Solvent extraction of In³⁺ with microreactor from leachant containing Fe²⁺ and Zn²⁺

2.1 Materials and devices

The experimental material, an aqueous solution in a sulfuric system with a pH value of 0.44, was obtained from a hydrometallurgy zinc plant in Yunnan Province, China. Its chemical composition was analyzed and is listed in Table 1. The organic phase for solvent extraction contained 30% of D2EHPA (mass fraction, w=0.99, DeZhong Chemical factory, ZhengZhou, HeNan province, China) and 70% of 260# kerosene (mass fraction, w=0.98, DeZhong Chemical factory, ZhengZhou, HeNan province, China). The main extraction equation is as follows [21]:

The microchip used in this study is shown in Figures 1 and 2. It was made of Pyrex glass (Institute of Microchemical Technology, Japan), produced with combined methods of photolithography, wet-etching and thermal bonding. The extraction reaction takes place in a 120 mm long microchannel (160 μm×40 μm), which is shown in Figure 1B, with a guide structure for obtaining a stable flow interface of the two immiscible liquid-liquid phases.

Figure 1

Schematic of microreactor system used in this study: (A) microreactor system; and (B) cross-section of the extraction channel.

Figure 2

Microreactor channel conditions monitored by optical microscopy.

2.2 Procedures

During the experiments of microreactor extraction, the aqueous and oil phases were, respectively, pumped into the Pyrex glass microchip by two constant flow pumps (HLB-4015, Yansan). The flow stability of the contacting interface of the aqueous and oil phases in the microchip was monitored with an optical microscope (Leica DM 4000 M). The flow rate of the aqueous phase ranged from 1 to 6 ml·h^-1 at a fixed organic/aqueous flow rate ratio of 0.8. Along the contacting channel, there is an arched guide lug in its middle for producing a stable aqueous-oil interface. The extraction channel terminates at a second Y-junction, where the two phases are separated and flow out of the microchip for samples collection and analysis.

The experiments of conventional extraction were carried out in a 250 ml separating funnel. The phase ratio was set as O:A=2:1, 1:1, 1:2 and 1:4, respectively. Each time, 40 ml solution was added to the organic phase (30% D2EHPA in 260^# kerosene) in a 250 ml separating funnel. Then, the funnel was stirred on a shaking machine for 5 min at a speed of 200 rpm. After separation, the aqueous phase was diluted and analyzed by inductively coupled plasma atomic emission spectrometry (ICP-AES).

2.3 Analysis method

The concentrations of In³⁺, Fe²⁺ and Zn²⁺ in the aqueous phase before and after the extraction were determined by an ICP-AES (Leeman ICP-AES PS1000). The acidity of the aqueous solution was determined by an acidity meter. Equation (2) is an expression of extraction rate, where C_ao and C_al signify the elements contained before and after extraction, respectively:

3.1 Stability of liquid-liquid interface

Figure 3 contains two photos of aqueous and oil flows taken by an optical microscope. A stable liquid-liquid interface was observed during this experiment under the condition of optimal flow rate ratio (oil/aqueous), R=0.80, and flow was laminar in all cases. Compared with the emulsions which easily appeared in conventional extractions, this type of microreactor extraction radically abandoned the formation of emulsions of oil-in-water [22].

Figure 3

Photos of the merging and diverging of aqueous and oil flows.

3.2 Calculation of contact time in the microchannel

The changes of geometry structure of a microreactor have a great influence on the flow pattern and the shape of the interface between L-L phases [23]. According to the microchannel geometry in Figure 4, the cross-section area of the aqueous channel, S_a, was estimated by Eq. (3). The volume of aqueous channel, V_a, was calculated by Eq. (4):

Figure 4

The actual size of cross-section/μm.

In Eq. (3), L is the total length of the channel, which is equal to 12 cm. It is calculated that V_a=3.08×10^-10m³. When the flow rate of the aqueous phase was 2 ml·h^-1, which means volumetric flow rate of the aqueous phase v=(5.56×10^-10m³/s), the phase ratio R=0.8. The residence time, t, of the aqueous phase in the microchannel can be expressed as Eq. (5):

The value of t is calculated as 0.55 s, which indicated that the extractability could be measured for a short contact time of both phases in the microreactor. In traditional solvent extraction plants with a mixing settler system, the extraction time is usually about 4 min (1 min of mixing and 3 min of phase separating).

3.3 Microreactor extraction of In³⁺ in a complex solution

A complex solution with a composition in Table 1 was used in the experiment of microreactor extraction. When the flow rate of the aqueous phase was set as 2 ml·h^-1 and that of the organic phase was set as 1.6 ml·h^-1, the raffinate was collected and analyzed by ICP-AES. The results are shown in Table 2.

It is shown in Table 2 that the extraction percentage of In³⁺ is much higher than that of Fe²⁺ and Zn²⁺.

In an industry plant, after 2-step extraction of In³⁺ ion by D2EHPA, Fe²⁺ and Zn²⁺ concentration in the organic phase is much higher. Thus a 3-step eluting process was needed to remove the above impurities.

The microreactor provides a good separation effect of In³⁺ and impurities. The main reason lies in the fact that the mass transfer speed of In³⁺ is much higher than Fe²⁺ and Zn²⁺ [24].

During the microchannel extraction, the flow pattern is much different with the mixing settler system. The mass transfer of the former is by diffusion through a stable interface, which may lead to better selection of metal ions.

For checking the co-extraction effects of Fe²⁺ and Zn²⁺ during the traditional mixing settler system, the extraction experiment was carried out in separating funnels. The relationship between the extraction rate and phase ratio of this experiment is shown in Figure 5.

Figure 5

Extraction in different phase ratios during separating funnel experiments.

From Figure 5, we can see that the co-extraction ratio of Fe²⁺ and Zn²⁺ is about one magnitude higher than that of microreactor extraction.

3.4 Surface-to-volume ratio of the microchannel and its mass transfer character

The aqueous-organic interfacial area A (m²) can be calculated with follow equation:

where d is the height of the microchannel and h_g is height of the guide structure. Thus, A is calculated to be 4.2×10^-6m². In a contact time of 0.55 s, 3.08×10^-10m³ of solution involved in the reaction and 0.98×10^-6g of In³⁺ transferred into the organic phase calculated by the data shown in Table 2. The mass transfer speed can be as high as 0.34 g·m^-2·s^-1.

The surface-to-volume ratio was about 6.8×10³ m²/m³ for this channel, which is much higher than that of the traditional mixing settler process. The surface-to-volume ratio of conventional scale laboratory or industrial reactors are usually in the range of 100~1000 m²/m³ [25].