The potential of NAA and ICP-MS for determining soil dispersibility in relation to landslide incidents

Dispersive soils pose significant challenges due to their inherent tendency to rapidly respond to the presence of water, resulting in erosion and heightened susceptibility to landslides. The inclusion of sodium (Na) onto the surface of soil clay particles results in the attenuation of the electrochemical interactions between the dispersed soil, leading to the repulsion and subsequent migration of the clay particles away from one another. This study employed Neutron Activation Analysis (NAA) and Inductively Coupled Plasma Mass Spectrometry (ICP-MS) to assess the Sodium Adsorption Ratio (SAR) and Exchangeable Sodium Percentage (ESP) as indicators of soil dispersibility. The results indicate that soils containing montmorillonite clay mineral exhibit greater values of SAR > 11 and ESP > 20 in comparison to soils containing other clay minerals i.e. kaolinite, illite, and vermiculite. The calculated SAR and ESP values obtained from both analyses demonstrate a notable agreement, indicating a positive correlation. In conclusion, it is apparent that both NAA and ICP-MS method has the capability to be applied for the purpose of predicting soil dispersibility in landslide-prone regions of Malaysia.


Introduction
Dispersive soil or known as sodic soil is a soil type that poses challenges due to its propensity to readily interact with water, primarily attributed to the presence of sodium ions (Na + ) generated by mineralogy of the parent material, topography, and water table quality [1].While it is commonly believed that the occurrence of dispersive soil is limited to region with arid or semi-arid climates, other countries globally have encountered challenges associated with this phenomenon [2,3].Dispersive soil has been reported in various tropical nations, including Thailand, Vietnam, Indonesia, and Malaysia [4,5,6,7].Several regions of Malaysia, prior studies have documented the presence of this particular type of soil in Batu Gajah, Wangsa Maju, Shah Alam and East Sabah.These places are distinguished by the prevalence of high clay content and clay particles [6,8,9].The susceptibility of dispersive soil to develop erosion renders it a contributing factor to heightened slope instability.This erosion can manifest in the form of internal erosion or pipe erosion, which is often triggered by soil cracking [10].Consequently, the IOP Publishing doi:10.1088/1757-899X/1308/1/012018 2 occurrence of landslides is exacerbated [11].The desiccation process leads to the development of crusts and fracture networks, which subsequently diminish the pore area on the soil surface.This reduction in pore area has significant implications for several soil properties and processes, including infiltration rates, surface water runoff, bearing capacity, and permeability.Previous research has indicated that the dispersion of montmorillonite minerals in soil is influenced by the presence of sodium (Na) derived from mineral sources rich in Na-plagioclase, such as albite [12].The presence of Na-montmorillonite in dispersive soil leads to the development of a second electrical layer on its surface.This phenomenon has the effect of reducing the bonding strength between mineral layers and the interparticle interactions within the lamellar structure [13].
Dispersive clay soils constitute a distinct group of fine-grained soils that provide challenges in their identification using visual categorization or physical examination techniques, such as consistency tests and particle size analysis [14].The crumb and double hydrometer tests are well recognized as straightforward and preliminary techniques for evaluating the dispersive characteristics of soil.The pinhole test provides distinct results for the categorization of soils into dispersive and non-dispersive groups, which is determined by analyzing the effluent and diameter of the drilled specimen.However, the results of the tests resulted in a certain degree of uncertainty in the categorization procedure, primarily due to the presence of sand-clay particles percentage.In instances of this nature, it becomes imperative to employ chemical testing to validate the dispersion characteristics within the soil.Therefore, the assessment of soil dispersibility can be accomplished through the application of Soil Exchangeable Sodium Percentage (ESP) and Sodium Adsorption Ratio (SAR).Soils can be characterized as dispersive when the SAR > 13 and ESP > 15 [15].To provide more clarification, the objective of this research endeavor is to assess the viability of utilizing NAA (Neutron Activation Analysis) and ICP-MS (Inductively Coupled Plasma Mass Spectrometry) techniques for the purpose of ascertaining the dispersibility of soil by estimating the SAR and ESP values.Both analytical techniques are commonly employed to quantify the constituents in diverse applications.NAA was seen as a comparatively more convenient technique for handling soil samples in comparison to ICP-MS, as the latter entails the utilization of potent acids during the sample preparation phase.NAA enables the concurrent and precise identification of several elements with a minimal quantity of solid matrix, in contrast to ICP-MS, which is better suited for analyzing liquid samples (dissolved solid) [16,17].

Material and methodology
The process flow in the present study is depicted in Figure 1.A collection of 10 soil samples with disturbed clayey composition was obtained from locations that experienced landslides.The selection of these samples was based on the relative abundance of four primary clay minerals identified in the soil, namely kaolinite (K), illite (I), montmorillonite (smectite) (M), and vermiculite (V), as determined using mineral phase characterization.According to the results of physical tests, specifically the crumb and pinhole dispersion test, the montmorillonite soil samples were categorized as dispersive soil, indicating that they exhibited a tendency to disperse when in contact with water.In contrast, the other soil samples were classed as non-dispersive soil, as they did not exhibit any reaction when exposed to water.The soil samples underwent a process of oven-drying at a temperature of 105°C for a duration of 24 hours.Following this, the samples were crushed and then sieved.The subsequent step involved the preparation of the samples for NAA and ICP-MS in order to quantify the elemental composition and ascertain the SAR and ESP, thereby reaffirming the observed dispersibility phenomena.

2.1.Neutron activation analysis (NAA)
NAA is a very sensitive analytical technique that enables the reliable and precise determination of constituent concentrations in soil samples.The soil samples, each weighing 1 g, were prepared and thereafter placed in polyethylene tubes prior to the irradiation process.The specimens, along with standard reference materials (SRMs) with predetermined elemental concentrations, are subjected to irradiation by thermal neutrons within a nuclear reactor.The irradiation of all samples and SRMs was performed in the 500 kW PUSPATI TRIGA Mark II research reactor at the Malaysian Nuclear Agency.
The reactor provided a thermal neutron flux of 4.0×10 12 n cm -2 s -1 .Following a suitable decay time, the utilization of high-resolution gamma ray spectroscopy is employed to quantify the intensity and energy of gamma lines generated by radioactive isotopes.A comparative analysis was conducted between the induced activity in the SRMs and the sample in order to ascertain the elemental abundances.The concentrations of sodium (Na), magnesium (Mg), and calcium (Ca) were measured in milliequivalents per litre (meq/L).

Inductively coupled plasma mass spectrometry (ICP-MS)
Soil samples to be analyzed by ICP-MS require acid digestion to ensure element dissolution and extraction.This study used 0.5 g of dry soil powder that was dissolved in 2 mL hydrochloric acid, 3 mL nitric acid and 5 mL hydrofluoric acid in a vessel and allowed to react for 15 minutes before sealing it.The vessels were transferred to a vessel liner and ramp at 175°C for 15 minutes before holding for another 15 minutes and thereafter cooling them for 30 minutes.The digested content was then emptied to a plastic container, labeled and stored at 4°C for analysis purpose.

Sodium adsorption ratio (SAR) and exchangeable sodium percentage (ESP)
The SAR and ESP play a crucial role in assessing the existence of soil dispersion resulting from the presence of exchangeable sodium.The calculation of the SAR was performed using the measured concentrations of sodium (Na), calcium (Ca), and magnesium (Mg) in milliequivalents per millilitre (meq/L), as described by Figure 1.The SAR does not incorporate monovalent potassium (K) and assumes an equal treatment of magnesium (Mg) and calcium (Ca) [17].ESP was estimated by calculating the ratio of exchangeable sodium (Na) to the cation exchange capacity (CEC) measured in milliequivalents per 100 grams (meq/100g) using the methylene blue spot method [18].

Results and discussion
Table 1 displays the concentration of Na, Mg and Ca measured by NAA and ICP-MS.From the results, NAA detect Na and Mg for all samples except for Ca while ICP-MS is less sensitive to Na compared to Mg and Ca elements.This is probably by some elements in soil particles were not entirely dissolved by the acid mixtures.The results of SAR and ESP calculated from NAA and ICP-MS elemental quantification are presented in Table 2 and Figure 2. Montmorillonite soil samples mineral exhibit greater values of SAR > 11 and ESP > 20 in comparison to soils containing other clay minerals i.e. kaolinite, illite, and vermiculite.The results obtained from the standard test provide further confirmation that the soil exhibits increased dispersibility as a result of the presence of Na-plagioclase rich minerals.Results indicated a change in the SAR and ESP in the following order: montmorillonite, illite, kaolinite, vermiculite.These expanding clay minerals have a greater tendency to disperse compared to nonexpanding clay minerals.On the other hand, vermiculite exhibits a lower sensitivity to dispersion as a result of its relatively low Na content compared to other soluble salts (Mg and Ca) in the soil.Low Na concentrations increase the attractive Van der Waals interactions between soil particles and decrease osmosis potential, minimising soil erosion.Figure 3 shows the correlation of SAR and ESP value determined by NAA and ICP-MS.From the graph, the pearson correlation matrix value for SAR and ESP of both analytical methods are r = 0.9897; R 2 = 0.9795 and r = 0.93917; R 2 = 0.88205, respectively.The value showing a good agreement, whereas most of samples fall in the same categories (non-dispersive to highly dispersive).This is noteworthy considering the vast range of element concentrations observed within each respective method.

Conclusion
In conclusion, the utilization of NAA and ICP-MS can be employed as analytical techniques to assess the soil dispersibility caused by the presence of Na in soil.The calculated SAR and ESP values obtained from both analyses demonstrate a good agreement, indicating a positive correlation.The observed variations in element concentrations throughout the samples can be attributed to a variety of factors, including sample preparation techniques and instrumental limitations such as partial sample dissolution and detection limits.Soils that are predominantly composed of montmorillonite clay minerals typically exhibit higher values SAR > 11 and ESP > 20, when compared to soils containing other clay minerals such as kaolinite, illite, and vermiculite.This can be attributed to the presence of Na-plagioclase rich minerals in the soil.In addition to topography and hydrology, the onset of landslides is also impacted by soil factors, particularly the presence of Na in the soil, which enhances the erodibility of slopes.

Figure 1 .
Figure 1.Flow of methods in this study.

Figure 2 .
Figure 2. Relationship between ESP and SAR value from NAA and ICP-MS to soil dispersibility, respectively.

Figure 3 .
Figure 3.The figure illustrated: a) Correlation between SAR value from NAA and ICP-MS analysis; and b) Correlation between ESP value from NAA and ICP-MS analysis.

Table 1 .
Concentration of elements and CEC values.

Table 2 .
Value of SAR and ESP calculated from elemental detection by NAA and ICP-MS Note: K -Kaolinitic soil; I -Illite soil; M -Montmorillonite soil; V -Vermiculite soil