Analysis of Alkali in Bayer Red Mud: Content and Occurrence State in Different Structures

Abstract

The application of large amounts of red mud in the field of building materials is one of the main ways to reuse this material, but the high alkali content of red mud limits its application. In this paper, the washable alkali, removable alkali, and lattice alkali contents of Bayer red mud were studied, and the occurrence states of potassium and sodium in red mud were studied using XRD, IR, XPS, and NMR. On this basis, the removal mechanism for potassium and sodium in red mud was analyzed. The results showed that the Na in the red mud was mainly deposited in the shelf silicon voids of hydroxy sodalite (Na₈(AlSiO₄)₆(OH)₂(H₂O)₂) in the form of Si-O-Na or Al-O-Na. K is deposited in the shelf silico-oxygen void of potassium feldspar (KAlSi₃O₈) in the form of Si-O-K or Al-O-K. The washable Na and K contents of the mud were 13.7% and 4.47%; the alkali removal agent CaO removed 83.1% and 50.8% of Na and K in the red mud; and the lattice alkali Na and K contents were 3.20% and 44.8%, respectively. In the process of red mud dealkalization, Ca²⁺ ions can enter the internal voids of the hydroxyl sodalite and potassium feldspar silica skeleton and then replace Al³⁺ in the Si-O skeleton and Na⁺ and K⁺ in the skeleton voids. The replacement reaction changes the silica tetrahedron network structure, resulting in the disintegration of the frame-like silica tetrahedron in the hydroxyl sodalite and potassium feldspar, forming an isolated, island-like silica tetrahedron in hydrated garnet.

1. Introduction

Red mud is a solid waste generated by the alumina industry. As a major producer of alumina, China produced more than 120 million tons of red mud in 2022, with a cumulative stockpile of 1.7 billion tons. Red mud created with the Bayer process is used as an industrial by-product for other applications, such as building materials [1,2,3,4,5,6] and environmental [7,8,9,10], chemical [2,11,12,13,14,15], agronomic [16,17,18,19], metallurgical [20,21,22] processes, and so on. Currently, building materials consume reused red mud the most. However, there are some problems with the application of red mud in this area. Given the complex composition and high alkali content of red mud, when used as a building material, it is mixed with water. NaOH and KOH are precipitated during the hydration process of alkali, resulting in expansion and cracking in cement, bricks, and roads [23]. These alkalis will precipitate and form ‘efflorescence’ on the surface, affecting the quality of the cement, brick, and road. In addition, if red mud is directly added as a raw material for cement clinker, the alkali in it will corrode the cement kiln wall and affect the lifespan of the cement kiln. High alkali content considerably limits the use of red mud as a building material [24,25].

According to statistical data from 2022, the utilization rate of red mud in China is lower than 10%, which is far lower than the country’s average comprehensive utilization rate of bulk industrial solid waste. A large amount of red mud has been stored up for a long time. Because of the small particle size and poor agglomeration performance of red mud, it is easy to send dust into the air after weathering, pollute the surrounding atmospheric environment, reduce the visibility of the surrounding atmosphere, and seriously endanger the health of surrounding people [26,27]. Therefore, red mud alkali removal is important in its large-scale application in the field of building materials.

At present, the reported methods for removing alkali metals from red mud mainly include water washing [23,28], acid neutralization [29,30], replacement [31,32,33], and salting out [34]. The alkali in red mud is divided into free alkali and chemically bound alkali. The key to dealkalization is converting chemically bound alkali into free alkali. The main mechanisms in the alkali removal process for red mud include the neutralization reaction [35], the precipitation reaction [36], and the Na replacement reaction [37,38]. Because of the complexity and diversity of red mud chemically bound alkali in red mud, researchers have conducted less research on its removal mechanism. The alkalinity of red mud is the result of a complex interaction between the liquid phase and the solid phase. To find a better dealkalization method, it is necessary to have a better understanding of these interactions. The transformation, dissolution behavior, and mechanisms of chemically bound alkali in the process of red mud dealkalization should be studied continuously, and a simple and low-cost dealkalization method should be developed to achieve large-scale industrial applications.

Atmospheric lime dealkalization is a common method of red mud dealkalization, which involves adding lime under normal pressure so it can react with red mud without pressure; this process has low dealkalization condition requirements [39]. In this paper, red mud created using the Bayer aberration method was dealkalized using the atmospheric lime method, the occurrence state and removal mechanism of sodium and potassium alkali in red mud were studied, and the content of alkali in three occurrence states of red mud was calculated providing a theoretical basis for alkalinity minimization and the harmless utilization of red mud. Research on the occurrence state of alkali in red mud lays a theoretical foundation for its large-scale application in the field of building materials. This cannot only bring economic benefits but also solve the environmental problems caused by red mud and further promote the sustainable development of the alumina industry.

2. Experimental

2.1. Experimental Raw Materials

Three-year-old red mud created with the Bayer process was obtained from the Zhengzhou branch of the China Great Wall Aluminum Company. A photograph of the red mud is shown in Figure 1. The chemical compositions of the red mud are shown in

Table 1. The chemical composition of red mud.

Oxide (wt%)	SiO2	Al2O3	Fe2O3	CaO	MgO	K2O	Na2O	TiO2	Others	Loss on Ignition
Bayer red mud	19.1	22.8	16.3	13.2	0.41	1.37	6.97	3.29	0.95	15.6

. The CaO was analytically pure CaO produced by the Tianjin Beichen Reagent Company, of which the content was no less than 98%.

Table 1 shows that the Na₂O and K₂O content of the Bayer red mud was 6.79% and 1.37%, respectively. The high content of alkali content of red mud not only pollutes groundwater sources and endangers human health during the storage process but also is one of the obstacles to applying red mud to building materials.

2.2. Dealkalization Experimental Process

The red mud alkali removal experiment was carried out using the normal-pressure lime method [22]. The red mud used in this experiment was the same Bayer red mud sample introduced in Section 2.1. The experimental process is shown in Figure 2.

The red mud was dried and ground through a 100-mesh (0.15 mm) square-hole sieve. The liquid–solid ratio of the alkali removal experiment was 3:1, the reaction temperature was 90 °C, and the reaction time was 7 h. The steps of the experiment are as follows.

First, an appropriate amount of Bayer red mud (the original red mud) was dried in an oven at 105 °C. In total, 200 g of the original red mud was weighed in a conical flask, distilled water was added at a liquid–solid ratio of 3:1, and it was placed in a constant-temperature water bath shaker at 90 °C for 7 h. The conical flask was taken out and filtered. The Na and K content of the filtrate was determined using a flame photometer. The dried red mud filter cake dried in the oven at 105 °C was added to the conical flask, and the above operation was repeated until Na and K were not detected in the filtrate. The filter cake was dried, referred to as water-leaching red mud.

We took 100 g of water-leaching red mud, added 5% CaO of analytical purity, added distilled water at a ratio of 3:1, put it in a constant-temperature water bath, shook it at 90 °C for 7 h, placed it in an Erlenmeyer flask, filtered it, and measured the Na and K contents of the filtrate with a flame photometer. The dried filter cake was added to the conical flask, and the above operation was repeated until Na and K in the filtrate could not be detected. The dried filter cake was made of dealkalized red mud.

The original red mud, the water-leaching red mud, and the dealkalized red mud were analyzed via inductively coupled plasmon analysis (ICP), X-ray diffraction (XRD), IR, X-ray photoelectron spectroscopy (XPS), and nuclear magnetic resonance (NMR) to determine the washing alkali, alkali, and lattice alkali content; the occurrence state of the alkali; and the removal mechanism of alkali in red mud.

2.3. Experimental Instrument

Flame photometer

The Na⁺ and K⁺ content of the filtrate of the red mud dealkalization experiment was determined using an FP640 flame photometer produced by Shanghai Precision Scientific Instrument Co., Ltd., Shanghai, China.

ICP: Inductively coupled plasma detection

The Na and K contents of the red mud before and after the alkali removal experiment were detected and analyzed using an X Series inductively coupled plasma mass spectrometer from the Materials Research and Testing Center of Wuhan University of Technology.

XRD: X-ray diffraction

The mineral composition of red mud before and after the alkali removal experiment was analyzed using an Uitima IV. X-ray diffractometer from the Green Wall Material Center of Tianjin Chengjian University. Instrument parameters: Cu target and Ka radiation, scanning angle of 10~80° (2θ), accelerating voltage of 40 KV, current of 25 mA.

XPS: X-ray photoelectron spectroscopy analysis

The purpose of XPS analysis is to irradiate a sample with X-rays so that the inner electrons or valence electrons of atoms or molecules are stimulated and emitted. The kinetic energy/binding energy of photoelectrons (Eb = hv light energy-Ek kinetic energy-w work function) is used as an abscissa, and the relative intensity (pulse/s) was used as the ordinate; thus, the photoelectron energy spectrum can be determined, and the relevant information sample can be obtained. The XPS analysis of red mud before and after the alkali removal experiment was conducted with an AXIS Ultra DLD type X-ray photoelectron spectrometer from the Analytical Testing Center of Sichuan University.

Fourier transform infrared spectroscopy

Our Fourier transform infrared spectroscopy (FTIR) analysis used a Nexus intelligent Fourier transform infrared spectrometer from the Materials Research and Testing Center of Wuhan University of Technology to characterize the structure of the red mud before and after the alkali removal experiment. Infrared spectroscopy analyzes the structure and properties of molecules by measuring their vibrations and vibrational spectra.

Solid-state NMR spectroscopy

The chemical shifts of the Si and Al elements in red mud before and after the alkali removal experiment were analyzed using an AV400D nuclear magnetic resonance spectrometer from the Nanjing University of Technology. The technical indicators and performance characteristics were solid phase 109Ag-31P. Solid-phase NMR is a technology that uses magic angle rotation (MAS), cross-polarization (CP), and other methods to determine the chemical shift of elements and obtains the structural environment changes through the chemical displacement changes.

3. Results and Discussion

To study the occurrence state of alkali in red mud, the content of various bases, and the mechanism of alkali removal, ICP, XPS, XRD, infrared spectroscopy (FTIR), and solid-state NMR were used to analyze the following: its chemical composition, the chemical state of Na and K, the mineral composition, changes in mineral molecular structure, and the binding state of Si-O groups and Al-O groups in minerals before and after alkali removal.

3.1. Inductively Coupled Plasmon Analysis

An ICP analysis of Na and K in the original red mud [40], the water-leaching red mud, and the dealkalized red mud is shown in

Table 2. The alkali content in red mud (ion concentration/wt%).

Sample	Na	K
origin red mud	3.84	0.92
water-leaching red mud	3.32	0.88
dealkalized red mud	0.12	0.41

and Figure 3. Table 2 shows that washable Na and K contents in red mud are 13.7% and 4.47%, respectively; the alkali removal agent CaO can remove 83.1% and 50.8% of the Na and K content; and the lattice alkali Na and K contents are 3.20% and 44.8%, respectively. Figure 3 shows that the concentration of Na and K in red mud gradually decreases from red mud 1 # to dealkalized red mud 3 #, and the concentration of Na decreased much more than K in the alkali removal process. The calculation methods for washing alkali, C_W; removing alkali, C_R; and lattice alkali, C_L, are as follows:

$$C_w=\frac{C_1-C_2}{C_1}\times 100\%$$

(1)

$$C_R=\frac{C_2-C_3}{m_1}\times 100\%$$

(2)

$$C_L=\frac{C_3}{C_1}\times 100\%$$

(3)

where C₁ is the ion concentration of Na or K in the original red mud; C₂ is the ion concentration of Na or K in the water-leaching red mud; and C₃ is the ion concentration of Na or K in the dealkalized red mud.

3.2. X-ray Diffraction Analysis

Figure 4 shows the XRD spectrum of the original red mud, the water-leaching red mud, and the dealkalized red mud. Figure 4 shows that the mineral composition of the red mud’s XRD pattern before and after washing is basically unchanged, indicating that the alkali that was washed out with water was attached to the alkali and not part of the mud’s mineral composition. By comparing the maps of the dealkalized red mud and the original red mud, it can be seen that the characteristic peaks of hydroxy sodalite (Na₈(AlSiO₄)₆(OH)₂(H₂O)₂) become very weak, the characteristic peaks of potassium feldspar (KAlSi₃O₈) were significantly weakened, and the characteristic peaks of hydrofossil garnet (Ca₃Al₂SiO₄(OH)₈) significantly increased and enhanced. Therefore, Ca²⁺ in the alkali removal agent CaO replaced most of the Na⁺ in hydroxy sodalite (Na₈(AlSiO₄)₆(OH)₂(H₂O)₂) and most of the K⁺ in (KAlSi₃O₈), which was converted to hydrofossil garnet (Ca₃Al₂SiO₄(OH)₈) [41]. This shows that the occurrence state of Na in red mud is mainly in hydroxy sodalite, and K is mainly deposited in potassium feldspar.

3.3. X-ray Photoelectron Spectroscopy Analysis

XPS analysis is an analysis method that obtains binding energy using X-ray photoelectron spectroscopy and then identifies the atomic or ionic composition of a substance through binding energy [42,43,44]. To further determine the occurrence state of Na and K elements in red mud, a full-scan XPS measurement spectrum before and after alkali removal was determined, as shown in Figure 5. Figure 5 shows that there was a significant Na peak at 1070.5 eV in the original red mud, and there was no obvious Na peak in the dealkalized red mud. However, the peak of Ca at 437.5 and 345.5 eV was significantly enhanced in the dealkalized red mud compared with the original red mud. The Na content of the red mud was extremely low after alkali removal, and the Ca element content increased. According to the XPS analysis manual, hydroxy sodalite is a mineral corresponding to Na1s with an electron-binding energy of 1070.5 eV, which means that the Na element in the original red mud was mainly deposited in hydroxy sodalite. This is consistent with the results of the XRD assay.

3.4. Fourier Transform Infrared Spectroscopy Analysis

Fourier transform infrared spectroscopy analysis is an analysis method used to study the changes in material structures by measuring vibration frequency changes in anion groups in molecules. The vibrations of molecules in infrared spectroscopy are divided into two types: stretching vibrations with a changing bond length and bending vibrations with a changing bond angle. Infrared spectroscopy can not only reflect changes in vibration energy levels in molecules but also changes in molecular structure, and changes in molecular structure and chemical bonds can be identified via changes in the infrared spectrum [45,46]. The infrared analysis patterns of the original red mud and the dealkalized red mud are shown in Figure 6. Figure 6 shows that the absorption peak at 3416.4 cm⁻¹ was the telescopic vibration of the O-H group, and the bending vibration of the O-H group was 1638.07 cm⁻¹ and 1617.24 cm⁻¹. The absorption peaks at 1429.87 cm⁻¹ and 874.7 cm⁻¹ were the telescopic vibration and bending vibration of the C-O group, respectively. The absorption peaks at 1003.35 cm⁻¹ and 937.16 cm⁻¹ were the contraction vibrations of the Si-O group, and the absorption peaks at 558.98 cm⁻¹ were the flexural vibrations of the Si-O group. The absorption peaks at 1113.92 cm⁻¹ and 621.18 cm⁻¹ were the telescopic vibration and bending vibration of the S-O group, respectively.

The absorption peak at 3416.4 cm⁻¹ of the dealkalized red mud was weaker than that of the original red mud, which indicated that the hydroxysodalite (Na₈(AlSiO₄)₆(OH)₂(H₂O)) in red mud underwent a depolymerization reaction. The C-O absorption peak’s telescopic vibrations at 1429.87 cm⁻¹ and 874.7 cm⁻¹ were significantly enhanced, indicating that the C-O in the red mud polymerized after alkali removal, which was because some of the CaO added during the alkali removal process did not participate in the reaction. CaCO₃ was generated by reactions with CO₂ in water and air, and the CaCO₃ content of the red mud increased after alkali removal. The weakening absorption peak at 1003.35 cm⁻¹ indicates that the Si-O bonds were depolymerized during the alkali removal process; the enhancement of the absorption peak of 937.16 cm⁻¹ indicates that the Si-O bonds were polymerized in this process; and the enhancement of the absorption peak at 558.98 cm⁻¹ indicates that the structure of Si-O groups is more stable after alkali removal; that is, the structure of Si-O group not only recombined but also became more stable during the dealkalization of red mud. The structure was also more stable, and the specifics need to be further determined via NMR analysis.

3.5. Nuclear Magnetic Resonance Spectroscopy Analysis

Through infrared spectrum analysis, we found that the Si-O network structure of red mud may undergo depolymerization and then recombine into a new network structure. To verify this discovery and identify the depolymerization combination structure, chemical shifts and network structure changes in ²⁹Si/²⁷Al NM in the original red mud and alkali red mud were analyzed via solid-state NMR technology [47,48].

Figure 7 shows the ²⁹Si NMR spectrum of the original red mud and dealkalized red mud. From the XRD analysis results in Figure 4, it can be seen that the Si-O tetrahedron in both muds was mainly present in hydroxy sodalite, potassium feldspar, and hydrated garnet. The Si-O tetrahedron mainly had a rack-like in the hydroxy sodalite and potassium feldspar and mainly had an island structure in hydrated garnet. Figure 7 shows that there are two spectral peaks at −82.98 and −121.96 ppm in the original red mud, indicating that the ²⁹Si was tetracoordinated; the structural environment at 82.98 ppm corresponding to ²⁹Si is Q⁴(4Al); the structural environment at −121.96 ppm corresponding to ²⁹Si is Q⁴; and the spectral peak width and height were large at a chemical shift of −82.98 ppm are large, indicating that the structural environment of ²⁹Si in the original red mud is mainly rack-like Q⁴(4Al). The chemical displacements of ²⁹Si in red mud after alkali removal are −75.41 and −114.34 ppm; the structural environment of ²⁹Si at the chemical shift of −75.41 ppm is Q⁰, which indicates that the Si-O tetrahedron is an island structure, and the structural environment of ²⁹Si at a chemical shift of −114.34 ppm is Q⁴(0Al). Moreover, the spectral peak width and height were large at a chemical shift of −75.41 ppm, indicating that the structural environment of the Si-O tetrahedron in the red mud after alkali removal was mainly island-like Q⁰. In other words, in the alkali removal process, rack-like Q⁴(4Al) and Q⁴ siloxytetrahedron in the original red mud is added to Ca²⁺ ions by replacing Al³⁺, K⁺, and Na⁺ ions in the Si-O-Al, Si-O-K, and Si-O-Na structures. Thus, the silicon–oxygen tetrahedron undergoes a depolymerization and recombination reaction to form an island-like Q⁰ structure in hydrated garnet.

Figure 8 shows that the ²⁷Al NMR spectra in the original red mud and dealkalized red mud were very different. Al³⁺ in the original red mud was mainly four-coordinated, in which the main aluminosilicate forms are Al[OSi]₃ and (AlO)₄ and a small amount of (AlO)_n, while Al³⁺ in dealkalized red mud is mainly six-coordinated, in which the main form of aluminosilicate is (AlO)_n. This is consistent with the conclusion in Figure 7, which shows that the structural environment of ²⁹Si in the original red mud is mainly shelf Q⁴ (4Al) and a Q⁴-type silicon–oxygen tetrahedron without Al bridging. Only the Q⁰ and Q⁴ (0Al) types in dealkalized red mud were not bridged with an Al silicon–oxygen tetrahedron. This is because the Ca²⁺ added in the alkali removal process is a network structure coordination ion; Ca²⁺ can open Al-O bonds and change the coordination state of Si and Al in certain environments.

3.6. Dealkalization Mechanism Analysis of Bayer Red Mud

The lattice skeleton of sodalite and potash feldspar is mainly a Si-O tetrahedron framework; it is formed by arranging a Si and four oxides around it based on the shape of the tetrahedron. The silicon in the silicon–oxygen tetrahedron can be replaced by aluminum to form [AlO₄]; the internal void of the rack-like silicon–oxygen skeleton is large, especially when Al³⁺ replaces the Si⁴⁺ in the silicon–oxygen tetrahedron; the silicon–oxygen skeleton is negatively charged; and metal ions with a large radius can be introduced to the skeleton’s void, such as K⁺, Na⁺, Ca²⁺, etc., which become light silicates or aluminosilicates with less density. Na⁺ in red mud exists in the form of Na-O-Si or Na-O-Al in the voids of the hydroxyl sodalite silica skeleton, and K⁺ exists in the form of K-O-Si or K-O-Al in the voids of the potassium feldspar silica skeleton. The Ca²⁺ ions in the process of red mud dealkalization can enter the internal voids of the hydroxyl sodalite and potassium feldspar silica skeletons, replacing Al³⁺ in the Si-O skeleton and Na⁺ and K⁺ in the skeleton voids. A replacement reaction changes the silica tetrahedron network structure, resulting in the disintegration of the frame-like silica tetrahedron in the hydroxyl sodalite and potassium feldspar, forming an isolated island-like silica tetrahedron in hydrated garnet. Since the ability of Ca²⁺ to replace Na⁺ and K⁺ is limited, the replaced Na⁺ and K⁺ cannot completely escape from the silica skeleton, meaning that Na⁺ and K⁺ in red mud can only be partially removed. At the same time, the ionic radius of K⁺ is larger than that of Na⁺, so the replaced K⁺ is less likely to escape from the Si-O skeleton, meaning that CaO can remove 83.1% of Na in red mud, but it can only remove 50.8% of K.

4. Conclusions

To study the occurrence state and alkali removal mechanism for alkali in red mud, this study established a theoretical basis for the application of red mud to building materials. First, the washed alkali and removable alkali in red mud were removed using water washing and an alkali removal agent, and the washed alkali, removable alkali, and lattice alkali content were calculated. Then, the occurrence state of alkali in the red mud was further analyzed using ICP, XRD, XPS, FTIR, and NRM. This laid a theoretical foundation for the large-scale application of red mud to the field of building materials. The following conclusions were drawn from the experimental results:

Alkali in red mud can be divided into washed alkali, removable alkali, and lattice alkali. The washable Na and K contents of red mud are 13.7% and 4.47%, respectively. The alkali removal agent CaO can remove 83.1% and 50.8% of the Na and K contents, and the lattice alkali Na and K contents are 3.20% and 44.8%, respectively.

Na ⁺ in red mud exists in the void of hydroxy sodalite silica framework in the form of Na-O-Si or Na-O-Al, and K⁺ in the form of K-O-Si or K-O-Al exists in the large void of the potash feldspar silica framework.

Ca²⁺ ions enter the internal voids of the hydroxyl sodalite and potassium feldspar silica skeleton and then replace Al³⁺ in the Si-O skeleton and Na⁺ and K⁺ in the skeleton voids. The replacement reaction changes the silica tetrahedron network structure, resulting in the disintegration of the frame-like silica tetrahedron in hydroxyl sodalite and potassium feldspar, forming an isolated island-like silica tetrahedron in hydrated garnet.