**1. Introduction**

The demand for base metals such as zinc (Zn) and copper (Cu) has increased in recent years. Natural (primary) resources of both metals are dominated by sulfide ores, which unfortunately have declined both in quantity and grade causing discrepancies between supply and demand [1]. This in turn encourages the exploration of new resources, which not only includes primary resources but also secondary ones through recycling. Aside from helping address the resource depletion problem, the exploitation of secondary resources also mitigates environmental and resource sustainability problems. One of the secondary resources of base metal, which holds potential for further processing, is electric arc furnace (EAF) dust. The dust is a waste of steel making and is classified as hazardous [2]. The amount of dust produced during steel making is significant, where 11–20 kg of dust is generated for each ton of steel produced [3].

In addition to exploring new resources, extraction is also progressing toward green technology, which prioritizes the principles of efficiency, sustainability, safety and environmental impact. The EAF dust not only contains Zn and Cu as target elements but also others, e.g., lead (Pb), chromium (Cr), iron (Fe), aluminum (Al), calcium (Ca) and magnesium (Mg). As Pb and Cr (VI) represent pollutants, their release into the environment during the extraction of Zn and Cu must be minimized and as such, a comprehensive study is required to formulate applicable technology to maximize benefits from EAF dust and minimize the impact of pollutants (heavy metals) to the environment.

One of the most widely applied methods to recover Zn and Cu from EAF dust is hydrometallurgy, especially in the case where Zn in EAF dust exists as zinc oxide. Other common Zn phases, i.e., zinc ferrite requires pre-treatment (e.g. fusion) [4,5]. The hydrometallurgical approach for Zn and Cu recovery involves a leaching process using chemical lixiviants to dissolve these metals in aqueous solution. Dissolved Zn and Cu in the pregnant leach solution can then be further separated and purified using solvent extraction [6,7], ion exchange [8] or electrochemical methods [9,10]. Several leaching schemes that use various lixiviants have been previously proposed in order to recover Zn and/or Cu from EAF dust at ambient and elevated temperatures, respectively. Halli et al., 2017 [11] had screened potential lixiviants including nitric acid, citric acid, sodium hydroxide and ethanol for metal recovery from EAF dust. Proposed lixiviants by other researchers include sulfuric acid [12–14], hydrochloric acid [15], oxalic acid [16], citric acid [17], sodium hydroxide [18–20], sodium carbonate/sodium bicarbonate [21], ammonia [22], imminodiacetic acid (IDA) [23] and nitrilotriacetic acid (NTA) [24].

In general, the leaching schemes proposed in the literature can be classified into four categories:


The first and second schemes are advantageous in terms of recovery efficiency but both lack selectivity. Other elements, especially Fe and Ca which would cause fouling in further separation and purification steps, and Pb or Cr which are toxic, are also extracted [25]. The third scheme possesses an advantage in terms of selectivity but the extraction efficiency is low, requiring very low pulp density ratio, high lixiviant concentration and high leaching temperature to attain satisfying recovery. In addition, the caustic and corrosive nature of lixiviant requires special handling and equipment during operation and also poses a risk to the environment. The same is true of schemes using strong acid as lixiviant to certain extent, especially leaching using high concentration (concentrated) mineral acid, e.g., Halli et al., 2017 [11].

Considering these factors, the fourth scheme presented is preferable, due to its relatively selective recovery of Zn and Cu, higher leaching efficiency, safer handling in general compared to lixiviants in the other schemes and possibility to recover the lixiviant such as NTA. Although the leaching of base metal in this scheme is considered as selective, previous studies by Yang et al. (2016) [24] revealed relatively high extraction of iron, while the recovery of other elements such as Ca, Mg and Al during leaching was yet to be addressed, where these elements play an important role in further separation and purification steps. The lixiviants in the fourth category rely on their ability to form complexes with metal ions. Aminocarboxylic acid ligands, e.g., IDA, NTA or ethylenediaminetetraacetic acid (EDTA) possess relatively strong yet homogenous binding capacity to the metal ions, not only target elements such as Cu and Zn, but also matrix elements i.e. Fe, Al, Mg and Ca.

In order to partition the matrix elements into the solid phase and the valuable elements into the aqueous phase, it is suggested to use a ligand which complexes preferentially to Cu and Zn and shows weaker affinity to Ca, Mg, Fe and Al. Lixiviants such as amino acids broadly meet these criteria. Glycine has been extensively studied in copper [26] and gold [27] alkaline leaching, but to the best of our knowledge, no study on the efficacy of amino acid on Zn recovery has been carried out so far. Aside from glycine, another amino acid with potential to be developed as lixiviant is glutamic acid. In this research an amino acid sodium salt, i.e., monosodium glutamate (MSG) is proposed as a novel lixiviant. The advantages of MSG as a lixiviant include its wide availability and low price, and low risk to environment (biodegradable). MSG in alkaline conditions could serve as a powerful ligand to

bind several transition metals, e.g., Cu, which in turn increase the extraction efficiency. The complex formation of these metals with MSG have been reported and exploited in chemical analysis [28,29] and our preliminary research also confirmed its efficacy to extract Zn [30]. Another advantage of using MSG is its reusability, since MSG can be recovered from the pregnant leach solution as glutamic acid precipitate by acidification of pregnant leach solution (PLS). Considering these possible advantages of using MSG, a study to confirm its efficacy in Zn and Cu extraction from EAF dust is warranted. The effect of leaching parameters on the leaching efficiency such as pH, lixiviant concentration, pulp density and kinetics, including selectivity toward other elements are investigated herein.

#### **2. Species Distribution Modeling for Leaching E**ffi**ciency and Selectivity Prediction**

The leaching efficiency and selectivity are hypothetically controlled by the distribution of species in the aqueous phase. Factors controlling this distribution include pH and concentration of metal ions and ligands/lixiviants involved in the system. The effect of pH on the recovery and selectivity could be predicted using a species distribution diagram, which was constructed using Visual MINTEQ 3.1 (KTH Royal Institute of Technology, Sweden). [31]. Figure 1 shows the species distribution of Zn and Cu in glutamate-H2O system as a function of pH.

**Figure 1.** Species distribution of Zn (**a**) and Cu (**b**) in glutamate-H2O system as function of pH. Glutamate concentration 1 M, Zn2<sup>+</sup> 300 mM and Cu2<sup>+</sup> 8 mM.

Based on Figure 1, in acidic conditions the species are dominated by free Zn2<sup>+</sup> and Cu2<sup>+</sup> ions. Glutamate as chelate effectively binds Zn and Cu from weakly acidic to alkaline conditions, and prevents the precipitation of hydroxides at strong alkaline pH. The figure also shows that Zn starts to hydrolyze at pH 12, while Cu at pH 13. The program was also used to model the species distribution according to pH for other elements (Supplementary file: Figure S1), i.e., Fe, Mg, Ca and Al. In general Fe and Al start to hydrolyze at weakly acidic pH (±4), while Mg and Ca tend to hydrolyze at medium alkaline conditions (pH 12).

Aside from species distribution, leaching selectivity can be predicted using the complex formation constant (*Kf*) (Table 1). The constant was calculated for each complex, which might exist in the solution phase. Table 1 also lists the formation constants of metal ions with EDTA, NTA and glycinate as comparisons to assess the selectivity of organic compounds as lixiviants. If a lixiviant is expected to selectively bind one metal ion (α) and not another metal ion (β), then the value of *Kf*α/*Kf*<sup>β</sup> has to be higher than 106 (log *Kf*<sup>α</sup> <sup>−</sup> *log Kf*<sup>β</sup> <sup>≥</sup> 6) [32]. The value of complex formation constants for glutamate, which were calculated based on the model generated by Visual Minteq in Table 1, indicates the possibility of separation of Zn and Cu from Mg and Ca, and also Zn from Cu using glutamate, and demonstrates better selectivity over the other chelators such as EDTA, NTA and glycine (amino acid). The weaker complex formation constants for Fe-glutamate compared to Fe-EDTA and Fe-NTA indicate the possibility of separation of Zn and Cu from Fe through Fe hydrolysis in alkaline condition.


**Table 1.** Log complex formation constant of metal ions with three organic compounds as chelator/lixiviants.
