Journal Information
Download PDF
More article options
Original Article
DOI: 10.1016/j.jmrt.2018.05.017
Open Access
Potassium alum thermal decomposition study under non-reductive and reductive conditions
Rodrigo Souzaa,
Corresponding author

Corresponding author.
, Rogério Navarrob, Alexandre Vargas Grilloc, Eduardo Brocchid
a Assistant Professor at Pontifícia Universidade Católica do Rio de Janeiro, Chemical and Materials Engineering Department, Rio de Janeiro, Brazil
b Adjunct Professor at Pontifícia Universidade Católica do Rio de Janeiro, Chemical and Materials Engineering Department, Rio de Janeiro, Brazil
c Adjunct Professor at Instituto Federal do Rio de Janeiro, Physical-Chemistry, Research Group, Nilópolis, Brazil
d Professor at Pontifícia Universidade Católica do Rio de Janeiro, Chemical and Materials Engineering Department, Rio de Janeiro, Brazil
This item has received

Under a Creative Commons license
Article information
Full Text
Download PDF
Figures (9)
Show moreShow less
Tables (2)
Table 1. Theoretical (Δmt) and experimental (Δmexp) mass losses for thermal decomposition under non-reductive conditions.
Table 2. Theoretical (Δmt) and experimental (Δmexp) mass losses for thermal decomposition under reductive conditions.
Show moreShow less

Potassium sulfate (K2SO4) is a very important compound, mostly used as nutrient for plant growth. Potassium sulfate production can be accomplished through the Mannheim process. For potassium bearing silicate minerals, such as glauconite, one alternative is acid leaching followed by selective precipitation and thermal decomposition of potassium alum (KAl(SO4)2). This chemical process is responsible for the formation of soluble K2SO4 and insoluble Al2O3, which can be later separated after solubilization in water and filtration. In this pyrometallurgical reaction, the temperature control is very important. Through the addition of a reducing agent, the decomposition temperature could be significantly reduced. In the present work, the thermal behavior of synthetic samples of hydrated potassium alum (KAl(SO4)2·12H2O) is appreciated through thermogravimetric analysis (TGA), both in the absence as well as in the presence of a reducing agent (charcoal) under inert atmosphere (nitrogen) and dynamic analysis. The addition of a stoichiometric amount of the reducing agent stimulated considerably the decomposition, which started at a lower temperature in comparison with the pure alum sample. Based on the XRD characterization of selected samples, it is suggested that the decomposition process should happen in at least two stages, with Al2(SO4)3 as one of the intermediate reagents. After full decomposition, only Al2O3 and K2SO4 have been identified, as expected based on thermodynamic simulations. Finally, it was demonstrated that the K2SO4 formed could be totally transferred to aqueous solution after a solubilization carried out at 363K for two hours; the remaining solid was characterized as pure aluminum oxide (Al2O3).

Potassium alum
Thermal decomposition
Full Text

Potassium is an important constituent of essential fertilizers applied in agriculture. Typically, this nutrient is offered on a large scale as a chloride (KCl) or sulfate (K2SO4), and the former accounts for the majority share of total market consumption. The latter, however, is mainly applied in specific situations where the crops are sensitive to the presence of chlorides or have the necessity of sulfur ions. Potassium sulfate can be obtained through several processes, like Mannheim process, through catalytic method, neutralization and by solvent extraction [1].

The demand for more efficient fertilizers at the most competitive prices motivates the development of alternative chemical routes to obtain these compounds. Under this perspective, some mining companies have investigated the use of other raw materials for this application such as glauconite bearing ores. A chemical route to obtain K2SO4 and Al2O3 has been proposed previously through sulfuric acid leaching, water solubilization of the formed sulfates, controlled precipitation of potassium alum (KAl(SO4)2) and thermal decomposition under nitrogen atmosphere [2].

The thermal decomposition is one of the most used processes in materials synthesis. Pysiak and Glinka [3] have studied the basic aluminum potassium sulfate (K[AI3(OH)6] [SO4]2) thermal decomposition. The results led to the conclusion that the decomposition happens in three stages. The first one considers the partial dehydration at a temperature which varies between 470 and 670K. After that, between 670 and 870K it can be observed the completion of dehydration and also, the formation of sulfates which decompose between 870 and 1200K. One of these sulfates are potassium alum. The study showed that the alum molecule is transformed in potassium and aluminum sulfates. The latter subsequentially decomposes in its respective oxide in a parallel reaction. Kuçuk and Yildiz [4] have studied the thermal decomposition of alunite ore mechanically activated and non-activated by thermogravimetry using air. The final structure was analyzed by X-ray diffraction after being mechanically worked. Alunite decomposition occurs in two steps: the first through dehydration and the second by desulfurization. From these two steps, the energies of activation were calculated from thermogravimetric. Pacewska et al. [5] have studied the dependence on the degree of conversion over time during the process of potassium and aluminum basic sulfate thermal decomposition in the presence of carbon. The experiment was performed in a tube reactor. The results of this experiment showed that at 903K after 3.5h there was only aluminum oxide (Al2O3) and potassium sulfate (K2SO4). Prost et al. [6] have studied the thermal decomposition of three different alunite compounds, the following: potassium, sodium and ammonium alunite. The work also, indicates that alunites decompose after a series of steps, based in the following processes: dehydration, dehydroxylation and desulfurization. The temperatures applied in the experiment are dependent to the cation in the experiment, in the case of K-alunite it occurs at 957K.

Within the context of potassium alum thermal decomposition, the present study has its purpose associated with the investigation of potassium alum decomposition under non-reductive and reductive atmospheres. The manuscript covers the thermodynamics of thermal and reductive decomposition reaction systems, their thermogravimetric analysis (TGA) as well as the materials characterization through X-ray diffraction (XRD) of intermediate and final products.

2Methodology2.1Thermodynamics assessment

The thermodynamics calculations were performed using the HSC Chemistry software [7]. For each reaction system (Reactions (1) and (2)), the equilibrium compositions, as a function of temperature, were calculated considering The stoichiometric input of its reagents.


Therefore, for Reaction (1), associated with the non-reductive decomposition, the diagram was obtained for 1kmol of KAl(SO4)2. On the other hand, for Reaction (2), related with the reductive decomposition, the speciation was carried out for 1kmol of KAl(SO4)2 in association with 1kmol of carbon.

2.2Experimental procedure

Samples of synthetic hydrate potassium alum (KAl(SO4)2·12H2O) were submitted to thermogravimetric analysis in two conditions: pure (Reaction (3)) and mixed with charcoal (Reaction (4)).


For the latter, the fixed carbon content was considered in the mixture preparation in order to attend the stoichiometry of the respective chemical reaction.

Both thermogravimetric analyses were conducted in a pure nitrogen atmosphere applying a heating rate of 20Kmin−1 starting at room temperature until 1473K. In the 12 present study, a Netzsch equipment, model STA 449 F3 Jupiter®, has been employed. The remaining materials were then characterized by means of X-ray diffraction through a Bruker equipment, model D8 Discover, and quantitative Rietveld analysis for the identification of the main crystalline phases present, using TOPAS software [8].

2.3Potassium solubilization

The sample produced after TGA of the pure alum sample (without reducing agent) has been dissolved in distilled water at 363K for two hours and continuous stirring. After filtration, the remaining solid has been characterized through XRD and Rietveld analysis so as to evaluate the amount of the potassium content, which had been transferred to the solution probably as soluble K2SO4.

3Thermodynamic assessment

Fig. 1 provides a quantitative picture of the nature of the equilibrium composition associated with 1kmol of KAl(SO4)2 as function of the temperature.

Fig. 1.

Equilibrium composition as function of temperature considering the input of 1kmol of KAl(SO4)2.

It can be observed that potassium alum is stable until 773K. Above that temperature and below 973K, it is clear that the system tends toward a slight decomposition of this compound in Al2(SO4)3 and K2SO4. For temperatures, higher than that, the equilibrium composition of potassium sulfate increases remarkably and aluminum oxide starts its formation. After 1023K, there is no more aluminum sulfate in the system and K2SO4 and Al2O3, in the same proportion, have their quantities in equilibrium increased until 1173K. This temperature, in turn, is associated with the full conversion of potassium alum in the desired reaction products.

It is known, however, that carbon can be used as a reducing agent in order to diminish oxidizing potential of some reaction systems. This specific characteristic could also be responsible for lowering the decomposition temperatures of salts containing oxygen in a polyatomic anion. In order to assess this information, Fig. 2 presents equilibrium composition as function of temperature resulted of the input of 1kmol of KAl(SO4)2 and 1kmol of carbon.

Fig. 2.

Equilibrium composition as function of temperature considering the input of 1kmol of KAl(SO4)2 and 1kmol of carbon.

It is patent that, under this condition, that potassium alum is not stable in any temperature between 573 and 1273K. Therefore, it can be said that K2SO4 and Al2O3 are the sole products containing potassium and aluminum which is, of course, associated with lowering the thermal decomposition of KAl(SO4)2 temperature under reductive conditions.

These results, in turn, motivate the development of an experimental study of 28 the thermal behavior of potassium alum as presented next by means of thermogravimetric analysis.

4Results and discussion4.1Thermogravimetric analysis

In the present topic, both TGA results obtained for pure alum and alum plus 2 reducing agent (charcoal) are presented and discussed, together with XRD characterization data, which corroborates some of the facts included in the results discussion.

4.1.1Potassium alum thermal decomposition under non-reductive conditions

Fig. 3 presents the TGA signal for pure hydrated potassium alum heating in nitrogen atmosphere. The data obtained for the pure alum sample indicated the presence of two main transformations. The first one (373–523K) can be explained by the water removal from the crystalline structure of the initial compound, and the second one, which happens at much higher temperatures (1023–1303K), should be associated with the anhydrous salt decomposition.

Fig. 3.

TGA signal from pure hydrated alum.

It is worthwhile to mention that the temperature values associated with these intervals, where each reaction takes place, should be influenced by the applied heating rate. For instance, for the same mass, the observed starting temperatures of each interval should be dislocated to values 30K lower through a reduction in heating rate to 10Kmin−1. Therefore, the precise determination of the initial and final temperatures demand complementary thermal analysis methods, such as DTA or DSC, to be performed in a future study.

The calculated mass losses (absolute and relative) associated with each one of these phenomena based on the compound and global reaction stoichiometry (Reaction (3)) and an initial mass of 280.90mg are very close to the experimental values (Table 1), and are also consistent with the data obtained through Rietveld analysis applied to the XRD pattern from each one of the produced materials (Figs. 4 and 5). In both cases, the quality of the assessment can be attested by the very low difference between experimental and calculated signals (line in the bottom part of the figure).

Table 1.

Theoretical (Δmt) and experimental (Δmexp) mass losses for thermal decomposition under non-reductive conditions.

Transformation  Δmt (mg)  Δmexp (mg)  Δmtr (%)  Δmexpr (%) 
H2O removal  −128.01  −124.99  −45.57  −44.50 
Salt decomposition  −71.11  −69.78  −25.32  −24.84 
Fig. 4.

XRD pattern of synthesized anhydrous alum.

Fig. 5.

XRD pattern for final thermal decomposition product.

Where, Δmtr (%) and Δmexpr (%) represent, respectively, the relative theoretical and experimental mass losses for each transformation evidenced, considering as reference the total hydrated alum mass (280.9mg).

It can be observed that the first material (Fig. 4) is identified as pure anhydrous alum, as expected, because the first TGA steps are associated exclusively to the escape from H2O molecules entrapped in the alum crystal structure. In the second case (Fig. 5), the material is composed exclusively of 19 Al2O3 and K2SO4, the only two solid products expected from the thermal decomposition of the dehydrated alum (Reaction (3)). As consequence, for each mole of KAl(SO4)2 that decomposes, equal amounts of Al2O3 and K2SO4 are generated.

Therefore, based on the literature values for each molar weight, the final mixture should consist of pure Al2O3 and K2SO4, with mass fractions respectively equal to 36.96% and 63.04%, values which lies closely to the ones obtained through Rietveld refinement of the diffraction pattern of Fig. 5.

A careful observation of each one of the observed TGA steps reveals that both phenomena (water removal and salt decomposition), should happen not in one single step, but it should be comprised of multiple stages. In the case of water removal, at least four steps can be identified in the range between 373 and 523K, and for the salt decomposition, two well defined steps can be observed. The first observation makes sense, as the water molecules do not occupy the same positions in the crystal lattice [9], and as consequence, the energy required for their removal can vary, and so the temperature necessary to accomplish this. In the case of the salt decomposition, the XRD pattern for the material removed from TGA during the 12 beginning of this step enabled, through Rietveld analysis, the identification of plausible intermediate species, such as Al2(SO4)3 and K3H(SO4)2 (Fig. 6), and also a small amount of remaining anhydrous alum. Such evidence suggests that the formation of Al2O3 and K2SO4, the expected decomposition products after full conversion of potassium alum, should also be associated with a non-single step mechanism. Therefore, Fig. 7 presents a schematic depiction of the thermal decomposition of hydrate potassium alum as well as the theoretical mass balance, detailing the contribution of each step-in weight variation through the chemical processing of this material.

Fig. 6.

XRD pattern of intermediate thermal decomposition product.

Fig. 7.

Schematic depiction of the potassium alum decomposition process.

4.1.2Potassium alum thermal decomposition under reductive conditions

The incorporation of charcoal in the system, as expected from the thermodynamic simulations (Fig. 2), should enhance the driving force for the anhydrous alum thermal decomposition, and as a result, the reaction can be conducted at smaller temperatures. Indeed, the thermogravimetric analysis of the sample under the presence of the mentioned reducing agent (Fig. 8) shows that the decomposition reaction starts at a much lower temperature (853K), in comparison with the sample of pure alum, whose thermal decomposition should start at 1023K.

Fig. 8.

TGA signals of pure alum and alum plus reducing agent.

It is interesting to observe that on what touches water removal, no difference among samples (pure alum or alum plus reducing agent) can be detected. This is again expected, as the mass loss associated with the elimination of water molecules occurs without the participation of oxygen. On the other hand, in the case of the thermal decomposition (Reaction (4)), the oxygen produced in the course of salt decomposition combines with carbon, thereby forming CO/CO2, stimulating, in the end, the process thermodynamic driving force. As observed for the experiment with the pure alum sample, the experimental mass losses (absolute and relative) due to dehydration and decomposition are close to the theoretical values calculated for an initial hydrated alum mass of 188.60mg, considering 3.65% (w/w) of charcoal (Table 2), and global stoichiometry defined by Reaction (4).

Table 2.

Theoretical (Δmt) and experimental (Δmexp) mass losses for thermal decomposition under reductive conditions.

Transformation  Δmt (mg)  Δmexp (mg)  Δmtr (%)  Δmexpr (%) 
H2O removal  −82.81  −79.66  −43.91  −42.24 
Salt decomposition  −52.89  −50.95  −28.04  −27.01 

Where, Δmtr (%) and Δmexpr (%) represent, respectively, the relative theoretical and experimental mass losses for each transformation evidenced, considering as reference the sample mass of 188.60mg, containing hydrated alum and charcoal.

4.2Potassium mass transfer to solution assessment

The sample produced after a complete decomposition of pure potassium alum under non-reductive conditions (topic 4.1.1) was exposed to a distilled water bath in order to stimulate the solubilization of the K2SO4 formed. As Al2O3 should be very stable and insoluble at the conditions imposed, it is expected that given sufficient time and sufficient initial hot water, all potassium present would be transported to the final solution. Indeed, after filtering and calcining at 1273K for four hours, the XRD pattern of the solid indicates the presence of only Al2O3 crystals (Fig. 9), in accordance to the expectations.

Fig. 9.

XRD pattern of filtered solid after calcination at 1000°C/4h.

This result corroborates the XRD data obtained for the initial sample (before solubilization), as only Al2O3 and K2SO4 crystals have been detected (Fig. 5).

Therefore, it can be concluded that the imposed conditions enabled the total transfer of potassium to the aqueous solution. However, it should be pointed out, that the solubilization conditions can still be improved, so as to enhance the solubilization kinetics. To accomplish this task, accurate mass balance data are needed which is a topic that should be covered in a future publication.

5Final remarks

Thermal behavior of hydrated potassium alum (KAl(SO4)2·12H2O) can be 10 associated with two main stages, which have been identified through TGA experiments. First, water molecules are removed from the alum crystal structure, a phenomenon, which should happen in at least four stages between 373 and 523K.

Next, through continuous heating, the anhydrous alum decomposes between 1023 and 1303K, a process, which should also happen in at least two stages, whereas Al2(SO4)3 and K3H(SO4)2 participate as intermediate compounds, as identified through XRD data. Through incorporation of 3.65% (w/w) of charcoal to the initial hydrate potassium alum, the beginning of salt decomposition shifts to a considerable lower temperature range (853–1023K), in accordance to the expectations built from thermodynamic calculations. Carbon combines with the resulting oxygen molecules (Reactions (3) and (4)), and as a result, enhances the thermodynamic driving force for the decomposition reaction. In addition, the presence of the reducing agent does not affect the alum dehydration. Finally, solubilization (363K for two hours) in distilled water of the decomposition product after TGA of potassium alum under non-reductive conditions, resulted in the total transfer of the produced K2SO4 to the aqueous media, illustrating the significant driving force for separating aluminum and potassium, when full conversion of the initial anhydrous alum can be performed, thereby resulting in insoluble Al2O3 and soluble K2SO4.

Conflicts of interest

The authors declare no conflicts of interest.


The authors are extremely grateful to the Brazilian National Council for Scientific and Technological Development (CNPq) for the financial support and scholarships.

A. Mientka, B. Grzmil, M. Tomaszewska
Production of potassium sulfate from potassium hydrosulfate solutions
Chem Papers, 62 (2008), pp. 123-126
E.A. Brocchi, C.L.R. Fragoso, R.N. Bulcão, R.C.S. Navarro, R.F.M. Souza
Chemical processing of a glauconite concentrate in order to produce K2SO4 and Al2O3
European Metallurgical Conference of 2015, Düsseldorf (Germany). Proceedings of EMC 2015. Clausthal-Zellerfeld (Germany): GDMB, 1, pp. 313-324
J. Pysiak, A. Glinka
Thermal decomposition of basic aluminium potassium sulfate part. I. Stages of decomposition
Thermochim Acta, 44 (1981), pp. 21-28
F. Kuçuk, K. Yildiz
The decomposition kinetics of mechanically activated alunite ore in air atmosphere by thermogravimetry
Thermochim Acta, 448 (2006), pp. 107-110
B. Pacewska, J. Pysiak, A. Klepanska
The investigations of termal decomposition process decomposition of basic aluminium potassium sulfate in the reduced conditions
Thermochim Acta, 92 (1985), pp. 657-659
R.L. Prost, R.A. Wills, A. Musemeci, A. Musecemi, W.A. Martens
Thermogravimetric study of the alunites of sodium, potassium and ammonium
Thermochim Acta, 443 (2006), pp. 56-61
A. Roine
HSC-Chemistry for Windows
Outokumpu Research Oy, (2006)
A.A. Coelho
DIFFRAC. TOPAS for Windows
Bruker AXS GmbH, (2007)
S.C. Nyberg, J. Steed, S. Aleksovska, V.M. Petrusevski
Structure of the 20 alums. I. On the sulfate group disorder in the alfa alums
Acta Crystallogr, 56 (2000), pp. 204-209
Journal of Materials Research and Technology

Subscribe to our Newsletter

Article options
Cookies policy
To improve our services and products, we use cookies (own or third parties authorized) to show advertising related to client preferences through the analyses of navigation customer behavior. Continuing navigation will be considered as acceptance of this use. You can change the settings or obtain more information by clicking here.