**3.2 Effect of Ag (I)'s concentration on the recovery efficiency**

In order to evaluate the influence of Ag(I)'s concentration on the efficiency recovery of the proposed separation system, two tests with different concentrations of Ag(I) (25 and 100 mg L-1), were performed. For the first test we used a 0.25 mol L-1 of HNO3 as the feed solution

Silver Recovery from Acidic Solutions by Formation of

fewer and smaller particles on the membrane.

(a) (b)

(c) (d)

Stripping solution: [HA] =1 mol L-1

Fig. 10. Micrographs of the membrane after 120 minutes of contact. Feed solution: [Ag(I)] = 100 mg L-1, [HNO3]: a) without, pH 5.4, b) 0.1 mol L-1, c) 0.5 mol L-1 and d) 1 mol L-1.

Nanoparticles and Submicroparticles of Ag on Microfiltration Membranes 449

solutions as a function of time. Figure 10 shows the micrographs obtained in each condition. Figure 10a shows a dendritic shape of the silver particles when no HNO3 is added to the solution. When HNO3 is added to the feed solution, the particles have a decahedral structure (figures 9b, 9c, 9d); like those obtained earlier. Another important observation is that when the nitric acid's concentration increases, the number and size of silver particles on the membrane decreases. The shape in all cases does not change very much. In the absence of HNO3 acid, the feed solution becomes cloudy after 5 minutes, suggesting that the Ag (I) reduction process takes place not only in the feed solution membrane interface but also in the bulk of the feed solution. By increasing the H+'s concentration in the feed solution, the solution does not become cloudy and the silver is reduced on the membrane. The size distribution and dispersion of these particles is higher at 0.5 mol L-1 (Figure 9c) than at 0.1 mol L-1 (Figure 9b) of nitric acid. When the concentration of H+ is increased to 1 mol L-1 (Figure 9d), there are few particles left on the membrane, indicating that a decrease in pH acts negatively on the Ag(I) reduction process, decreasing the reduction rate and generating

and 1 mol L-1 of ascorbic acid as the stripping solution. Figure 9 shows the variation of Ct/Co with the Ag (I) in the feed solution as a function of time (being Co the initial concentration of Ag (I) and Ct the concentration at time t).

Fig. 9. Silver's Ct/Co variation in the feed solution as a function of time. () [Ag (I)] = 25 mg L-1;(z)[Ag (I)] = 100 mg L-1. Stripping solution [HA] = 1 mol L-1

Silver Ct/Co values decrease as function of time, in the feed solution in both tests. Silver recovery efficiency is near 95%. For initial silver's concentration of 25 and 100 mg L-1, the recovery efficiency is 99%. Silver's concentrations decrease faster when the initial concentration is 100 mg L-1. The concentration of Ag (I) in the stripping solution was practically negligible after 120 minutes of contact (no more than 0.2 mg L-1). Therefore, we can consider that the transfer of Ag (I) from the feed solution to the stripping solution is negligible. In both cases the membrane has a silver deposit on the surface in contact with the feed solution, so that the absence of silver ions in the feed solution is due to its reduction induced by the ascorbic acid.

When a 25 mg L-1 concentration of silver is used in the feed solution, the quantity of silver particles on the surface of the membrane, are scarce and show a less uniform distribution (data not shown). Nevertheless, the silver particles morphology is quiet similar to the observed previously, namely in decahedra shapes.

#### **3.3 Effect of the H+ ions' and the ascorbic acid's concentrations**

The effect of H+'s concentration in the reduction of silver by ascorbic acid and its deposition on the microfiltration membrane, was performed by varying the HNO3 concentration between 0 (pH 5) to 1 mol L-1 into the feed phase, while the ascorbic acid concentration was 1 mol L-1 into the stripping phase. The stirring speed was kept constant, in both compartments, at 600 rpm. In all cases the experimental time was 120 minutes. The recovery efficiency was evaluated by analyzing the amount of Ag (I) in the feed and in stripping

and 1 mol L-1 of ascorbic acid as the stripping solution. Figure 9 shows the variation of Ct/Co with the Ag (I) in the feed solution as a function of time (being Co the initial concentration of

Fig. 9. Silver's Ct/Co variation in the feed solution as a function of time. () [Ag (I)] = 25 mg

Silver Ct/Co values decrease as function of time, in the feed solution in both tests. Silver recovery efficiency is near 95%. For initial silver's concentration of 25 and 100 mg L-1, the recovery efficiency is 99%. Silver's concentrations decrease faster when the initial concentration is 100 mg L-1. The concentration of Ag (I) in the stripping solution was practically negligible after 120 minutes of contact (no more than 0.2 mg L-1). Therefore, we can consider that the transfer of Ag (I) from the feed solution to the stripping solution is negligible. In both cases the membrane has a silver deposit on the surface in contact with the feed solution, so that the absence of silver ions in the feed solution is due to its reduction

When a 25 mg L-1 concentration of silver is used in the feed solution, the quantity of silver particles on the surface of the membrane, are scarce and show a less uniform distribution (data not shown). Nevertheless, the silver particles morphology is quiet similar to the

 **ions' and the ascorbic acid's concentrations**  The effect of H+'s concentration in the reduction of silver by ascorbic acid and its deposition on the microfiltration membrane, was performed by varying the HNO3 concentration between 0 (pH 5) to 1 mol L-1 into the feed phase, while the ascorbic acid concentration was 1 mol L-1 into the stripping phase. The stirring speed was kept constant, in both compartments, at 600 rpm. In all cases the experimental time was 120 minutes. The recovery efficiency was evaluated by analyzing the amount of Ag (I) in the feed and in stripping

L-1;(z)[Ag (I)] = 100 mg L-1. Stripping solution [HA] = 1 mol L-1

induced by the ascorbic acid.

**3.3 Effect of the H+**

observed previously, namely in decahedra shapes.

Ag (I) and Ct the concentration at time t).

solutions as a function of time. Figure 10 shows the micrographs obtained in each condition. Figure 10a shows a dendritic shape of the silver particles when no HNO3 is added to the solution. When HNO3 is added to the feed solution, the particles have a decahedral structure (figures 9b, 9c, 9d); like those obtained earlier. Another important observation is that when the nitric acid's concentration increases, the number and size of silver particles on the membrane decreases. The shape in all cases does not change very much. In the absence of HNO3 acid, the feed solution becomes cloudy after 5 minutes, suggesting that the Ag (I) reduction process takes place not only in the feed solution membrane interface but also in the bulk of the feed solution. By increasing the H+'s concentration in the feed solution, the solution does not become cloudy and the silver is reduced on the membrane. The size distribution and dispersion of these particles is higher at 0.5 mol L-1 (Figure 9c) than at 0.1 mol L-1 (Figure 9b) of nitric acid. When the concentration of H+ is increased to 1 mol L-1 (Figure 9d), there are few particles left on the membrane, indicating that a decrease in pH acts negatively on the Ag(I) reduction process, decreasing the reduction rate and generating fewer and smaller particles on the membrane.

Fig. 10. Micrographs of the membrane after 120 minutes of contact. Feed solution: [Ag(I)] = 100 mg L-1, [HNO3]: a) without, pH 5.4, b) 0.1 mol L-1, c) 0.5 mol L-1 and d) 1 mol L-1. Stripping solution: [HA] =1 mol L-1

Silver Recovery from Acidic Solutions by Formation of

stripping solution).

of the stripping phase.

Nanoparticles and Submicroparticles of Ag on Microfiltration Membranes 451

mass transfer process of Ag+ ions from the feed solution towards the specific zone where the redox process takes place with the ascorbic acid (which is also transported from the

The analysis of the concentration profiles of [Ag+] at different concentrations of HNO3 in the feed phase and with a two different ascorbic acid's concentrations in the stripping phase, allows to see that the rate of decrease of [Ag+] in the feed solution as a function of time (figures 12a and 12b), strongly depends on the conditions of the feed phase as much as those

Fig. 12. Variation of Ag (I) concentration as a function of time in the feed compartment

() [HNO3] = 0.1 mol L-1; () [HNO3] = 0.5 mol L-1; (z) [HNO3] = 1 mol L-1. Stripping

for different nitric acid's concentrations. (S) In the absence of HNO3, pH = 5.4;

solution: (A) [HA] = 1 mol L-1, (B) [HA] = 0.5 mol L-1

Figure 11 shows the morphology of the silver nanoparticles at different HNO3's concentration in the feed solution when the ascorbic acid concentration is reduced from 1 mol L-1 to 0.5 mol L-1 in the stripping solution. The silver deposits obtained, show a similar morphology that in Figure 10. The only difference is the amount of reduced silver; in this case the amount is less. In the absence of HNO3, the morphology of the particles is dendritic (Figure 11a), but in the presence of HNO3 acid a decahedra structure was obtained (Figure 11b, 11c and 11d). A more uniform particle size and better distribution occurs at low concentrations of acid in the feed compartment.

From this study it was found that a greater amount of ascorbic acid increases the reduction of Ag (I) on the membrane. The increase of H+´s concentration results in a reduced amount of silver but on a more uniform shape and size.

(a) (b)

Fig. 11. Micrographs of the membranes after 120 minutes of contact. Feed solution: [Ag+] = 100 mg L-1; [HNO3]: a) without, pH 5.4, b) 0.1 mol L-1, c) 0.5 mol L-1 and d) 1 mol L-1. Stripping solution: [HA] = 0.5 mol L-1

#### **3.4 Analysis of the mass transfer process of Ag (I)**

The morphology of Ag particles obtained is directly related to the hydrodynamic and chemical conditions of the system proposed. But also, the morphology is connected with the

Figure 11 shows the morphology of the silver nanoparticles at different HNO3's concentration in the feed solution when the ascorbic acid concentration is reduced from 1 mol L-1 to 0.5 mol L-1 in the stripping solution. The silver deposits obtained, show a similar morphology that in Figure 10. The only difference is the amount of reduced silver; in this case the amount is less. In the absence of HNO3, the morphology of the particles is dendritic (Figure 11a), but in the presence of HNO3 acid a decahedra structure was obtained (Figure 11b, 11c and 11d). A more uniform particle size and better distribution occurs at low concentrations

From this study it was found that a greater amount of ascorbic acid increases the reduction of Ag (I) on the membrane. The increase of H+´s concentration results in a reduced amount

of acid in the feed compartment.

of silver but on a more uniform shape and size.

(a) (b)

(c) (d)

Stripping solution: [HA] = 0.5 mol L-1

**3.4 Analysis of the mass transfer process of Ag (I)** 

Fig. 11. Micrographs of the membranes after 120 minutes of contact. Feed solution: [Ag+] = 100 mg L-1; [HNO3]: a) without, pH 5.4, b) 0.1 mol L-1, c) 0.5 mol L-1 and d) 1 mol L-1.

The morphology of Ag particles obtained is directly related to the hydrodynamic and chemical conditions of the system proposed. But also, the morphology is connected with the mass transfer process of Ag+ ions from the feed solution towards the specific zone where the redox process takes place with the ascorbic acid (which is also transported from the stripping solution).

The analysis of the concentration profiles of [Ag+] at different concentrations of HNO3 in the feed phase and with a two different ascorbic acid's concentrations in the stripping phase, allows to see that the rate of decrease of [Ag+] in the feed solution as a function of time (figures 12a and 12b), strongly depends on the conditions of the feed phase as much as those of the stripping phase.

Fig. 12. Variation of Ag (I) concentration as a function of time in the feed compartment for different nitric acid's concentrations. (S) In the absence of HNO3, pH = 5.4; () [HNO3] = 0.1 mol L-1; () [HNO3] = 0.5 mol L-1; (z) [HNO3] = 1 mol L-1. Stripping solution: (A) [HA] = 1 mol L-1, (B) [HA] = 0.5 mol L-1

Silver Recovery from Acidic Solutions by Formation of

concentrations in both phases.

Nanoparticles and Submicroparticles of Ag on Microfiltration Membranes 453

were performed by replacing HA by different concentrations of HNO3 at the stripping phase (pH from 0 to 5), and a feed solution of Ag (I) 100 mg L-1 and HNO3 0.5 mol L-1. The Ag+ concentration in the stripping phase after all the experiments, shows values around 10%, indicating that Ag+ diffusion through the stripping phase is very low in all tests performed. A pH decrease in the stripping phase is also detected and this effect is more pronounced when the initial pH in the stripping phase is higher. Figure 14 shows Ag+

Fig. 14. Variation of Ag (I) concentration as a function of time in absence of HA. () Feed solution: [Ag (I)] = 100 mg L-1; [HNO3] = 0.5 mol L-1. () Stripping solution: [HNO3] = 1 mol L-1

Fig. 15. Relative Ag+ (I)'s concentration as a function of [H+]strip/[H+]feed

With a higher pH (without a HNO3 addition) at the feed phase, we reach the highest Ag reduction rate, as shown in figure 12A. Due to the low Ag+ concentration found at the stripping phase after the test, the Ag+ mass transfer through the microfiltration membrane cannot be considered during the reduction process, indicating that the redox process takes place inside of the microfiltration membrane. Final Ag+ concentrations in the feed solutions are close to 2% of the initial value, indicating a high yield process.

When the pH in the feed phase is 5, the [Ag+] in the feed solution diminishes to less than 1%, and a precipitate becomes apparent in the feed solution. This one can be associated to the saturation of Ag precipitate in the membrane and due to the stirring process the Ag particles come to the feed solution. However, when pH is reduced in the feed solution, the Ag precipitated in the feed solution does not appear anymore and all the Ag particles are retained in the membrane. When the HNO3 concentrations are higher than 0.5 mol L-1, the reduction rates are slow and the curves trends of [Ag+] remain constant during the first 30 minutes of the experiment.

To explain the results obtained it is necessary to analyse the various phenomena that take place at the membrane. There are two important aspects: 1) Ag+ and HA mass transfer process, and 2) the redox process of both compounds connected with the formation of silver particles.

In analysing the mass transfer process of Ag+ and HA, it is necessary to consider the different zones existing in the system. Close to the membrane, a non-stirring zone exists. In non-stirring areas Ag+ and HA movements are controlled by a diffusion process because the convection process is negligible. Figure 13 shows the several areas formed due to the stirring process. Zones named *a)* and *e)* correspond to non-stirring areas (diffusion region) in feed and stripping phases. Zones named *b)* and *d)* represent the interphases feed-membrane phase and stripping-membrane phase respectively, and *c)* is the membrane phase. The thickness of the diffusion regions are represented by *da* and *de*, corresponding to the thickness of feed and stripping diffusion areas respectively and *do* is the membrane thickness.

Fig. 13. Scheme of the mass transfer of Ag (I) and ascorbic acid (HA)

In order to correctly define the mass transfer process, it is necessary to determine the areas where the redox process could take place. The experimental results show that the redox process takes place mainly at the feed-membrane interface and Ag particles are formed on the feed side of the membrane. Based on this result, the Ag+ and HA mass transfers were studied in the absence of one of the two compounds, alternately Ag+ mass transfer studies

With a higher pH (without a HNO3 addition) at the feed phase, we reach the highest Ag reduction rate, as shown in figure 12A. Due to the low Ag+ concentration found at the stripping phase after the test, the Ag+ mass transfer through the microfiltration membrane cannot be considered during the reduction process, indicating that the redox process takes place inside of the microfiltration membrane. Final Ag+ concentrations in the feed solutions

When the pH in the feed phase is 5, the [Ag+] in the feed solution diminishes to less than 1%, and a precipitate becomes apparent in the feed solution. This one can be associated to the saturation of Ag precipitate in the membrane and due to the stirring process the Ag particles come to the feed solution. However, when pH is reduced in the feed solution, the Ag precipitated in the feed solution does not appear anymore and all the Ag particles are retained in the membrane. When the HNO3 concentrations are higher than 0.5 mol L-1, the reduction rates are slow and the curves trends of [Ag+] remain constant during the first 30

To explain the results obtained it is necessary to analyse the various phenomena that take place at the membrane. There are two important aspects: 1) Ag+ and HA mass transfer process, and 2) the redox process of both compounds connected with the formation of silver particles. In analysing the mass transfer process of Ag+ and HA, it is necessary to consider the different zones existing in the system. Close to the membrane, a non-stirring zone exists. In non-stirring areas Ag+ and HA movements are controlled by a diffusion process because the convection process is negligible. Figure 13 shows the several areas formed due to the stirring process. Zones named *a)* and *e)* correspond to non-stirring areas (diffusion region) in feed and stripping phases. Zones named *b)* and *d)* represent the interphases feed-membrane phase and stripping-membrane phase respectively, and *c)* is the membrane phase. The thickness of the diffusion regions are represented by *da* and *de*, corresponding to the thickness of feed and stripping diffusion areas respectively and *do* is the membrane

are close to 2% of the initial value, indicating a high yield process.

Fig. 13. Scheme of the mass transfer of Ag (I) and ascorbic acid (HA)

In order to correctly define the mass transfer process, it is necessary to determine the areas where the redox process could take place. The experimental results show that the redox process takes place mainly at the feed-membrane interface and Ag particles are formed on the feed side of the membrane. Based on this result, the Ag+ and HA mass transfers were studied in the absence of one of the two compounds, alternately Ag+ mass transfer studies

minutes of the experiment.

thickness.

were performed by replacing HA by different concentrations of HNO3 at the stripping phase (pH from 0 to 5), and a feed solution of Ag (I) 100 mg L-1 and HNO3 0.5 mol L-1. The Ag+ concentration in the stripping phase after all the experiments, shows values around 10%, indicating that Ag+ diffusion through the stripping phase is very low in all tests performed. A pH decrease in the stripping phase is also detected and this effect is more pronounced when the initial pH in the stripping phase is higher. Figure 14 shows Ag+ concentrations in both phases.

Fig. 14. Variation of Ag (I) concentration as a function of time in absence of HA. () Feed solution: [Ag (I)] = 100 mg L-1; [HNO3] = 0.5 mol L-1. () Stripping solution: [HNO3] = 1 mol L-1

Fig. 15. Relative Ag+ (I)'s concentration as a function of [H+]strip/[H+]feed

Silver Recovery from Acidic Solutions by Formation of

PVDF Hydrophobic /H2O

hydrophobic membrane.

to the AgNO3/HNO3 solutions.

oxireduction reaction is carried out.

Time (s)

Contact

time

Nanoparticles and Submicroparticles of Ag on Microfiltration Membranes 455

for a PVDF hydrophilic membrane and as means of comparison we included a PVDF

angle (°) 143.1 72.9 115.2 114.58 58.47

It can be seen that the values of water contact angles on a hydrophobic PVDF membrane are higher than those obtained with a hydrophilic PVDF membrane. When the polarities are very close, the contact angles are smaller. Also the analysis of Table 1 shows that the contact angle values of the solutions of Ag+ are higher than those obtained with solutions of HA and water, and are similar to the contact angle values of a hydrophobic membrane with water (large difference in polarity). This clearly indicates that there is a rejection of the membrane

Analyzing the variation of the contact angle as a function of time (see Table 2), we found that in the case of solutions AgNO3/HNO3 the value of the contact angle after 6 seconds, remained around 114º. In the case of ascorbic acid, the value of the contact angle ranged from 50.48º to 58.47º, which is a considerable variation. The decrease in the contact angle value indicates that the membrane gets impregnated by the ascorbic acid. Thus, the HA is transported into the membrane to reach the interfacial membrane area, where the

> AgNO3 100 mg L-1 HNO3 0.5 mol L-1 Contact angle (°) (Average)

1.00 114.58 58.47 2.00 113.29 55.17 3.00 114.09 53.66 4.00 114.29 52.41 5.00 114.64 51.51 6.00 114.11 50.48

Table 2. Contact angle values of a microfiltration membrane in contact with water and solutions containing AgNO3 (100 mg L-1)/HNO3 and ascorbic acid 1 mol L-1 as a function of

According to the above, the formation process speed of silver nanoparticles depends on the diffusion speed of Ag+ ions to the membrane. If this speed is lower than the oxidation-

Table 1. Contact angle values of the microfiltration membrane in contact with water and

PVDF Hydrophilic/ AgNO3 /HNO3 0.1 mol L-1

PVDF Hydrophilic/ AgNO3 /HNO3 0.5 mol L-1

HA 1 mol L-1 Contact angle (°) (Average)

PVDF Hydrophilic/ HA

PVDF Hydrophilic/ H2O

solutions containing AgNO3 (100 mg L-1)/HNO3 and ascorbic acid 1 mol L-1

Analyzing the value of H+'s concentration in the feed and in the stripping solutions, we found that there is a proton transfer from the feed solution to the stripping solution due to a concentration gradient. The silver ions diffused to the stripping compartment are related with the H+ transfer. The diffusion of Ag+ increases with the increasing of the H+'s concentration in the feed solution. This effect can be observed in Figure 15.

The transference of ascorbic acid trough the membrane has been studied using a feed solution containing HNO3 at different concentrations (0.1, 0.25 y 0.5 mol L-1) and a HA's concentration of 1 mol L-1 in the stripping solution. The results show that the concentration of ascorbic acid in the feed phase after 120 minutes is minimal for each of the conditions studied. Figure 16 shows the results of the transfer of HA through the microfiltration membrane when HNO3 in the feed solution was 0.5 mol L-1.

Fig. 16. Variation of the ascorbic acid's concentrations in the stripping () and feed () phases as a function of time

In summary both Ag+ and HA have a minimum transfer through the microfiltration membrane. This behavior can be explained considering that the membrane is made of a polyvinylidene fluoride polymer whose surface has been modified to increase the hydrophilicity. This modification produces electrical charges on the membrane surface. These charges could be positive or negative depending on the nature of aqueous solutions that are in contact with the membrane. Thus, there may be a rejection of the membrane to the charged species present in the interface membrane/feed or membrane/stripping solution. Although no transfer occurs of any of both species through the membrane, we have shown that the oxireduction reaction takes place in the interface (feed solution/membrane) generating silver nanoparticles on the membrane.

In order to determine the degree of rejection of the microfiltration membrane to Ag+ and ascorbic acid, contact angle measurements were performed. Table 1 shows the contact angle

Analyzing the value of H+'s concentration in the feed and in the stripping solutions, we found that there is a proton transfer from the feed solution to the stripping solution due to a concentration gradient. The silver ions diffused to the stripping compartment are related with the H+ transfer. The diffusion of Ag+ increases with the increasing of the H+'s

The transference of ascorbic acid trough the membrane has been studied using a feed solution containing HNO3 at different concentrations (0.1, 0.25 y 0.5 mol L-1) and a HA's concentration of 1 mol L-1 in the stripping solution. The results show that the concentration of ascorbic acid in the feed phase after 120 minutes is minimal for each of the conditions studied. Figure 16 shows the results of the transfer of HA through the microfiltration

Fig. 16. Variation of the ascorbic acid's concentrations in the stripping () and feed ()

solution/membrane) generating silver nanoparticles on the membrane.

In summary both Ag+ and HA have a minimum transfer through the microfiltration membrane. This behavior can be explained considering that the membrane is made of a polyvinylidene fluoride polymer whose surface has been modified to increase the hydrophilicity. This modification produces electrical charges on the membrane surface. These charges could be positive or negative depending on the nature of aqueous solutions that are in contact with the membrane. Thus, there may be a rejection of the membrane to the charged species present in the interface membrane/feed or membrane/stripping solution. Although no transfer occurs of any of both species through the membrane, we have shown that the oxireduction reaction takes place in the interface (feed

In order to determine the degree of rejection of the microfiltration membrane to Ag+ and ascorbic acid, contact angle measurements were performed. Table 1 shows the contact angle

concentration in the feed solution. This effect can be observed in Figure 15.

membrane when HNO3 in the feed solution was 0.5 mol L-1.

phases as a function of time

for a PVDF hydrophilic membrane and as means of comparison we included a PVDF hydrophobic membrane.


Table 1. Contact angle values of the microfiltration membrane in contact with water and solutions containing AgNO3 (100 mg L-1)/HNO3 and ascorbic acid 1 mol L-1

It can be seen that the values of water contact angles on a hydrophobic PVDF membrane are higher than those obtained with a hydrophilic PVDF membrane. When the polarities are very close, the contact angles are smaller. Also the analysis of Table 1 shows that the contact angle values of the solutions of Ag+ are higher than those obtained with solutions of HA and water, and are similar to the contact angle values of a hydrophobic membrane with water (large difference in polarity). This clearly indicates that there is a rejection of the membrane to the AgNO3/HNO3 solutions.

Analyzing the variation of the contact angle as a function of time (see Table 2), we found that in the case of solutions AgNO3/HNO3 the value of the contact angle after 6 seconds, remained around 114º. In the case of ascorbic acid, the value of the contact angle ranged from 50.48º to 58.47º, which is a considerable variation. The decrease in the contact angle value indicates that the membrane gets impregnated by the ascorbic acid. Thus, the HA is transported into the membrane to reach the interfacial membrane area, where the oxireduction reaction is carried out.


Table 2. Contact angle values of a microfiltration membrane in contact with water and solutions containing AgNO3 (100 mg L-1)/HNO3 and ascorbic acid 1 mol L-1 as a function of time

According to the above, the formation process speed of silver nanoparticles depends on the diffusion speed of Ag+ ions to the membrane. If this speed is lower than the oxidation-

Silver Recovery from Acidic Solutions by Formation of

formation of hexagonal plates (Jixiang et al. 2007).

**4. Conclusions** 

microorganisms contained in water.

drastic being an advantage over other methods reported.

Nanoparticles and Submicroparticles of Ag on Microfiltration Membranes 457

two well-defined zones, one with a value for K of 0.0033 cm s-1 and another with an overall mass transfer coefficient of 0.0135 cm s-1. It is clear that an increase of the HNO3's concentration, has a negative effect on the speed of the oxireduction reaction between Ag+ ions and HA. Moreover, to understand better the results shown in Figure 17, it is necessary to consider the rate of impregnation of the membrane by HA. In the absence of HNO3, the rate of impregnation of the membrane by the HA, appears to be faster than in the acidic media. This is why the K value diminishes with the increase of HNO3's concentration and at 1 mol L-1 of HNO3, the variation of [Ag+] remains almost constant during the first 30 minutes

It is possible to correlate the values of the global mass transfer coefficient with the morphology of silver particles deposited in the microfiltration membrane under each of the studied conditions. When the ascorbic acid rapidly permeates the microfiltration membrane (low concentration of HNO3 in the feed solution) the value of K is high (0.1779 cm s-1). The silver particles obtained in this case, have a dendritic shape (figure 10a and 11a). If the value of K decreases, the silver crystals grow as decahedra (Figures 10b and 11b). Finally, the no agitation of the feed and stripping phases make transference process very slow, and under these conditions the crystallization time is sufficient for the formation of metallic silver hexagonal plates (figure 8). These observations agree with those reported in the literature regarding the process of crystallization of metallic silver, which in a first stage involves the formation of dendrites trees that slowly form decahedra shaped particles leading to the

The methodology proposed is suitable for obtaining silver nanoparticles and submicroparticles on microfiltration membranes with different shapes and sizes. The control of mass transference can be carried out by changes in the stirring solutions, the pH, and the concentrations of Ag+ and ascorbic acid. The conditions used in this methodology are not

We have developed a methodology for the recovery of silver (I) from aqueous solutions on a microfiltration membrane using ascorbic acid as a reducing agent. Under certain conditions, it is possible to recover about 99% of the silver contained in the aqueous solutions. The silver particles are deposited in nanometric and submicron sizes. The shape of these particles depends on the hydrodynamic and chemical conditions of the system. Silver particles can be obtained as dendrites, decahedra and hexagonal plates. We have analyzed the mass transfer process of the species involved in the system in order to explain the observed phenomena and to correlate the morphology of the particles obtained, with the mass transfer process. We can conclude that the reaction between silver and ascorbic acid occurs at the interface membrane-feed solution. The permeation rate of ascorbic acid into the membrane is linked to the Ag+ mass transfer process. Finally, the global coefficient of mass transfer is related to the morphology of the particles obtained. At high K values, silver dendrites nanoparticles are obtained; whereas if the value of K decreases the deposit of silver particles corresponds to a slow crystallization process. The methodology proposed allows the efficient recovery of Ag (I) ions and allows the obtaining of microfiltration membranes modified by Ag particles, which can be used as filters for the removal of

of contact between the phases and the membrane (Figure 12a and 12b).

reduction speed, the diffusion process takes place and it is possible to calculate the overall mass transfer coefficient of the Ag+ using the equation (1).

$$\begin{aligned} \, \_L \text{Ln} \frac{[Ag^+]\_t}{[Ag^+]\_0} = -K \frac{Q}{V} t \end{aligned} \tag{1}$$

Where:

K = overall mass transfer coefficient

t = time (sec)

Q = effective area of the membrane (11.34 cm2)

V = volume (250 cm3)

The overall mass transfer coefficient depends on the diffusion rate as well as on the chemical reaction between silver ions and ascorbic acid.

The variation of Ln [Ag+]t/[Ag+]0 as a function of t\*Q/V, is a straight line with a slope equal to –K. Figure 17 shows the results obtained in the case of a system containing a feed solution of Ag(I) 100 mg L-1 with different HNO3 concentrations and 1 mol L-1 of HA in the stripping solution.

Fig. 17. Variation of de Ln[Ag+]t/[Ag+]0 as function of t\*Q/V in the feed compartment for different nitric acid's concentrations. (S) Without HNO3, pH = 5.4; () [HNO3] = 0.1 mol L-1; () [HNO3] = 0.5 mol L-1; (z) [HNO3] = 1 mol L-1. Stripping solution: [HA] = 1 mol L-1

The value of K (slope) obtained for the system in absence of HNO3 is 0.1779 cm s-1,with a R2 value of 0.94. For a HNO3 concentration of 0.1 mol L-1, in the feed solution, the value of K (0.1175 cm s-1) is very close to the former. In these two cases the K values are high for a system with a diffusion control. It is important to note that under these conditions, the formation of silver nanoparticles not only occurs in the feed phase-membrane interface but also in the bulk of the feed solution, then the process is controlled by the redox reaction. On the other hand, when the concentration of HNO3 is 0.5 mol L-1, the value of the overall mass transfer coefficient is 0.0483 cm s-1. When the HNO3's concentration is 1 mol L-1, there are two well-defined zones, one with a value for K of 0.0033 cm s-1 and another with an overall mass transfer coefficient of 0.0135 cm s-1. It is clear that an increase of the HNO3's concentration, has a negative effect on the speed of the oxireduction reaction between Ag+ ions and HA. Moreover, to understand better the results shown in Figure 17, it is necessary to consider the rate of impregnation of the membrane by HA. In the absence of HNO3, the rate of impregnation of the membrane by the HA, appears to be faster than in the acidic media. This is why the K value diminishes with the increase of HNO3's concentration and at 1 mol L-1 of HNO3, the variation of [Ag+] remains almost constant during the first 30 minutes of contact between the phases and the membrane (Figure 12a and 12b).

It is possible to correlate the values of the global mass transfer coefficient with the morphology of silver particles deposited in the microfiltration membrane under each of the studied conditions. When the ascorbic acid rapidly permeates the microfiltration membrane (low concentration of HNO3 in the feed solution) the value of K is high (0.1779 cm s-1). The silver particles obtained in this case, have a dendritic shape (figure 10a and 11a). If the value of K decreases, the silver crystals grow as decahedra (Figures 10b and 11b). Finally, the no agitation of the feed and stripping phases make transference process very slow, and under these conditions the crystallization time is sufficient for the formation of metallic silver hexagonal plates (figure 8). These observations agree with those reported in the literature regarding the process of crystallization of metallic silver, which in a first stage involves the formation of dendrites trees that slowly form decahedra shaped particles leading to the formation of hexagonal plates (Jixiang et al. 2007).

The methodology proposed is suitable for obtaining silver nanoparticles and submicroparticles on microfiltration membranes with different shapes and sizes. The control of mass transference can be carried out by changes in the stirring solutions, the pH, and the concentrations of Ag+ and ascorbic acid. The conditions used in this methodology are not drastic being an advantage over other methods reported.
