Thermoluminescence and electron spin resonance dating of freshwater fossil shells from Pa Toh Roh Shelter archaeological site in southern Thailand.

This research aims to compare the dating results by thermoluminescence in comparison with the electron spin resonance techniques. We also intended to resolve the discrepancy on the precision being measured between these two. Based upon the archaeological evidence, six portions of freshwater fossil shells, used as samples, contain calcium carbonate polymorphs acquired from six locations in Pa Toh Roh Shelter archaeological site, Khao Han cave, Satun province, Thailand. The approach requires two important quantities for accurate dating, namely the annual dose and the accumulated dose. Using neutron activation analysis, the annual doses of the shells from different depths were found to be 2.64±0.07 to 7.27±0.11 mGy per year. The accumulated dose is evaluated from calibrated TL and ESR intensity versus accumulated γ-dose. Linear regression was used to fit the dependence of TL intensity on dose with a linear saturation function. The TL intensity at 350∘C was used to estimate ages of the fossil shells from all depths. The ESR dating would be also available for aragonite calcite shells using the ESR signal of g = 2.0016. The accumulated dose was found to vary from 10.49±2.09 to 22.50±3.99 Gy and the calculated ages of the shells were between 3,094±551 to 4,479±666 years old. These results confirm that the freshwater fossil shells in the area date back to the Neolithic period are similar to those found in the nearby archaeological sites. In addition, the dates are in agreement with those results from the relative dating as reported by a government agency.

Introduction

Thailand has various archaeological sites located sporadically throughout the country, both prehistoric and historic. The study of cultural evolution in prehistoric times in southern Thailand was started by archaeologists as early as 1912. Archaeological evidence in the area dates back 40,000 years to the Pleistocene period in the Satun province in southern Thailand. The archaeological evidence indicates human settlement and ancient communities in the prehistoric time around 6,000-3,000 years ago, the Neolithic era. Moreover, there is direct evidence of human presence, namely human and animal bones, color paintings, earthenware, stoneware, and fossil shells [1], [2], [3]. Various items such as pottery, coarse-grained and fine-grained for both basic and decorative patterns on the surface of the container. This was also found in a striped rope pattern and a trace of soil watering, including a striped rope pattern and a trace of soil watering. There were also polished and smoked animal bones, both fired and unburned bone beads and shell bracelets. However, absolute dating is yet to be done in this area. There are a variety of absolute dating techniques that yield accurate dates for fossil shells, e.g., radiocarbon, electron spin resonance (ESR), thermoluminescence (TL), and optically stimulated luminescence (OSL) dating. Radiocarbon dating is a destructive technique that is widely used for dating fossils whose ages range from a few tens of years to tens of thousands of years. Unlike the luminescence techniques, it requires organic compounds. The number of unpaired electrons trapped by inherent crystal defects is critical for TL, OSL, and ESR. Nonetheless, while these unpaired electrons are caused by radiation exposure in the environment, they are evicted by other external sources of energy such as heat, light, and/or pressure. Therefore, it is essential to control our samples in the light of environmental protection [4].

Each of these techniques uses a different approach in counting the number of free electrons. In ESR, microwave radiation propagates through a sample and undergoes the absorption process causing the electron to change its spin state. Hence the intensity of transmitted microwave is correlated to the number of trapped electrons [5], [6], [7], [8], [9]. TL and OSL, on the other hand, are similar in their underlying processes, i.e., stimulated emission where trapped electrons in crystal undergo transitions from metastable states to the lower energy states as a result of thermal and optical excitation, respectively [6], [9], [10]. However, TL has advantages both in sample preparation and lower the instrumentation cost as opposed to those of OSL and ESR [11]. TL is commonly used to date pottery due to its composition, mainly comprising of quartz, that has a pronounced TL signal. While quartz shows a clear and strong signal, calcium carbonate also displays a strong signal.

Nevertheless, its contribution comes from two crystalline forms, calcite, and aragonite. While calcite is thermodynamically stable, aragonite can transform into calcite under the appropriate heating condition. The phase change is seen in a glow peak around 400∘C [12]. Even though these structures give off the thermoluminescence signals, the natural aragonite itself is a poor TL source [13]. Therefore, most TL studies on calcium carbonate have mainly relied on calcite. The glow curve of aragonite without pre-annealing has the most substantial peak around 340∘C These TL signals were found in shells, corals, and lithic objects. On the top of that these crystals were essential parts in determining the absolute age of these samples [14]. TL was possibly suggested for dating shells as early as 1967 [15]. The work focused on the feasibility of TL for dating by using mollusk shells as samples. Later in 1971, an attempt to utilize TL for defining the age difference of soil layers using paludina shells [16]. This study was aimed to understand the physics of sedimentary formation. One of the earliest TL-dating works dedicated to Pecten shells was conducted to compare with the ESR result. It showed that they were all in agreement with each other [17]. Besides, the results also revealed the feasible dating range, from years to recent. Nonetheless, the TL results done by the experts were on the dating of various shell species, including Pacific oysters, from multiple places in Japan [18], [19], [20]. Nonetheless, the TL dating is more suitable for structures with quartz found in pottery. While for shells or others with calcium carbonate structures like calcites or aragonites, ESR seems to provide a more precise result in a specific age range. However, several papers use both techniques, TL and ESR [17], [18], [19]. The degree of precision does not only depend on the structures of material but also the age range.

In this paper, we aim to measure a sample with TL compared to the ESR. This will answer the discrepancy between these two techniques, provided it will give us a guideline on how to use each type of measurement to determine the exact age of the material. We apply TL dating, and ESR is dating to aragonite-calcite crystals in freshwater fossil shells, e.g., a bracelet found on the female remains, in Pa Toh Roh Shelter archaeological site in Khao Han cave of Satun province, in southern Thailand. Scattered around that area are shell remains which are inappropriate for dating due to possible alteration of soil layers. Other grave goods also include potteries, animal bones, and stone ornaments. However, they are conserved by the government agency due to scarcity. On the contrary, tip-removed fossil shells were abundant and generally used as ornaments for noble ladies. The results of absolute dating found in the fossil shells indicate that the area is in the Neolithic era, around 10,200-4500 BC when people started to gather in a group of communities either temporarily or permanently. The outcome also agrees with a relative date given by an authorized agency, the 13th Fine Arts Department. Moreover, this result is also comparable to those nearby archaeological sites [3]. This study will help us fill up the gap of human being in the prehistoric period.

Historic site

The samples consist of freshwater fossil shells and surrounding sediments that were collected by personnel of the 13th Regional Office of Fine Arts Department Songkhla at the archaeological site of Khao Han cave, moo 7, Ban Pa Toh Roh, Khuan Don District, Satun Province, as shown in Fig. 1. The inset of Fig. 1 shows the map of Thailand, while Satun province is shown in red while the enlarged on the right is Khao Han cave indicated by a red star located 46 meters above the sea level at the geographic location 6∘48'20.2”N and 100∘6'20.3”E. The site was unearthed for the sample collection on May 27th, 2010. The excavation plan was displayed in Fig. 2. The inset of Fig. 2 shows a side view of the sample collecting area, where TP1 is the sample collection hole. Fig. 3 shows a close up of the female skeleton remain lying supine in the extended position oriented in the east-west direction. Locations and types of offerings are indicated in Fig. (a) and a real photo is illustrated in Fig. (b). This kind of inhumation can be found in prehistoric archaeological sites throughout mainland in southeast Asia. Grave goods being found included pottery, animal bones, fossil shells, stone flakes, stone tools, and indigenous soil. These suggested that the burial site was for a noble lady in the society at that time. Geographically, Khao Han Cave was embedded in a limestone mountain. The cavern is about 30 m long and 16 m in height. The north and south ends are connected to the shelter and natural forest. The eastern boundary is next to a natural forest and rubber farms, while the western border is adjacent to the cave. Nowadays this location is active with local villagers who earn their living by collecting bat guano. However, this exploration was carried out before the settlement of those villagers.

Figure 1

Archaeological Site of Khao Khan Cave, Satun province.

Figure 2

Location of Archaeological excavation site of Khao Khan Cave, Satun province [3].

Figure 3

The female skeleton lying supine in the east-west direction with its grave goods. (a) a sketch and (b) a photo of the remain are depicted with the positions of the offerings in the grave [3].

Materials and methods

Sample collection

Archaeologists have discovered large numbers of artefacts at Pa Toh Rock-Shelter archaeological site in Khao Han cave Satun province. Such archaeological evidence is found from the depth of 60-100 cm. It includes pottery fragments, beads, animal bones and fossil shells. Shell bracelets were discovered at the depth of 70 cm. As the excavation went deeper to around 80 cm, human bones were located as well as shell and lithic ornaments. The actual depth of the site is 160 cm but there were nothing achaeologically interesting.

The excavation site is displayed in Fig. 4(a). The site is 2 m wide and 2 m in depth. The excavation was carried out layer by layer in 10 cm steps. In this study the freshwater shells were used for dating. We collected samples at Levels 2-4. The samples were then labelled for simple identification, as shown in the first column of Table 1. In addition, the sample codes assigned by the officials from the regional office of fine arts department in Songkhla are also included in column 3 for future reference.

Figure 4

Freshwater fossil shells collected from Khao Han cave archaeological site. (a) the excavation area and (b) the freshwater fossil shells. [3].

Table 1

Sample information and codes used in this study and shown for reference, as assigned by The 13th Regional Office of Fine Arts Department Songkhla.

In Table 1, summarizes the samples (fossil shells) as shown in Fig. 4. SH1-3 are the sample codes located in level 2 at 70-80 cm in depth. SH4-5 are the sample codes located at the 80-90 cm in-depth level, and SH6 is the sample code at level 4 of 90-100 cm in depth. This table also found that level 2 has the most enormous amount of fossil shells in the same excavation site.

Preparation of specimens

Freshwater fossil shells from each site were divided into two parts. The first part was prepared under dim red-light for the determination of accumulated dose (AD). The second was prepared under ambient light for evaluating internal annual dose ().

The freshwater shells were washed and cleaned in an ultrasonic bath, followed by etching in 5% HCl for 1 hour to remove soil and the surface portion, which were exposed to α-particle irradiation in the sediment. Then they were cleansed with distilled water. After that, the samples were gently ground with a mortar, and the grains were sieved in order to obtain the fraction between 90-150 μm. The granules were again etched in 0.5% acetic acid for a few minutes in order to suppress spurious TL emission [21], [22]. Specimens with 90-150 μm particle size were then washed repeatedly in distilled water and allowed to dry at a temperature of 40∘C. All the sample preparation steps were performed under a dim red light. Each sample was divided into 9 aliquots, approximately 300-500 mg each, and all the aliquots were irradiated for TL measurements. Then Each sample was divided into 10 aliquots, approximately 0.50 g each, and all the aliquots were irradiated for ESR measurements.

Artificial γ-irradiation was carried out with a cobalt-60 source (GC-220E), which delivered 3.404 Gy/s of gamma-ray radiation in the Office of Atoms for Peace in Thailand. As for samples of the accumulated dose, nine artificial aliquots of freshwater fossil shells were exposed to γ-irradiation with 10 Gy increments from 0 to 80 Gy. The weight of a sample was about 150-250 mg. All the TL measurements were carried out at room temperature. The TL was induced by heating the sample at heating rate of 5 ∘C/s up to 400 ∘C in a high-purity nitrogen atmosphere. The emitted TL was recorded using a Harshaw-3500 TL reader. Before entering the detector, the photons were filtered by a neutral density filter with an optical density of 0.5–1.5. All the samples were analyzed for response at 350 ∘C due to the heating trap of aragonite [23], [24], [25]. As for samples of the accumulated dose, ten artificial aliquots of freshwater fossil shells were exposed to γ-irradiation with 10 Gy increments from 0 to 120 Gy. The weight of a sample was about 0.50 g. ESR measurements were performed at room temperature utilizing a pulsed ESR Bruker EMX premium spectrometer. The ESR spectra were recorded at 9.8 GHz (X-band) microwave frequency, 1.002 mW microwave power, field modulation of 0.1 mT.

Signal intensity from TL and ESR measurement for each accumulated known dose was used to construct a calibration curve. The accumulated dose is determined from the signal intensity assuming a linear calibration relationship: Here and I are the observed signal intensity before and after irradiation, respectively. AD is accumulated dose, D is the annual dose and t is irradiation time. The term represents the additive dose that is crucial for the determination of accumulated dose [5], [6], [8].

The freshwater shells grains were instead sieved with a 90 μm mesh sieve to obtain a particle size range of 0-90 μm. Uranium, thorium, and potassium contents in the shells were determined by neutron activation analysis (NAA) [26]. These contents allow us to calculate the internal annual dose ().

The surrounding sediment was prepared in an ambient light for external annual dose () determination. The sediment samples were gently ground with a mortar, and the grains were sieved with a 90 μm mesh in order to obtain the particle size range 0-90 μm. Neutron activation analysis (NAA) was used to estimate external annual and internal annual doses [26]. The sample codes for surrounding sediment were SD1-SD6. They were used for determining concentrations of U-238, Th-232, and K-40 in SH1-SH6. The results are displayed in Table 2.

Table 2

The contents of U-238, Th-232, and K-40 in nearby soil (SD) and shells (SH) required for estimating annual doses (D). Calculated D and D are included [50].

The annual dose (D) is composed of three parts, namely internal annual dose (), external annual dose (), and cosmic ray dose (). The internal annual dose is the result of radiation originated from U-238, Th-232, and K-40 contents in the shell itself, while the external annual dose is from radiation by the surrounding sediment. Each contribution was evaluated by measuring the concentrations of radioactive elements: decay products of U-238 series, Th-232 series, and K-40 are alpha (α), beta (β) and gamma (γ) radiation. Moreover, the cosmic ray dose () was also included in the annual dose. The cosmic rays were governed by longitude, latitude, and elevation of the area of interest and these three contributions were lessened by the depth of the sample [27]. The annual dose can be determined using Eq. 2 [8], [22].

The contribution of internal γ-rays can be neglected since the shells were too thin to absorb γ-rays emitted by themselves. For the internal annual dose calculation, we assumed the efficiency of the defect production (k-value) by α-particles in aragonite to be negligible because of the coarse grain nature of our samples. As a result, the internal dose solely depends on the β dose rate [5]. For estimating the external annual dose, alpha particles can also be neglected since the shell surface was etched [28], [29], [30], [31]. The internal and external annual dose rates can then be written as

The effects of moisture or water content of the surrounding sediment and shell must be taken into account since the materials were dried before the analysis [8], [22]. The dose rates of alpha (α), beta (β) and gamma (γ) (, and respectively) with the water content (%W) can be written as In Eq. (5), , and can be determined using Eq. (6). In the following equation, for example, is the concentration and , a.k.a. a conversion factor, [5], [28], is the α-particle rate emitted by U-238 [28], [29], [30], [31]. The water content in this work was zero since the weight of each specimen was constant.

The concentrations of those radioactive elements are then used to estimate and based on the online Dose Rate and Age Calculator (DRAC) using the conversion factors from G. Adamiec and M. J. Aitken [5], α- and β-grain size attenuation factors respectively from W. T. Bell [32] and B. J. Brennan, et al. [33], and β-etch attenuation factors from V. Mejdahl [34]. , the cosmic dose rate, is derived from the geographical location and elevation of the site. Its value for each collecting site is nearly identical since they are in the same cosmic exposure area, with the mean square deviation of 1% The combination of the three doses is the total annual dose of the fossil shells. The TL age of aragonite-calcite was estimated by dividing the accumulated dose (AD) by the annual dose (D).

Results and discussion

The X-ray diffraction (XRD) result

The structure of aragonite is orthorhombic, with higher density and better durability than those of calcite, while calcite is rhombohedral [35], [36], [37]. Typically, calcium carbonate has three structural polymorphs: vaterite, aragonite and calcite. All three polymorphs can occur simultaneously in some types of shells. In the nature, aragonite and calcite are the most common forms. Therefore, expectedly these freshwater shells mainly contain aragonite and calcite. Fig. 5 shows the X-ray diffraction (XRD) result of our shell sample. It consists of mostly aragonite and calcite. The aragonite structure dominates in the shell sample, This can be seen in the peak response of the glow curve.

Figure 5

The freshwater fossil shells are composed of aragonite and calcite. (Aragonite, A; Calcite, C).

The thermoluminescence (TL) results

The glow curves of a gamma-irradiated shell sample, SH6, are presented in Fig. 6(a). Dose usage ranges from 0 to 80 Gy. The inset, Fig. 6(b), shows the plateau, which could be derived from the ratio of the unirradiated sample to the irradiated ones at different dose level. The plateau shown in here is corresponded to the temperature between 325∘C and 375∘C. In this paper, we selected the signal responded to 350∘C to normalize the signal responses to the irradiation varying between 10 Gy to 80 Gy as shown in Fig. 7. All curves have peaked around 160∘C and 350∘C, where the maximum TL signal is at the latter, possibly the response from aragonite [38], [39]. The former is likely to be the response from puny calcite. However, these temperature responses are tricky to identify for their origin since many studies have stated differently [40], [41], [42], [43], [44]. In many cases, shell samples include not only of calcite, aragonite, and vaterite, but also dolomite (CaMg(CO3)2), strontianite (SrCO3), or smithsonite (ZnCO3), in addition to additional minerals that cannot be identified by XRD (its threshold is 3-5 percent). All of them (dolomite, strontianite, and smithsonite) exhibit a significant TL emission that is consistent with the CGD results of the Ca-rich samples that were used in this study, with maxima peaked at 175∘C, 275∘C, and 350∘C (as shown in Fig. 6). [45], [46], [47] Deconvoluting all glow curves yield three peak responses around 175∘C, 275∘C, and at 350∘C. A glow-curve deconvolution (GCD) for general orders of the kinetics is then employed to fit the curves based on those temperatures [48]. The area under the curve of such peak gives the total emitted photons. Plotting the area under the curve at 350∘C against the additive doses allows one to determine the accumulated dose, as shown in the x-interception, according to Eq. (1). The linear relation between the mass normalized intensities and the additive doses dominates in all samples. We, hence, obtain the accumulated dose for the samples as illustrated in Fig. 7.

Figure 6

(a) The TL spectra of laboratory irradiated SH6 sample with 0 to 80 Gy doses. (b) shows the plateau, which could be derived from the ratio of the unirradiated sample to the irradiated ones at different dose levels. (c) A glow curve deconvolution (GCD) for general orders of the kinetics is then employed to fit the curves based on 175∘C, 275∘C, and at 350∘C.

Figure 7

TL analyses of accumulated dose (AD) for SH1-SH6 are displayed from (a) to (f), respectively, based on TL responses at 350∘C.

Electron Spin Resonance (ESR) Result

Response results on freshwater fossil shells powdered samples from electron spin resonance, the ESR spectra corresponding to a laboratory radiation dose of 0 Gy and 120 Gy are shown in Fig. 8 The characteristic signal for dating is observed at about g = 2 (Field Center 348 mT)in the middle of sextet of hyperfine line. Scan range is 70 mT and 1.002 mW in microwave power.

Figure 8

EPR spectrum of the shell samples after irradiated up to 120 Gy and ESR spectrum of irradiated shell (120 Gy) showing the dating signal and the A, peak-to-peak amplitude used to build the growth curve.

The g-factors and assignments of these centers are related to orthorhombic CO (g = 2.0016 and g = 1.9973). The signal with g=2.0057 is related to the free rotating SO, the g=2.0031 is related to isotropic SO, while for g = 2.0056 is related to free rotating CO. The signal used for the dating is that at g=2.0016. Our results are in good agreement with those of previous works [22], [49] Fig. 8. The signal amplitudes of these centers increase with dose. According to the X-ray diffraction results, Fig. 9 shows the peak-to-peak amplitudes at g=2.0016 (CO) versus dose. The CO line was used as an indicator of dose because this signal is characterized by a long lifetime. Usually, the mean lifetime of the trapped electrons for the signal at g = 2.0016 is estimated to be about 2x108 years, the average temperature of 15C. This estimation is good enough for equal to/more than one million years [22] (Table 3).

Figure 9

Growth curve in shell sample (SH1): experimental, show the best fit curve. The intersection of the best fit curves with the x-axis provides the values of the accumulated dose which are reported in each subplots.

Table 3

Cosmic dose rate (D), annual dose (D), accumulated dose (AD) and TL age as a result of the online trapped charge dating [50].

Neutron Activation Analysis (NAA) result

Table 2 gives a summary of all the data from experiment in comparison with the calculation [50]. The radionuclide contents, U-238, Th-232, and K-40, are the results of the NAA performed on the surrounding soil and shell samples. The abundance of uranium in earth crust ranges from 1.5 to 6 ppm, with an average of around 2.8 ppm [51]. For thorium abundance, the range is from 6 to 20 ppm, with an average of 10 ppm, the percentage of potassium abundance is generally from 0.24 to 2.23, with an average of 1.18 [52]. Our NAA results also yield accordingly. On the average, the uranium being found in soil samples is 2.27 ppm, 10.42 ppm thorium, and 1.93% for potassium. These numbers are in reasonable agreement with those of earth crust. Though, the amount of thorium and uranium in SD1 and SD4 show a distinct difference from others. There are possibilities that the soil sample may contain earthenware debris that might carry additional uranium and thorium as indicated in the values of . The amount of potassium, on the other hand, is slightly more than that of the world average because the soil samples are partly clay of which is a good source of potassium [53]. for SD5 is significantly higher than those of others since the radionuclides 3.02 for uranium, 7.75 for thorium, and 5.70 for potassium at this location are more concentrated, 6.39 mGy/y, than most, particularly potassium. The impact of this phenomenon mainly causes the predicted age to be younger, even though it was on the same layer as SD4, as seen in Table 2. For the fossil shell samples, SH1-6, all radionuclides are markedly low since natural food uptakes of the shells in the area contain minute amounts of the radioactive nuclei [54].

The electron spin resonance (ESR) results were compared with the TL (Table 4). In this research, ESR techniques were applied to fossil shell samples SH1 and SH2 to compare these with TL techniques. Plotting the peak-to-peak amplitudes at g=2.0016 (CO) against the additive doses allows one to determine the accumulated dose, as shown in the x-interception. For SH1, the accumulated dose was 14.91±0.92 Gy and 11.32±1.09 Gy for SH2, result of age value with the ESR technique are shown in Table 4.

Table 4

Comparison between TL and ESR ages.

Because fossil shell behaves as an open system, the radioactive disequilibrium is exhibited Preusser et al. [55]. The leached elements can produce the over-estimation of dose rate. Therefore, ages in this study were possibly underestimated. Nevertheless, our dating results show that the excavation site dates back to about the mid-second millennium to the late beginning of the first millennium BC, the New Stone Age or Neolithic era, in agreement with the archaeological evidence indicating the sedentary settlements. Though animal farming and agriculture are believed to exist in this era, we could not find any clear evidence to support this. The lack is common in most Neolithic sites located near caves or rock shelters in the mainland of Southeast Asia [56]. They were used as campsites, asylums, or burial grounds for foragers since the findings incorporated charred pottery fragments, bones, and a female skeleton, as well as an infant remains [57]. Occupations were likely occasional. This argument is in contrast to that given in the government excavation report that states the area was residential [3].

Conclusion

The aragonite and calcite in freshwater fossil shells from Pa Toh Roh Shelter archaeological site in Khao Han cave, Satun province, in southern Thailand, were studied by Thermoluminescence (TL) dating and Electron Spin Resonance (ESR) techniques. The TL sensitivity is better than the ESR by one order of magnitude. The TL applies to date fossil calcite shells younger than year [19]. Therefore, the age of fossil shells found in this paper is in the order of thousand years. The result from TL is quite reliable based upon our data collections which are reasonably sufficient for the precision in this measurement.

The samples were found to be to 4,479±666 years old for SH1-SH6. These age estimates are in agreement with those reported by the 13th regional office of the fine art department, Songkhla, Thailand.

The results from ESR and TL here also agree with another relative dating in the archeology [3]. The age of the freshwater shells at the archaeological site of Khao Han Cave, Satun Province, was found to be in the Neolithic period. Within this approach, the dating of the fossil shells at the same level as the human bones should be a descent criteria to conclude it was in the Neolithic epoch. In other words, we investigate the dating of the fossil shell to infer the human epoch in our study. Furthermore, the evidence found at this archaeological site, such as the female skeleton, potteries, and personal ornaments, also confirms that the site is in the New Stone Age. The settlement, however, might be occasional. The area was possibly a regular campsite or asylum for hunting expeditions since it has an abundant natural resource close to the water reservoir [3], [58]. Besides, such locations, rock shelters, or caves were typically used as burial grounds. The prehistoric period we obtained from the measurement is in accordance with those found in nearby provinces in Thailand, even in Malaysia.

Declarations

Author contribution statement

Tidarut Vichaidid: Conceived and designed the experiments; Performed the experiments; Analyzed and interpreted the data; Contributed reagents, materials, analysis tools or data; Wrote the paper.

Pathipat Saeingjaew: Performed the experiments.

Funding statement

Tidarut Vichaidid was supported by Prince of Songkla University.

Data availability statement

This research is ongoing and not complete yet and therefore not to be disclosed unless being published elsewhere at the moment.

Declaration of interests statement

The authors declare no conflict of interest.

Additional information

No additional information is available for this paper.