Experimental and theoretical evaluation of thermodynamic properties of low melting chlorides suitable for liquefaction of biomass

Abstract

For many decades, molten salts have been employed in various industries due to their thermodynamic and physical properties. Climate change and the importance of renewable energy to stave off the worsening effects of rising global temperatures have opened the door for molten salts to be employed in new technologies. Molten salts have attracted a lot of attention in the energy and renewable energy sectors due to their high heat capacity, low vapor pressure, low viscosity, stability at high temperatures, low cost, etc. Biofuel production from waste stream lignocellulosic, as a renewable energy source, can take advantage of molten salts. This study is part of an EU project named “ABC-Salt” that aims to produce sustainable liquid biofuels through hydro-pyrolysis from various lignocellulosic waste streams. Molten salts, due to their ability to dissolve organic (and inorganic) materials, can be employed in liquefying biomass as a feed stream for biofuel production. Liquefying biomass in molten salt can make the feed pumpable. Moreover, the rapid heating of biomass in molten salt can enhance product quality. The spent molten salt should be purified and reused in the cycle to make the process more economically and environmentally attractive. The main aim of this study was to select and characterize molten salts suitable for ABCSalt. This work consists of experimental studies and theoretical modeling. The crucial thermodynamic properties of the suitable molten salt are a relatively low melting point (~200℃), high thermal stability at the high-temperature hydro-pyrolysis unit (500℃), low hydrolysis level in contact with water in biomass, and good transport properties. In the current study, two molten salt systems were investigated and compared. Four compositions of ZnCl2:KCl:NaCl (Salt #1:60:20:20, Salt #2: 59.5:21.9:18.6, Salt #3: 52.9: 33.7: 13.4, and Salt #4: 44.3:41.9:13.8 mol%) and three compositions of KCl:CuCl (Salt #5: 32:68, Salt #6: 34:66, and Salt #7: 36:64 mol%) were selected for the experimental investigations. Melting points using the cooling curve method and thermal stability employing thermal gravimetry analysis (TGA) up to 500℃ were examined for all seven compositions. The hydrolysis level of each composition was assessed by adding water vapor to the molten salt. Chlorides can react with water to generate highly corrosive hydrochloric acid (HCl). The formed HCl was measured by analyzing the outlet gas from the reactor at intervals of 50℃ up to 500℃. The results show that KCl:CuCl with a melting point of around 146℃, very high thermal stability, and no HCl formation of up to 500℃ is an interesting alternative for liquefying biomass. However, the appearance of the stainless steel/nickel setup indicates that this molten salt is highly corrosive and can ruin the construction material. Although the results for the binary system showed higher stability in terms of mass loss and hydrolysis, the thermodynamic properties of the ternary system still make it an interesting alternative to the ABC-Salt project. Among the salts from the ternary system, Salt #4 has been found to be the most promising alternative, with a melting point of 205℃ and mass loss of 0.2% up to 500℃. No HCl formation was detected for Salt #4 up to 350℃ and it slightly increased by increasing the temperature reaching 1300 ppmv at 500℃. To inhibit HCl generation, the effect of metal oxide was investigated, which showed that adding zinc oxide (ZnO) to the salt can significantly decrease HCl formation by reaction between ZnO and HCl. Moreover, the modeling results show that increasing the pressure suppresses the HCl and ZnCl2(g) formation meaning a lower amount of corrosive gas and a lower vapor pressure and mass loss in the high-pressure system. After liquefying the biomass, the spent molten salt needs to be purified from the contaminants after it leaves the hydro-pyrolysis unit before being reused in the system. In this study, the electrolysis method was used to examine the molten salt purification, with a focus on Salt #4. Although molten salts can be contaminated with many components, theoretically, it is impossible to remove all of them with electrolysis. Therefore, a few impurities with electrode potential within the electrochemical window of the salt were selected for further investigation. To simulate the spent salt, impurities were added to the molten salt, and cyclic voltammetry and chronoamperometry were carried out in a few experiments. Cu, Fe(II), Fe(III), and Mn were introduced to the molten salt by adding CuCl, FeCl2, FeCl3, and MnO2. The results showed that Cu and Fe(III) were fully extracted from the molten salt, whereas only partial removal of Fe(II) and Mn was achieved after a few hours. Hydrolysis was one of the main focuses of this work due to the lack of theoretical and experimental data in the literature. In the present work, potassium chloride (KCl) was selected for further study because it is the salt used in both ZnCl2:KCl:NaCl and KCl:CuCl systems and is also popular in many other salt mixtures. The liquid phase of a hydrolyzed KCl solution was modeled using the Calphad technique. The thermodynamic modeling of the liquid phase was carried out based on the subregular solution model using the Redlich– Kister polynomials in FactSage. The binary subsystems of H2O–KCl, H2O–HCl, H2O– K2O, and KCl–KOH were critically assessed. To the authors’ knowledge, this is the first time that chlorides, oxides, and water are modeled as a single liquid phase. The thermodynamic parameters were optimized considering all experimental data available from the literature, and binary phase diagrams were generated. ix Although binary KCl:CuCl was not selected as the best alternative for ABC-Salt, it is still a very interesting molten salt for many purposes. There is a phase diagram for this system in the literature based on the experimental data, but no thermodynamic parameters have been calculated for this system, and no database was found for the KCl:CuCl solution. Therefore, Calphad modeling of the binary KCl:CuCl molten salt was performed in this study, and the solution was modeled with a subregular solution model using Redlich–Kister polynomials. A phase diagram of the system was generated based on thermodynamic data and experimental results. The results show that the predicted eutectic point of the binary system was located at T = 145.9°C and 64.9 CuCl mol%. The calculated results are in excellent agreement with the values measured in the current work and in the literature.