Characterization of microstructural features of Tamusu mudstone

Hongdn Yu,Chen Lu,b,Weizhong Chen,*,Dinsen Yng,**,Honghui Li

a State Key Laboratory of Geomechanics and Geotechnical Engineering,Institute of Rock and Soil Mechanics,Chinese Academy of Sciences,Wuhan,430071,China

b University of Chinese Academy of Sciences,Beijing,100049,China

c China National Nuclear Corporation,Key Laboratory on Geological Disposal of High-level Radioactive Waste,China Institute for Radiation Protection,Taiyuan,030006,China

Keywords:Tamusu mudstone Pore structure Field emission scanning electron microscopy (FE-SEM)N2 physisorption Mercury intrusion porosimetry (MIP)

ABSTRACT Tamusu mudstone formation,located in the Alxa area in western Inner Mongolia,is considered a potential host formation for high-level radioactive waste (HLW) underground disposal in China.In this study,complementary analyses with X-ray diffraction (XRD),field emission scanning electron microscopy(FE-SEM),mercury intrusion porosimetry(MIP),and N2 physisorption isotherm were conducted on the Tamusu mudstone to characterize its physical characteristics and microstructural features,such as mineral compositions and pore structure.Several minerals,including carbonates,feldspar,clays and analcime,were identified in Tamusu mudstone by XRD.Images from FE-SEM show that pores in the Tamusu mudstone were dominantly on nanometer scale and generally located within their mineral matrix or at the interface with non-porous minerals.The combination of the MIP and N2 physisorption curves indicated that the Tamusu mudstone has diverse pore sizes,a porosity varying from 2.34% to 2.84%,and a total pore volume in the range of 0.0065-0.0222 cm3/g with the average pore diameter ranging from 9.6 nm to 19.23 nm.The specific surface area measured by MIP (2.572-5.861 m2/g) was generally higher than that by N2 physisorption (1.29-3.04 m2/g),due to the pore network effect,pore shape (e.g.ink-bottle shape),or technique limits.The results related to pore information can be applied as an input in the future to model single-or multi-phase fluid flow and the transport of radionuclides in porous geomedium by migration and diffusion.

1.Introduction

In many countries,deep geological disposal is a promising and widely accepted option for managing high-level radioactive waste(HLW) (Bish,1999; ˊSlizowski et al.,2003; ONDRAF/NIRAS,2013;Charles et al.,2013; Andra,2015; Hart et al.,2015; McEvoy et al.,2016; Bossart et al.,2017; Zhou et al.,2020).To contain or isolate HLW from human and natural dynamic processes for a long duration,a multi-barrier with an engineered barrier and a geological barrier is often recommended(Kang et al.,2003;SKB,2004;Li et al.,2006; Zhang et al.,2006).The natural host formation,i.e.the last barrier,is a vital component of this multi-barrier system.Mudstones,which account for about 66%of all sedimentary rocks in the Earth’s crust,are considered natural barriers for HLW repositories(Kim et al.,2011; Lahn et al.,2020; Lu et al.,2022).Several countries,including Belgium,Switzerland,France and Netherlands,choose mudstones as their potential host media for the HLW disposal (Li et al.,2006; Corkuma and Martin,2007; Hemes et al.,2015; Andra,2015; Plúa et al.,2021).In China,besides the Beishan granite (Chen et al.,2017; Zuo et al.,2017; Wang et al.,2018; Zhao et al.,2018; Sun et al.,2020a; Wu et al.,2021),the Tamusu mudstone,located in western Inner Mongolia,is a preselected host medium for the HLW disposal repositories(An and Di,2016; He et al.,2019; Xiang et al.,2020).

In mudstones,pore sizes are usually distributed in a wide range,i.e.from a few nanometers (＜2 nm) up to several hundred nanometers or even micrometers (both mesopores and macropores)(Tsakiroglou et al.,2009;Loucks et al.,2009,2012;Kim et al.,2011;Ko et al.,2016).Their pore structures,including pore size distribution(PSD) and pore morphology,are directly related to grain size,composition and arrangement of the matrix minerals,degree of diagenesis,etc.(Borst,1982; Yang and Aplin,1998; Desbois et al.,2010).For HLW underground disposal,to gain a deeper understanding of the fluid-geomaterial interaction and transport properties,the characterization of the relationships between nanoand micro-structures and macro-properties is essential.The key points in relating porosity and hydraulic properties are the geometrical (e.g.shape,pore size and distribution) and topological(e.g.tortuosity,connectivity and fractality)properties of the pores,as well as the associated pore connectivity and percolating network.With the rapid development of science and technology,we can visualize the nanoscale pore structure in mudstones.Various indirect and direct methods allow for this visualization.Indirect methods include mercury intrusion porosimetry(MIP),gas physisorption(CO2or N2),nuclear magnetic resonance,helium porosimetry,small-angle and ultrasmall-angle neutron scattering (SANS/USANS),and spontaneous imbibition; direct methods include computed tomography(CT),scanning electron microscopy,X-ray microtomography,and helium ion microscopy (Ougier-Simonin et al.,2016; Yang et al.,2017;Sun et al.,2020b;Zhang et al.,2020).

With a growing concern in nuclear waste underground disposal,much research on the microstructure in mudstones has been reported with great advancements.In China,a significant focus has been on the safety assessment of HLW disposal sites located within granite since 1985(Wang,2014).However,research on mudstones is still rare concerning HLW underground disposal.In addition,little work has been done on the microstructure analysis of Tamusu mudstone.Therefore,X-ray diffraction (XRD),field emission scanning electron microscopy (FE-SEM),MIP,and N2physisorption analyses are applied to investigating the microstructural features of Tamusu mudstone,in particular the morphology and structure of the pore space,representing end-members concerning mineralogy and the distribution of grain size for this formation.

2.Materials and methods

2.1.Sample collection

Core samples were taken from a vertical borehole deeper than 500 m in the Tamusu area,a preselected disposal site for HLW in the west of Alxa,Inner Mongolia(Fig.1).A small region of volcanic outcrops in the Permian period,lacustrine units in the Late Mesozoic era,and Holocene Aeolian deposits in the Cenozoic era constitute the strata of the Alxa Block(Du et al.,2017).The Tamusu mudstone in this study is dark gray with a natural water content of roughly 1% and a natural density of approximately 2.5 g/cm3.This material is found in the upper member of the Bayingebi Formation and belongs to the lacustrine units in the Late Mesozoic era.The test samples were cut from massive and well-preserved cores.

2.2.Mineralogy

A hammer was used to obtain the rock slices.The cores of the rock slices were ground and passed through a 200-mesh sieve and then separated for clay minerals by a gravity method.Using a Bruker instrument (D8 Discover X-ray diffractometer) with a step of 0.02°from 2°to 70°,XRD analyses were performed on the powdered Tamusu mudstones and clay minerals at a temperature of 20°C.Jade 6.5 software was employed to semi-quantify the bulk mineral composition and the relative content of clay minerals.

2.3.FE-SEM

For the FE-SEM analysis,samples were first roughly cut to the size of cubic centimeter.Subsequently,Ilion II(Gatan 697)was used to smooth the samples’ surfaces.Finally,the microstructure morphology of the Tamusu mudstone was observed with FE-SEM(Hitachi SU8010); the surface of the sample was coated with a 10-nm thickness platinum layer to enhance the electrical conductivity.

2.4.MIP

Fig.1.Geological map of the Yingejing sub-basin and the borehole location (marked by a red multi-pointed star) (after An and Di,2016; Wei and Jiang,2019).

In the MIP analysis,mudstone samples were first processed into 1-cm3cubes,then oven-dried for at least 48 h at a temperature of 60°C to remove the volatile matter and moisture,and finally cooled to ambient temperature(around 23°C)in a desiccator with relative humidity less than 10%.MIP tests were conducted on a micromeritics AutoPore IV 9510 at pressures up to 413 MPa(60,000 psia),equivalent to a minimum limit detection of a pore throat diameter of roughly 3 nm according to the Washburn equation (Katz and Thompson,1986).The distribution of pore-throat diameter and total pore volume could be directly obtained through the physical constants for mercury on geomaterials,where the surface tension and the contact angle are 485 mN/m and 130°,respectively(Zhang et al.,2016; Sun et al.,2020b).

2.5.N2 physisorption

Before the experiment,mudstone samples were crushed through a 40-80 mesh sieve and degassed for about 5 h under vacuum at approximately 110°C to remove adsorbed moisture and capillary water.At-195.85°C,the pressure was increased to the saturated vapor pressure of N2and then gradually decreased.N2adsorption/desorption isotherms were obtained through the measurement of the N2adsorption under different relative pressures P/P0(P is pressure,P0is initial pressure).The N2physisorption experiments were carried out with an Automatic gas adsorption analyzer (Autosorb iQ).

3.Results and discussion

3.1.Mineral compositions of Tamusu mudstone

The mineral compositions of Tamusu mudstone(at the depth of 505-534 m) were measured by XRD,and the typical results are shown in Fig.2,Tables 1 and 2.The main mineral compositions of Tamusu mudstone were found to be carbonates (dolomite,ankerite),feldspar (plagioclase,K-feldspar),clays and analcime.In addition,small portion of quartz,calcite,pyrite and anhydrite were also detected in the Tamusu mudstone.Carbonate minerals are abundant in Tamusu mudstone.The total amounts of both dolomite and ankerite for each sample varied from 21.4%to 45.9%.Xiang et al.(2020) analyzed the spatial distribution (geological homogeneity)and the mineralogical characteristics of Tamusu mudstone at the depth of 381.82-800.6 m.Their study showed that the content of dolomite and ankerite was in the range of 3%-50%for one borehole.Our results clearly fell into this zone.Among the clay minerals,illite and illite-smectite mixed layers were the most common.According to Misaelides (2019),analcime in Tamusu mudstone may be beneficial for radionuclide retention during the HLW underground disposal,owing to its ion-exchange properties.

Fig.2.XRD patterns of a typical Tamusu mudstone sample.

Based on the terminology and classification for mudstones from Gamero-Diaz et al.(2013),the normalized WCAR(total carbonate),WQFM (quartz-feldspar-mica),and WCLA (total clay) data for Tamusu mudstone on average are about 44%,35% and 21%,respectively.Following the conventions for naming the mudstone facies proposed by Gamero-Diaz et al.(2013),the Tamusu mudstone in this study is intermediate between mixed carbonate and carbonate siliceous mudstones.In the study of Xiang et al.(2020),the Tamusu mudstone was named a mixed mudstone.In addition to the different specific locations of the drilling boreholes,the differences in the definition of Tamusu mudstone between Xiang et al.(2020) and this study might also be attributed to the classification methods.Although the two classification methods were similar,the axes of the triangle diagram for Xiang et al.(2020)were QFM (quartz-feldspar-mica),carbonate and analcime,while those of this study were WCAR,WQFM and WCLA.

3.2.Pore morphology

The PSD of the studied Tamusu mudstone covers a wide range,from a few nanometers to tens or even hundreds of micrometers(Fig.3).Following the standard of the International Union of Pure and Applied Chemistry (IUPAC) (Sing et al.,1985),pores are classified as micropores,mesopores and macropores according to their diameter dimensions.As defined by Rouquerol et al.(1994),the diameters of micropores,mesopores and macropores are less than 2 nm,2-50 nm,and greater than 50 nm,respectively.As indicated in Fig.3,the dominant pores in the Tamusu mudstones are mesopores.

There are various types of nano-scale pores in the Tamusu mudstone.Based on the relationships between pores and particles,Loucks et al.(2012) proposed that pores in mudstones can be classified into three categories: interparticle,intraparticle and organic-matter pores.Generally speaking,interparticle pores in Tamusu mudstone were usually developed between mineral particles(Fig.3),such as crystalline particles and clay minerals.These clay-related pores were commonly distributed along the boundary of the particles,as shown in Fig.3b.A considerable number of intraparticle pores were developed in the clay mineral layers,especially in the illite and illite/smectite mixed interlayers,along with numerous inter-crystalline pores.Many interparticle or intraparticle pores were filled with carbonate or silica cement,as shown in Fig.3c.The absence of organic matter in the analysis of mineral composition indicates that the content of organic matter pores in the Tamusu mudstone might be low.Fortunately,the analysis of pore morphology by FE-SEM tests located these pores.Some isolated pores with diameters up to tens of nanometers were located either within or at the edges of the organic matter that infilled pre-existing interparticle pores or microfractures.Fig.3d also shows that various pore shapes were observed,including triangular,linear,ellipsoid and wedge-shaped pores.The intrapores in clays were commonly sheet-like and parallel to one another(Fig.3b),while the shape of the intrapores in organic matter was usually elongated,ellipsoidal and irregular (Fig.3d).These pores were generally defined by the arrangement of mineral particles in diagenesis processes.Some fracture-related pores were also observed,as shown in Fig.3a and d.The types of minerals detected by the energy dispersive X-ray spectroscopy were consistent with the results obtained by XRD (Table 1) on the whole.Fissures have been found in the Tamusu mudstone,which can be also detected in other HLW underground disposal host rocks,like Opalinus clay(Seiphoori et al.,2017) and Boom Clay (Ma,2017).The mineral composition and microstructure of these mudstones will vary due to different tectonic movements and diagenesis.However,abundant microcracks and fissures existed in these rocks,which may form a huge spatial network together with pores,influencing the long-term stability of the HLW geological repository,and also the migration of nuclides in the host medium.The images,as shown in Fig.3,are representatives of the Tamusu mudstone microstructure in the vicinity at the depth of 500 m.The microstructure can be varied due to the limitations of SEM specimen size as well as the nonuniformity and the random particle arrangement of the Tamusu mudstone.

Table 1 Quantitative analysis of whole-rock minerals in the Tamusu mudstone.

Table 2 Relative content of clay minerals in the Tamusu mudstone.

3.3.N2 physisorption isotherms and mercury intrusion-extrusion curves

The N2adsorption/desorption isotherms for Tamusu mudstone samples (i.e.TMS N1,TMS N2,TMS N3,TMS N4) at liquid N2temperature(-195.85°C)are presented in Fig.4.The isotherms exhibit a hysteresis pattern but with no plateau at high relative pressures.Strictly speaking,these isotherms do not fall within the IUPAC classification.Nevertheless,they are combinations of types I and IV.At low relative pressures,adsorption was found from the isotherms,indicating the presence of micropores (type I).While at high relative pressures (P/P0＞0.45),the curves show hysteresis loops,which indicate the multilayer range related to capillary condensation in mesopores (type IV).However,unlimited adsorption is evident at large relative pressures.At the relative pressure P/P0≈1,the shape of the isotherms is still hyperbolic,indicating that a range of macropores was contained in the Tamusu mudstone,which could not be detected by N2physisorption.The adsorption amount of this relative pressure varied from 4.31 cm3/g to 6.77 cm3/g for different samples.

Based on the IUPAC classification,a careful examination of the hysteresis loops for Tamusu mudstone revealed that they are between types H2 and H3,suggesting the occurrence of several pore shapes,such as inkbottle-shaped pores,which are typically within clay minerals and organic matters,and slit-shaped pores,which are typically from the aggregates of plate-like particles (Sing et al.,1985; Tian et al.,2013; Wang et al.,2020).However,this interpretation may be subject to error(Tian et al.,2013,2015;Schmitt et al.,2013).Caution was also recommended by Clarkson et al.(2012a,b,2013),who used gas adsorption and USANS/SANS to study tight gas sandstones.The authors found that the assumption of slit-shaped pores inferred from the shape of the hysteresis loops was inconsistent with the SANS scattering results.The isotherms at the desorption branch showed the “forced closure” at a relative pressure close to 0.45,which was caused by the hemispherical meniscus’s instability during desorption in pores with critical diameters of about 4 nm (Groen et al.,2003),referred to as the “tensile strength effect” (Tian et al.,2013).

The mercury intrusion-extrusion curves for Tamusu mudstone samples (i.e.TMS M1,TMS M2,TMS M3,TMS M4) obtained from the MIP tests are presented in Fig.5.All test samples exhibited small pore volumes,i.e.0.011-0.022 cm3/g,at the end of the intrusion stage.The intrusion-extrusion curves show clear hysteresis between the intrusion and extrusion branches,which means that when the extrusion finished,a large amount of intruding mercury remained in the samples.More specifically,0.005-0.015 cm3/g of mercury remaining in the pore system after the extrusion pressure reached atmospheric pressure,which was verified by weighing the test samples after the experiment.Kaufmann et al.(2009) attributed this phenomenon to the role of the pore network connectivity during the processes of filling and emptying mercury,even after equilibrium,regardless of time.As the compressibility of mercury is too low for force extrusion by pressure,the inkbottle-shaped pores remained,which were filled with mercury during the extrusion.This result is supported by Porcheron and Monson (2004) and Schmitt et al.(2013),who proposed that the network effects and ink-bottle snap-off most likely caused the entrapment of the mercury in the analyzed geomaterials.This conclusion is consistent with the observation from the N2adsorption/desorption isotherms,which exhibited hysteresis loops varying between types H2 and H3.Capillary pressure curves with bimodal or multimodal distributions indicate that the Tamusu mudstone has complex and heterogeneous lithologies and a wide range of pore sizes.

3.4.PSD

Fig.3.Pore morphology and energy dispersive X-ray spectroscopy for Tamusu mudstone.Note:Only one energy spectrum analysis chart is given for the same mineral.(a)and(b)Pore morphology of Tamusu mudstone; (c) Pores filled with carbonate or silica cement; (d) Organic matter and microfractures in Tamusu mudstone.

The Barrett-Joyner-Halenda(BJH)model is commonly used to calculate PSD curves in experiments of N2physisorption (Clarkson et al.,2012a).The PSD can be derived from N2physisorption isotherms,from either adsorption or desorption branches.However,the application of the BJH model to the desorption branch tends to be more influenced by the pore network than the adsorption branch,as the formation of the hysteresis loops vacillates between types H2 and H3 (Sing et al.,1985).Significant variances between the calculation from adsorption in comparison with that from desorption would be obtained(Groen et al.,2003;Kaufmann et al.,2009).The plot of dV/dD (V and D represent the pore volume and pore diameter,respectively) versus D for Tamusu mudstone is shown in Fig.6,with the comparison of PSD derived from the desorption and adsorption branches of isotherms.The desorption branches in the BJH model caused large,artificial pores to peak clearly in the vicinity of 4 nm,while this peak was not present in the PSD curves derived from adsorption branches.Thus,to avoid the above-mentioned phenomenon during desorption,the PSD determined with the BJH method applied from the adsorption isotherm was preferable.Fig.6 also shows that the Tamusu mudstone was mainly composed of mesopores less than 50 nm in diameter,with numerous pores measuring less than 30 nm.However,the diameters of considerable pores were larger than 50 nm.Additionally,micropores less than 6 nm were abundantly present in the Tamusu mudstone.However,these pores contribute little to the total pore volume.

Fig.4.N2 physisorption isotherms.

Fig.5.Mercury intrusion-extrusion cycles for Tamusu mudstone.

Fig.6.PSD curves derived from N2 adsorption/desorption isotherms.

The pore volume distribution with respect to the pore size of the Tamusu mudstone,as obtained in the MIP experiment,is presented in Fig.7.Numerous pores ranged from 3 nm to 50 nm in diameter for the Tamusu mudstone,which are aligned with the observation in the N2adsorption/desorption experiment.N2physisorption resulted in PSD information with pore diameter from approximately 1.7 nm to around 300 nm,whereas MIP measurements were between 3.6 nm and 100 μm.A common pore size interval was included by both N2adsorption/desorption and MIP testing; thus,the data in this range are shown together in Fig.8.The modes exhibited in MIP experiments are at 3-4 nm and 20-30 nm.Similar modes are observed in N2physisorption,although the peak position does not match accurately.N2physisorption suggested a higher pore diameter of 40-60 nm.Some researchers associated this discrepancy with different testing conditions of the two methods(Labani et al.,2013; Hinai et al.,2014).The PSD of the N2adsorption measurement was conducted at nearly ambient pressure,while the MIP measurement was performed under different intrusion pressures (up to 414 MPa).Indeed,the high mercury intrusion pressure might result in the underestimation of large pores and the overestimation of small pores due to the complex pore shapes and the compressibility of grains (Hildenbrand and Urai,2003; Kuila and Prasad,2011; Klaver et al.,2015).The PSD curves derived from N2adsorption and MIP tests correlate well in general.This consistency was further verified by Kuila and Prasad(2011).

Data derived from N2physisorption and MIP measurements are listed in Table 3.The total pore volume of the Tamusu mudstone varied from 0.0065 cm3/g to 0.0103 cm3/g from the calculation of N2physisorption,which generally resulted in lower values than that of MIP.This difference may be caused by the existence of large macropores,which cannot be detected by the N2adsorption measurement.The pore volume obtained from N2physisorption was calculated by taking the volume uptake at the relative pressure of 0.994.This relative pressure corresponds to a pore diameter of around 320 nm,and hence,the pore volume indicated by N2physisorption includes pores smaller than this pore diameter.Meanwhile,the pores with a diameter smaller than 3.6 nm,which only have a small influence on the total pore volume,are not considered in the MIP measurement.The average pore diameters calculated from N2physisorption are comparable to that from MIP as shown in Table 3.

Some researchers question mercury intrusion data obtained at pressures larger than 70 MPa,as the potential particles may break down and some closed pores may open up at high mercury intrusion pressure (Bustin et al.,2008).However,this concern may be invalid,as the existing studies have shown that the pore sizes larger than 10 nm are almost unaffected under high mercury intrusion pressure and the pore-throat size drift caused by pore compression is insignificant (Peng et al.,2017).The permeability for Tamusu mudstone predicted from MIP test results ranges from 0.11 mD to 0.79 mD.The Tamusu mudstone porosity,derived from the total volume of intruded mercury,was in the range of 2.34%-2.84%.The percentage of the pore volume was calculated using the BJH model by N2physisorption,as shown in Fig.9.Mesopores accounted for the greater part of the total pore volume in all Tamusu mudstone samples,ranging from 71% to 84%.

Fig.7.PSD curves derived from MIP.

Fig.8.PSD curves derived from both N2 adsorption and MIP.

Fig.9.Percentages of pore volume.

3.5.Specific surface area

The specific surface area was determined differently for the N2physisorption and MIP experiments.For the calculations from N2adsorption isotherms,the specific surface area was determined by means of the Brunauer-Emmett-Teller (BET) and t-plot methods.A relative pressure range between 0.069 and 0.199 (Sing et al.,1985),where pure adsorption is assumed to be predominant and no Kelvin condensation is expected(Kaufmann et al.,2009;Schmitt et al.,2013),and a cross-sectional area of 0.162 nm2for the N2molecule were used for calculations by the BET theory.The calculation from the MIP data was carried out based on a cylindrical pore model.The specific surface areas calculated from the N2adsorption isotherms(using both the BET equation and the t-plot method)and from MIP are presented in Table 3.

Table 3 Pore structure parameters by low-pressure N2 adsorption and MIP.

For the N2physisorption analysis,the specific surface area using the BET equation varies from 1.29 m2/g to 2.86 m2/g.This range is comparable to that calculated by the t-plot method,which is 1.37-3.04 m2/g.The specific surface area derived from MIP varies from 2.572 m2/g to 5.861 m2/g.These higher values may correlate to the pore network effects and large pore volume(ink-bottles),which is attributed to smaller sizes(neck entrances)with a large surface-tovolume ratio.Additionally,the technique limits may be the cause.For example,the high injection pressure may destroy the pore structure by damaging pore walls or compressing soft material(Bustin et al.,2008; Huang et al.,2019; Yu et al.,2019),or the hypothesis of cylindrical pore shape is inaccurate.

Fig.10.Percentages of specific surface area.

The percentage of the specific surface area is derived using the BJH model by N2physisorption,as shown in Fig.10.Similar to their percentage of pore volume,mesopores still have overwhelming superiority in surface area over micropores or macropores,ranging from 76% to 89%.However,the contribution of micropores to specific surface area is much greater than that of pore volume,with 7%-18%for the specific surface area compared to 5%-10%for pore volume.In contrast,macropores account for about 2%of the specific surface area,which is far less than its contribution(around 10%)in pore volume.

Clay minerals may have a great influence on the mudstone’s pore structure (Rutherford et al.,1997; Kuila and Prasad,2013;Rexer et al.,2014; Obour et al.,2019; Liu et al.,2020).Liu et al.(2020) indicated that abundant modified pores in mudstone may be closely associated with clay minerals,and the increase of clay mineral content may lead to the increases of pore volume and specific surface area.Among the clay minerals,illite/smectite (I/S)contributes more to the specific surface area than any other type of clay mineral (Kuila and Prasad,2013).

4.Conclusions

In the present work,several main mineral types were first identified in Tamusu mudstone,including carbonates (dolomite,ankerite),feldspar(plagioclase,K-feldspar),clays and analcime.The type of Tamusu mudstone was ascertained to vary between mixed carbonate mudstone and carbonate siliceous mudstone,according to the terminology and classification for mudstones from Gamero-Diaz et al.(2013).Secondly,the pore structure of Tamusu mudstone was investigated by the implementation of both direct (FE-SEM)and indirect(high-pressure mercury and low-pressure adsorption)methods.Our studies showed that the pores in Tamusu mudstone are dominantly nanoscale,and the majority of nanopores are located within their mineral matrix or at its interface with nonporous minerals.Numerous pores in this mudstone are slot-or inkbottle-shaped,as evidenced by the shape of the adsorption isotherm hysteresis loops,which align with the N2adsorption-and MIP-derived dominant pore size and FE-SEM imaging.For the first time,the PSD obtained from BJH has been compared with that from MIP.The dominant pore size interpreted from each method had a reasonable agreement,suggesting that pore throats and pore bodies have similar dimensions.

Our FE-SEM,MIP and low-pressure adsorption-based characterizations of nanopores make up a crucial first step toward a better understanding of the distribution and inducement of pore development in Tamusu mudstone.Further research is necessary to well interpret the relationships between pore information,nanopore distribution and fluid flow in Tamusu mudstone.These meaningful results can be used in future studies as input to model single-and multi-phase fluid flow and radionuclide transport in porous geomedium.Moreover,these findings can be used to evaluate fluid flow characteristics of mudstones,such as permeability.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

The financial support of the National Natural Science Foundation of China (Grant Nos.51979266,51879258 and 51991392) is greatly acknowledged.Thanks are extended to Dr.M.D.Sun,from the Department of Earth Sciences,Northeast Petroleum University,China,for his help in the analysis of N2physisorption isotherm and mercury injection capillary pressure.