(424d) Purification of Concentrated Brines Used in the Chlor-Alkali Industry By Ion Exchange: Experimental Study and Simulation

The previously widespread mercury cell technology for chlorine production has now been replaced by more environmentally friendly membrane cell electrolysis, which is considered BAT technology. However, this technology requires a rock salt solution that is much cleaner than the previous technology and contains polluting metals (Al, Ca, Mg, etc.) in the order of ppb at most. Mined rock salt and sea salt contain a significant amount of these metal contaminants, so it is a very important task nowadays to remove the polluting metal content of the concentrated salt solution produced as a raw material of the chlor-alkali industry as completely as possible. Metal contaminants damage the membrane of the electrolysis cell resulting in higher electricity consumption and poorer quality products. The purification of this raw salt and brine is partially solved with current technology, but more efficient purification procedures promise many advantages. Thus, for example, they can increase the lifetime of the membrane of the electrolyzing cell, reduce specific electricity consumption, and contribute to the production of cleaner products, especially cleaner sodium hydroxide. Among the polluting metal ions, aluminum is currently the most significant problem.

The first step of the treatment process currently commonly used for concentrated (26 wt%) brine is precipitation of various metal contaminants, followed by sedimentation and filtration. After this, however, concentrations of contaminants at or near ppm remain in the system, predominantly in dissolved ionic form, which is removed by multi-step ion exchange. Resin beds with different functional groups and different optimum operating pH are used to remove alkaline earths, iron, and aluminum. The concentrated, multi-purified brine from the ion exchange process, containing contaminants only up to a level of ppb, is fed into the electrolysis cells.

Our work aimed to develop and optimize the ion exchange process in an industrial environment, in cooperation with BorsodChem Ltd., the leading company of the chlor-alkali industry in Hungary. We paid special attention to the removal of aluminum content, which is a knowledge gap in contrast to other major metal contaminants. However, we also investigated other metal contaminants. Our experimental work was complemented by computer simulations using Aspen Plus, Aspen Chromatography, and Aspen Adsorption software.

Following the research work presented at last year's Annual Meeting, we have carried out further dynamic experiments on aluminum removal and also investigated the elimination of alkaline earth metals. The removal of aluminum was carried out with an aminomethylphosphonic acid functional group resin at a pH of 2.5-3.0, which we had previously found to be optimal. In the tests, brines with different aluminum contents, nearly saturated (26 w%) with NaCl, were continuously flowed through a 30 ml resin bed at a rate of 30 BV/h (this means that 30 times the volume of the resin bed is flowing through the resin bed per hour, i.e. the residence time of 2 minutes). These experiments were carried out with resins at different degrees of saturation in proportion to their total capacity to investigate the effect of saturation degree on aluminum binding efficiency by measuring the aluminum content of the outlet brine. For this purpose, we applied the ICP-AES method and our previously developed and published [1] spectrophotometric method for the determination of the aluminum content of concentrated brines at ppb level. Additionally, the effect of the aluminum content of the inlet brine on the removal efficiency was investigated, and the effect of temperature and the column height of the resin were also studied. The analyses were carried out on resins from several different suppliers. Based on our results, we are able to provide recommendations for the operating cycles of ion exchange resins, specifying at which degree of saturation of the resin is required to regenerate at each temperature at the maximum allowed aluminum content in the outlet brine. This avoids the possibility of a more contaminated brine being introduced into the electrolysis cell than is appropriate, as well as avoiding that too frequent regeneration leads to unnecessary chemical consumption, i.e. increased costs and higher environmental impact.

In addition to the aluminum-binding aminomethylphosphonic acid functional group ion exchange resin, iminodiacetate functional group resins for the removal of alkaline earth metals and iron were also investigated by a similar method in dynamic experiments. For these resin types, the aim was to identify the optimal operating cycle to determine the degree of saturation, measured as a percentage of total capacity, to which these resins can be operated while maintaining the calcium and magnesium content of the outlet brine below 20 ppb.

Our experimental results were extended by computer simulations of the ion exchange operation. Simulations have been performed both for system-level analysis of brine purification and other related technologies and for detailed investigation of ion exchange. In all our studies, the ELECNRTL thermodynamic model of the Aspen software, which describes electrolyte systems accurately [2-4], was used.

The system level analysis with the Aspen Plus software package also extended to other brine purification steps commonly used in the chlor-alkali industry: precipitation with NaOH and Na₂CO₃, sedimentation, filtration to remove metal contaminants, membrane separation, crystallization to remove Na₂SO₄. This simulation allowed us to analyze the relationship of the ion exchange columns as technological units with other parts of the technology, investigating their interactions. The outlet flow of the upstream purification steps, i.e. precipitation and sedimentation, is exactly the inlet flow of the ion exchange columns, the composition of which has an impact on the ion exchange efficiency.

The ion exchange was simulated in detail using Aspen Adsorption and Aspen Chromatography software. Applying these programs, we also used our experimental data to investigate the effect of some of the parameters that were also experimentally investigated (e.g. column height, temperature, inlet concentration, degree of saturation). Additional input parameters to our experimental data and the thermodynamic data of the software were taken from the literature. The simulation allowed us to obtain isochrones, breakthrough curves, and other isoplans with different parameter settings. The results obtained by computer simulation were also compared with our experimental results.

We believe that our results can contribute to the further development of brine purification technologies that are so important in membrane cell chlor-alkali electrolysis. We are confident that a better understanding of ion exchange operations will open up the possibility for optimal operation of these processes, contributing to a lower contaminant content of the inlet brine on the electrolysis membrane. This, in addition to the longer lifetime of the very expensive membrane, can lead to lower specific electricity consumption for the products, which is beneficial from both a sustainability and an economic point of view. Furtheromore, preventing the damage of the membrane by metal contaminants will also lead to better quality products, especially in the case of NaOH, which is contaminated by the brine from the anode side through membrane damage.

Project no. C1340882 has been implemented with the support provided by the Ministry of Culture and Innovation of Hungary from the National Research, Development and Innovation Fund, financed under the KDP-2021 funding scheme.

[1] B. Csorba, L. Farkas, A. Mihalkó, R. Z. Boros and I. L. Gresits (2023) ”Photometric Determination of Trace Amounts of Aluminum in Nearly Saturated Rock Salt Solutions Used by Chlor-alkali Industry” Periodica Polytechnica Chemical Engineering 67(3):442-451. https://doi.org/10.3311/PPch.22051

[2] A. A. S. Gallindo, R. A. Silva Junior, M. G. F. Rodrigues and W. B. Ramos (2021) ”Modelling and simulation of the ion exchange process for Zn²⁺(aq) removal using zeolite NaY” Research, Society and Development 10(12):e310101220362. http://dx.doi.org/10.33448/rsd-v10i12.20362

[3] J. L. Valverde, V. R. Ferro and A. Giroir-Fendler (2022) ”Estimation of e-NRTL binary interaction parameters and its impact on the prediction of thermodynamic properties of multicomponent electrolyte systems” Fluid Phase Equilibria 551:113264. https://doi.org/10.1016/j.fluid.2021.113264

[4] E. García, L. Rodriguez, V. Ferro and J. L. Valverde (2019) ” Prediction of multicomponent ION exchange equilibria by using the e-NRTL model for computing the activity coefficients in solution” Fluid Phase Equilibria 498:132-143. https://doi.org/10.1016/j.fluid.2019.07.002