ag again and recover template features to keep the AgNCs stable in the solution. A further study was tried to solve the question. We obtained deprotonated PEI by adding 1 M NaOH to the above solution. Here, to avoid hydrolysis of ionic Ag(I) in highly alkaline solution, the pH value was only adjusted to about 9. Subsequently, 7 μL of formaldehyde (3.5 wt%) was added and the solution was continuously stirred until it was mixed thoroughly. The solution was then stored at room temperature and avoided light. Next, an exciting scene appeared, blue fluorescence of the solution reproduced under ultraviolet light and the characteristic absorption band of PEI-AgNCs showed up again in the UV-vis absorption spectrum. This miraculous phenomenon was constantly monitored for 6 days with an F-2700 spectrofluorophotometer and UV-2450 spectrophotometer. Fig. 2 shows the change of fluorescence and UV-vis absorption spectra of the regenerated PEI-AgNCs in the process over time. As shown in Fig. 2, in the first 6 days, the solution would produce more intense fluorescence and the characteristic absorption band of PEI-AgNCs showed more obvious increase with increasing aging time in air. One possible reason for this phenomenon is that the deprotonated PEI would have good coordination ability with Ag again and recover template features to keep the AgNCs stable in aqueous solution and protect the AgNCs from oxygen in air. Thus, longer aging time led to the production of more AgNCs with more intense fluorescence and more obvious characteristic absorption band in the first 6 days. The fluorescence intensity of regenerated PEI-AgNCs could then remain relatively stable for at least a month, as shown in Fig. S7 in the ESI.†

Fig. 2 Fluorescence (a) and UV-vis (b) absorption spectra of the regenerated PEI-AgNCs over time in 6 days, and the corresponding photographs (c) under the daylight and ultraviolet lamp, respectively.

Furthermore, we compared the fluorescence spectra and characteristic absorption peaks of original and regenerated PEI-AgNCs. It can be seen from Fig. 3 that remarkably similar excitation and emission peaks were observed for the two PEI-AgNCs and their characteristic adsorption peaks were also similar in UV-vis region. However, the fluorescence intensity of the regenerated PEI-AgNCs is only about 75% of the original nanoclusters. There may be several factors that contribute to the lower fluorescence intensity of regenerated PEI-AgNCs. Firstly, the protonated amino groups on the PEI can not be deprotonated completely in the solution when the pH value is about 9,^36,37 which will inevitably reduce the number of ligands reacting with Ag atom and decrease the amount of regenerated PEI-AgNCs compared to original PEI-AgNCs. Secondly, a lot of formic acid is produced from the redox reaction between formaldehyde and Ag(I) ion during the synthesis of original PEI-AgNCs, which may have some effects on the yield of regenerated PEI-AgNCs due to the interactions of carboxyl groups and silver. Thirdly, regenerated PEI-AgNCs is produced in dilute solution, which is diluted 100 times, the different reaction conditions probably influence the productivity of AgNCs.

Fig. 3 Comparison of the fluorescence (a) and UV-vis spectra (b) of original and regenerated PEI-AgNCs on the 6th day.

Cyclic regeneration of polyethyleneimine-templated silver nanoclusters

More characterizations about whole experiment procedure were done to further confirm that the circulation among PEI-AgNCs, AgNPs, and Ag(I) ions in aqueous solution. The TEM images in Fig. 4 were obtained with same rulers and indicated that the original PEI-AgNCs with a size of ∼2 nm (Fig. 4a) aggregated to form larger AgNPs with the particle size range from several to several ten nanometers (Fig. 4b) after adjusting the pH to 1.5, and when all of the AgNPs finally dissolved to form Ag(I) ions, no particles was observed in the TEM image, as shown in Fig. 4c. However, when NaOH (360 μL, 1 M) and formaldehyde (7 μL, 3.5 wt%) were added to the solution, the fluorescence of solution from PEI-AgNCs appeared again. Fig. 4d shows the morphology of regenerated PEI-AgNCs with the average diameter of about 1.6 nm on the 6th day and it dispersed better and particle size was more uniform compared to the original PEI-AgNCs.

Fig. 4 TEM images of original PEI-AgNCs (a), AgNPs (b), dissolved Ag(I) ions (c), and regenerated PEI-AgNCs on the 6th day (d). The insets show the particle size distributions.

Meanwhile, the corresponding fluorescence and UV-vis absorption spectra of cyclic transformation provided another proof for the above circulation (Fig. 5). When the small AgNCs agglomerated into AgNPs, fluorescence of solution was quenched, the characteristic absorption peaks of PEI-AgNCs disappeared in the UV-vis absorption spectra and the intrinsic absorption band of AgNPs at about 400 nm appeared correspondingly. Then the absorption band at around 400 nm simply faded away whereas the fluorescence had little recovery as the AgNPs dissolved to form Ag(I) ions. Next, the fluorescence recovered largely and the characteristic absorption peaks of PEI-AgNCs appeared again when the PEI-AgNCs were regenerated.

Fig. 5 Fluorescence (a) and UV-vis (b) absorption spectra of the circulation. Numbers 1, 2, 3, and 4 represent original PEI-AgNCs, AgNPs, Ag(I) ions, and regenerated PEI-AgNCs on the 6th day, respectively.

In addition, as we all know, larger metal nanoparticles possessing metallic properties will show the characteristic peak of metal, whereas the small nanoclusters only show a broad peak in XRD patterns.^35,48–50 Therefore, this cyclic regeneration of PEI-AgNCs could be also confirmed by XRD measurements. Although the noise/signal ratios for the XRD measurements were higher, which maybe result from large amount of polymers PEI and the low level of Ag in the XRD samples, we can still find the evidence of cyclic regeneration of PEI-AgNCs in XRD patterns. As shown in Fig. 6, we can only observe a broad peak between 25° to 35° in XRD patterns of the original (Fig. 6a) and regenerated PEI-AgNCs (Fig. 6d), whereas the metallic AgNPs exhibited four sharper peaks at 38°, 44°, 64°, and 78° (peaks at 38°, 64°, 78° correspond to the three peaks at (111), (200), and (311) reported in ref. 48 and 50) in its XRD pattern (Fig. 6b) and the XRD pattern has no peak when AgNPs dissolved completely to form Ag(I) ions (Fig. 6c). In addition, we can also find the evidence of the cyclic conversion in FT-IR spectra. Fig. S8† depicts the FT-IR spectra of original PEI-AgNCs (a), AgNPs (b), Ag(I) ions (c), and regenerated PEI-AgNCs (d) and the results showed that the original and regenerated PEI-AgNCs are almost identical.

Fig. 6 XRD patterns of Ag situated in different states in the circulation. (a) original PEI-AgNCs; (b) AgNPs; (c) Ag(I) ions; (d) regenerated PEI-AgNCs on the 6th day.

Moreover, the circulation among PEI-AgNCs, AgNPs, and Ag(I) ions could be realized more than once under the same operational conditions. Fig. S9 in the ESI† shows the fluorescence spectra of original and regenerated PEI-AgNCs after 1 and 2 cycles. Therefore, we finally confirm the recycled transformation of AgNCs → AgNPs → Ag(I) ions → AgNCs based on the well-defined PEI-AgNCs in aqueous solution. The whole process can be briefly stated as follows: first, HNO₃ was added to the original PEI-AgNCs solution to adjust the solution with high acidity, which made the PEI be protonated quickly and no longer keep the AgNCs stable, leading to the aggregation of AgNCs and formation of larger AgNPs. Second, The AgNPs would gradually dissolve in the presence of oxygen with stirring in acidic solution. At last, when the AgNPs completely dissolved, the pH value of solution was adjusted to about 9 with NaOH solution to restore the template features of PEI, and right amount of formaldehyde was added as the reductant. Thus, the PEI-AgNCs were obtained again.

Conclusion

In summary, this work demonstrates an interesting state transformation of PEI-AgNCs based on the pH-induced conformational change of PEI in aqueous solution. We not only explored the aggregation and dissolution of the AgNCs, but also attempted and fulfilled the cyclic regeneration of the AgNCs capped by PEI based on the reversibility of PEI protonation under experimental conditions. Meanwhile, the cyclic transformation among AgNCs, AgNPs, and Ag(I) ions enriches the theory and research on fundamental properties and behavior of metal nanoclusters in solution. And what's more, this finding has great potential for applications by providing a very unique perspective on cyclic regeneration of PEI-AgNCs and the transformation among AgNCs, AgNPs, and Ag(I) ions.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (nos 20975083 and 21273174) and the Municipal Science Foundation of Chongqing City (no. CSTC-2013jjB00002).

Notes and references

1. R. C. Jin, Nanoscale, 2010, 2, 343.
2. I. Diez and R. H. A. Ras, Nanoscale, 2011, 3, 1963.
3. Q. B. Zhang, J. P. Xie, Y. Yu and J. Y. Lee, Nanoscale, 2010, 2, 1962.
4. L. Shang, S. Dong and G. U. Nienhaus, Nano Today, 2011, 6, 401.
5. C. I. Richards, S. Choi, J. C. Hsiang, Y. Antoku, T. Vosch, A. Bongiorno, Y. L. Tzeng and R. M. Dickson, J. Am. Chem. Soc., 2008, 130, 5038.
6. J. Belloni, Catal. Today, 2006, 113, 141.
7. C. Marambio-Jones and E. M. V. Hoek, J. Nanopart. Res., 2010, 12, 1531.
8. L. Shang, N. Azadfar, F. Stockmar, W. Send, V. Trouillet, M. Bruns, D. Gerthsen and G. U. Nienhaus, Small, 2011, 7, 2614.
9. H. C. Yeh, J. Sharma, J. J. Han, J. S. Martinez and J. H. Werner, Nano Lett., 2010, 10, 3106.
10. W. Guo, J. Yuan, Q. Dong and E. Wang, J. Am. Chem. Soc., 2009, 132, 932.
11. S. W. Yang and T. Vosch, Anal. Chem., 2011, 83, 6935.
12. J. Yu, S. Choi and R. M. Dickson, Angew. Chem., Int. Ed., 2009, 48, 318.
13. P. Tarasankar, K. S. Tapan and R. J. Nikhil, Langmuir, 1997, 13, 1481.
14. G. Q. Wang, T. Nishio, M. Sato, A. Ishikawa, K. Nambara, K. Nagakawa, Y. Matsuo and K. N. K. Ijiro, Chem. Commun., 2011, 47, 9426.
15. M. Ozaki, J. Kato and S. Kawata, Appl. Opt., 2013, 52, 6788.
16. T. M. Petrova, B. Tzaneva, L. Fachikov and J. Hristov, Chem. Eng. Process., 2013, 71, 83.
17. J. W. Rigoli, C. D. Weatherly, J. M. Alderson, B. T. Vo and J. M. Schomaker, J. Am. Chem. Soc., 2013, 135, 17238.
18. M. Özyürek, N. Güngör, S. Baki, K. Güçlü and R. Apak, Anal. Chem., 2012, 84, 8052.
19. H. K. Lee, Y. H. Lee, Q. Zhang, I. Y. Phang, J. M. R. Tan, Y. Cui and X. Y. Ling, ACS Appl. Mater. Interfaces, 2013, 5, 11409.
20. C. F. Juan, A. P. Luis and A. C. Eduardo, J. Phys. Chem. C, 2013, 117, 23090.
21. X. Y. Li, J. Li, X. M. Zhou, Y. Y. Ma, Z. P. Zheng, X. F. Duan and Y. Q. Qu, Carbon, 2014, 66, 713.
22. A. K. Karumuri, D. P. Oswal, H. A. Hostetler and S. M. Mukhopadhyay, Mater. Lett., 2013, 109, 83.
23. T. Li, L. B. Zhang, J. Ai, S. J. Dong and E. K. Wang, ACS Nano, 2011, 5, 6334.
24. X. Yuan, M. I. Setyawati, A. S. Tan, C. N. Ong, D. T. Leong and J. P. Xie, NPG Asia Mater., 2013, 5, e39.
25. J. J. Li, X. Q. Zhong, H. Q. Zhang, X. C. Le and J. J. Zhu, Anal. Chem., 2012, 84, 5170.
26. J. Sharma, R. C. Rocha, M. L. Phipps, H. Yeh, K. A. Balatsky, D. M. Vu, A. P. Shreve, J. H. Werner and J. S. Martinez, Nanoscale, 2012, 4, 4107.
27. Y. Xiao, F. Shu, K. Y. Wong and Z. H. Liu, Anal. Chem., 2013, 85, 8493.
28. J. Liu and R. H. Hurt, Environ. Sci. Technol., 2010, 44, 2169.
29. X. Li, J. J. Lenhart and H. W. Walker, Langmuir, 2010, 26, 16690.
30. S. Wojtysiak and A. Kudelski, Colloids Surf., A, 2012, 410, 45.
31. Z. T. Luo, X. Yuan, Y. Yu, Q. B. Zhang, D. T. Leong, J. Y. Lee and J. P. Xie, J. Am. Chem. Soc., 2012, 134, 16662.
32. Y. Yu, Z. T. Luo, D. M. Chevrier, D. T. Leong, P. Zhang, D. E. Jiang and J. P. Xie, J. Am. Chem. Soc., 2014, 136, 1246.
33. F. Qu, N. B. Li and H. Q. Luo, Langmuir, 2013, 29, 1199.
34. F. Qu, N. B. Li and H. Q. Luo, Anal. Chem., 2012, 84, 10373.
35. F. Qu, N. B. Li and H. Q. Luo, J. Phys. Chem. C, 2013, 117, 3548.
36. Z. K. Wu and R. C. Jin, Nano Lett., 2010, 10, 2568.
37. J. Nagaya, M. Homma, A. Tanioka and A. Minakata, Biophys. Chem., 1996, 60, 45.
38. M. Borkovec and G. J. M. Koper, Macromolecules, 1997, 30, 2151.
39. B. G. Ershov, E. Janata, A. Henglein and A. Fojtik, J. Phys. Chem., 1993, 97, 4589.
40. A. Henglein, P. Mulvaney and T. Linnert, Faraday Discuss., 1991, 92, 31.
41. B. G. Ershov and A. Henglein, J. Phys. Chem. B, 1998, 102, 10663.
42. T. Linnert, P. Mulvaney, A. Henglein and H. Weller, J. Am. Chem. Soc., 1990, 112, 4657.
43. Y. Q. Qin, X. H. Ji, J. Jing, H. Liu, H. L. Wu and W. S. Yang, Colloids Surf., A, 2010, 372, 172.
44. E. J. W. Verwey and J. T. G. Overbeek, Theory of the Stability of Lyophobic Colloids, Elsevier, Amsterdam, 1948.
45. Z. Q. Yuan, N. Cai, Y. Du, Y. He and E. S. Yeung, Anal. Chem., 2014, 86, 419.
46. X. Yang, L. F. Gan, L. Han, E. K. Wang and J. Wang, Angew. Chem., Int. Ed., 2013, 52, 2022.
47. B. Manuel, C. O. Alejandro and O. Beatriz, J. Chil. Chem. Soc., 2009, 54, 93.
48. T. U. B. Rao and T. Pradeep, Angew. Chem., Int. Ed., 2010, 49, 3925.
49. T. U. B. Rao, B. Nataraju and T. Pradeep, J. Am. Chem. Soc., 2010, 132, 16304.
50. M. S. Bootharaju and T. Pradeep, Langmuir, 2011, 27, 8134.

---

Footnote

† Electronic supplementary information (ESI) available: Fig. S1 to S9. See DOI: 10.1039/c4ra14812f

---