Stability and Aggregation Kinetics of Silver Nanoparticles in Water in Oil Microemulsions of Cetyltrimethylammonium Bromide and Triton X-100

Water in oil (W/O) microemulsions are simple preparative route for nanoparticles where water droplets dispersed in oil stabilized by surfactant or surfactant and cosurfactant monolayer act as nanoreactors to carry out chemical reactions. In this work, silver nanoparticles (AgNPs) were prepared in W/O microemulsions of cetyltrimethylammonium bromide (CTAB) and triton X-100 (TX-100) by using AgNO3 and NaBH4 as a precursor salt and reducing agent, respectively. To prepare microemulsions, CTAB or TX-100, 1-pentanol, cyclohexane and water were mixed with different molar ratio. AgNPs were prepared with different [AgNO3] in microemulsions of CTAB with fixed water to surfactant ratio (Wo). Average particle sizes were determined from dynamic light scattering (DLS) measurements. AgNPs prepared from microemulsions of CTAB were unstable while from TX-100, NPs were stable. Aggregation kinetics was investigated by measuring the absorbance at definite time intervals at the absorption maximum, λmax of AgNPs in different media under pseudo-first-order conditions. The aggregation behavior was studied at different [AgNO3]:[NaBH4] and Wo and the parameters were optimized to ensure formation of stable AgNPs without aggregation in microemulsions. This would help tuning the size, stability, and aggregation kinetics of AgNPs by controlling the nature of the surfactant and composition of the microemulsions.


Introduction
Nanoparticles (NPs) are of great scientific interest as they bridge the gap between bulk materials and atomic or molecular structures [1].Among various NPs, silver nanoparticles (AgNPs) are the most promising ones.The physical and chemical properties of the AgNPs change with change in the size or surface of the NPs [2,3].Good conductivity, optical properties, and chemical stability of AgNPs made them promising for a range of applications for instances, in industries as catalysts, in medical sciences to deal with HIV virus and in food industries as anti-bacterial agents for food packing, etc [4][5][6][7][8][9][10].
The usage of AgNPs more often becomes the concern for their toxicity at high concentrations.The production of goods containing AgNPs, in general, increases release to the environments with possible harmful effect on human beings [11,12].Once released into the environment, AgNPs may bring environmental transformations which inter alia include: aggregation, sedimentation, oxidation, interaction with organic matter and pollutants.Thus, it is important to investigate factors that determine the stability and aggregation of AgNPs in addition to intrinsic properties like diameter, charge, shape, concentration and coating [13][14][15][16].
There are numerous reports in the literature on the preparation of AgNPs from microemulsions using a range of surfactants [33][34][35][36][37][38][39].Surprisingly, stability and aggregation of AgNPs prepared from different W/O microemulsions still remains an unsettled issue.In this work, we prepared AgNPs from microemulsions of a cationic surfactant, cetyltrimethylammonium bromide (CTAB) and a nonionic surfactant, triton X-100 (TX-100) with different water to surfactant ratio (W o ).The stability and aggregation kinetics of prepared AgNPs have been studied in detail.The effects of ratio of AgNO 3 with NaBH 4 and W o on stability and aggregation of AgNPs have been examined to find suitable means to prepare stable AgNPs avoiding aggregation.

Preparation of water in oil microemulsions of CTAB and TX-100
Water in oil (W/O) microemulsions were prepared by mixing 10.0 %wt.surfactant (for both CTAB and TX-100) with cyclohexane as the oil phase (62.3 %wt.and 83.0 %wt.for CTAB and TX-100, respectively) followed by addition of water and n-pentanol to have clear microemulsions with varying water and n-pentanol contents.These microemulsions were prepared by using deionized double distilled water and the water to surfactant ratio (W o ) was maintained in the range of 3.90 ~ 20.0.The required amount of CTAB/TX-100, n-pentanol, cyclohexane and water were added one after another in a vial and sonicated to prepare clear W/O microemulsions.

Preparation and characterization of AgNPs
Aqueous AgNO 3 solution and NaBH 4 were used, respectively, as a metal salt precursor and a reducing agent for the preparation of AgNPs.The characteristic pale yellow color after the addition of NaBH 4 solution indicated the formation of AgNPs in aqueous solution.
AgNPs were synthesized using double microemulsions reactant addition method. 2 mL of two microemulsions, one solubilized with AgNO 3 and the another with NaBH 4 as the reducing agent were mixed and sonicated for about 30 min.An immediate color change; from colorless to golden yellow solution indicated the formation of AgNPs [17].
The optical absorption spectra for AgNPs in water and microemulsions with different surfactants were recorded using a double beam Shimadzu UV-visible spectrophotometer, UV-1800.Rectangular quartz cells of path length 1 cm were used throughout the investigation.Size and size distribution of the AgNPs synthesized in water and the water pools of the W/O microemulsion were measured using a Zetasizer Nano ZS90 (ZEN3690, Malvern Instruments Ltd., UK) using dynamic light scattering method (DLS) method.The particle size detection limit was about 0.3 nm -5 m (diameter) and accuracy of the average diameter determined has been ±2%.A He-Ne laser beam of 632.8 nm wavelength was used and the measurements were made at a fixed scattering angle of 90.The average diameters were determined from cumulants mean of the intensity average of 50 runs using Stokes-Einstein equation, R h = kT/6πηD eff , where R h is hydrodynamic radius, k is the Boltzmann constant, T is the temperature, η is the solvent viscosity, and D eff represents the diffusion coefficient.The reproducibility was checked from at least 3 measurements.The temperature of the apparatus was controlled automatically within ± 0.01K by a built-in Peltier device.

Kinetic study
The rate of the formation/aggregation AgNPs was measured by solubilizing appropriate amount of AgNO 3 in water or W/O microemulsions of CTAB or TX-100 and adding calculated amount of NaBH 4 solution ([NaBH 4 ] >> [AgNO 3 ]).The progress of the reaction was followed spectrophotometrically by pipetting out aliquots at definite time intervals and measuring the absorbance of silver sol formed at the absorption maximum, λ max corresponding to the pale-yellow color.Water or microemulsion containing all the constituents except AgNO 3 and NaBH 4 was used as the reference.In the cases of kinetic measurements, absorbance at the  max of AgNPs in microemulsions of CTAB/TX-100 was monitored since the  max was found to shift slightly in these media compared to that in water.Replicate measurements gave results that were reproducible within ± 4%.The apparent rate constants (k obs , s -1 ) were calculated from the slopes of the semilogarithimc plots of concentration and time [42].

Formation/aggregation and characterization of AgNPs in water
The AgNPs were prepared from water following the procedure described by Solomon et al [17].The stable yellow AgNPs could be prepared from water by controlling reaction conditions including stirring time and relative amounts of reagents.Upon continuous stirring for 2 h after the addition of all AgNO 3 , the color of the solution did not show appreciable change indicating stability of AgNPs.In contrast, after 10 days, darker yellow solution turned into grayish solution due to aggregation of AgNPs (vide infra).

Characterization
AgNPs showed yellowish brown color in water due to excitation of surface plasmon resonance band in the UV-visible region or interband transition of AgNPs [43,44].The UV-Vis spectroscopy could therefore be used to examine size and shape of the AgNPs.Fig. 1 represents the UV-visible spectra of synthesized AgNPs stirring at 40 min and 2 h more after the formation of AgNPs.
The absorption spectra of AgNPs in aqueous solution consist of a single sharp surface plasmon resonance band at 395 nm (Fig. 1).The most characteristic part of silver sol is a narrow plasmon absorption band observable in the 350-600 nm regions.A broad surface plasmon resonance band is due to aggregation of AgNPs.The wavelength of the plasmon absorption maximum in a given solvent can be used to indicate particle size [45].The value of the peak width at half maximum (PWHM) for freshly prepared AgNPs are 67 and 70 nm for AgNPs stirred at 40 min and 2 h 40 min, respectively.After 10 days the value of the absorption maximum changes from 395 to 410 and 405 nm for AgNPs stirred for 40 min and 2 h 40 min, respectively.The corresponding values of PWHM of freshly prepared AgNPs and 10 days after preparation stirred at 40 min and 2 h 40 min are listed in Table 1.Table 1.Peak width at half maximum (PWHM) of AgNPs.

Particle size of AgNPs
Particle sizes of the prepared AgNPs were determined by DLS method.Table 2 represents sizes of freshly prepared AgNPs with their scattering intensity, polydispersity index (PDI) and standard deviations (Std.Dev.) for NPs stirred for 40 min.
From Table 2, it is apparent that, the Z-Average diameters of freshly prepared AgNPs are lower than those for particles after 10 days of preparation to indicate aggregation of AgNPs.Size distribution shows bimodal and trimodal distributions for freshly prepared AgNPs and those after 10 days of preparation.The % intensity indicates that the amount of larger NPs is more than the smaller one.The PDI value suggests that the AgNPs prepared from water are polydispersed with broader size distribution: aggregation for a prolonged period for instance, after 10 days of preparation gives rise to larger aggregates of different sizes.

Formation/aggregation and characterization of AgNPs in W/O microemulsions of CTAB
The formation of AgNPs in microemulsion could be visually observed from characteristic color change from colorless to golden yellow in the microemulsions [17,46,47] and therefore UV-visible spectral analysis could be used for the study of formation and aggregation behavior in microemulsions.It is interesting to note a considerable amount of large particles (Table 3) at lower [AgNO 3 ].At higher [AgNO 3 ] on the other hand, there were substantial amount of smaller AgNPs which is indicative of a faster nucleation of the AgNPs.Small AgNPs with a high specific surface area could be formed in a short formed which is in good agreement with literature [48].

Stability and aggregation of AgNPs in water and W/O microemulsions of CTAB
The absorption spectrum of AgNPs changes due to transition from a well-dispersed state to an aggregated state.The absorbance vs time profile in Fig. 3 shows an increase in absorbance (up to 55~60 min for water and 10~15 min for microemulsion of CTAB (W o = 7.10)) followed by decrease with increasing time.The increase in absorbance indicates an increase in the no. of AgNPs due to faster nucleation; while decrease in absorbance with time corresponds to     481V, [51]) follows a four step growth mechanism proposed by Polte (Fig. 6) [52].Reduction of Ag + to Ag 0 atoms occurs in the first step and different types of clusters like dimers, trimers are formed.Within ≈ 5 s, the clusters coalesce to generate small particles in a second stage; followed by a metastable state, where the particles maintain a constant size for around 5-10 min.A second and last coalescence phase takes place (within 30-60 s) to render the final AgNPs (Fig. 6) [51].All of these stages are completed within 5-10 min during the formation of AgNPs.Wuithschick et al. studied the metastable state and final coalescence phase by timeresolved SAXS in detail and showed that colloidal stability during the intermediate state is enough to prevent further growth due to particle aggregation [53].It is well known that each coalescence process implies an aggregation process indicating a decrease in particle stability.The total consumption of BH 4 − is a triggering event for the initiation of the second coalescence phase due to loss of stability reasonable by oxidation of the metal surface [53].Although the reaction stoichiometry with Ag + is usually 1:1, BH .Hydrolysis is slower compared to the reductive reaction, however, becomes the dominant process once all the Ag + is reduced.

Aggregation kinetics of AgNPs
The partial oxidation of the AgNPs surface forming AgO also takes place.The electrostatic stabilization of AgNPs decreases through the presence of these oxides provoking their aggregation.The oxidation of the AgNPs surface can be reversed when there is still BH 4 − left in solution that avoids the aggregation process.But once all BH 4 − ions are consumed, the final coalescence stage commences until the stable size is obtained [53].Thus, using a strong reductant, like NaBH 4 a faster reduction of Ag + compared to the growth process can be ensured and the growth mechanism is governed by coalescence closely related to the electrostatic stabilization of the AgNPs.In microemulsions with low water content, the electrostatic repulsions between surface charges of AgNPs with hydrophilic part of CTAB are so strong to lower the stability of AgNPs; the AgNPs, containing positive charge in their surface and the hydrophillic part, trimethylammonium ions, (-N + (CH 3 ) 3 ) of CTAB repels each other as well as the AgNPs release from the water droplets and start to aggregate quickly (after ~ 30-35 min).In contrast, microemulsions of high water content having larger water droplets decreases the electrostatic repulsion between positive surface charge of AgNPs and the -N + (CH 3 ) 3 of CTAB which gives more stability (~ 80-85 min) of the AgNPs and aggregation starts very slowly (after ~ 80-85 min) in microemulsions.Thus, high water content in the micremulsion ensures more stable AgNPs and slower aggregation then low water content microemulsions during the preparation of AgNPs.

Stability and aggregation of AgNPs in W/O microemulsions of TX-100
AgNPs have also been prepared from microemulsions of TX-100 to study the influence of nature of surfactants on the formation/aggregation and stability.Fig. 8 represents size distribution (determined from DLS measurements) of AgNPs prepared from microemulsions of TX-100 at different W o .The average sizes of AgNPs prepared from micremulsions of TX-100 with W o = 3.5 are ~81 nm (DLS method) while from spectroscopic method this value is ~ 66 nm (PWHM).This is due to the fact that DLS measures the sizes of NPs along with the hydrodynamic diameters of microemulsion droplets.It may be worth noting that correlograms (correlation intensity vs time) in the DLS measurements were characteristic of stable species in the system without any sedimentation and aggregation during the course of measurements.For all the W o , the size distribution profiles give bimodal distributions with varying sizes of particles.This may be due to the presence of AgNPs and/or aggregated one.Here, the AgNPs with smaller size are only considered for discussion.As the droplet sizes of microemulsions increases with increasing W o [54], the size of AgNPs also increases.On the other hand, from Fig. 8, surprisingly, it has been found that the average sizes of AgNPs decreases with increases W o of the microemulsions and give narrower size distribution with higher %intensity value for W o = 15.0.This may be due to the fact that with increasing W o , the sizes of water droplets increases in which more reactant species (Ag + and BH 4 -) are solubilized into the water core compared to the microemulsions with lower water content.Thus, the AgNPs with smaller sizes may be observed due to faster nucleation and faster reduction [48] for W o = 15.0.It is well-known that microemulsions can change the rates, pathway of chemical reactions etc. by virtue of their medium effect due to the incorporation and/or solubilization of reactants into the small volume of water droplets through hydrophobic, electrostatic, hydrogen bonding and Van der Waals forces [55][56][57].The presence of positive charge on Ag + and AgNPs must be considered to explain the different behavior of water droplets in CTAB and TX-100 microemulsions.In microemulsions of CTAB, Ag + and BH 4

Role of surfactants in the stabilization of AgNPs
 are expected to be mainly in the oil phase and micellar pseudophases, respectively due to electrostatic repulsion and interaction between Ag + and BH 4  with positive -N + (CH 3 ) 3 of CTAB micelles, respectively.On the other hand, AgNPs stabilized by hydrophilic oxyethylene groups of TX-100 correspond to the steric as well as hydrophobic interactions in water droplets of microemulsions [50].One key factor is the reaction rate which is faster in presence of CTAB than that of TX-100.Aggregation of AgNPs in microemulsions of TX-100 is much slower than that of CTAB due to the fact that the sterically and hydrophobically stabilized AgNPs may be more resistant to aggregation compared to the electrostatically stabilized AgNPs [50].Thus, the extent of the stability of AgNPs having the same stabilization mechanisms may differ.

Conclusions
Stability and aggregation can be controlled by using microemulsions with proper choice of surfactants as templates for preparation of AgNPs.In microemulsions of CTAB, with increasing the concentration of AgNO 3 , the number of AgNPs increases.On the contrary, AgNPs prepared from CTAB microemulsions are unstable and precipitate out into the bulk and aggregate depending on the concentration of AgNO 3 .With increasing the ratio of AgNO 3 and NaBH 4 , stability of AgNPs prepared from microemulsions of CTAB remains constant (~ 30-35 min); however the rate of aggregation decreases.With change in water to surfactant ratio of CTAB microemulsions, stability of AgNPs can be increased up to ~ 80-85 min.Silver nanoparticles prepared from TX-100 are more stable (~ 3 days) than that of CTAB.In addition, aggregation occurs very slowly in TX-100 compared to CTAB.Thus, a suitable surfactant, compositions, ratio of AgNO 3 and NaBH 4 or time is essential for the production of stable silver nanoparticles.

Fig. 1 .
Fig. 1.UV-visible spectra of synthesized AgNPs in aqueous solution stirring at 40 min and 2 h more after the formation of AgNPs.

Fig. 3
shows the change of absorbance at 395 nm with increasing time during the formation of AgNPs from water and W/O microemulsions of CTAB (W o = 7.10) with [AgNO 3 ]:[NaBH 4 ] = 1:6.

[Fig. 3 .
Fig. 3. Absorbance at 395 nm as a function of time during the formation of AgNPs from water and W/O microemulsions of CTAB with AgNO 3 :NaBH 4 = 1:6.Absorbance in the left corresponds to that for aqueous solution and in the right for microemulsions.
AgNPs have also been prepared by changing [AgNO3 ]:[NaBH 4 ] (from 1:10.8 to 1:30.8) from microemulsions of CTAB (W o = 7.10) to study the effect of the strong reducing agent, NaBH 4 on the stability and aggregation.The stability and aggregation of AgNPs were measured by monitoring the change of absorbance at 445 nm (λ max ) with time.Fig. 4 represents the absorbance vs time profile for AgNPs with different [AgNO 3 ]:[NaBH 4 ] in microemulsions of CTAB (W o = 7.10).It is apparent from Fig. 4 that the absorbance increases until ~10 min and remains constant up to ~30 min and then decreases.The AgNPs are stable for ~15-20 min and aggregation starts after ~ 30 min for all [AgNO 3 ]:[NaBH 4 ].
Stability and aggregation kinetics have been studied by measuring the absorbance at definite time intervals at λ max of AgNP in microemulsions of CTAB.Aggregation kinetics have been measured after 30 min (when aggregation started) under pseudo-first order conditions using [NaBH 4 ] >> [AgNO 3 ] during the preparation of AgNPs [24,25,49,50].Different [AgNO 3 ]:[NaBH 4 ] were used to prepare AgNPs and hence to monitor the effect of [NaBH 4 ] on aggregation.Fig. 5 shows the rate of aggregation of AgNPs with different [AgNO 3 ]:[NaBH 4 ] prepared from microemulsions of CTAB (W o = 7.10).

Fig. 6 .
Fig. 6.Four step growth mechanism during the formation of AgNPs synthesized by the reduction of AgNO 3 using NaBH 4 as reducing agent [52].

3. 3 . 3 .
Effect of W o of microemulsions of CTAB on stability and aggregation Stability and aggregation of AgNPs have also been investigated at different W o .The stability increases and the extent of aggregation decreases with increasing W o .AgNPs, prepared from microemulsions of CTAB with W o = 15.0 and 20.0 were stable for ~ 80-85 min (Fig. 7) while they were stable only for ~ 30-35 min prepared from microemulsions of CTAB with W o = 7.10 (Fig. 4) for [AgNO 3 ]:[NaBH 4 ] = 1:23.1.In addition, the rate of aggregation of AgNPs is very slow for microemulsions of CTAB with W o = 15.0 and 20.0 compared to that of microemulsions of CTAB with W o = 7.10.This may be due to the fact that with increasing W o , the size of core of W/O microemulsions increases [54] by swelling to result in decrease in electrostatic repulsions of surface charge of AgNPs with hydrophillic part, -N + (CH 3 ) 3 of CTAB.With change in water content in the W/O microemulsions of CTAB, the solubilization capacity of AgNO 3 /NaBH 4 inside the water droplets of the microemulsion of varying sizes differs and consequently brings about changes in electrostatic repulsions of surface charge of AgNPs with hydrophillic part of CTAB.

Fig. 8 .
Fig. 8. Size distribution of AgNPs prepared from microemulsions of TX-100 for different W o by DLS method.

Fig. 9
shows the stability and aggregation of AgNPs in W/O microemulsions of TX-100 with (W o = 3.50) for [AgNO 3 ]:[NaBH 4 ] = 1:23.1.After 30 min stirring, the absorbance still increases with time which is due to either the growth of AgNPs or the increase in the no. of AgNPs.The change in absorbance is almost constant after 1 day with time and indicates the stability of AgNPs.After 3 days, the absorbance decreases very slowly due to the aggregation of AgNPs.Thus, AgNPs prepared from microemulsions of TX-100 are more stable (~ 3 days) then the AgNPs prepared from microemulsions of CTAB.
Scheme 1.The stabilization mechanisms of AgNPs prepared from microemulsions of CTAB (a) and TX-100 (b).

Table 2 .
Size for AgNPs stirred for 40 min.The PDI, a width parameter, has been calculated from a Cumulants analysis of the DLS measured intensity autocorrelation function.In general, values smaller than 0.1 indicate reasonably narrow distribution and values greater than 0.7 indicate that the sample has a very broad size distribution. 1

Table 3 .
The absorption intensity increases with increase in [AgNO 3 ] increases to suggest formation of more AgNPs.Table 3 represents concentration and PWHM values of AgNPs with increasing [AgNO 3 ].Concentrations and PWHM values of AgNPs with increasing [AgNO 3 ].