Platinum recycling going green via induced surface potential alteration enabling fast and efficient dissolution

The recycling of precious metals, for example, platinum, is an essential aspect of sustainability for the modern industry and energy sectors. However, due to its resistance to corrosion, platinum-leaching techniques rely on high reagent consumption and hazardous processes, for example, boiling aqua regia; a mixture of concentrated nitric and hydrochloric acid. Here we demonstrate that complete dissolution of metallic platinum can be achieved by induced surface potential alteration, an ‘electrode-less' process utilizing alternatively oxidative and reductive gases. This concept for platinum recycling exploits the so-called transient dissolution mechanism, triggered by a repetitive change in platinum surface oxidation state, without using any external electric current or electrodes. The effective performance in non-toxic low-concentrated acid and at room temperature is a strong benefit of this approach, potentially rendering recycling of industrial catalysts, including but not limited to platinum-based systems, more sustainable.

Platinum dissolution has been extensively studied by our group in relation to PEM fuel cell electrocatalysts stability. Although a complete understanding of this complex process has still not been achieved, several valuable insights have been gained recently. On the basis of our electrochemical perspective, in the current study we present a conceptually different chemical approach of dissolving Pt at substantially milder conditions compared with state-of-the-art processes. We utilize only low-concentrated hydrochloric acid (0.3 M, to provide a sufficient pH and presence of chlorides), atmospheric pressure, room temperature, and minimal amounts of gases like ozone and carbon monoxide. Specifically, our approach stands on the following crucial concepts: (i) transient platinum dissolution is an aggressive corrosion process that occurs upon oxidation and reduction of a Pt surface, also referred to as anodic and cathodic Pt dissolution 21,22 . Note that the dissolution at constant electrode potentials above the Pt dissolution onset, which could be compared with soaking in acid, is much lower compared with the transient process [23][24][25] ; (ii) the rate of Pt transient dissolution is accelerated by the presence of chlorides and other halides. We show that, by slowing the process of Pt passivation, chlorides dramatically increase Pt dissolution during anodic and cathodic potential excursions and, at the same time, stabilize Pt ions in the solution 25,26 ; and (iii) the presence of dissolved CO gas increases Pt cathodic dissolution due to strong Pt-CO interaction, which physically prevents any potential Pt redeposition 27 that may occur when CO (or some other reducing species such as H 2 ) oxidation potential gets below Pt depostition potential. The electrochemical potential of Pt, that is, the Galvani potential difference between metal surface/electrolyte, can be controlled by exposure to an appropriate gas, either oxidative or reductive in nature, most important, without the use of external potential control (potentiostat) 28 . Thus, the core of our approach is the so-called transient electrochemical dissolution 21 process triggered by repetitive cycling between two gases, leading to an electrodeless-induced surface potential alteration (Fig. 1).

Results
On-line electrode potentials and dissolution measurements. To demonstrate and explore the effectiveness of this new approach, we initially focus on a commercial Pt black catalytic system (transmission electron microscopy images in Supplementary  Fig. 1). The study firstly verifies our conceptual framework by monitoring the electrode potentials and dissolution of Pt in the defined system of scanning flow cell (SFC) 21,29 . Second, the accumulated knowledge is transferred to the more application-relevant configuration of a small-scale reactor. We employ a SFC as an advanced three electrode electrochemical cell, in which the flowing electrolyte can be fed to different analytical equipment, for example, Inductively coupled plasmamass spectrometer (ICP-MS) 21,29 or online electrochemical mass spectrometer (OLEMS) 30,31 , to obtain time-resolved information on the dissolved ions and/or evolved gases during electrochemical reactions. Generally, the three electrode set-up would allow controlling the potential of the working electrode, for example, a polycrystalline surface or high-surface catalysts. However, in the present case, it is only employed to monitor the alteration of the surface potential, that is, the open circuit potential (OCP), upon exposure to gases. The OCP is a measure of the metal/solution potential difference and corresponds to the condition of equilibrium when both cathodic and anodic reactions have the same rate. Figure 2a,b confirms the effect of different gases on the surface potential of Pt. In brief, ozone is effectively oxidizing Pt and adjusts the surface potential to slightly above 1.3 V versus RHE, where ozone reduction and oxygen/chlorine evolution are in equilibrium. In contrast, CO acts as a reductive agent that effectively reduces Pt-oxide and sets the potential to ca. 0.8 V versus RHE, the equilibrium between CO oxidation and reduction of residual oxygen. Note that as the OCP only reaches the onset potential of the anodic evolution reactions, the amount of hazardous gases (Cl 2 ) evolved is minimal (see OLEMS data in Supplementary Fig. 2). Figure 2c,   potential, while the dissolution signal quickly drops close to the detection limit upon prolonged exposure to the gases. Figure 2a displays the OCP and Fig. 2c Pt dissolution profile for a cycle of 10 min O 3 followed by 10 min CO purging. Figure 2b,d includes the dissolution profile upon three cycles of 3 min O 3 and 3 min CO, showing that increasing the frequency of gas exchanges effectively enhances the dissolution yield in the same time interval more than twofolds, that is, from 4% to 9% of the total mass of the catalyst. Lastly, it is interesting to observe that the extent of the dissolution is not only related to variations in potential, but also the nature of the oxidizing and reducing gases. Supplementary  Fig. 3 reports the comparison of the Pt dissolution profiles for 3/3 min O 3 /CO cycles and the potentiostatically simulated OCP variations, showing a largely enhanced dissolution when the gases are employed. On one hand, the presence of carbon monoxide, as mentioned before, has the additional effect of preventing redeposition of dissolved species during the surface reduction process. On the other hand, the presence of oxygen radicals related to ozone decomposition most likely have an aggravating effect on dissolution during the oxidation process.
Complete dissolution measurements. So far, the dissolution of Pt has been studied in the defined conditions of a three electrode assembly with the catalyst immobilized onto the working electrode in a flow-type electrochemical cell, which allowed us to analyse and understand the underlying fundamental dissolution processes. In a further step, the induced surface potential alteration approach was investigated in a more applicationrelevant system, that is, a small-scale batch-type reactor (Fig. 4). In this set-up, conversely to the SFC, the Pt black is present in larger quantities (10 mg) and not deposited on any electrode, but rather dispersed in 100 ml of 0.3 M HCl. Moreover, to further increase the amount of chlorides without having to increase the HCl concentration, the solution has been brought to 1 M NaCl (the effect of NaCl molarity on the dissolution yield is reported in Supplementary Fig. 4).
To provide a firm basis on this new system, we first confirmed and optimized our control over the electrochemical potential of the solution by in operando monitoring the OCP with a Pt wire electrode. Figure 3 displays the OCP variations induced by a sequence of 5/5 min O 3 /CO cycles, the optimized protocol for this system. The upper and lower potentials observed for the reactor are in complete agreement with generally expected processes of Pt oxidation and reduction ( Supplementary Fig. 5). Figure 4 demonstrates how effective the procedure in dissolving platinum can be, both visually and quantitatively. Figure 4a includes the concentration of dissolved Pt species in solution versus the number of O 3 /CO cycles (Supplementary  Table 1). Clearly, 20 cycles are more than enough to completely dissolve 10 mg of Pt black. This result is further confirmed by Fig. 4b, showing a distinct evolution of the initial black suspension of metallic Pt black to a yellow-colored solution upon O 3 /CO cycling, which is a strong indication of the presence of hexachloroplatinic complex. Conversely, exposing the suspension solely to ozone is ineffective at these mild conditions and yields only 0.6% over the same timeframe (Fig. 4a). This is in complete agreement with a previous work by Viñals et al. 32 , showing that pure ozone treatment even in 6 M aqueous HCl solution is not strong enough to dissolve Pt. These findings are the direct confirmation of the effectiveness of our novel transient dissolution approach and completely agree with our proposed mechanism in Fig. 1.

Discussion
Interestingly, a recent work by Lutsuzbaia et al. 33 also utilizes the concept of Pt transient dissolution for recycling platinum from PEM fuel cells. A similar concept, however, with electrochemical pulsed wave methods was already used before 34,35 . Nevertheless, as they apply an external potential by means of a potentiostat, their approach requires electrical contact. Thus, it is limited to conductive catalytic systems deposited on electrodes. Provided that with our system, we easily dissolve state-of-the-art PEM catalysts within few cycles ( Supplementary Fig. 6), O 3 /CO cycling has none of these limitations. As evidence, we have also extended the recycling of platinum and palladium contained in an end-of-life automotive catalytic converter, that is, 0.38 wt.% Pt and 0.24 wt.% Pd on a 99.32 wt.% honeycomb alumina support. Provided some small adjustments in the cycling protocol ( Supplementary Fig. 7  cycles both metals are completely leached from the nonconductive support. To conclude, the intention of this study is to present a completely new, environmentally friendly, safe, cheap, scalable and efficient hydrometallurgical concept of platinum leaching/ recycling. We utilize the electrochemical transient dissolution mechanism by alternatively employing oxidative and reductive conditions without the use of external potential control (electrodes). The induced surface potential alteration approach opens up a whole new technology to leach and recycle platinum, respectively, from metal ores to end-of-life products at much milder and safer conditions than current state-of-the-art processes. We prove on a fundamental level a complete new concept for Pt chemistry and we pave the way to further engineering optimization like different samples pretreatment, choice of gases (for example, H 2 and syngas,), gas flows, pH (alkaline electrolytes), concentrations of chlorides or some other lixiviants (for example, Br À and I À ), hydrodynamic conditions, reactor design, degassing and so on. Finally, as also other noble metals exhibit major transient dissolution 36 , we strongly believe the same concept can be applied for the recycling of other CRMs, including the ones out of reach for aqua regia, that is, Ru 37,38 , Ir 39 and Os 20 , which are in addition to Pt also included in our recent German patent application (app. no.: 10 2015 118 279.3).

Methods
Scanning flow cell. The online monitoring of dissolution was achieved by coupling a SFC with an ICP-MS (NexION 300X, Perkin Elmer) with a flow of 180 ml min À 1 . The catalyst films for analysis with the fully automated SFC system were deposited with a drop-on-demand printer (Nano-PlotterTM 2.0, GeSim) in the form of circular spots of ca. 500 mm in diameter onto a glassy carbon substrate. The electrolyte flowing through the SFC is mixed downstream with an internal standard ( 186 Re, 10 p.p.b.) in a Y-connector and the resulting electrolyte stream is continuously fed into the ICP-MS, where the dissolved metal ions during the electrochemical treatment are detected online. Coupling of SFC-OLEMS, as a combinatorial technique combines the parameter screening abilities of the SFC with the possibility for direct detection of volatile products and their correlation to applied potentials or currents. A mass spectrometer with electron impact ionization method is connected to the SFC over a porous Teflon membrane with a pore size of 20 nm that is positioned 50-100 mm above the catalysts surface and is therefore in a region of high product concentration. The small distance between the membrane and the electrode, because of that not only results in good sensitivity, but also leads to a fast response time between 1 and 3 s that is mainly determined by the diffusion coefficient of the analysed species. The system allows the simultaneous semi quantitative detection of several products and is therefore well suited for the parallel measurement of oxygen and chlorine.
Reactor. A laboratory 250 ml glass beaker with 100 ml of deionized water, 6 g of NaCl (1 M) and 3 ml of 30% HCl (0.3 M). The solution is stirred with a Teflon covered magnetic stirrer (3 cm) at 1,000 r.p.m. The ozone is on-site produced with a concentration of 70 g per m 3 of O 2 by an ozone generator (Innotec OGVi-8G Lab) and purged with a flow of 1 l min À 1 , while carbon monoxide is purged with a flow of 56 ml min À 1 . The dissolution yield is measured by sampling and filtering small amounts of the solution at every five-cycle interval. The diluted samples were then fed to either an ICP-MS or ICP-OES depending on the concentration. The OCP was measured by immersing a Pt wire as working electrode, graphite rod as counter electrode and Ag/AgCl as reference electrode and simply tracking the potential by setting current to zero. To avoid interference with the leaching yield results, the in operando OCP (Fig. 3) and Pt black dissolution (Fig. 4)   The car catalyst has been proven to be more resilient to dissolution compared to Pt black probably due to some additional effects related to the alumina support (ca. 99% of the all mass). Specifically, it is known that alumina acts as pH buffer and it can also physically protect the metals.