Size Control of Ag3PO4 Nanoparticles Using Monoethanolamine and Oleylamine Chelating Agents

INTRODUCTION
Nowadays, population growth and human development as well as industrial activities have led to increasing pressure on water supplies. In front of this growing shortage of water, the treatment of wastewater for reuse represents an encouraging alternative. Therefore, some studies are focused on advanced oxidation processes (AOPs) for the degradation of organic pollutants resistant to the conventional methods [1]. Among them, the photocatalytic process is a viable technology for the wastewater treatments, because it is efficient, simple and inexpensive [2]. Currently, many semiconductor-based photocatalysts have been investigated to meet the requirements of the pollution remediation. However, the photocatalytic activity of these semiconductors such as ZnO [3-4] and TiO2 [5-6] is limited to the use of UV light of wavelengths under 400 nm, which corresponds to only 4% of the solar spectrum, and 46% of the visible sunlight between 400 and 800 nm is then not exploited. Therefore, the search for photocatalysts directly acting under visible-light irradiation is needed. Recently, Yi et al. [7-8] have reported that silver orthophosphate, Ag₃PO₄, has shown an excellent photo-oxidative ability for O₂ generation from water splitting under visible-light irradiation, with a quantum efficiency of about 90%. However, its crystallite size remained relatively large (0.5–2 µm) thus limiting the photocatalytic performance, which is still correlated with the specific surface area of the photocatalyst [9]. The use of nanoscale particles is a way to improve the surface better. In this context, several studies have been carried out to prepare Ag₃PO₄ using different chemical methods; such as deposition–precipitation [10], an oleate-assisted method [11], co-precipitation method [12] and simple solid phase ion exchange method [13]. However, the sol–gel method remains very common for nanoparticles fabrication due to its simplicity, reliability, cheapness and high quality products. By controlling the sol–gel process and delaying crystal growth, a good control of the size and morphology of the particles can be attained. This strategy can be achieved using a surfactant implantation (chelating agent) onto the surface of the particles preventing the occurrence of agglomeration phenomena [14]. In the present paper, a systematic study on the dependence of the nature and amount of this chelating agent (capping-ligand) on the size control is presented. Two chelating agents, monoethanolamine and oleylamine, were selected by their complexing behavior and the length of carbon chains. We demonstrate that the sol-gel process based on these chelating agents is compatible to product Ag₃PO₄ nanoparticles. Due to the high content of grafted chelating agents, these materials can be chemically treated to obtain a porous photocatalyst which then opens up highly photocatalytic activities.

MATERIALS AND METHODS
Materials 
All chemicals were used as received (form Aldrich Co.); silver nitrate (AgNO₃), sodium hydrogen phosphate (Na₂HPO₄, 2H₂O), oleylamine (70%, OLA), monoethanolamine (99%, MEA), phosphoric acid (85%, H₃PO₄), ammonia (33%, NH₃), absolute ethanol and toluene solvents (analytical grade).

Synthesis of Ag₃PO₄ powders using MEA as a complexing agent
Silver phosphate Ag₃PO₄ powders were synthesized by the sol-gel method using a mixture of AgNO₃ and Na₂HPO₄,2H₂O solutions. A first solution was prepared by dissolving a giving concentration of AgNO₃ in 25 ml distilled water, followed by a controlled addition of the MEA agent aiming to complex the Ag⁺ ion inhibiting therefore the growth of Ag-based nucleons. Then, a second solution was achieved by dissolving Na2HPO4 with an appropriate concentration in 25 mL of deionized water. This second solution was afterwards added in drops into the first one under continuous stirring for 1 h at room temperature. This solution-mixture immediately gave rise to yellow precipitates that were collected by centrifugation and successively washed with distilled water and ethanol to remove the unreacted reagents and/or by-products. Finally, the as-synthesized yellowish powders were dried at 60°C overnight in the dark.

Synthesis of Ag₃PO₄ powders using OLA
As described above and to study the effect of OLA on the crystal growth of Ag₃PO₄, two solutions were also separately prepared and then mixed in the presence of OLA molecules. The first solution of AgNO₃ with OLA was prepared and dissolved in 15 ml of toluene then stirred for about 1h at room temperature. A second solution of H₃PO₄ dissolved in 5 ml of ethanol was added to the first solution under vigorous stirring at room temperature. After 30 min, the resulting solution-mixture immediately gave rise to yellow precipitates which were ultrasonically re-dispersed in toluene and re-precipitated in ethanol and then centrifuged. The dark-yellow precipitate was then oven-dried. To remove OLA surfactant from Ag₃PO₄ matrix, the Ag₃PO₄ powders were gradually dispersed in a mixture of solvents, ammonia/water/ ethanol/toluene, with stirring at room temperature. After stirring for 1 hour, the mixture was centrifuged and the precipitates were then re-dispersed in 100 ml of ethanol. To this mixture, a solution of 50 ml of ethanol containing 1 mmol of ammonia has added and then left under stirring for another hour. Finally, the centrifuged powder is only dispersed in water and the resulting powder was dried at 60°C overnight.

Techniques 
The crystalline phase of Ag₃PO₄ was identified using an X-ray diffractometer (XRD) (Panalytical Expert Pro) in Bragg-Brentano geometry with Cu-Kα radiation source, while morphology and size of the particles were determined by a transmission electron microscopy (TEM, Thermo-Fisher, Talos) at an operational voltage of 200 kV, equipped with an energy dispersive spectrometer (EDS). FT-IR measurements have served to characterize and identify the functional groups of the powders as performed using a VERTEX 70 spectrometer.

RESULTS AND DISCUSSION 
In the present study, the particles size of the as-synthesized Ag₃PO₄ powders was scrupulously investigated according to the precursor’s concentration and the nature of chelating agent (MEA and OLA).

Effect of Mono-ethanolamine (MEA) 
To better control the overall reaction conditions effect on the particles size, a series of Ag+ concentrations (0.005M, 0.007M, 0.01M, 0.05M and 0.1M) were used. For each Ag⁺ concentration, five different amounts of MEA were tested following these MEA: Ag+ molar ratios: 0:10, 1:10, 3:10, 5:10 and 10:10.
A series of XRD patterns were performed as presented in Fig. 1. As can clearly be shown, indexation of all diffraction peaks shows the presence of the body-centered cubic structure of Ag₃PO₄ (JCPDS no. 01-071-1836). In addition, the intense and sharp peaks of all patterns indicate a good crystallinity of Ag₃PO₄ powders. No diffraction peaks resulting from impurities or secondary phases were detected, the lattice parameter “a” of the as-synthesized Ag₃PO₄ powders was determined according to:

where d=λ∕2sinɵ is the (khl) interplanar spacing and λ is the wavelength of the used x-ray radiation (λ = 1.54056 Å for Cu-Kα radiation).The average crystallite size was estimated using the Debye Scherrer’s formula [17]:

where K is a constant equal to 0.94, β is the peak full-width at half-maximum, and θ is the Bragg peak position. Results are reported in Table 1, indicating that the crystallite size of the Ag₃PO₄ decreases with increasing the MEA amount up to 5:10 ratio and then the size of the crystallites begins to increase beyond this ratio.
In order to examine the local molecular binding of Ag₃PO₄ particles and to verify the existence of MEA residues related by-products, FTIR spectra of selected samples are shown in Fig. 2. The absorption band at 940 cm^-1 is due to stretching vibration of P-O bonding, while the band at 540 cm^-1 comes from the bending vibration of O-P-O [18-19]. A broad peak of low intensity between 3000 and 3500 cm^-1 corresponds to the vibration of OH of the adsorbed water [20]. It can clearly be seen that there are no absorption band related to the MEA molecules (Stretching vibration N-H at 3300 cm^-1, deformation vibration -CH₂ at 670 cm^-1 and stretching vibration C-N at 1100 cm^-1). This result is also confirmed using TEM images (Fig. 3a) showing clean surfaces indicating a complete removal of MEA from powders. Similarly, nanoscale particle size has been detected, but their agglomeration prevents their clear visualization. For all samples, the chemical analyses by the energy dispersive X-ray (EDS) only confirms the presence of Ag, P and O elements (Fig. 3b). These analyzes are in good correlation with the XRD results within single Ag₃PO₄ phase.

Effect of oleylamine (OLA) 
As in the case of MEA chelation, a series of Ag⁺ concentrations (0.005M, 0.007M, and 0.01M) were used. For each Ag⁺ concentration, five OLA amounts were tested according to the OLA:Ag⁺ molar ratio (0:1, 1/2:1, 1:1, 2:1 and 3:1), recalling that OLA has been removed from Ag₃PO₄ powders using a dilute ammonia solution. 
The OLA surfactant effect was first investigated using XRD (Fig. 4) showing that all diffraction peaks were indexed in accordance with the cubic Ag₃PO₄ (JCPDS no. 01-071-1836). According to OLA content, the crystallite sizes are given in Table 2, showing that the size of the Ag₃PO₄ crystallites decreases as when the OLA amount increases to 2:1 ratio. Beyond this ratio, the size of the crystallites begins to increase. Fig. 5 shows IR spectra of the as-synthesized Ag₃PO₄ samples prepared at different concentrations of Ag+ and OLA:Ag⁺ molar ratios. As previously described two strong absorption bands were observed at 540 cm^–1 assigned to the O-P-O bending vibration and at 940 cm^–1 related to the asymmetric P-O stretching [18-19]. Peaks observed at 2920 and 2847 cm^-1 are due to C-H asymmetric and symmetric stretching modes of the methylene group (-CH₂), respectively [21-22]. The appearance of these two bands indicates that a very small residual amount of C–H groups from OLA is probably present in resulting powders. This was confirmed by TEM observations (Fig. 6a-b), clearly showed the presence of an amorphous-like surface layer (organic matter) around the nanoparticles. It should be noted that this layer was irremovable even after several washing steps. As previously described, chemical analyzes by EDS lead to the detection of elements Ag, P and O constituting the Ag₃PO₄ phase whatever the synthesis method used in good agreement with the Ag₃PO₄ formulation (molar ratio Ag/P close to 3).

Comparaison between MEA and OLA effects
Before comparing the effect of MEA and OLA, we first checked the effect of the solvents used (water and toluene), without chelating agent, during the synthesis and depending on the Ag+ concentration taken initially. As can clearly be observed from Fig. 7a, the crystallite size of the Ag₃PO₄ powders in both cases remained unchanged. This result is of great importance which justifies our strategies of the use of MEA and OLA as chelating agents for the nanoparticles size control. However, the variation trend of the crystallite size is very significant up to an MEA:Ag⁺ ratio of 5:10, above which the size slightly starts to increase (Fig. 7b). Although this trend is obtained for all Ag⁺ concentrations and it is important to note that the best slope is obtained for low concentrations due to their effects on the condensation kinetics during the chemical reaction [23], where Ag₃PO₄ nuclei are small enough favoring therefore a better size control. The Fig. 7c exhibits the trend of the crystallites size as a function of OLA:Ag⁺ ratio. It seems that the particles size of the Ag₃PO₄ powders decreases with increasing the OLA:Ag⁺ molar ratio up to 2:1 ratio, above which the size slightly increases. Although this trend is obtained for all Ag⁺ concentrations, it is important to note that the best slope is again obtained for low concentrations due to the small Ag₃PO₄ nuclei. Therefore, the size of Ag₃PO₄ particles can efficiently be controlled using these two chelating agents at low Ag+ concentration. However, the use of OLA provides even smaller nanoparticles. The observed difference between the two surfactants (MEA and OLA) in terms of their ability to control particle size is related their alkyl chain lengths. Both surfactants were used to block the crystalline growth and obtain nanoparticles. First, one can notice that the particles size control is best achieved using low Ag⁺ concentrations and a relatively high surfactant fraction as compared to Ag⁺. However, OLA effect seems to be more efficient since our results show that particles around 20 nm in size are successfully obtained. This result can be explained on the basis of alkyl-amines interaction with the Ag₃PO₄ nanoparticles. Silver–amine complexes obtained with different alkyl-amines (MEA and OLA) are formed via the following chemical reaction process:

The lone pair of electrons on nitrogen of amines coordinates with Ag⁺ cations and then forms the complex. As shown in previous studies, amines having a short alkyl chain as ligands (such as MEA) have a stronger activating effect, which could bring more neighbouring nuclei at low temperature due to its weak capping effect. Therefore, amines with a long alkyl chain as ligands (such as OLA) are expected to prevent a nuclei assemblage and consequently limit the silver nanoparticles growth and yield to a narrow particles size distribution. Our arguments are indeed confirmed by calculating the energy interaction between the Ag+ cations and MEA/OLA using Molecular Dynamics simulations. These simulations were performed with Lammps [26] by using the Verlet integration method based on Newton’s motions equation with a time step of 1 fs. The temperature and pressure were controlled using the isothermal-isobaric Nose-Hoover style non Hamiltonian equations of motions. The coupling constants used for temperature control were chosen as 100 fs during the heating and relaxation runs and 500 fs during data sampling at constant temperature. A cubic simulation cell of 60*60*60 nm3 in volume is used using periodic boundary conditions to mimic an infinite solution. Intermolecular potentials were evaluated up to a distance cut-off of 12 Å. MEA simulation consisted of 1100 molecules while OLA of 200 due to its large morphology (Fig. 8). Table 3 exhibits the total energy interaction between the Ag⁺ cation and MEA/Oleylamine, separately. It is found that the interaction energy of the Ag+ cation with Oleylamine (ET = -36.222 Kcal/mol) is greater than that with MEA (ET = -86.33 Kcal/mol). Therefore, the Oleylamine molecule is strongly attached to the surface of Ag+, which is directly related to the chain length of the OLA alkyl.
Finally, it is important to compare the sizes of the prepared particles with those reported in the literature. The sizes of Ag₃PO₄ particles prepared in the presence of MEA and OLA as chelating agents used in this study are significantly finely divided about 20 nm for the OLA and 42 nm for the MEA, compared to those prepared elsewhere by varying the chalking agent (Table 4). It is obvious that the affinity of OLA and MEA is more efficient for complexing the Ag+ ions and consequently retarding their reactivity with the PO4^3- anions, which blocks better the crystal growth of the grains.

CONCLUSIONS
In this study, Ag₃PO₄ nanoparticles have been successfully synthesized via a simple sol gel method using monoethanolamine and oleylamine as chelating agents to control the particles size. The size distribution of the Ag₃PO₄ particles becomes narrower with increasing the surfactant content (MEA or OLA) with Ag⁺. Using XRD, FT-IR, and TEM characterization tools, we showed that OLA provides nanoparticles with 20 nm in size, while MEA gives rise to nanoparticles with a size of 42 nm, but both associated with low concentration of Ag precursor. This result indicates that the ratio between the surfactant and Ag+ is of great importance the synthesis of nanoscale Ag₃PO₄ particles.

ACKNOWLEDGEMENTS
The Authors would like to acknowledge the support through the R&D Initiative – Call for projects APPHOS phosphates– sponsored by OCP (OCP Foundation, R&D OCP, Mohammed VI Polytechnic University, National Center of Scientific and technical Research CNRST, Ministry of Higher Education, Scientific Research and Professional Training of Morocco MESRSFC) under the project entitled *Development of a phosphate-based photocatalytic reactor prototype for the treatment and recycling of wastewater*, project ID * TRT-NAJ-01/2017*”. We thank our sponsors who provided insight, expertise and follow-up that greatly assisted this research.

CONFLICT OF INTEREST 
The authors declare that there is no conflict of interests regarding the publication of this paper.