Adsorption characteristics of Cs⁺ and IO₃⁻ by organic-rich weathered volcanic ash deposit containing imogolite in southern Kyushu, Japan before and after H₂O₂ and DCB treatments

Abstract

Mineralogy of the Kaimondake volcanic ash deposit and its adsorption characteristics of Cs⁺ and IO₃⁻ were examined to reveal the adsorption mechanisms of these ions and the effects of coexisting humic substances and amorphous Fe-oxides on the adsorption. The Kaimondake volcanic ash deposit is widely distributed in the southern Kyushu, Japan. It is highly weathered, rich in humic substances, and contains imogolite as the predominant weathering product. This volcanic ash was collected and bulk (<1.0 mm size) and clay samples (<1.0 µm size) were prepared for the experiments before and after the decomposition of humic substances by H₂O₂ treatment and the subsequent selective dissolution of amorphous Fe-oxides by DCB treatment. Results confirmed that the humic substance content in the bulk and clay samples was approximately 14.9 and 37.7%, respectively, and the remainder was mostly composed of amorphous materials including imogolite, with a chemical composition of Al_3.8Si_2.1O₆(OH)₈, and amorphous Fe-oxides in a weight ratio of approximately 85:15. The PZC of the clay sample containing humic substances was pH 4.2, which increased to pH 8.7 after H₂O₂ treatment, and then decreased to pH 6.5 after DCB treatment. Both bulk and clay samples containing humic substances exhibited the highest Cs⁺ adsorption, which decreased significantly after H₂O₂ treatment and then increased greatly after DCB treatment. In contrast, the samples with humic substances showed the lowest IO₃⁻ adsorption, which increased to the highest amount after H₂O₂ treatment and then decreased to about half after DCB treatment. Thus, the adsorption of Cs⁺ and IO₃⁻ by the Kaimondake volcanic ash is mainly controlled by imogolite, while the humic substances and amorphous Fe-oxides play an important role in these adsorption behaviors. The main effects of humic substances and amorphous Fe-oxides and their possible mechanisms are as follows: (1) Humic substances promote the Cs⁺ adsorption through binding Cs⁺ to their negatively charged sites. (2) Humic substances reduce the IO₃⁻ adsorption by forming complexes with imogolite and blocking the positively charged sites on the outer walls of imogolite. (3) Amorphous Fe-oxides inhibit the Cs⁺ adsorption by preferentially binding to the negatively charged sites on the inner walls of imogolite.

INTRODUCTION

Imogolite is hydrous aluminum silicate mineral that exhibits short-range ordered structure with the structural formula of Al₄Si₂O₆(OH)₈ and an Al/Si atomic ratio of approximately 2.0 (Wada and Yoshinaga, 1969). It has a unique hollow tubular morphology characterized by inner and outer diameters of approximately 1.0 and 2.0 nm, respectively, and a length of 2-3 µm (Wada and Greenland, 1970; Cradwick et al., 1972). This mineral primarily forms during the weathering process of volcanic glass through the interaction of hydroxyaluminum ions and orthosilicic acid under slightly acidic conditions with pH <5 (Yoshinaga and Aomine, 1962; Farmer and Fraser, 1979). Thus, imogolite frequently occurs as a weathering product in volcanic ash deposits and is the predominant clay mineral found in soils derived from volcanic ash (Wada, 1989). It is also known to rarely form by the weathering of various rocks such as basalt and sandstone (Wada et al., 1972; Matsue et al., 1991). Similar to other short-range ordered clay minerals, imogolite exhibits a high reactivity with a variety of dissolved inorganic ions, including both cationic and anionic species, as well as with various organic molecules, from monomers to polymers (Gustafsson, 2001). Thus, imogolite is considered to play an important role in the behavior of dissolved ionic species in volcanic ash deposits and soils where this mineral is present as a major weathering product. However, in these weathering environments, imogolite usually occurs as a complex or mixture with organic matter derived from the remnants of plants, microorganisms, and other organisms, and also with amorphous Fe-oxides that form surface coatings and complexes with organic matter.

Various weathered volcanic ash deposits originating from andesitic to basaltic volcanic eruptions during the late Pleistocene are widely distributed in southern Kyushu, Japan. These deposits consist primarily of volcanic glass and some minor quantities of rock-forming minerals such as quartz, feldspar, pyroxene, and other minerals, as well as some rock fragments. Additionally, they typically contain a variety of clay minerals, including smectite, halloysite, allophane, imogolite, ferrihydrite, and poorly ordered Al-Si-Fe minerals, which vary depending on the extent of weathering reactions and their physicochemical conditions (Kawano and Tomita, 2001, 2002; Kawano and Nanamura, 2024). Among these volcanic ash deposits, this study used the Kaimondake volcanic ash deposit as the target sample because it contains imogolite as the main weathering product, small amounts of amorphous Fe-oxides, and a significantly higher amount of organic matter compared to other volcanic ash deposits distributed in this region. Consequently, to understand the ion adsorption characteristics of this volcanic ash deposit, it is essential to investigate the ion adsorption properties of imogolite and assess the effects of organic matter and amorphous Fe-oxides on the adsorption behavior of this volcanic ash.

The organic matter present in these volcanic ash deposits and soils consists mainly of humic substances with various molecular sizes ranging from 10² to 10⁶ Da (Tipping, 2002). This organic matter exhibits a diverse range of chemical compositions and structures, and typically contains several mmol/g of functional groups such as carboxyl groups and phenolic groups (Perdue et al., 1980). These functional groups tend to deprotonate, generating negative charges under weakly acidic to alkaline conditions, depending on the pH of the solution. Thus, humic substances readily complex with dissolved cationic species and also with positively charged sites on the surface of clay minerals, including imogolite. Therefore, it can be inferred that the humic substances in the Kaimondake volcanic ash deposit directly interact with dissolved cationic species, significantly influencing their mobility and behavior. Additionally, these substances likely form complexes with imogolite, changing the reactivity of its surface and indirectly affecting the behavior of dissolved species. Previous experiments on Cs⁺ adsorption from dilute Cs⁺ solutions using mica, related clay minerals, and soils containing these clay minerals have shown that humic substances can form complexes with the clay minerals, thereby blocking Cs⁺ uptake into their interlayers and inhibiting Cs⁺ adsorption by the clay minerals and soils (Dumat and Staunton, 1999; Staunton et al., 2002). Thus, the effects of humic substances on ion adsorption are expected to vary based on the structure of clay minerals and the presence of coexisting amorphous Fe-oxides. However, ion adsorption characteristics of imogolite and the effects of humic substances and amorphous Fe-oxides on these adsorptions are not fully understood.

In this study, Cs⁺ and IO₃⁻ were selected as target ions for investigating ion adsorption by the Kaimondake volcanic ash deposit. The Cs⁺ ion is a monovalent cation with a hydrated ionic radius of 3.29 Å (Volkov et al., 1997), which is very similar to that of K⁺ (3.31 Å). In contrast, IO₃⁻ is a monovalent anion with a hydrated ionic radius of 3.74 Å (dos Santos et al., 2010), which is larger than that of NO₃⁻ (3.35 Å) and comparable to that of SO₄²⁻ (3.79 Å). Following the radionuclide contamination of soils owing to the Fukushima nuclear disaster in 2011, numerous studies have explored the adsorption of Cs⁺ and IO₃⁻ by clay minerals, focusing mainly on the mica group and related phyllosilicates, as well as various soils and sediments that contain these clay minerals. However, there has been limited investigation into the adsorption of Cs⁺ and IO₃⁻ by imogolite and volcanic ash deposits and soils that contain imogolite. Additionally, the effects of humic substances and amorphous Fe-oxides on Cs⁺ and IO₃⁻ adsorption by imogolite and imogolite-containing volcanic ash deposits and soils remain underexplored.

To investigate the adsorption characteristics of Cs⁺ and IO₃⁻ by the Kaimondake volcanic ash deposit, the following sample pretreatments were performed: (1) no treatment; (2) decomposition of humic substances; and (3) selective dissolution of amorphous Fe-oxides following the decomposition of humic substances. The adsorption of Cs⁺ and IO₃⁻ was examined using samples before and after each treatment, clarifying the impact of humic substances and amorphous Fe-oxides on the adsorption. Adsorption characteristics of Cs⁺ and IO₃⁻ were evaluated through batch experiments with solutions of 100 and 1000 µmol/L of Cs⁺ or IO₃⁻ at pH range of 3-10, and isotherm experiments with various concentrations ranging from 20 to 2000 µmol/L at three fixed pH of 4, 6, and 8.

MATERIALS AND METHODS

Materials

A volcanic ash sample was collected from the Kaimondake volcanic ash deposit erupted ∼ 4040 years ago from Kaimondake volcano (Ishikawa et al., 1979), which is located approximately 10 km south of the sampling point (Fig. 1). The volcanic ash deposit is widely distributed around the southern areas of the Satsuma and Osumi Peninsulas. It is highly weathered, rich in organic matter, contains over 10 wt% humic substances, and has a black color. The sample used in this study was collected from a deposit in the Ikeda area of Ibusuki City, Kagoshima Prefecture, at a depth of 30-40 cm from the ground surface, and was stored in a polyethylene bag to prevent drying. This deposit is 60-70 cm thick and overlies the Kikai Akahoya ash fall deposit, which erupted 7300 years ago from the Kikai caldera located approximately 50 km south of the Satsuma Peninsula (Machida and Arai, 2003). The sample consisted mainly of fine particles and showed almost no visible rock or mineral fragments. It was also rich in humic substances, black in color, retained a considerable amount of moisture, and exhibited a soft and plastic texture.

Figure 1. Map showing the sampling points of the volcanic ash deposit, and outcrop photograph of the Kaimondake volcanic ash deposit where the sample was collected. The dark area on the map indicates the distribution area of the Kaimondake volcanic ash deposit.

Sample preparation

Volcanic ash samples of particles sized <1.0 mm and <1.0 µm were prepared and pretreated with hydrogen peroxide (H₂O₂) and dithionite-citrate-bicarbonate (DCB) for characterization and adsorption experiments. Each sample was labeled based on the pretreatment, as shown in Table 1. The collected volcanic ash sample was dispersed in distilled water, and particles >1.0 mm in size were removed using a stainless-steel sieve. This process confirmed that approximately 98% of the fractions were <1.0 mm in size, while rock fragments and mineral particles sized >1.0 mm constituted less than 2%. The fractions of <1.0 mm in size were washed multiple times with 0.1 M NaCl, and excess salts were removed with distilled water. They were then freeze-dried in a vacuum and stored as a bulk sample (KaB). The <1.0 µm fractions were also separated from the collected volcanic ash sample using the sedimentation method in distilled water with the pH adjusted to approximately 9 to enhance dispersion. This <1.0 µm sample was subsequently saturated with Na⁺ following the same procedure used for the bulk sample and was freeze-dried in a vacuum for storage as a clay sample (KaC).

Table 1. Sample pretreatment and names of the samples used in this study

Pretreatment	Bulk sample (<1.0 mm size)	Clay sample (<1.0 µm size)
No treatment	KaB	KaC
H2O2 treatment	KaB(H2O2)	KaC(H2O2)
DCB treatment*	KaB(DCB)	KaC(DCB)

^*This pretreatment involves the decomposition of humic substances by H₂O₂ treatment followed by selective dissolution of amorphous Fe-oxides by DCB treatment.

For the H₂O₂ treatment, a portion of each bulk and clay sample was incubated multiple times with 10% H₂O₂ at pH 5-7 for 2 days at 80 °C to decompose the humic substances. Subsequently, the samples were saturated with Na⁺, washed with distilled water, freeze-dried in a vacuum, and stored as H₂O₂-treated samples [KaB(H₂O₂) and KaC(H₂O₂)]. After the H₂O₂ treatment, a portion of the samples underwent DCB treatment to remove Fe-oxide coatings on the clay mineral surfaces and other amorphous Fe-oxides in the samples (Mehra and Jackson, 1960; Jackson et al., 1986). DCB treatment involved aging a 50 mL centrifuge tube containing 1.0 g of sample, 0.1 g of sodium dithionite (Na₂S₂O₄), 20 mM citric acid [C₃H₄(OH)(COOH)₃], and 20 mM sodium carbonate (NaHCO₃) at 80 °C for 15 min. This process was repeated five times, followed by multiple washes with distilled water. Next, 10% H₂O₂ was added, and the samples were incubated at 80 °C at pH 7 for 24 h to decompose the citric acid adsorbed on the clay mineral surfaces. Finally, the samples were saturated with Na⁺, washed with distilled water, and freeze-dried in a vacuum to prepare them for storage as DCB-treated samples [KaB(DCB) and KaC(DCB)].

Characterization

The contents of adsorbed water and humic substances in the bulk and clay samples were determined through weight loss after heating at 110 °C and treatment with 10% H₂O₂, respectively. Subsequently, the samples were characterized using various methods, including scanning electron microscopy (SEM), transmission electron microscopy (TEM), X-ray diffraction analysis (XRD) with Rietveld quantification, attenuated total reflectance Fourier-transform infrared absorption spectroscopy (ATR-FTIR), X-ray fluorescence analysis (XRF), and acid-base titration.

SEM observations were performed on the bulk sample using a VE-9800 scanning electron microscope (Keyence Co., Japan), operating at an accelerating voltage of 10 kV. TEM was conducted with a Hitachi H-700H transmission electron microscope (Hitachi High-Tech, Japan), operating at an accelerating voltage of 200 kV using carbon-coated clay sample dropped on Cu grids covered with a collodion film. XRD was performed using a RINT2000 diffractometer (Rigaku Co., Japan) with CuKα radiation generated at 40 kV and 30 mA employing a scanning speed of 2° 2θ/min. Both bulk and clay samples were homogenized in an agate mortar and subsequently packed into an aluminum holder for measurement. The Rietveld quantification of constituents, including amorphous materials, was performed using the powder diffraction profile analysis software PDXL integrated into the XRD instrument. XRD data were collected at a scanning speed of 0.5° 2θ/min for each sample, which included 20% corundum as an internal standard. The structural data for the minerals used were corundum (ICDD: 01-071-1123), quartz (ICDD: 00-005-0490), anorthite (ICDD: 00-020-0528), cristobalite (ICDD: 01-071-0785), and magnetite (ICDD: 01-072-2303). ATR-FTIR spectra were recorded using a JASCO FT/IR4700 spectrometer (JASCO Co., Japan) equipped with a single reflection diamond prism ATR module (ATR PRO450-S) at an incident angle of 45 degrees. XRF analysis was performed with a ZSX 100e (Rigaku Co, Japan), using an X-ray tube with a Rh anode operating at 50 kV and 60 mA. Major elements were measured using fused glass beads made from powder samples heated to 1000 °C, and the weight loss owing to this heating was assigned to H₂O %.

Acid-base titrations were conducted to assess the surface charge of clay samples before and after treatment with H₂O₂ and DCB. The titration procedures followed the methodology outlined by Kawano and Nanamura (2024) for three clay samples: KaC, KaC(H₂O₂), and KaC(DCB). Briefly, 0.1 g of each sample was combined with 100 mL of a decarbonated background solution containing 0.1, 1.0, and 10 mM NaCl in a polyethylene reaction vessel, which was equipped with a glass electrode pH meter and a nitrogen gas supply valve. The titration was carried out at 25 °C with stirring and a continuous flow of nitrogen gas, adding 0.1 mL of titration solution (0.1 M HCl and 0.1 M NaOH) every 15 min. The point of zero charge (PZC) of each sample was determined at the intersection of the three titration curves at different NaCl concentrations in the background solution.

Adsorption experiments

The adsorption experiments for Cs⁺ and IO₃⁻ were performed in a batch method using both bulk and clay samples before and after H₂O₂ and DCB treatments. In each 10 mL glass test tube, 0.01 g of the sample was combined with 10.0 mL of 1.0 mM NaCl solution containing specific concentrations of Cs⁺ or IO₃⁻. First, experiments were carried out at fixed concentrations of 100 and 1000 µmol/L at ten different pH ranging approximately from pH 3 to 10, to evaluate the influence of solution pH on the adsorption of Cs⁺ and IO₃⁻. The pH of the solution was adjusted using HCl or NaOH solutions during the aging process. The test tubes were shaken for 24 h at 25 °C, then the supernatant solutions were separated through centrifugation, and the concentrations of Cs⁺ and IO₃⁻, as well as the solution pH, were measured using high-performance liquid chromatography (HPLC) and a glass electrode pH meter, respectively. Additionally, adsorption isotherms for Cs⁺ and IO₃⁻ were determined at concentrations of 20, 50, 100, 200, 500, 1000, 1500, and 2000 µmol/L at three different pH of 4, 6, and 8. The solution pH was adjusted several times during the aging for 24 h using HCl or NaOH solutions. After aging for 24 h, the concentrations of Cs⁺ and IO₃⁻, as well as the solution pH, were measured using the same procedure as described earlier. The concentrations of Cs⁺ and IO₃⁻ were measured by HPLC with a Hitachi LaChrom Elite (Hitachi High-Tech Co., Japan), using a TSKgel IC-Cation I/II HR column for Cs⁺ and a TSKgel IC-Anion-PW_XL column for IO₃⁻.

RESULTS

Characterization of volcanic ash

The Kaimondake volcanic ash is characterized by its black color and a substantially higher content of humic substances. Table 2 shows the amounts of adsorbed water and humic substances in the bulk and clay samples, which were determined by heating at 110 °C and treating with 10% H₂O₂. Following the H₂O₂ treatment, both bulk and clay samples changed color from black to brown, and the humic substance content calculated from weight loss was found to be 14.9 and 37.7%, respectively.

Table 2. Contents of adsorbed water and humic substances in the bulk and clay samples

Phase (wt%)	Bulk sample (<1.0 mm size)	Clay sample (<1.0 µm size)
Adsorbed water	11.8	16.3
Humic substances	14.9	37.7
Inorganics	73.3	46.0

Weight loss after heating at 110 °C and H₂O₂ treatment was used as the contents of adsorbed water and humic substances, respectively. The remaining weight was taken as the content of inorganics.

SEM and TEM observations. SEM of the KaB revealed that the volcanic ash consisted mainly of irregularly shaped micro- to nano-sized aggregates of humic substances and micro-sized fragments of volcanic glass (Fig. 2a). The identification of imogolite and amorphous Fe-oxide minerals proved difficult, likely owing to these minerals forming complexes or aggregates with the humic substances. TEM of the KaC confirmed that the clay fractions were predominantly composed of imogolite fibers and irregularly shaped aggregates of humic substances (Fig. 2b). The imogolite fibers showed varying widths of less than approximately 40 nm and were believed to result from the bundle-like assembly of imogolite nanotubes.

Figure 2. SEM micrograph (a) of the bulk sample (KaB) showing volcanic glass fragments and humic substances in the Kaimondake volcanic ash deposit, and TEM micrograph (b) of the clay sample (KaC) showing bundles of imogolite tubes and irregularly shaped aggregates of humic substances. The label HS indicates humic substances, which may be partly complexed with imogolite tubes and amorphous Fe-oxides.

XRD analysis. XRD of the bulk samples before and after H₂O₂ and DCB treatments revealed that the samples were primarily composed of quartz, feldspar, cristobalite, magnetite, and amorphous materials, which may include volcanic glass and short-range ordered clay minerals (Fig. 3a). For the clay samples, KaC exhibited no distinct Bragg reflection, indicating that it consisted largely of amorphous materials such as humic substances and some clay minerals with short-range ordered structures (Fig. 3b). After the H₂O₂ and DCB treatments, broad peaks corresponding to imogolite (Wada, 1989) and small peaks of quartz, cristobalite, and magnetite were detected in the KaC(DCB) sample. Rietveld quantification indicated that KaB(DCB) consisted of 64% amorphous materials, including volcanic glass and imogolite, 9% quartz, 23% feldspar, 1.3% cristobalite, and 2.4% magnetite. In contrast, KaC(DCB) comprised 95% amorphous materials, predominantly imogolite, possibly with trace contamination from other amorphous materials such as volcanic glass. It also contained 1.6% quartz, 0.5% feldspar, 1.5% cristobalite, and 1.8% magnetite.

Figure 3. XRD of the bulk samples (a) collected from the Kaimondake volcanic ash deposit before and after H₂O₂ and DCB treatments, and the clay samples (b) before and after the same treatments.

ATR-FTIR spectroscopy. ATR-FTIR spectrum of KaB demonstrated pronounced absorption peaks at 954, 1392, and 1587 cm⁻¹ (Fig. 4a). The absorption band at 954 cm⁻¹ is attributed to the Si-O vibrations of certain silicate minerals and volcanic glass (Farmer, 1974). The bands at 1392 and 1587 cm⁻¹ are characteristic of humic substances, which can be attributed to the symmetric and asymmetric stretching vibrations of carboxyl group (COO⁻) in the organic molecules, respectively (Stevenson and Goh, 1971; Vinkler et al., 1976). After the decomposition of humic substances by H₂O₂ treatment, the absorption bands of COO⁻ nearly disappeared, while a small absorption band around 1640 cm⁻¹, associated with the bending vibration of H₂O, appeared. The KaC displayed relatively strong absorption bands for humic substances at 1394 and 1589 cm⁻¹ (Fig. 4b), indicating a greater presence of humic substances compared to the bulk sample (Table 2). In the KaC(H₂O₂) and KaC(DCB) samples, these two absorption bands were absent, suggesting almost complete decomposition of humic substances after H₂O₂ treatment. Additionally, KaC(DCB) exhibited two characteristic adsorption bands at 936 and 998 cm⁻¹, which are consistent with the Si-O stretching vibrations observed in natural and synthetic imogolite samples (Farmer et al., 1979; Bishop et al., 2013). Therefore, KaC(DCB) is primarily composed of imogolite, with trace amounts of quartz, cristobalite, and magnetite confirmed by XRD (Fig. 3b). In contrast, the KaC and KaC(H₂O₂) samples likely contain substantial amounts of amorphous Fe-oxides, which may be present as coatings on mineral surfaces and complexes with humic substances. This presence could reduce the detection sensitivity of imogolite in both XRD and ATR-FTIR analyses.

Figure 4. ATR-FTIR of the bulk samples (a) collected from the Kaimondake volcanic ash deposit before and after H₂O₂ and DCB treatments, and the clay samples (b) before and after the same treatments.

XRF analysis. Table 3 presents the chemical composition of the bulk and clay samples after H₂O₂ and subsequent DCB treatments, as measured by XRF. The KaB(H₂O₂) exhibited higher contents of Al₂O₃ and Fe₂O₃, along with considerably lower contents of MgO and CaO compared to certain fresh andesitic and basaltic volcanic ashes (Vogel et al., 2017). This suggests that relatively intense weathering occurred in the organic-rich environment after the deposition of the volcanic ash. In the KaB(DCB) sample, DCB treatment confirmed a notable decrease in Fe₂O₃ content owing to the dissolution of amorphous Fe-oxides in the bulk sample. However, a small amount of Fe₂O₃ persisted after the DCB treatment, which is consistent with the detection of a minor quantity of magnetite by XRD. In the clay samples, KaC(H₂O₂) showed considerably higher contents of Al₂O₃ and Fe₂O₃, along with lower amounts of MgO and CaO compared to the bulk samples. Whereas, the Fe₂O₃ content in KaC(DCB) was markedly reduced owing to the dissolution of Fe-oxides that coated the imogolite surface and formed complexes with humic substances. The chemical composition of imogolite was derived by correcting the composition of KaC(DCB) based on the amount of contaminating minerals quantified by the Rietveld method. The chemical compositions used for this correction were An₉₄Ab₆ for feldspar (Matsui et al., 2003) and ideal compositions for quartz, cristobalite, and magnetite. The structural formula of imogolite, calculated from this corrected composition, is Al_3.8Si_2.1O₆(OH)₈, which is close to the ideal composition of the mineral [Al₄Si₂O₆(OH)₈].

Table 3. Chemical composition of the bulk and clay samples after H₂O₂ and subsequent DCB treatments

	KaB(H2O2)	KaB(DCB)	KaC(H2O2)	KaC(DCB)
SiO2	40.11	54.36	19.83	27.89
TiO2	1.22	0.87	1.39	0.72
Al2O3	21.12	20.72	33.89	36.77
Fe2O3	10.88	4.10	13.38	1.67
MgO	1.18	1.58	0.76	0.97
CaO	1.11	1.66	0.12	0.07
Na2O	1.07	2.00	0.42	0.50
K2O	0.57	0.85	0.24	0.43
H2O	22.73	13.86	29.97	30.99
Total (%)	100.00	100.00	100.00	100.00

The oxide % data of the major elements were obtained by XRF analysis, and the weight loss values after heating at 1000 °C were used for H₂O %. Total Fe was expressed as Fe₂O₃, and each oxide % was normalized to 100% total.

Acid-base titrations. The titration curves and PZC values of the clay samples obtained from the acid-base titration experiments are shown in Figure 5. These clay samples exhibited notable differences in their titration curves before and after treatment with H₂O₂ and DCB. KaC shifted considerably toward the acidic side, while KaC(H₂O₂) moved toward the alkaline side. The titration curves verified that the PZC values of each sample were pH 4.2, 8.7, and 6.5 for KaC, KaC(H₂O₂), and KaC(DCB), respectively. These results are consistent with the high content of humic substances in KaC with an amount of 37.7% (Table 2), and relatively high content of amorphous Fe-oxides in KaC(H₂O₂) at 13.38% Fe₂O₃ (Table 3). While KaC(DCB) contains trace impurities, most of the humic substances and amorphous Fe-oxides have been either decomposed or selectively dissolved through H₂O₂ and DCB treatments. Consequently, the titration curve of KaC(DCB) likely reflects the surface charge characteristics of imogolite in this volcanic ash sample.

Figure 5. Titration curves of the clay samples before and after H₂O₂ and DCB treatments in background solutions of 0.1, 1.0, and 10 mM NaCl solutions. The PZC of each sample was obtained as the pH value at the intersection of the titration curves for three different NaCl concentrations, as indicated by the arrows.

Adsorption of Cs⁺ and IO₃⁻ ions

The adsorption experiments revealed that the adsorption of Cs⁺ by the bulk samples increased gradually with increasing solution pH in response to the pretreatment (Fig. 6a). In the 1000 µmol/L Cs⁺ system, KaB exhibited the highest Cs⁺ adsorption among the bulk samples, reaching approximately 230 µmol/g at pH 8. In contrast, KaB(H₂O₂) showed a lower Cs⁺ adsorption of approximately 60 µmol/g compared to the other samples. Meanwhile, KaB(DCB) demonstrated a higher Cs⁺ adsorption of approximately 120 µmol/g than KaB(H₂O₂) under the same pH conditions. Therefore, the order of Cs⁺ adsorption by the bulk samples was confirmed to be KaB > KaB(DCB) > KaB(H₂O₂). Similar trends in Cs⁺ adsorption were observed in the 100 µmol/L Cs⁺ system before and after H₂O₂ and DCB treatments. The clay samples also exhibited pH-dependent Cs⁺ adsorption, which is consistent with the order of the bulk samples: KaB > KaB(DCB) > KaB(H₂O₂) (Fig. 6b). On the other hand, IO₃⁻ adsorption gradually increased as pH decreased depending on the pretreatment (Fig. 6c). The order of IO₃⁻ adsorption by the bulk samples was KaB(H₂O₂) > KaB(DCB) > KaB in both the 100 and 1000 µmol/L systems, which was the reverse of the trend observed for Cs⁺ adsorption. The IO₃⁻ adsorption by the clay samples exhibited a slight increase compared to the bulk samples, with the order of magnitude of adsorption remaining consistent with that of the bulk samples (Fig. 6d). These findings suggest that humic substances in the volcanic ash deposit promote Cs⁺ adsorption while inhibiting IO₃⁻ adsorption. Conversely, amorphous Fe-oxides tend to reduce Cs⁺ adsorption and promote IO₃⁻ adsorption.

Figure 6. Adsorption of Cs⁺ and IO₃⁻ by the bulk [(a) and (c)] and clay samples [(b) and (d)] before and after H₂O₂ and DCB treatments at initial concentrations of 100 and 1000 µmol/L in the solution pH range of 3-10.

The adsorption isotherms for Cs⁺ on the bulk and clay samples before and after H₂O₂ and DCB treatments, performed at concentrations up to 2000 µmol/L and solution pH of 4, 6, and 8, are presented in Figure 7. The results clearly demonstrate that the Cs⁺ adsorption increased with increasing equilibrium Cs⁺ concentrations and solution pH. Additionally, Cs⁺ adsorption by both bulk and clay samples was highest before H₂O₂ and DCB treatments, decreased significantly after the H₂O₂ treatment, and slightly increased after the subsequent DCB treatment. However, IO₃⁻ adsorption was the lowest before H₂O₂ and DCB treatments, increased significantly after H₂O₂ treatment, and then declined again after the subsequent DCB treatment (Fig. 8). The adsorption behaviors of Cs⁺ and IO₃⁻ by the bulk and clay samples before and after H₂O₂ and DCB treatments were evaluated quantitatively using the linear equation of the Langmuir isotherm presented below.

  

\begin{equation} C/Q = (1/Q_{\text{m}})C + 1/KQ_{\text{m}} \end{equation} (1),

where Q is the amount of adsorption (µmol/g), K is the equilibrium constant (L/µmol), C is the equilibrium concentration of the adsorbate (µmol/L), and Q_m is the maximum amount of adsorption (µmol/g). The Q_m and K values for Cs⁺ and IO₃⁻ adsorption by the bulk and clay samples, calculated using this method, are listed in Table 4. The Q_m values for Cs⁺ adsorption by the bulk and clay samples before H₂O₂ and DCB treatments ranged from approximately 68-463 and 130-780 µmol/g, respectively. After H₂O₂ treatment, the Q_m values of both samples at the same pH conditions were 0.3-0.1 times lower than those before H₂O₂ treatment. After DCB treatment, the Q_m values for the bulk and clay samples were approximately 65-170 and 67-330 µmol/g, respectively, which were about 0.7-0.4 times lower than those before H₂O₂ and DCB treatments. On the other hand, the Q_m values for IO₃⁻ adsorption by the bulk and clay samples before H₂O₂ and DCB treatments were approximately 68-9 and 117-31 µmol/g, respectively. After H₂O₂ treatment, these Q_m values increased by approximately 3.0-5.5 and 1.9-2.7 times, and then by approximately 1.0-1.5 and 1.3-1.7 times, respectively, after subsequent DCB treatment.

Figure 7. Adsorption isotherms for Cs⁺ on the bulk [(a)-(c)] and clay samples [(d)-(f)] before and after H₂O₂ and DCB treatments up to an initial concentration of 2000 µmol/L at solution pH of 4, 6, and 8.

Figure 8. Adsorption isotherms for IO₃⁻ on the bulk [(a)-(c)] and clay samples [(d)-(f)] before and after H₂O₂ and DCB treatments up to an initial concentration of 2000 µmol/L at solution pH of 4, 6, and 8.

Table 4. Langmuir parameters for Cs⁺ and IO₃⁻ adsorption on the bulk and clay samples at pH 4, 6, and 8 before and after H₂O₂ and DCB treatments

Sample	Cs+ adsorption		IO3− adsorption
	Qm (µmol/g)	K (L/µmol)	Qm (µmol/g)	K (L/µmol)
KaB
pH 4	67.5	0.0032	67.8	0.0102
pH 6	178.3	0.0052	23.3	0.0094
pH 8	463.0	0.0035	8.7	0.0171
KaB(H2O2)
pH 4	18.1	0.0029	201.2	0.0052
pH 6	53.6	0.0020	146.0	0.0027
pH 8	65.6	0.0047	78.8	0.0029
KaB(DCB)
pH 4	65.2	0.0017	92.9	0.0033
pH 6	133.5	0.0043	28.7	0.0051
pH 8	170.4	0.0051	22.5	0.0023
KaC
pH 4	129.7	0.0012	117.3	0.0049
pH 6	359.7	0.0015	49.5	0.0038
pH 8	781.3	0.0010	31.4	0.0067
KaC(H2O2)
pH 4	32.1	0.0109	325.7	0.0101
pH 6	76.7	0.0057	109.9	0.0065
pH 8	113.1	0.0142	59.8	0.0070
KaC(DCB)
pH 4	66.5	0.0019	149.2	0.0035
pH 6	198.0	0.0029	76.2	0.0017
pH 8	330.0	0.0037	54.1	0.0016

DISCUSSION

Adsorption of Cs⁺ and IO₃⁻ on imogolite surface

Characterization of the Kaimondake volcanic ash deposit indicated that imogolite was the primary component responsible for the adsorption of Cs⁺ and IO₃⁻ by this volcanic ash. Additionally, humic substances and amorphous Fe-oxides significantly affect their adsorption behavior. Titration experiments revealed that the PZC of imogolite was pH 6.5, which aligned closely with previously reported titration values and was considerably lower than the values obtained via electrophoresis (Su et al., 1992; Su and Harsh, 1993). It is widely recognized that the PZC of imogolite varies significantly based on the measurement method. The electrophoretic method yields values around pH 9-10, which are considerably higher than those obtained through titration. This discrepancy is believed to be due to the positive charges of residual cations present in the hollow tubes of imogolite (Harsh et al., 1992). In contrast, the PZC determined by titration reflects the point of zero net proton charge, arising from the transfer of H⁺ and OH⁻ between the imogolite surface and the solution. Thus, this PZC is crucial for understanding the surface charge involved in the electrostatic adsorption of external ions onto the imogolite surface. Essentially, the structure of imogolite comprises a gibbsite-like Al octahedral network sheet curled into a tube shape, with Si tetrahedra bonding to the vacant Al octahedral sheet from the inside of the hollow tube by sharing three oxygen atoms. Notably, the outer wall surface of imogolite carries some positive charge, despite sharing a similar structure with the (001) surface of gibbsite. This is attributed to the distortion of the gibbsite-like sheet of imogolite caused by the three oxygen-sharing bonds of the Si tetrahedron in the hollow. This distortion leads to unsaturated bonds in the Al₂OH on the outer walls of the tube, resulting in the generation of positive charges (Gustafsson, 2001). The edge surface area of imogolite tubes accounts for approximately 0.1% of the total surface area (Arai et al., 2006), making the influence of edge surface functional groups on the overall surface charge virtually negligible. Consequently, it can be concluded that the adsorption of Cs⁺ by imogolite mainly results from the negatively charged sites of the SiOH groups on the inner wall surface. In contrast, the adsorption of IO₃⁻ is predominantly attributed to the positively charged sites of the Al₂OH groups on the outer wall surface.

As a result of the adsorption experiment, both KaB(DCB) and KaC(DCB) samples, which removed humic substances and amorphous Fe-oxides, exhibited a continuous increase in Cs⁺ adsorption with increasing pH. This suggests that the SiOH functional groups on the inner walls of imogolite play a considerable role in Cs⁺ adsorption. Typically, the PZC of SiOH groups is approximately 2 (Dove and Rimstidt, 1994). Thus, they become deprotonated, forming negatively charged SiO⁻ sites on the inner walls. This process enhances the electrostatic attraction between Cs⁺ and these sites at pH >2. This is the same mechanism as the adsorption of Cs⁺ to the SiOH functional groups on the edge surface of phyllosilicate minerals such as kaolinite and illite (Wahlberg and Fishman, 1962; Kim et al., 1996). Conversely, the KaB(DCB) and KaC(DCB) samples demonstrated a gradual increase in IO₃⁻ adsorption as pH decreased. The PZC of the Al₂OH functional groups on the outer wall surface of imogolite is usually pH 9-10 (Gustafsson, 2001). Consequently, these groups become protonated at pH <9-10, resulting in the formation of positively charged sites that likely enhance the adsorption of IO₃⁻ through electrostatic complexation. A similar pH-dependent adsorption behavior of IO₃⁻ was observed in experiments using γ-Al₂O₃. Surface complexation modeling analysis revealed that IO₃⁻ predominantly formed outer-sphere complexes and adsorbed through electrostatic interactions with the AlOH groups at pH <∼ 9. In contrast, adsorption through inner-sphere complexation via ligand exchange increases significantly under lower pH conditions (Nagata and Fukushi, 2010). Additionally, XAFS analysis of IO₃⁻ adsorption on allophane, which has a similar structure to imogolite, confirmed that IO₃⁻ primarily formed inner-sphere complexes, binding through monodentate bonds via ligand exchange with Al₂OH functional groups on the outer surface at pH below the PZC (Wang et al., 2024). Therefore, for imogolite as well, IO₃⁻ is likely to be adsorbed mainly through electrostatic binding or ligand exchange with the Al₂OH functional groups on the outer wall, depending on the solution pH.

Effect of amorphous Fe-oxides on Cs⁺ and IO₃⁻ adsorption

The Kaimondake volcanic ash contains significant amounts of amorphous Fe-oxides, which form coatings on mineral surfaces and may include ferrihydrite complexed with humic substances. Adsorption experiments demonstrated that these amorphous Fe-oxides significantly influenced the adsorption of Cs⁺ and IO₃⁻ by the volcanic ash. When comparing the adsorption of Cs⁺ and IO₃⁻ before and after DCB treatment, Cs⁺ adsorption after DCB treatment increased to nearly twice that before DCB treatment, while IO₃⁻ adsorption decreased to about half of that before DCB treatment. To assess the impact of the amorphous Fe-oxides on the adsorption of Cs⁺ and IO₃⁻, the content of amorphous Fe-oxides in the clay sample was calculated using the following method. First, the chemical composition of KaC(H₂O₂) sample (Table 3) was corrected using the impurity crystalline mineral contents quantified by the Rietveld method. This correction allowed for determination of the chemical composition of the amorphous materials, which mainly consist of imogolite and amorphous Fe-oxides. The contents and chemical compositions of the crystalline minerals used were 0.1% quartz, 0.56% feldspar, 0.32% cristobalite, and 1.9% magnetite, and An₉₄Ab₆ for feldspar and ideal compositions for other minerals. The corrected chemical composition was then used to calculate the weight ratio of imogolite to amorphous Fe-oxides, assuming all Fe₂O₃ originated from the amorphous Fe-oxides, while SiO₂ and Al₂O₃ were derived from imogolite. The chemical composition of the amorphous Fe-oxides used was Fe₁₀O₁₄(OH)₂, which is a typical composition of ferrihydrite (Michel et al., 2007). As a result, the chemical composition of imogolite was obtained as Al_4.05Si_1.95O₆(OH)₈, and the weight ratio of imogolite to amorphous Fe-oxides was found to be 85:15. This implies that the amorphous Fe-oxides make up approximately 15 wt% of the amorphous materials in the clay sample before DCB treatment. Typically, ferrihydrite contains small and variable amounts of Si and Al as impurity ions (Cismasu et al., 2011), and humic substances readily form complexes with Al ions as well as Fe³⁺ (Parfitt, 2009). These compositional variations were not considered in this weight ratio calculation.

The PZC of natural Fe-oxides, such as ferrihydrite, ranges from pH 7 to 9 and is influenced by impurity ion content, including Si ions (Schwertmann and Fechter, 1982). In the alkaline region, the FeOH functional groups of Fe-oxides become deprotonated, resulting in a surface negative charge that enhances the adsorption of cationic species such as Cs⁺ on the Fe-oxide surfaces. Conversely, in the weakly alkaline to acidic regions, the functional groups become gradually protonated, creating a surface positive charge that increases the adsorption of anions on the surfaces. Previous studies have demonstrated that pH-dependent adsorption of Cs⁺ and IO₃⁻ on the Fe-oxide surfaces. It has been shown that Cs⁺ mainly adsorbs by forming outer-sphere complexes with positively charged sites on the Fe-oxide surface (Kikuchi et al., 2019). In contrast, IO₃⁻ adsorption occurs through both outer-sphere and inner-sphere complex formation with negatively charged sites on the Fe-oxide surface (Nagata and Fukushi, 2010).

The ion adsorption ability of amorphous Fe-oxides appears to be lower for cations and higher for anions compared to imogolite, which is largely attributed to the considerably higher PZC of Fe-oxides relative to imogolite. Before DCB treatment, the volcanic ash sample was primarily composed of imogolite and amorphous Fe-oxides with the weight ratio of 85:15. Therefore, the samples before DCB treatment are expected to show lower Cs⁺ adsorption and higher IO₃⁻ adsorption than the samples after DCB treatment, reflecting the influence of amorphous Fe-oxides. However, the differences in Cs⁺ and IO₃⁻ adsorption before and after DCB treatment were much greater than those expected from the weight ratio of amorphous Fe-oxides (Figs. 6 and 7). This phenomenon is believed to be attributed to the preferential deposition of Fe-oxide coatings on the inner walls of imogolite. In weathering environments, the formation of Fe-coatings on the surfaces of clay minerals is common, and these coatings tend to accumulate preferentially on negatively charged surfaces (Gallez et al., 1976). Imogolite generates negative charges through the deprotonation of SiOH functional groups on its inner walls. It is inferred that the Fe octahedra bind preferentially to these negatively charged sites, as illustrated in Figure 9a, which inhibits the binding of Cs⁺ to the SiO⁻ sites and reduces the adsorption of Cs⁺ by imogolite. This is consistent with the observation that the PZC of clay samples was significantly higher before DCB treatment (pH 8.7) than after this treatment (pH 6.5). Additionally, the FeOH functional groups of amorphous Fe-oxides become positively charged through protonation at pH <∼ 8, which facilitates the electrostatic binding of IO₃⁻ and increases its adsorption. Previous ion adsorption experiments using kaolinite and allophane have confirmed a similar effect of Fe-coatings in increasing the PZC and enhancing anion adsorption (Escudey and Galindo, 1983; Zhuang and Yu, 2002; Kawano and Nanamura, 2024).

Figure 9. Schematic diagram showing deposition of Fe octahedra on the inner walls of imogolite (a) and complexes of humic substances with imogolite tube bundles (b).

Effect of humic substances on Cs⁺ and IO₃⁻ adsorption

The bulk and clay samples before H₂O₂ treatment contain approximately 14.9 and 37.7 wt% humic substances, respectively, excluding adsorbed water. During sample preparation, these samples were repeatedly washed with 0.1 M NaCl solution and distilled water, which likely removed most soluble organic molecules, except for those adsorbing on the mineral surfaces. Adsorption experiments on samples containing these humic substances demonstrated considerably higher Cs⁺ adsorption and markedly lower IO₃⁻ adsorption compared to the samples after decomposition of humic substances (Figs. 6 and 7). These results indicate that humic substances substantially enhance the adsorption of Cs⁺ by volcanic ash, while strongly inhibiting the adsorption of IO₃⁻. Humic substances in terrestrial surface environments typically exhibit a wide range of molecular weights, approximately between 10² and 10⁶ Da (Tipping, 2002), They feature functional groups, such as phenolic and carboxyl groups, present in total amounts of approximately 2-7 mmol/g (Andjelkovic et al., 2006; Janoš et al., 2008). The PZC of humic substances, determined using the titration method, is highly acidic at pH <2. When the pH increases above the PZC value, these functional groups become deprotonated, resulting in a negative charge on the surfaces of the humic substances (Coles and Yong, 2006; Maguey-González et al., 2023). Therefore, humic substances have a strong affinity for clay mineral surfaces, often binding to positively charged sites via electrostatic bonding or ligand exchange to form surface complexes (Cornejo and Hermosín, 1996; Tombácz et al., 2004; Iorio et al., 2022). In the case of imogolite, humic substances form surface complexes by electrostatically binding to the Al₂OH₂⁺ sites on its outer walls, as shown in Figure 9b. This complex formation likely blocks the positively charged sites on the imogolite surface, thereby inhibiting the adsorption of dissolved anions such as IO₃⁻. Similarly, the complexation of humic substances with the FeOH₂⁺ sites of amorphous Fe-oxides can block these sites, which inhibits the adsorption of IO₃⁻ on the amorphous Fe-oxides. As a result, Kaimondake volcanic ash before decomposition of humic substances shows considerably lower IO₃⁻ adsorption compared to after H₂O₂ treatment. Besides, owing to their negative surface charge, humic substances have a strong affinity for dissolved cationic species (Tipping, 2002; Galicia-Andrés et al., 2021), enabling them to form surface complexes with dissolved cationic species such as Cs⁺. The humic substances present in the Kaimondake volcanic ash are primarily insoluble, which suggests that their complexation with Cs⁺ is likely to enhance the adsorption of Cs⁺ by the volcanic ash. In fact, the samples containing humic substances before H₂O₂ treatment showed significantly greater Cs⁺ adsorption than the samples after humic substance decomposition. This likely explains why the Kaimondake volcanic ash with humic substances demonstrates greater Cs⁺ adsorption than that after H₂O₂ treatment.

CONCLUSION

The mineralogy of the Kaimondake volcanic ash deposit and its adsorption characteristics of Cs⁺ and IO₃⁻ were examined before and after H₂O₂ and DCB treatments. The volcanic ash comprised approximately 14.9% humic substances in the bulk samples and 37.7% in the clay samples. The remaining inorganic components, excluding humic substances, were approximately 64 and 95% amorphous materials in the bulk and clay samples, respectively. These consisted mainly of imogolite, with a chemical composition of Al_3.8Si_2.1O₆(OH)₈, and amorphous Fe-oxides in a weight ratio of approximately 85:15.

The PZC value of the clay sample containing humic substances was pH 4.2, which increased to pH 8.7 after the decomposition of humic substances and then decreased to pH 6.5 after the removal of amorphous Fe-oxides. In the adsorption experiments, both bulk and clay samples with humic substances demonstrated the highest Cs⁺ adsorption, which significantly declined after H₂O₂ treatment but subsequently increased to nearly twice the level of the H₂O₂-treated samples after DCB treatment. In contrast, both bulk and clay samples containing humic substances showed the lowest adsorption of IO₃⁻ compared to the samples treated with H₂O₂ and DCB. However, the IO₃⁻ adsorption increased to the highest values after H₂O₂ treatment and then decreased to approximately half that of the H₂O₂-treated samples by subsequent DCB treatment.

These ion adsorption characteristics of the Kaimondake volcanic ash are mainly controlled by imogolite, but the humic substances and amorphous Fe-oxides coexisting with imogolite have significant effects on these ion adsorptions. Humic substances contribute to increasing Cs⁺ adsorption by electrostatically binding Cs⁺ to their negatively charged sites. Conversely, they significantly reduce IO₃⁻ adsorption, likely by forming complexes with imogolite and blocking the positively charged sites (Al₂OH₂⁺) on the outer walls of imogolite. Whereas, amorphous Fe-oxides reduce Cs⁺ adsorption, likely because the Fe octahedra preferentially bind to the negatively charged sites (SiO⁻) on the inner wall of imogolite. However, they also contribute to increased IO₃⁻ adsorption by electrostatically binding to the positively charged surface sites (FeOH₂⁺) of the Fe-oxides.

ACKNOWLEDGMENTS

The authors are grateful to Professor Emeritus K. Tomita of Kagoshima University for his advice on clay minerals in Kagoshima Prefecture. We would like to thank N. Yamada, Kagoshima University, for his help and cooperation. We also thank the anonymous reviewers for their useful comments and suggestions.

REFERENCES

•  Andjelkovic,  T.,  Perovic,  J.,  Purenovic,  M. and  Blagojevic,  S. (2006) A direct potentiometric titration study of the dissociation of humic acid with selectively blocked functional groups. Ecletica Quimica, 31, 39-46.
•  Arai,  Y.,  McBeath,  M.,  Bargar,  J.R.,  Joye,  J. and  Davis,  J.A. (2006) Uranyl adsorption and surface speciation at the imogolite-water interface: Self-consistent spectroscopic and surface complexation models. Geochimica et Cosmochimica Acta, 70, 2492-2509.
•  Bishop,  J.L.,  Rampe,  E.B.,  Bish,  D.L.,  Abidin,  Z., et al (2013) Spectral features were similar but not identical for allophane and imogolite. Clays and Clay Minerals, 61, 57-74.
•  Cismasu,  A.C.,  Michel,  F.M.,  Tcaciuc,  A.P.,  Tyliszczak,  T. and  Brown,  Jr, G.E. (2011) Composition and structural aspects of naturally occurring ferrihydrite. Comptes Rendus Geoscience, 343, 210-218.
•  Coles,  C. and  Yong,  R.N. (2006) Humic acid preparation, properties and interactions with metals lead and cadmium. Engineering Geology, 85, 26-32.
• Cornejo, J. and Hermosín, M.C. (1996) Interaction of humic substances and soil clays. In Humic substances in terrestrial ecosystems (Piccolo, A. Ed.). pp. 675, Elsevier Science, Amsterdam, 595-624.
•  Cradwick,  P.D.G.,  Farmer,  V.C.,  Russell,  J.D.,  Masson,  C.R., et al (1972) Imogolite, a hydrated aluminium silicate of tubular structure. Nature Physical Science, 240, 187-189.
•  dos Santos,  A.P.,  Diehl,  A. and  Levin,  Y. (2010) Surface tensions, surface potentials, and the hofmeister series of electrolyte solutions. Langmuir, 26, 10778-10783.
• Dove, P.M. and Rimstidt, J.D. (1994) Silica-water interface. In Silica, physical behavior, geochemistry and materials applications (Heaney, P.J. and Prewitt, C.T. Eds.). pp. 624, Reviews in Mineralogy, Vol. 29, Mineralogical Society of America, Washington DC, 259-308.
•  Dumat,  C. and  Staunton,  S. (1999) Reduced adsorption of caesium on clay minerals caused by various humic substances. Journal of Environmental Radioactivity, 46, 187-200.
•  Escudey,  M. and  Galindo,  G. (1983) Effect of iron oxide coatings on electrophoretic mobility and dispersion of allophane. Journal of Colloid and Interface Science, 93, 78-83.
• Farmer, V.C. (1974) The Infrared spectra of minerals. pp. 539, Mineralogical Society, Bartholomew Press, London.
• Farmer, V.C. and Fraser, A.R. (1979) Synthetic imogolite, A tubular hydroxyaluminium Silicate. In Developments in sedimentology, Vol. 27 (Mortland, M.M. and Farmer, V.C. Eds.). pp. 662, Elsevier, Oxford, 547-553.
•  Farmer,  V.C.,  Fraser,  A.R. and  Tait,  J.M. (1979) Characterization of the chemical structures of natural and synthetic aluminosilicate gels and sols by infrared spectroscopy. Geochimica et Cosmochimica Acta, 43, 1417-1420.
•  Galicia-Andrés,  E.,  Escalona,  Y.,  Oostenbrink,  C.,  Tunega,  D. and  Gerzabek,  M.H. (2021) Soil organic matter stabilization at molecular scale: the role of metal cations and hydrogen bonds. Geoderma, 401, 115237.
•  Gallez,  A.,  Juo,  A.S.R. and  Herbillon,  A.J. (1976) Surface and charge characteristics of selected soils in the tropics. Soil Science Society of America Journal, 40, 601-608.
•  Gustafsson,  J.P. (2001) The surface chemistry of imogolite. Clays and Clay Minerals, 49, 73-80.
•  Harsh,  J.B.,  Traina,  S.J.,  Boyle,  J. and  Yang,  Y. (1992) Adsorption of cations on imogolite and their effect on surface charge characteristics. Clays and Clay Minerals, 40, 700-706.
•  Iorio,  E.D.,  Circelli,  L.,  Angelico,  R.,  Torrent,  J., et al (2022) Environmental implications of interaction between humic substances and iron oxide nanoparticles: A review. Chemosphere, 303, 135172.
•  Ishikawa,  H.,  Arimura,  K.,  Oki,  K. and  Maruno,  K. (1979) ¹⁴C ages of the Ata pyroclastic flow and the Kaimon volcanic ash bed in the Kagoshima Prefecture. Journal of the Geological Society of Japan, 85, 695-697 (in Japanese).
• Jackson, M.L., Lim, C.H. and Zelazny, L.W. (1986) Oxides, hydroxides and aluminosilicates. In Methods of soil analysis, Part 1, Physical and mineralogical methods, 2nd Edition (Klute, A. Ed.). pp. 1188, Agronomy Monogram 9, American Society of Agronomy and Soil Science Society of America, Madison, 101-150.
•  Janoš,  P.,  Kříženecká,  S. and  Madronová,  L. (2008) Acid-base titration curves of solid humic acids. Reactive and Functional Polymers, 68, 242-247.
•  Kawano,  M. and  Tomita,  K. (2001) Microbial biomineralization in weathered volcanic ash deposit and formation of biogenic minerals by experimental incubation. American Mineralogist, 86, 400-410.
•  Kawano,  M. and  Tomita,  K. (2002) Microbiotic formation of silicate minerals in the weathering environment of a pyroclastic deposit. Clays and Clay Minerals, 50, 99-110.
•  Kawano,  M. and  Nanamura,  K. (2024) Allophanes in the weathered volcanic ash deposits distributed in southern Kyushu, Japan and their ion adsorption characteristics. Journal of Mineralogical and Petrological Sciences, 119, 240729.
•  Kikuchi,  S.,  Kashiwabara,  T.,  Shibuya,  T. and  Takahashi,  Y. (2019) Molecular-scale insights into differences in the adsorption of cesium and selenium on biogenic and abiogenic ferrihydrite. Geochimica et Cosmochimica Acta, 251, 1-14.
•  Kim,  Y.,  Kirkpatrick,  R.J. and  Cygan,  R.T. (1996) ¹³³Cs NMR study of cesium on the surfaces of kaolinite and illite. Geochimica et Cosmochimica Acta, 60, 4059-4074.
• Machida, H. and Arai, F. (2003) Atlas of tephra in and around Japan (revised edition). pp. 360, University of Tokyo Press, Tokyo (in Japanese).
•  Maguey-González,  J.A.,  de Jesús Nava-Ramírez,  M.,  Gómez-Rosales,  S.,  de Lourdes Ángeles,  M., et al (2023) Humic acids preparation, characterization, and their potential adsorption capacity for aflatoxin B1 in an in vitro poultry digestive model. Toxins, 15, 83.
•  Matsue,  N.,  Minagawa,  T. and  Yoshinaga,  N. (1991) Imogolite and Allophane in Shattered Sandstone. Soil Science and Plant Nutrition, 37, 171-173.
• Matsui, T., Kimata, M. and Nishida, N. (2003) Megacrystallization process of anorthite from Kaimon-dake volcano. Abstracts for Annual Meeting of the Mineralogical Society of Japan. pp. 154 (in Japanese).
• Mehra, O.P. and Jackson, M.L. (1960) Iron oxide removal from soils and clays by a dithionite-citrate system buffered with sodium bicarbonate: In Clays and clay minerals, Proc. 7th Natl. Conf., Washington, D.C., 1958 (Swineford, A. Ed.). pp. 369, Pergamon Press, New York, 317-327.
•  Michel,  F.M.,  Ehm,  L.,  Antao,  S.M.,  Lee,  P.L., et al (2007) The structure of ferrihydrite, a nanocrystalline material. Science, 316, 1726-1729.
•  Nagata,  T. and  Fukushi,  K. (2010) Prediction of iodate adsorption and surface speciation on oxides by surface complexation modeling. Geochimica et Cosmochimica Acta, 74, 6000-6013.
•  Parfitt,  R.L. (2009) Allophane and imogolite: role in soil biogeochemical processes. Clay Minerals, 44, 135-155.
•  Perdue,  E.M.,  Reuter,  J.H. and  Ghosal,  M. (1980) The operational nature of acidic functional group analyses and its impact on mathematical descriptions of acid-base equilibria in humic substances. Geochimica et Cosmochimica Acta, 44, 1841-1850.
•  Schwertmann,  U. and  Fechter,  H. (1982) The point of zero charge of natural and synthetic ferrihydrites and its relation to adsorbed silicate. Clay Minerals, 17, 471-476.
•  Staunton,  S.,  Dumat,  C. and  Zsolnay,  A. (2002) Possible role of organic matter in radiocaesium adsorption in soils. Journal of Environmental Radioactivity, 58, 163-173.
•  Stevenson,  F.J. and  Goh,  K.M. (1971) Infrared spectra of humic acids and related substances. Geochimica et Cosmochimica Acta, 35, 471-483.
•  Su,  C.,  Harsh,  J.B. and  Bertsch,  P.M. (1992) Sodium and chloride sorption by imogolite and allophanes. Clays and Clay Minerals, 40, 280-286.
•  Su,  C. and  Harsh,  J.B. (1993) The electrophpretic mobility of imogolite and allophane in the presence of inorganic anions and citrate. Clays and Clay Minerals, 41, 461-471.
• Tipping, E. (2002) Cation binding by humic substances, Cambridge environmental chemistry series-12. pp. 434, Cambridge University Press, Cambridge.
•  Tombácz,  E.,  Libor,  Illés,  E.,  Majzik,  A. and  Klumpp,  E. (2004) The role of reactive surface sites and complexation by humic acids in the interaction of clay mineral and iron oxide particles. Organic Geochemistry, 35, 257-267.
•  Vinkler,  P.,  Lakatos,  B. and  Meisel,  J. (1976) Infrared spectroscopic investigations of humic substances and their metal complexes. Geoderma, 15, 231-242.
•  Vogel,  A.,  Durant,  A.J.,  Diplas,  S.,  Azar,  A., et al (2017) Reference dataset of volcanic ash physicochemical and optical properties for atmospheric measurement retrievals and transport modelling. Journal of Geophysical Research: Atmospheres, 122, 9485-9514.
•  Volkov,  A.G.,  Paula,  S. and  Deamer,  D.W. (1997) Two mechanisms of permeation of small neutral molecules and hydrated ions across phospholipid bilayers. Bioelectrochemistry and Bioenergetics, 42, 153-160.
• Wada, K. (1989) Allophane and imogolite. In Minerals in soil environments (Dixon, J.B. and Weed, S.B. Eds.). pp. 1244, Soil Science Society of America, Madison, Wisconsin, 1051-1087.
•  Wada,  K. and  Yoshinaga,  N. (1969) The structure of imogolite. American Mineralogist, 54, 50-71.
•  Wada,  K. and  Greenland,  D.J. (1970) Selective dissolution and differential infrared spectroscopy for characterization of “amorphous” constituents in soil clays. Clay Minerals, 8, 241-254.
•  Wada,  K.,  Henmi,  T.,  Yoshinaga,  N. and  Patterson,  S.H. (1972) Imogolite and allophane formed in saprolite of basalt on Maui, Hawaii. Clays and Clay Minerals, 20, 375-380.
• Wahlberg, J.S. and Fishman, M.J. (1962) Adsorption of cesium on clay minerals. Geological Survey Bulletin, USA Government Printing Office, Washington, 1140-A.
•  Wang,  S.,  Zhang,  Y.,  Liu,  D.,  Yuan,  P., et al (2024) Adsorption behaviors and atomistic mechanisms of iodate and iodide on hollow spherical allophane nanoparticles. Applied Clay Science, 250, 107293.
•  Yoshinaga,  N. and  Aomine,  S. (1962) Imogolite in some ando soils. Soil Science and Plant Nutrition, 8, 22-29.
•  Zhuang,  J. and  Yu,  G.R. (2002) Effects of surface coatings on electrochemical properties and contaminant sorption of clay minerals. Chemosphere, 49, 619-628.